Conceptoscillators 1b - Pitch Experiment FInal.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    }
   ],
   "source": [
    "\n",
    "%pylab inline\n",
    "from IPython.display import Audio\n",
    "import librosa\n",
    "import scipy as sp\n",
    "\n",
    "figsize(20,6)\n",
    "\n",
    "import tensorflow as tf\n",
    "tf.enable_eager_execution()\n",
    "\n",
    "prefix=\"osc_rect\"\n",
    "def filepre(nm):\n",
    "    return \"tmp/\"+prefix+\"_\"+nm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def nrmse(output,target):\n",
    "    combinedVar = 0.5 * (np.var(target, ddof=1) + np.var(output, ddof=1))\n",
    "    errorSignal = output - target\n",
    "    return np.sqrt(np.mean(errorSignal ** 2) / combinedVar)\n",
    "\n",
    "def generateInternalWeights(nInternalUnits, connectivity):\n",
    "    success = False\n",
    "    internalWeights = 0\n",
    "    while success == False:\n",
    "        try:\n",
    "            internalWeights = np.random.randn(nInternalUnits,nInternalUnits) * (np.random.random((nInternalUnits,nInternalUnits)) < connectivity)\n",
    "            specRad = max(abs(np.linalg.eig(internalWeights)[0]))\n",
    "            if (specRad > 0):\n",
    "                internalWeights = internalWeights / specRad\n",
    "                success = True\n",
    "        except e:\n",
    "            print(e)\n",
    "    return internalWeights\n",
    "\n",
    "pLoop = lambda n,p: p[n%p.size]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "def compareMelSpec(seq1, seq2):\n",
    "    fftSize=2048\n",
    "    hop=1024  \n",
    "    melspec = librosa.feature.melspectrogram(y=seq1, sr=sr, n_fft=fftSize,hop_length=hop)\n",
    "    melspec2 = librosa.feature.melspectrogram(y=seq2, sr=sr, n_fft=fftSize,hop_length=hop)\n",
    "    return nrmse(melspec.flatten(), melspec2.flatten())\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#using digital bandlimited oscillators, from https://gist.github.com/endolith/407991 - thanks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "def bl_sawtooth(t): # , width=1\n",
    "    \"\"\"\n",
    "    Return a periodic band-limited sawtooth wave with\n",
    "    period 2*pi which is falling from 0 to 2*pi and rising at\n",
    "    2*pi (opposite phase relative to a sin)\n",
    "    Produces the same phase and amplitude as scipy.signal.sawtooth.\n",
    "    Examples\n",
    "    --------\n",
    "    >>> t = linspace(0, 1, num = 1000, endpoint = False)\n",
    "    >>> f = 5 # Hz\n",
    "    >>> plot(bl_sawtooth(2 * pi * f * t))\n",
    "    \"\"\"\n",
    "    t = asarray(t)\n",
    "\n",
    "    if abs((t[-1]-t[-2]) - (t[1]-t[0])) > .0000001:\n",
    "        raise ValueError(\"Sampling frequency must be constant\")\n",
    "\n",
    "    if t.dtype.char in ['fFdD']:\n",
    "        ytype = t.dtype.char\n",
    "    else:\n",
    "        ytype = 'd'\n",
    "    y = zeros(t.shape, ytype)\n",
    "\n",
    "    # Get sampling frequency from timebase\n",
    "    fs =  1 / (t[1] - t[0])\n",
    "    #    fs =  1 / amax(diff(t))\n",
    "\n",
    "    # Sum all multiple sine waves up to the Nyquist frequency\n",
    "\n",
    "    # TODO: Maybe choose between these based on number of harmonics?\n",
    "\n",
    "    # Slower, uses less memory\n",
    "    for h in range(1, int(fs*pi)+1):\n",
    "        y += 2 / pi * -sin(h * t) / h\n",
    "\n",
    "    # Faster, but runs out of memory and dies\n",
    "#    h = arange(1, int(fs * pi) + 1)\n",
    "#    phase = outer(t, h)\n",
    "#    y = 2 / pi * -sin(phase) / h\n",
    "#    y = sum(y, axis=1)\n",
    "\n",
    "    return y\n",
    "\n",
    "\n",
    "def bl_triangle(t):\n",
    "    \"\"\"\n",
    "    Return a periodic band-limited triangle wave with\n",
    "    period 2*pi which is falling from 0 to pi and rising from\n",
    "    pi to 2*pi (same phase as a cos)\n",
    "    Produces the same phase and amplitude as scipy.signal.sawtooth(width=0.5).\n",
    "    Examples\n",
    "    --------\n",
    "    >>> t = linspace(0, 1, num = 1000, endpoint = False)\n",
    "    >>> f = 5 # Hz\n",
    "    >>> plot(bl_triangle(2 * pi * f * t))\n",
    "    \"\"\"\n",
    "    t = asarray(t)\n",
    "\n",
    "    if abs((t[-1]-t[-2]) - (t[1]-t[0])) > .0000001:\n",
    "        raise ValueError(\"Sampling frequency must be constant\")\n",
    "\n",
    "    if t.dtype.char in ['fFdD']:\n",
    "        ytype = t.dtype.char\n",
    "    else:\n",
    "        ytype = 'd'\n",
    "    y = zeros(t.shape, ytype)\n",
    "\n",
    "    # Get sampling frequency from timebase\n",
    "    fs =  1 / (t[1] - t[0])\n",
    "\n",
    "    # Sum all odd multiple sine waves up to the Nyquist frequency\n",
    "\n",
    "    # Slower, uses less memory\n",
    "    for h in range(1, int(fs * pi) + 1, 2):\n",
    "        y += 8 / pi**2 * -cos(h * t) / h**2\n",
    "\n",
    "    # Faster, but runs out of memory and dies\n",
    "#    h = arange(1, int(fs * pi) + 1, 2)\n",
    "#    phase = outer(t, h)\n",
    "#    y = 8 / pi**2 * -cos(phase) / h**2\n",
    "#    y = sum(y, axis=1)\n",
    "\n",
    "    return y\n",
    "\n",
    "\n",
    "def bl_square(t, duty=0.5):\n",
    "    \"\"\"\n",
    "    Return a periodic band-limited square wave with\n",
    "    period 2*pi which is +1 from 0 to pi and -1 from\n",
    "    pi to 2*pi (same phase as a sin)\n",
    "    Produces the same phase and amplitude as scipy.signal.square.\n",
    "    Similarly, duty cycle can be set, or varied over time.\n",
    "    Examples\n",
    "    --------\n",
    "    >>> t = linspace(0, 1, num = 10000, endpoint = False)\n",
    "    >>> f = 5 # Hz\n",
    "    >>> plot(bl_square(2 * pi * f * t))\n",
    "    >>> sig = np.sin(2 * np.pi * t)\n",
    "    >>> pwm = bl_square(2 * np.pi * 30 * t, duty=(sig + 1)/2)\n",
    "    >>> plt.subplot(2, 1, 1)\n",
    "    >>> plt.plot(t, sig)\n",
    "    >>> plt.subplot(2, 1, 2)\n",
    "    >>> plt.plot(t, pwm)\n",
    "    >>> plt.ylim(-1.5, 1.5)\n",
    "    \"\"\"\n",
    "    return bl_sawtooth(t - duty*2*pi) - bl_sawtooth(t) + 2*duty-1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.python.client import device_lib\n",
    "device_lib.list_local_devices()\n",
    "tfdevice='/gpu:0'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 5 µs, sys: 0 ns, total: 5 µs\n",
      "Wall time: 7.63 µs\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "def computeConceptor(p, net, i_pattern, alpha):\n",
    "#     print('Computing conceptor, alpha: ', alpha)\n",
    "    with tf.device(tfdevice):    \n",
    "        Cs = np.zeros((4, 1), dtype=np.object)\n",
    "        R = net['patternRs'][0,i_pattern]\n",
    "#         [s,U,V] = tf.svd(R)\n",
    "        [U,s,V] = svd(R)\n",
    "\n",
    "#         S = diag(s)\n",
    "        S = tf.diag(s)\n",
    "    #     U = U.numpy()\n",
    "\n",
    "        sinv = tf.matrix_inverse(tf.add(S, tf.multiply(double(pow(alpha, -2)), tf.eye(p['N'], dtype=float64))))\n",
    "        Snew = tf.matmul(S,sinv)\n",
    "    #     Snew = tf.matmul(Snew, tf.eye(p['N'], dtype=float64))\n",
    "        Snew = Snew.numpy()\n",
    "    #     Snew = (S * linalg.inv(S + pow(alpha, -2) * np.eye(p['N'])))\n",
    "\n",
    "        C = tf.matmul(U,Snew)\n",
    "        C = tf.matmul(C,U, adjoint_b=True).numpy()\n",
    "#     C =  U.dot(Snew).dot(U.conj().T);\n",
    "\n",
    "    Cs[0,0] = C\n",
    "    Cs[1,0] = U\n",
    "    Cs[2,0] = diag(Snew)\n",
    "    Cs[3,0] = diag(S)\n",
    "\n",
    "    return locals()\n",
    "\n",
    "def testConceptor(p, cNet, recallTestLength):\n",
    "    with tf.device(tfdevice):    \n",
    "        trials = 1\n",
    "        attens = np.zeros(trials)\n",
    "        C = cNet['Cs'][0,0]\n",
    "        for i_trial in range(trials):\n",
    "            x_CTestPL = np.zeros((p['N'], recallTestLength))\n",
    "            z_CTestPL = np.zeros((p['N'], recallTestLength))\n",
    "            x = 0.5 * np.random.randn(p['N'],1)\n",
    "            z = x\n",
    "    #         matW = tf.constant(cNet['net']['W'])\n",
    "            for n in range(recallTestLength + p['washoutLength']):\n",
    "                xOld = x\n",
    "                Wtarget = (cNet['net']['W'].dot(x))\n",
    "    #             Wtarget = tf.matmul(matW, x).numpy()\n",
    "\n",
    "                z = ((1.0-p['LR']) * xOld) + (p['LR'] * tanh(Wtarget + cNet['net']['Wbias']))\n",
    "\n",
    "    #             xOldLR = tf.multiply(tf.constant(1.0-p['LR'], dtype=float64), xOld)\n",
    "    #             biasedTarget = tf.add(Wtarget, cNet['net']['Wbias'])\n",
    "    #             biasedTarget = tf.tanh(biasedTarget)\n",
    "    #             biasedTargetLR = tf.multiply(tf.constant(p['LR'], dtype=float64), biasedTarget)\n",
    "    #             z = tf.add(xOldLR, biasedTargetLR)\n",
    "\n",
    "                x = C.dot(z)\n",
    "    #             x = tf.matmul(C,z).numpy()\n",
    "                xPrev = x\n",
    "                if (n > p['washoutLength']):\n",
    "                    x_CTestPL[:,n-p['washoutLength']] = x.T\n",
    "    #                 z_CTestPL[:,n-p['washoutLength']] = z.numpy().T\n",
    "                    z_CTestPL[:,n-p['washoutLength']] = z.T\n",
    "    #                 p_CTestPL[:,n-p['washoutLength']] = cNet['net']['Wout'].dot(np.concatenate((x[:,0], np.array([1]))))\n",
    "\n",
    "    #         attenuation = np.mean(pow(np.sum(z_CTestPL[:,:] - x_CTestPL[:,:], axis=1),2)) / np.mean(pow(np.sum(z_CTestPL[:,:], axis=1),2))\n",
    "            attenuation = np.mean(pow(np.linalg.norm(z_CTestPL[:,:] - x_CTestPL[:,:], axis=1),2)) / np.mean(pow(np.linalg.norm(z_CTestPL[:,:], axis=1),2))\n",
    "            attens[i_trial] = attenuation\n",
    "        \n",
    "    return np.mean(attens)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "\n",
    "#create a random reservoir and load it with patterns using Jaeger's method\n",
    "\n",
    "def makeLoadedNetwork_v2(p):\n",
    "    with tf.device(tfdevice):\n",
    "        Netconnectivity = 1\n",
    "        if p['N'] > 20:\n",
    "            Netconnectivity = 10.0/p['N'];\n",
    "        WstarRaw = generateInternalWeights(p['N'], Netconnectivity)\n",
    "        WinRaw = 2 * (np.random.rand(p['N'], 1) - 0.5)\n",
    "        WbiasRaw = 2 * (np.random.rand(p['N'], 1) - 0.5)\n",
    "\n",
    "        #Scale raw weights     \n",
    "        Wstar = p['NetSR'] * WstarRaw;\n",
    "        Win = p['NetinpScaling'] * WinRaw;\n",
    "        Wbias = p['BiasScaling'] * WbiasRaw;  \n",
    "        I = np.eye(p['N'])\n",
    "        xCollector = np.zeros((p['N'], p['learnLength']))\n",
    "        pCollector = np.zeros((1, p['learnLength']))\n",
    "        x = np.zeros((p['N'],1))\n",
    "\n",
    "        allTrainxArgs = np.zeros((p['N'] + 1, 0));\n",
    "        allTrainOldxArgs = np.zeros((p['N'], 0));\n",
    "        allTrainWtargets = np.zeros((p['N'], 0));\n",
    "        allTrainOuts = np.zeros((1, 0));\n",
    "        patternRs =  np.zeros((1, p['patts'].shape[0]), dtype=np.object)\n",
    "        print('Loading patterns: ', end='')\n",
    "        for i_pattern in range(p['patts'].shape[0]):\n",
    "            print(i_pattern, \" \", end='')\n",
    "            patt = p['patts'][i_pattern]\n",
    "            xCollector = np.zeros((p['N'] + 1, p['learnLength']));\n",
    "            xOldCollector = np.zeros((p['N'], p['learnLength']));\n",
    "            WTargetCollector = np.zeros((p['N'], p['learnLength']));\n",
    "            pCollector = np.zeros((1, p['learnLength']));\n",
    "            x = np.zeros((p['N'], 1));\n",
    "\n",
    "            for n in range(p['washoutLength'] + p['learnLength']):\n",
    "                u = patt.take(n, mode='wrap')\n",
    "                xOld = x\n",
    "    #             Wtarget = (Wstar.dot(x)) + (Win.dot(u))\n",
    "                wstarx=tf.matmul(Wstar,x)\n",
    "                winu = tf.multiply(Win,u)\n",
    "                Wtarget = tf.add(wstarx, winu)\n",
    "    #             x = ((1.0-p['LR']) * xOld) + (p['LR'] * tanh(Wtarget + Wbias))\n",
    "                xOldLR = tf.multiply(tf.constant(1.0-p['LR'], dtype=float64), xOld)\n",
    "                biasedTarget = tf.add(Wtarget, Wbias)\n",
    "                biasedTarget = tf.tanh(biasedTarget)\n",
    "                biasedTargetLR = tf.multiply(tf.constant(p['LR'], dtype=float64), biasedTarget)\n",
    "                x = tf.add(xOldLR, biasedTargetLR)\n",
    "                if n >= p['washoutLength']:\n",
    "                    xCollector[:, n - p['washoutLength']] = np.concatenate((x[:,0], np.array([1])))\n",
    "                    xOldCollector[:, n - p['washoutLength']] = xOld[:,0]\n",
    "                    WTargetCollector[:, n - p['washoutLength']] = Wtarget[:,0]\n",
    "                    pCollector[0, n - p['washoutLength']] = u\n",
    "                uOld = u\n",
    "\n",
    "            R = xCollector[0:-1].dot(xCollector[0:-1].T) / p['learnLength']\n",
    "            patternRs[0,i_pattern] = R\n",
    "            allTrainxArgs = np.concatenate((allTrainxArgs, xCollector), axis=1)\n",
    "            allTrainOldxArgs = np.concatenate((allTrainOldxArgs, xOldCollector), axis=1)\n",
    "            allTrainOuts = np.concatenate((allTrainOuts, pCollector), axis=1)\n",
    "            allTrainWtargets = np.concatenate((allTrainWtargets, WTargetCollector), axis=1)\n",
    "\n",
    "\n",
    "        Wout = (linalg.inv(allTrainxArgs.dot(allTrainxArgs.conj().T) +\n",
    "                          (p['TychonovAlphaReadout'] * np.eye(p['N'] + 1))).dot(allTrainxArgs).dot(allTrainOuts.conj().T)).conj().T\n",
    "\n",
    "        outsRecovered = Wout.dot(allTrainxArgs);\n",
    "        NRMSE_readout = mean(nrmse(outsRecovered, allTrainOuts))\n",
    "        absSize_readout = mean(mean(abs(Wout), axis=0))\n",
    "        print(\"\\nNRMSE readout: \", NRMSE_readout, \" :: \", end='')\n",
    "        print(\"absSize readout: \", absSize_readout)\n",
    "\n",
    "        W = (linalg.inv(allTrainOldxArgs.dot(allTrainOldxArgs.conj().T) +\n",
    "                          (p['TychonovAlpha'] * np.eye(p['N']))).dot(allTrainOldxArgs).dot(allTrainWtargets.conj().T)).conj().T\n",
    "        NRMSE_W = mean(nrmse(W.dot(allTrainOldxArgs), allTrainWtargets))\n",
    "        absSize_W = mean(mean(abs(W), axis=0))\n",
    "        print(\"NRMSE W: \", NRMSE_W, \" :: \", end='')\n",
    "        print(\"absSize W: \", absSize_W)\n",
    "\n",
    "        data ={k: v for k, v in locals().items() if k in \n",
    "               ('p','Win','Wstar', 'Wbias','NRMSE_W', 'absSize_W','patternRs','W',\n",
    "                'Wout','NRMSE_readout', 'absSize_readout')}\n",
    "        return data\n",
    "\n",
    "\n",
    "    \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def conceptorAnd(C1, C2):\n",
    "    return np.linalg.inv(np.linalg.inv(C1) + np.linalg.inv(C2) - np.eye(C1.shape[0]))\n",
    "\n",
    "def render(model, srCtl, lrCtl, oversample=1, rl=8000):\n",
    "    audio = np.zeros(rl)\n",
    "    with tf.device(tfdevice):\n",
    "        cNet = model[0]\n",
    "        p = cNet['net']['p']\n",
    "        x = tf.contrib.eager.Variable(0.5 * np.random.randn(p['N'],1))\n",
    "    #     x = 0.5 * np.random.randn(p['N'],1)\n",
    "\n",
    "        C = tf.constant(cNet['Cs'][0,0])\n",
    "        srCtl = tf.constant(srCtl)\n",
    "        orgLR = p['LR']\n",
    "        LR = orgLR\n",
    "        matW = tf.constant(cNet['net']['W'])\n",
    "        matBias = tf.constant( cNet['net']['Wbias'])\n",
    "        matWout = tf.constant(cNet['net']['Wout'])  \n",
    "        tfone = tf.constant(double(1.0), dtype=float64)\n",
    "        #run to washout\n",
    "        for n in range(500):\n",
    "            xOld = x\n",
    "            Wtarget = tf.matmul(matW, x)\n",
    "            xOldLR = tf.multiply(tf.constant(1.0-p['LR'], dtype=float64), xOld)\n",
    "            biasedTarget = tf.add(Wtarget, matBias)\n",
    "            biasedTarget = tf.tanh(biasedTarget)\n",
    "            biasedTargetLR = tf.multiply(tf.constant(p['LR'], dtype=float64), biasedTarget)\n",
    "            z = tf.add(xOldLR, biasedTargetLR)\n",
    "            x = tf.matmul(C,z)\n",
    "\n",
    "\n",
    "        for n in range(rl):\n",
    "            LR = double(orgLR * lrCtl[n])\n",
    "#             if n % 500 == 0:\n",
    "#                 print(n, \" / \", rl)\n",
    "#                 print(LR)\n",
    "            newSample=0\n",
    "            for i_oversample in range(oversample):\n",
    "                xOld = x\n",
    "                Wtarget = tf.matmul(tf.multiply(tf.constant(srCtl[n]),matW), x)\n",
    "\n",
    "                xOldLR = tf.multiply(tf.constant(1.0-LR, dtype=float64), xOld)\n",
    "                biasedTarget = tf.add(Wtarget, matBias)\n",
    "                biasedTarget = tf.tanh(biasedTarget)\n",
    "                biasedTargetLR = tf.multiply(tf.constant(LR, dtype=float64), biasedTarget)\n",
    "                z = tf.add(xOldLR, biasedTargetLR)\n",
    "                x = tf.matmul(C,z)\n",
    "                xBias = tf.concat([tf.transpose(x)[0],[1]],0)\n",
    "                xBias = tf.expand_dims(xBias,1)\n",
    "                netOutput = tf.matmul(matWout, xBias)\n",
    "                newSample = newSample + netOutput.numpy()\n",
    "\n",
    "            audio[n] = newSample/oversample\n",
    "    return audio\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fbc58d14b70>]"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pitchMults = pow(2,arange(13)/12)\n",
    "freqMuls = np.concatenate((0.5 * pitchMults, pitchMults[1:]))\n",
    "figsize(20,4)\n",
    "plot(freqMuls)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "testAlgos = [bl_sawtooth, bl_square, bl_triangle]\n",
    "testFreqs = [40,80,160,320,640, 1280, 2560]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====== running test:  40 <function bl_sawtooth at 0x7fc012258e18>\n",
      "LR:  0.01\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.33586244282768807  :: absSize readout:  1.017320151456425\n",
      "NRMSE W:  0.06437871527726818  :: absSize W:  0.5162673498288932\n",
      "NRMSEs: best:  0.06437871527726818 avg:  0.06437871527726818\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  9.44643e-08        initial\n",
      "    2        618.034  2.64958e-07        golden\n",
      "    3        236.068  1.24298e-07        golden\n",
      "    4        351.159  9.85595e-08        parabolic\n",
      "    5        472.136  8.94057e-08        golden\n",
      "    6        471.219  1.25224e-07        parabolic\n",
      "    7        527.864  2.56669e-07        golden\n",
      "    8        493.422  1.16016e-07        golden\n",
      "    9        480.267  1.21499e-07        golden\n",
      "   10        475.242  8.98853e-08        golden\n",
      "   11        473.322  1.42617e-07        golden\n",
      "   12        472.589  1.27877e-07        golden\n",
      "   13        471.786  1.44934e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42976608 1.23498363 1.40020446 1.01034748 1.40231559]\n",
      "Trial error:  1.4002044642229308\n",
      "Best so far ***************** \n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.33913515496728586  :: absSize readout:  1.0904446911824792\n",
      "NRMSE W:  0.06542454730044471  :: absSize W:  0.5872911881085477\n",
      "NRMSEs: best:  0.06542454730044471 avg:  0.06542454730044471\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    6.964e-07        initial\n",
      "    2        618.034  1.83076e-06        golden\n",
      "    3        236.068  3.29246e-07        golden\n",
      "    4         99.026   2.6764e-07        parabolic\n",
      "    5        136.779  3.12721e-07        parabolic\n",
      "    6        61.2014  2.85523e-07        golden\n",
      "    7         90.832   2.5811e-07        parabolic\n",
      "    8        84.3957  2.36699e-07        parabolic\n",
      "    9        75.5363  2.63456e-07        golden\n",
      "   10        83.6053   2.5298e-07        parabolic\n",
      "   11        87.1114  2.69803e-07        parabolic\n",
      "   12         85.433  3.40261e-07        golden\n",
      "   13         84.792  3.03804e-07        golden\n",
      "   14        84.0624  2.71997e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.30217506 1.02429875 1.11890678 1.18463295 0.69995148]\n",
      "Trial error:  1.1189067835114814\n",
      "Best so far ***************** \n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.3358299383673135  :: absSize readout:  1.055925934984017\n",
      "NRMSE W:  0.06598353191865851  :: absSize W:  0.5518146741862073\n",
      "NRMSEs: best:  0.06598353191865851 avg:  0.06598353191865851\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  2.63712e-06        initial\n",
      "    2        618.034  6.56995e-06        golden\n",
      "    3        236.068  6.72439e-07        golden\n",
      "    4        145.898   4.6649e-07        golden\n",
      "    5        166.874  6.64448e-07        parabolic\n",
      "    6        90.1699  4.66018e-07        golden\n",
      "    7            118  5.33819e-07        parabolic\n",
      "    8        55.7281  5.44726e-07        golden\n",
      "    9        77.0143  5.06196e-07        golden\n",
      "   10          100.8  5.63436e-07        golden\n",
      "   11        85.1449  5.93637e-07        golden\n",
      "   12        94.2302  5.70034e-07        golden\n",
      "   13        88.2506  4.94355e-07        golden\n",
      "   14        91.7208  5.12189e-07        golden\n",
      "   15        89.4368   5.4027e-07        golden\n",
      "   16        90.7623  6.21243e-07        golden\n",
      "   17        89.8366  5.92932e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.07558078 0.91013256 1.03794684 1.27732622 1.38664811]\n",
      "Trial error:  1.075580779418215\n",
      "Best so far ***************** \n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.3415437300389686  :: absSize readout:  1.0546335643830793\n",
      "NRMSE W:  0.06636421060976525  :: absSize W:  0.5618269190491858\n",
      "NRMSEs: best:  0.06636421060976525 avg:  0.06636421060976525\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  8.60867e-07        initial\n",
      "    2        618.034  1.85333e-06        golden\n",
      "    3        236.068  4.00092e-07        golden\n",
      "    4        145.898  3.32537e-07        golden\n",
      "    5        154.275   3.1022e-07        parabolic\n",
      "    6        182.007  3.78563e-07        parabolic\n",
      "    7        164.868  3.32842e-07        golden\n",
      "    8        155.351  3.64209e-07        parabolic\n",
      "    9        153.942  3.77602e-07        parabolic\n",
      "   10        154.686  4.17082e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.78489531 0.73711019 0.79651035 1.05148429 0.50096645]\n",
      "Trial error:  0.7848953139464125\n",
      "Best so far ***************** \n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.3385852839560804  :: absSize readout:  1.0433178989311682\n",
      "NRMSE W:  0.06578133515010787  :: absSize W:  0.5706350480900982\n",
      "NRMSEs: best:  0.06578133515010787 avg:  0.06578133515010787\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  3.52015e-07        initial\n",
      "    2        618.034  3.35596e-06        golden\n",
      "    3        236.068  2.20495e-07        golden\n",
      "    4        294.456  1.38208e-07        parabolic\n",
      "    5        291.948  1.77762e-07        parabolic\n",
      "    6        327.882  1.79261e-07        golden\n",
      "    7        309.871  1.96292e-07        parabolic\n",
      "    8        294.789   1.4636e-07        parabolic\n",
      "    9        293.759  1.71413e-07        parabolic\n",
      "   10        294.123  2.00714e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.40706529 0.92953508 0.85188634 1.31853841 1.04275278]\n",
      "Trial error:  1.0427527784758817\n",
      "LR:  0.07999999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.08276774111472156  :: absSize readout:  0.34679706148981765\n",
      "NRMSE W:  0.025052364725250402  :: absSize W:  0.2526684680974075\n",
      "NRMSEs: best:  0.025052364725250402 avg:  0.025052364725250402\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00253937        initial\n",
      "    2        618.034   0.00227764        golden\n",
      "    3        763.932   0.00214703        golden\n",
      "    4        854.102   0.00201892        golden\n",
      "    5         909.83   0.00197831        golden\n",
      "    6        958.788   0.00194193        parabolic\n",
      "    7        940.088   0.00197916        golden\n",
      "    8         974.53   0.00196663        golden\n",
      "    9        944.705   0.00197224        parabolic\n",
      "   10        960.372   0.00193648        parabolic\n",
      "   11        964.442   0.00194438        parabolic\n",
      "   12        961.927   0.00194691        golden\n",
      "   13        960.039   0.00194951        parabolic\n",
      "   14        960.966   0.00196385        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42497976 1.42494918 1.42506155 1.42467526 1.42466442]\n",
      "Trial error:  1.4249491781242067\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.08649465566285887  :: absSize readout:  0.3559767715138915\n",
      "NRMSE W:  0.031204139571532546  :: absSize W:  0.2839005526797521\n",
      "NRMSEs: best:  0.031204139571532546 avg:  0.031204139571532546\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000131476        initial\n",
      "    2        618.034   0.00105384        golden\n",
      "    3        236.068  0.000174391        golden\n",
      "    4        322.388  0.000161679        parabolic\n",
      "    5        472.136   0.00084946        golden\n",
      "    6        416.408  0.000611677        golden\n",
      "    7        359.209  0.000147568        golden\n",
      "    8        395.122  0.000619329        golden\n",
      "    9        373.274  0.000139396        golden\n",
      "   10        386.991  0.000113904        golden\n",
      "   11        390.097  0.000512518        golden\n",
      "   12        385.072  0.000519817        golden\n",
      "   13        388.177  0.000510638        golden\n",
      "   14        386.258  0.000120905        golden\n",
      "   15        387.444  0.000510088        golden\n",
      "   16        386.658  0.000123163        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43856454 1.40547673 1.43626155 1.43651111 1.43639114]\n",
      "Trial error:  1.436391140359972\n",
      "Trial:  2\n",
      "0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.08187779882578904  :: absSize readout:  0.349974309465539\n",
      "NRMSE W:  0.027103975887337647  :: absSize W:  0.2674047003029042\n",
      "NRMSEs: best:  0.027103975887337647 avg:  0.027103975887337647\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000158187        initial\n",
      "    2        618.034  0.000341655        golden\n",
      "    3        236.068  3.00868e-05        golden\n",
      "    4        145.898  3.01813e-05        golden\n",
      "    5        191.124  3.01518e-05        parabolic\n",
      "    6        291.796  4.56759e-05        golden\n",
      "    7        257.354  3.00788e-05        golden\n",
      "    8        258.422  3.00469e-05        parabolic\n",
      "    9         271.17  3.00961e-05        golden\n",
      "   10        263.292  3.00519e-05        golden\n",
      "   11         260.76   3.0063e-05        parabolic\n",
      "   12        259.315  3.00697e-05        golden\n",
      "   13        258.014  3.00807e-05        golden\n",
      "   14        258.763  3.00667e-05        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43337863 1.43439454 1.43394557 1.43334011 1.43307149]\n",
      "Trial error:  1.4333786333373033\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.08204549709483068  :: absSize readout:  0.3349226900725281\n",
      "NRMSE W:  0.028362292123346716  :: absSize W:  0.28477341618136526\n",
      "NRMSEs: best:  0.028362292123346716 avg:  0.028362292123346716\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00138977        initial\n",
      "    2        618.034   0.00161292        golden\n",
      "    3        236.068  0.000330618        golden\n",
      "    4        145.898  3.66121e-06        golden\n",
      "    5        73.1937  3.62109e-06        parabolic\n",
      "    6        109.533  3.40389e-06        parabolic\n",
      "    7        108.009  3.40119e-06        parabolic\n",
      "    8        104.799  3.38889e-06        parabolic\n",
      "    9        92.7266  3.38059e-06        golden\n",
      "   10        97.0922  3.36967e-06        parabolic\n",
      "   11        97.9324  3.35318e-06        parabolic\n",
      "   12        100.555  3.36652e-06        golden\n",
      "   13        98.8873  3.36226e-06        parabolic\n",
      "   14        98.2971  3.35667e-06        golden\n",
      "   15        97.5991   3.3591e-06        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.26152029 0.24684413 0.33307481 0.24430686 0.24738684]\n",
      "Trial error:  0.24738684024378893\n",
      "Best so far ***************** \n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.08336287484020204  :: absSize readout:  0.369579705377021\n",
      "NRMSE W:  0.024578320432717626  :: absSize W:  0.24757091070752968\n",
      "NRMSEs: best:  0.024578320432717626 avg:  0.024578320432717626\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000491078        initial\n",
      "    2        618.034  0.000745187        golden\n",
      "    3        236.068  0.000199098        golden\n",
      "    4        145.898  1.34107e-06        golden\n",
      "    5        90.1699  1.49846e-06        golden\n",
      "    6        118.128  1.38539e-06        parabolic\n",
      "    7        150.178   1.3396e-06        parabolic\n",
      "    8        152.425  1.35462e-06        parabolic\n",
      "    9        148.197  1.34977e-06        parabolic\n",
      "   10        151.036  1.34666e-06        golden\n",
      "   11        149.421  1.34075e-06        golden\n",
      "   12        150.511  1.32858e-06        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.39034785 0.30403933 0.28358523 0.3112439  0.32500179]\n",
      "Trial error:  0.31124390373164823\n",
      "LR:  0.15\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.03762187817049321  :: absSize readout:  0.20467876308783\n",
      "NRMSE W:  0.02205036791507495  :: absSize W:  0.2536413770096736\n",
      "NRMSEs: best:  0.02205036791507495 avg:  0.02205036791507495\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00114517        initial\n",
      "    2        618.034   0.00117339        golden\n",
      "    3        236.068  0.000842785        golden\n",
      "    4        145.898  0.000297526        golden\n",
      "    5        90.1699   0.00011897        golden\n",
      "    6        35.8186  4.34157e-05        parabolic\n",
      "    7        20.8153  2.13163e-05        parabolic\n",
      "    8        12.8646  2.35202e-05        golden\n",
      "    9        18.6577  2.09014e-05        parabolic\n",
      "   10        18.3243   2.0906e-05        parabolic\n",
      "   11         18.991   2.1032e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.32174531 0.32093684 0.32676586 0.38407798 0.32108204]\n",
      "Trial error:  0.32174530636063176\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.03869342653405471  :: absSize readout:  0.2203800563266859\n",
      "NRMSE W:  0.02265770871571014  :: absSize W:  0.2522369812487934\n",
      "NRMSEs: best:  0.02265770871571014 avg:  0.02265770871571014\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00516201        initial\n",
      "    2        618.034   0.00553486        golden\n",
      "    3        236.068   0.00409876        golden\n",
      "    4        145.898   0.00141636        golden\n",
      "    5        90.1699  0.000826322        golden\n",
      "    6         77.723   0.00102663        parabolic\n",
      "    7        104.507   0.00115355        parabolic\n",
      "    8        89.4843  0.000790638        parabolic\n",
      "    9        85.3354   0.00082405        parabolic\n",
      "   10        87.8996  0.000589423        golden\n",
      "   11        87.4863  0.000757331        parabolic\n",
      "   12        88.4541   0.00104748        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.40534906 1.36722091 1.35022954 1.34705209 1.39401747]\n",
      "Trial error:  1.367220912999117\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.038373383949194886  :: absSize readout:  0.21508597480775388\n",
      "NRMSE W:  0.02063552659368658  :: absSize W:  0.24730666790611772\n",
      "NRMSEs: best:  0.02063552659368658 avg:  0.02063552659368658\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00406534        initial\n",
      "    2        618.034   0.00508713        golden\n",
      "    3        236.068   0.00183387        golden\n",
      "    4        145.898  0.000810507        golden\n",
      "    5        90.1699  0.000297168        golden\n",
      "    6        55.7281  0.000297737        golden\n",
      "    7        73.0298   0.00030177        parabolic\n",
      "    8        111.456  0.000303507        golden\n",
      "    9        98.3006  0.000146276        golden\n",
      "   10        103.326  0.000200389        golden\n",
      "   11        98.6339  0.000149802        parabolic\n",
      "   12        95.1949  0.000114642        golden\n",
      "   13        93.2756  9.44726e-05        golden\n",
      "   14        92.0893  0.000297837        golden\n",
      "   15        94.0087  0.000102178        golden\n",
      "   16        92.8225  8.89804e-05        golden\n",
      "   17        92.4891  0.000296494        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.6988735  1.37631357 1.39114212 1.39614506 0.68017298]\n",
      "Trial error:  1.3763135712532033\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.035700755303333866  :: absSize readout:  0.20442637737496852\n",
      "NRMSE W:  0.019295179771617516  :: absSize W:  0.22592860019878258\n",
      "NRMSEs: best:  0.019295179771617516 avg:  0.019295179771617516\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00285385        initial\n",
      "    2        618.034   0.00317931        golden\n",
      "    3        236.068    0.0017634        golden\n",
      "    4        145.898  0.000590359        golden\n",
      "    5        90.1699  2.65565e-05        golden\n",
      "    6        55.7281  1.33102e-05        golden\n",
      "    7        71.1674  1.53113e-05        parabolic\n",
      "    8        58.6183  1.36566e-05        parabolic\n",
      "    9        34.4419  1.15073e-05        golden\n",
      "   10        15.9546   1.4923e-05        parabolic\n",
      "   11        42.5725   1.1936e-05        golden\n",
      "   12         27.659   1.1622e-05        parabolic\n",
      "   13         32.862  1.14856e-05        parabolic\n",
      "   14        32.4862  1.14845e-05        parabolic\n",
      "   15        32.1529  1.14848e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.35852418 0.23748113 0.23426552 0.22173097 0.21337863]\n",
      "Trial error:  0.23426551902471393\n",
      "Best so far ***************** \n",
      "Trial:  4\n",
      "0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.034211765564864924  :: absSize readout:  0.20711204624446655\n",
      "NRMSE W:  0.019098548333991647  :: absSize W:  0.2241962071523932\n",
      "NRMSEs: best:  0.019098548333991647 avg:  0.019098548333991647\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00407576        initial\n",
      "    2        618.034   0.00487161        golden\n",
      "    3        236.068   0.00189924        golden\n",
      "    4        145.898  0.000212367        golden\n",
      "    5        90.1699  5.09627e-06        golden\n",
      "    6         99.932   0.00181104        parabolic\n",
      "    7        55.7281  5.91115e-06        golden\n",
      "    8         73.234  5.18404e-06        parabolic\n",
      "    9         84.157   0.00168058        parabolic\n",
      "   10        90.5033  4.59328e-06        parabolic\n",
      "   11        94.1047   0.00177217        golden\n",
      "   12        91.8789   0.00175481        golden\n",
      "   13        91.0287  4.82339e-06        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.3150485  0.3461442  0.27846307 0.37349812 1.3150485 ]\n",
      "Trial error:  0.3734981165299042\n",
      "LR:  0.21999999999999997\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.016475965661981364  :: absSize readout:  0.12161712615075394\n",
      "NRMSE W:  0.01664883411644627  :: absSize W:  0.2438772290435616\n",
      "NRMSEs: best:  0.01664883411644627 avg:  0.01664883411644627\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00664383        initial\n",
      "    2        618.034   0.00592637        golden\n",
      "    3        763.932   0.00534534        golden\n",
      "    4        854.102   0.00500152        golden\n",
      "    5         909.83   0.00480084        golden\n",
      "    6        944.272    0.0046822        golden\n",
      "    7        965.558   0.00461104        golden\n",
      "    8        978.714   0.00456789        golden\n",
      "    9        986.844    0.0383971        golden\n",
      "   10        973.689   0.00458429        golden\n",
      "   11        981.819    0.0334693        golden\n",
      "   12        976.794   0.00457414        golden\n",
      "   13          979.9    0.0304932        golden\n",
      "   14        977.981   0.00457028        golden\n",
      "   15        979.167   0.00487058        golden\n",
      "   16         978.38   0.00456897        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.36707588 1.3715347  1.36703707 1.36655116 1.36653859]\n",
      "Trial error:  1.3670370660240911\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.016573564144104616  :: absSize readout:  0.11962296434364102\n",
      "NRMSE W:  0.016538208853908883  :: absSize W:  0.26646008058236886\n",
      "NRMSEs: best:  0.016538208853908883 avg:  0.016538208853908883\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00231119        initial\n",
      "    2        618.034     0.349895        golden\n",
      "    3        236.068   0.00256701        golden\n",
      "    4        309.244   0.00245031        parabolic\n",
      "    5        472.136     0.266831        golden\n",
      "    6        416.408    0.0022462        golden\n",
      "    7        437.694     0.195787        golden\n",
      "    8        403.252   0.00227088        golden\n",
      "    9        424.538   0.00223104        golden\n",
      "   10        429.563      0.15538        golden\n",
      "   11        421.433   0.00223683        golden\n",
      "   12        426.458   0.00222747        golden\n",
      "   13        427.644     0.138177        golden\n",
      "   14        425.725   0.00222888        golden\n",
      "   15        426.911     0.128146        golden\n",
      "   16        426.125   0.00890803        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42532445 1.42533599 1.42532972 1.28872535 1.42535099]\n",
      "Trial error:  1.4253297214453144\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.017498680578954522  :: absSize readout:  0.123191698302109\n",
      "NRMSE W:  0.016442876340478058  :: absSize W:  0.28229194135701446\n",
      "NRMSEs: best:  0.016442876340478058 avg:  0.016442876340478058\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00633156        initial\n",
      "    2        618.034     0.422777        golden\n",
      "    3        236.068    0.0109547        golden\n",
      "    4        312.387   0.00964961        parabolic\n",
      "    5        472.136     0.339286        golden\n",
      "    6        416.408    0.0138633        golden\n",
      "    7        355.389   0.00761134        golden\n",
      "    8        395.122   0.00741867        golden\n",
      "    9        375.992   0.00644549        parabolic\n",
      "   10        380.772   0.00664937        parabolic\n",
      "   11        386.991   0.00662421        golden\n",
      "   12        383.885   0.00633793        golden\n",
      "   13        382.341   0.00678324        parabolic\n",
      "   14        381.633   0.00648467        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43536096 1.43183143 1.4379412  1.43936408 1.43818685]\n",
      "Trial error:  1.437941202126397\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.016691087750683706  :: absSize readout:  0.12167886361927197\n",
      "NRMSE W:  0.01813503612072509  :: absSize W:  0.2678979114078257\n",
      "NRMSEs: best:  0.01813503612072509 avg:  0.01813503612072509\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00272282        initial\n",
      "    2        618.034     0.193641        golden\n",
      "    3        236.068   0.00236972        golden\n",
      "    4        308.444   0.00219638        parabolic\n",
      "    5         290.54   0.00244965        parabolic\n",
      "    6        336.527   0.00271023        golden\n",
      "    7        319.171   0.00212495        golden\n",
      "    8          325.8   0.00215373        golden\n",
      "    9        318.837   0.00216809        parabolic\n",
      "   10        322.372   0.00215302        parabolic\n",
      "   11        320.393   0.00221401        golden\n",
      "   12        319.638   0.00206348        golden\n",
      "   13        319.971   0.00221831        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.35841643 1.36430343 1.35082084 1.36235647 1.36196394]\n",
      "Trial error:  1.3619639431567911\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.016743654824291635  :: absSize readout:  0.12650416778908868\n",
      "NRMSE W:  0.017884337903129648  :: absSize W:  0.2535675982220942\n",
      "NRMSEs: best:  0.017884337903129648 avg:  0.017884337903129648\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00608669        initial\n",
      "    2        618.034     0.293056        golden\n",
      "    3        236.068    0.0036688        golden\n",
      "    4        306.377   0.00560598        parabolic\n",
      "    5        145.898    0.0022422        golden\n",
      "    6        82.7668   0.00099404        parabolic\n",
      "    7        51.1527  0.000124264        golden\n",
      "    8        31.6141  0.000126416        golden\n",
      "    9        41.4854  0.000121601        parabolic\n",
      "   10        42.7932  0.000121024        parabolic\n",
      "   11         44.712  0.000120223        parabolic\n",
      "   12        47.1721  0.000120996        golden\n",
      "   13        45.0453  0.000122699        parabolic\n",
      "   14        44.3786   0.00012188        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.76937496 0.76416879 0.72993422 0.71159116 0.77686024]\n",
      "Trial error:  0.7641687916351345\n",
      "LR:  0.29\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.007772092083717648  :: absSize readout:  0.07077562748154413\n",
      "NRMSE W:  0.013884159983763095  :: absSize W:  0.22236531698844986\n",
      "NRMSEs: best:  0.013884159983763095 avg:  0.013884159983763095\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00673593        initial\n",
      "    2        618.034     0.270949        golden\n",
      "    3        236.068   0.00748357        golden\n",
      "    4        309.887   0.00718889        parabolic\n",
      "    5        472.136     0.193018        golden\n",
      "    6        416.408   0.00652748        golden\n",
      "    7        437.694   0.00640595        golden\n",
      "    8         450.85   0.00633397        golden\n",
      "    9         458.98     0.159233        golden\n",
      "   10        445.825   0.00636118        golden\n",
      "   11        453.955   0.00631733        golden\n",
      "   12        455.875   0.00630712        golden\n",
      "   13        457.061   0.00630084        golden\n",
      "   14        457.794     0.144095        golden\n",
      "   15        456.608   0.00630324        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   16        457.394   0.00629908        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.37600024 1.38024366 1.38024366 1.37793215 1.37785519]\n",
      "Trial error:  1.37793214885354\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.007763405784086909  :: absSize readout:  0.07332658718022292\n",
      "NRMSE W:  0.014398106903630848  :: absSize W:  0.24017061829497346\n",
      "NRMSEs: best:  0.014398106903630848 avg:  0.014398106903630848\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.278371        initial\n",
      "    2        618.034      0.38773        golden\n",
      "    3        236.068   0.00218372        golden\n",
      "    4        145.898    0.0022042        golden\n",
      "    5        190.997   0.00243206        parabolic\n",
      "    6        291.796  0.000213173        golden\n",
      "    7        326.238  0.000220984        golden\n",
      "    8         308.73  0.000223797        parabolic\n",
      "    9         270.51   0.00146609        golden\n",
      "   10        283.665  0.000211475        golden\n",
      "   11         278.64  0.000226492        golden\n",
      "   12        286.771  0.000205918        golden\n",
      "   13        287.468  0.000207336        parabolic\n",
      "   14        286.108  0.000205718        parabolic\n",
      "   15        285.775  0.000205411        parabolic\n",
      "   16        284.969  0.000209509        golden\n",
      "   17        285.441  0.000208756        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.30579027 1.30745517 1.30836779 1.30786359 1.30601067]\n",
      "Trial error:  1.3074551711317346\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.007449233041165715  :: absSize readout:  0.07038019101771761\n",
      "NRMSE W:  0.014216121960919753  :: absSize W:  0.2318786582290538\n",
      "NRMSEs: best:  0.014216121960919753 avg:  0.014216121960919753\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0111163        initial\n",
      "    2        618.034      0.29245        golden\n",
      "    3        236.068    0.0109492        golden\n",
      "    4        308.833    0.0114133        parabolic\n",
      "    5        145.898   0.00329665        golden\n",
      "    6        90.1699  0.000924451        golden\n",
      "    7        44.6258    0.0010606        parabolic\n",
      "    8        70.7206   0.00144697        parabolic\n",
      "    9        111.456   0.00132476        golden\n",
      "   10         71.981   0.00149171        parabolic\n",
      "   11        98.3006   0.00129535        golden\n",
      "   12        83.2224   0.00147924        golden\n",
      "   13        93.2756   0.00138069        golden\n",
      "   14        87.5162   0.00136903        golden\n",
      "   15        91.3562   0.00112091        golden\n",
      "   16        89.1563  0.000983694        golden\n",
      "   17         90.623   0.00087335        golden\n",
      "   18        90.9564   0.00101105        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.90701339 0.74901073 0.84013636 0.74551523 0.79416926]\n",
      "Trial error:  0.7941692569066917\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.006931077643399002  :: absSize readout:  0.0699488630331736\n",
      "NRMSE W:  0.013406851883025692  :: absSize W:  0.2509589939671517\n",
      "NRMSEs: best:  0.013406851883025692 avg:  0.013406851883025692\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966       0.2481        initial\n",
      "    2        618.034     0.418965        golden\n",
      "    3        236.068   0.00268637        golden\n",
      "    4        145.898   0.00157986        golden\n",
      "    5        190.116   0.00223939        parabolic\n",
      "    6        90.1699  0.000921325        golden\n",
      "    7        55.7281   6.7216e-05        golden\n",
      "    8        34.4419  0.000248424        golden\n",
      "    9        52.2058  9.15091e-05        parabolic\n",
      "   10        68.8837  0.000293303        golden\n",
      "   11        60.7531  0.000197811        golden\n",
      "   12        54.8632  0.000163511        parabolic\n",
      "   13        57.6475  0.000173058        golden\n",
      "   14        56.4612  0.000109614        golden\n",
      "   15        55.3948  6.89829e-05        golden\n",
      "   16        56.0614  7.04908e-05        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.63139097 0.70563882 0.64958045 0.63021253 0.68552577]\n",
      "Trial error:  0.6495804490903205\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.008079417403264592  :: absSize readout:  0.0744822135382296\n",
      "NRMSE W:  0.014718551158112988  :: absSize W:  0.23003180936063686\n",
      "NRMSEs: best:  0.014718551158112988 avg:  0.014718551158112988\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.173964        initial\n",
      "    2        618.034      0.30543        golden\n",
      "    3        236.068   0.00394445        golden\n",
      "    4        145.898   0.00391798        golden\n",
      "    5        190.953   0.00399815        parabolic\n",
      "    6        90.1699    0.0007212        golden\n",
      "    7        55.7281  0.000444445        golden\n",
      "    8        65.6048  0.000367778        parabolic\n",
      "    9        66.7016  0.000391032        parabolic\n",
      "   10        62.1368  0.000375448        parabolic\n",
      "   11        64.2802  0.000379412        golden\n",
      "   12        65.0988  0.000452764        golden\n",
      "   13        66.0237  0.000508393        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.06328259 1.06265789 0.46609282 0.72184354 0.63965366]\n",
      "Trial error:  0.7218435442542777\n",
      "LR:  0.36\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0033714682135204924  :: absSize readout:  0.03911434883462131\n",
      "NRMSE W:  0.01304213469521262  :: absSize W:  0.22658023924030743\n",
      "NRMSEs: best:  0.01304213469521262 avg:  0.01304213469521262\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.301711        initial\n",
      "    2        618.034     0.339263        golden\n",
      "    3        236.068    0.0150132        golden\n",
      "    4        145.898    0.0012522        golden\n",
      "    5        181.044   0.00132768        parabolic\n",
      "    6        163.078    0.0015496        parabolic\n",
      "    7        90.1699  0.000912623        golden\n",
      "    8        55.7281   0.00116465        golden\n",
      "    9        97.5491  0.000891495        parabolic\n",
      "   10          107.3  0.000823141        parabolic\n",
      "   11        122.043  0.000817303        golden\n",
      "   12        115.405  0.000817103        parabolic\n",
      "   13        118.437  0.000849351        parabolic\n",
      "   14        115.072  0.000827453        parabolic\n",
      "   15        116.563  0.000837041        golden\n",
      "   16        115.847  0.000813235        golden\n",
      "   17        116.181  0.000815895        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.76726026 0.65559101 0.69314529 0.69697745 0.77653021]\n",
      "Trial error:  0.696977452150994\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0031419885834253914  :: absSize readout:  0.039149532012684486\n",
      "NRMSE W:  0.009114534924269996  :: absSize W:  0.1981440471941793\n",
      "NRMSEs: best:  0.009114534924269996 avg:  0.009114534924269996\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.255777        initial\n",
      "    2        618.034     0.303361        golden\n",
      "    3        236.068   0.00844917        golden\n",
      "    4        145.898   0.00184039        golden\n",
      "    5        185.649   0.00229465        parabolic\n",
      "    6        161.117   0.00203974        parabolic\n",
      "    7        90.1699  0.000315542        golden\n",
      "    8        55.7281  0.000334693        golden\n",
      "    9        73.8469  0.000325721        parabolic\n",
      "   10        111.456  0.000260951        golden\n",
      "   11        124.612   0.00023798        golden\n",
      "   12        132.742   0.00020161        golden\n",
      "   13        137.767  0.000239834        golden\n",
      "   14        131.113  0.000206202        parabolic\n",
      "   15        134.662  0.000208566        golden\n",
      "   16        132.409  0.000198661        parabolic\n",
      "   17        132.076  0.000222972        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.29712177 0.51490455 0.51357595 0.52233119 0.50434473]\n",
      "Trial error:  0.5149045535624546\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0030785514816133916  :: absSize readout:  0.03704989303712604\n",
      "NRMSE W:  0.011742202165527908  :: absSize W:  0.22834805504456937\n",
      "NRMSEs: best:  0.011742202165527908 avg:  0.011742202165527908\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.354832        initial\n",
      "    2        618.034     0.399087        golden\n",
      "    3        236.068   0.00333411        golden\n",
      "    4        145.898   0.00117006        golden\n",
      "    5        189.795   0.00251278        parabolic\n",
      "    6        90.1699   0.00137174        golden\n",
      "    7        123.304    0.0010731        parabolic\n",
      "    8        125.614   0.00104768        parabolic\n",
      "    9        131.757  0.000974498        parabolic\n",
      "   10        137.158  0.000900841        golden\n",
      "   11        140.497  0.000847558        golden\n",
      "   12         142.56  0.000808276        golden\n",
      "   13        143.835  0.000778296        golden\n",
      "   14        144.623  0.000754173        golden\n",
      "   15         145.11  0.000732931        golden\n",
      "   16        145.443  0.000706492        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.36710396 1.36713674 1.36723043 1.36718657 1.36724419]\n",
      "Trial error:  1.3671865704209216\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0034688258847326483  :: absSize readout:  0.04032754954781492\n",
      "NRMSE W:  0.01100028811125026  :: absSize W:  0.22302017597246376\n",
      "NRMSEs: best:  0.01100028811125026 avg:  0.01100028811125026\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0022294        initial\n",
      "    2        618.034     0.290803        golden\n",
      "    3        236.068   0.00184104        golden\n",
      "    4          308.6   0.00142988        parabolic\n",
      "    5        297.295   0.00154166        parabolic\n",
      "    6        336.623   0.00201453        golden\n",
      "    7        319.304   0.00191148        golden\n",
      "    8        304.282   0.00193407        golden\n",
      "    9        312.689   0.00140444        golden\n",
      "   10        315.216   0.00137131        golden\n",
      "   11        316.777   0.00209047        golden\n",
      "   12         314.25    0.0013882        golden\n",
      "   13        315.812   0.00188456        golden\n",
      "   14        314.847   0.00151722        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.80192172 1.01560221 0.99837752 0.77930352 0.96421262]\n",
      "Trial error:  0.964212622268875\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.003372973450825695  :: absSize readout:  0.04186094832166226\n",
      "NRMSE W:  0.011536721761798307  :: absSize W:  0.2148385716200921\n",
      "NRMSEs: best:  0.011536721761798307 avg:  0.011536721761798307\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.364628        initial\n",
      "    2        618.034     0.400919        golden\n",
      "    3        236.068   0.00787845        golden\n",
      "    4        145.898   0.00657864        golden\n",
      "    5        190.283   0.00782809        parabolic\n",
      "    6        90.1699   0.00351808        golden\n",
      "    7        55.7281   0.00267614        golden\n",
      "    8        36.7838   0.00203573        parabolic\n",
      "    9        22.7336    0.0015092        golden\n",
      "   10        14.0502  0.000718651        golden\n",
      "   11        8.68348  0.000415963        golden\n",
      "   12        5.36668   0.00049574        golden\n",
      "   13        8.32308  0.000416441        parabolic\n",
      "   14        9.01681  0.000396055        parabolic\n",
      "   15        10.9394  0.000388813        golden\n",
      "   16         10.054  0.000400299        parabolic\n",
      "   17        12.1276  0.000372508        golden\n",
      "   18        12.8619  0.000371202        golden\n",
      "   19        13.1953  0.000366633        parabolic\n",
      "   20        13.5286  0.000336849        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.0070649  0.98082606 1.00678272 1.00705193 0.95542552]\n",
      "Trial error:  1.0067827208014404\n",
      "LR:  0.42999999999999994\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0013179016556289149  :: absSize readout:  0.02056270043344922\n",
      "NRMSE W:  0.009885714931004968  :: absSize W:  0.18976686459253245\n",
      "NRMSEs: best:  0.009885714931004968 avg:  0.009885714931004968\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.358801        initial\n",
      "    2        618.034     0.385689        golden\n",
      "    3        236.068   0.00602101        golden\n",
      "    4        145.898   0.00143318        golden\n",
      "    5        188.446   0.00341421        parabolic\n",
      "    6        90.1699   0.00123699        golden\n",
      "    7        114.015   0.00134815        parabolic\n",
      "    8        55.7281  0.000525961        golden\n",
      "    9        34.4419  0.000841158        golden\n",
      "   10        56.7232  0.000844007        parabolic\n",
      "   11        45.5783  0.000405258        parabolic\n",
      "   12        41.3245  0.000476394        golden\n",
      "   13        47.6602  0.000486689        parabolic\n",
      "   14        44.4002   0.00060088        parabolic\n",
      "   15        46.3735  0.000605815        golden\n",
      "   16        45.1283  0.000432081        golden\n",
      "   17        45.9116  0.000463484        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.81134418 0.97332899 0.77009387 0.98074504 0.6802552 ]\n",
      "Trial error:  0.811344176010325\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.001247180907634173  :: absSize readout:  0.020721821874948986\n",
      "NRMSE W:  0.010810904996185255  :: absSize W:  0.19523404262104876\n",
      "NRMSEs: best:  0.010810904996185255 avg:  0.010810904996185255\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.340321        initial\n",
      "    2        618.034      0.35704        golden\n",
      "    3        236.068     0.268866        golden\n",
      "    4        145.898   0.00167973        golden\n",
      "    5        90.1699  0.000745214        golden\n",
      "    6        117.619  0.000898929        parabolic\n",
      "    7         96.805  0.000716205        parabolic\n",
      "    8        98.0502  0.000714214        parabolic\n",
      "    9        99.6999  0.000706021        parabolic\n",
      "   10        106.544  0.000685317        golden\n",
      "   11        109.741  0.000653372        parabolic\n",
      "   12         112.75  0.000617521        golden\n",
      "   13         114.61  0.000621005        golden\n",
      "   14        113.349  0.000647919        parabolic\n",
      "   15        111.601  0.000656856        golden\n",
      "   16        112.311  0.000628051        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.87551692 0.86533207 0.87235114 1.08717536 0.93057279]\n",
      "Trial error:  0.8755169164325077\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0015553436505666516  :: absSize readout:  0.021341232750682324\n",
      "NRMSE W:  0.01028990676403293  :: absSize W:  0.16812035792364888\n",
      "NRMSEs: best:  0.01028990676403293 avg:  0.01028990676403293\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.268588        initial\n",
      "    2        618.034     0.293421        golden\n",
      "    3        236.068     0.175531        golden\n",
      "    4        145.898   0.00138984        golden\n",
      "    5        90.1699   0.00183525        golden\n",
      "    6        118.335   0.00119884        parabolic\n",
      "    7        125.577   0.00122902        parabolic\n",
      "    8        106.632   0.00168456        parabolic\n",
      "    9        113.865   0.00161605        golden\n",
      "   10        121.101   0.00133633        golden\n",
      "   11        116.627    0.0012396        golden\n",
      "   12        119.391    0.0012195        golden\n",
      "   13        117.682   0.00118158        golden\n",
      "   14        117.279   0.00119361        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.87112934 0.71080267 0.88909957 0.90237655 0.98926462]\n",
      "Trial error:  0.8890995681120248\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0012980560827467683  :: absSize readout:  0.021251527645358766\n",
      "NRMSE W:  0.007377380165793483  :: absSize W:  0.17135306837257486\n",
      "NRMSEs: best:  0.007377380165793483 avg:  0.007377380165793483\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.305413        initial\n",
      "    2        618.034     0.321226        golden\n",
      "    3        236.068     0.236551        golden\n",
      "    4        145.898   0.00144537        golden\n",
      "    5        90.1699   0.00121016        golden\n",
      "    6        117.916   0.00131648        parabolic\n",
      "    7        55.7281    0.0008348        golden\n",
      "    8        34.4419  0.000586753        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    9        21.2862  0.000907227        golden\n",
      "   10        39.5127  0.000992519        parabolic\n",
      "   11        34.1085  0.000548435        parabolic\n",
      "   12        29.2108  0.000261413        golden\n",
      "   13        28.8775  0.000237556        parabolic\n",
      "   14        25.9779  0.000584136        golden\n",
      "   15          27.77   0.00046487        golden\n",
      "   16        28.5442  0.000221375        parabolic\n",
      "   17        28.2108  0.000208172        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.38246569 0.69073346 0.28553146 0.35254666 0.39167124]\n",
      "Trial error:  0.38246568965711814\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.001130226273252735  :: absSize readout:  0.020479396749055433\n",
      "NRMSE W:  0.010214387871492838  :: absSize W:  0.20085441549077912\n",
      "NRMSEs: best:  0.010214387871492838 avg:  0.010214387871492838\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.381988        initial\n",
      "    2        618.034     0.398145        golden\n",
      "    3        236.068     0.315235        golden\n",
      "    4        145.898    0.0052558        golden\n",
      "    5        90.1699   0.00404382        golden\n",
      "    6         117.57   0.00508818        parabolic\n",
      "    7        55.7281  0.000905731        golden\n",
      "    8        34.4419  0.000213476        golden\n",
      "    9        29.6189  0.000553198        parabolic\n",
      "   10        40.9615  9.57697e-05        parabolic\n",
      "   11        39.6562  7.59057e-05        parabolic\n",
      "   12        39.1164  6.38801e-05        parabolic\n",
      "   13        37.3309  6.97137e-05        golden\n",
      "   14        38.3724  6.32364e-05        parabolic\n",
      "   15         38.039  6.44845e-05        parabolic\n",
      "   16        38.7057  6.28762e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.5029225  0.42852877 0.41096471 0.43820216 0.43076955]\n",
      "Trial error:  0.4307695514915875\n",
      "LR:  0.49999999999999994\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0005765715915698039  :: absSize readout:  0.012273775721209973\n",
      "NRMSE W:  0.009163378579889317  :: absSize W:  0.13994340878325903\n",
      "NRMSEs: best:  0.009163378579889317 avg:  0.009163378579889317\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.176639        initial\n",
      "    2        618.034     0.193267        golden\n",
      "    3        236.068   0.00606886        golden\n",
      "    4        145.898   0.00229283        golden\n",
      "    5        186.598   0.00314463        parabolic\n",
      "    6        141.536   0.00254506        parabolic\n",
      "    7         160.26   0.00274158        parabolic\n",
      "    8        151.384   0.00247629        golden\n",
      "    9        146.836   0.00245083        parabolic\n",
      "   10        144.232   0.00256261        golden\n",
      "   11        145.262   0.00241279        golden\n",
      "   12        146.231   0.00258537        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.1033797  1.12351704 1.04003118 1.13046294 1.13878574]\n",
      "Trial error:  1.1235170373100456\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0004355747938114188  :: absSize readout:  0.010870785153946159\n",
      "NRMSE W:  0.008624039279548721  :: absSize W:  0.13482460969403698\n",
      "NRMSEs: best:  0.008624039279548721 avg:  0.008624039279548721\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.323222        initial\n",
      "    2        618.034     0.347803        golden\n",
      "    3        236.068     0.256213        golden\n",
      "    4        145.898  6.75099e-05        golden\n",
      "    5        90.1699  3.36143e-05        golden\n",
      "    6        118.018  3.33874e-05        parabolic\n",
      "    7        104.278  3.33615e-05        parabolic\n",
      "    8         109.82  3.16247e-05        parabolic\n",
      "    9        111.124  3.16915e-05        parabolic\n",
      "   10        110.153  3.10963e-05        parabolic\n",
      "   11        110.486  3.14147e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.27717295 0.29456251 0.25480197 0.29759857 0.30134142]\n",
      "Trial error:  0.2945625130649157\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0004836203619694808  :: absSize readout:  0.012659796860847143\n",
      "NRMSE W:  0.008849856157214427  :: absSize W:  0.16946098179160718\n",
      "NRMSEs: best:  0.008849856157214427 avg:  0.008849856157214427\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.371622        initial\n",
      "    2        618.034     0.384861        golden\n",
      "    3        236.068     0.320524        golden\n",
      "    4        145.898   0.00299352        golden\n",
      "    5        90.1699    0.0028676        golden\n",
      "    6        117.987   0.00303309        parabolic\n",
      "    7        55.7281   0.00108122        golden\n",
      "    8        34.4419  0.000960904        golden\n",
      "    9         41.677  0.000850764        parabolic\n",
      "   10        43.1827  0.000703107        parabolic\n",
      "   11        47.9746   0.00105262        golden\n",
      "   12         45.013   0.00101741        golden\n",
      "   13        43.8818   0.00062057        golden\n",
      "   14        44.3139  0.000842443        golden\n",
      "   15        43.5485  0.000662944        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.72284976 0.72514676 0.70302136 0.62039934 0.574273  ]\n",
      "Trial error:  0.7030213581140847\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00039236410748709884  :: absSize readout:  0.010945591010224496\n",
      "NRMSE W:  0.008315677285621995  :: absSize W:  0.12631853104694804\n",
      "NRMSEs: best:  0.008315677285621995 avg:  0.008315677285621995\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.241886        initial\n",
      "    2        618.034     0.262306        golden\n",
      "    3        236.068  0.000282333        golden\n",
      "    4        145.898  0.000270091        golden\n",
      "    5        190.973  0.000285493        parabolic\n",
      "    6        90.1699  0.000184972        golden\n",
      "    7        55.7281  0.000156918        golden\n",
      "    8        21.4359  0.000581785        parabolic\n",
      "    9        68.8837  0.000147828        golden\n",
      "   10        67.1909  0.000148579        parabolic\n",
      "   11        77.0143  0.000145538        golden\n",
      "   12        81.4602  0.000129387        parabolic\n",
      "   13        84.7871  0.000146368        golden\n",
      "   14        80.8531  0.000125366        parabolic\n",
      "   15        79.9172  0.000117507        parabolic\n",
      "   16        78.8084  0.000145393        golden\n",
      "   17        79.4937  0.000112366        golden\n",
      "   18        79.1604  0.000106198        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.38250616 1.41968832 1.38201932 1.38534209 1.38204743]\n",
      "Trial error:  1.3825061614173095\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0003907145580309635  :: absSize readout:  0.011411962164954433\n",
      "NRMSE W:  0.009111550605637554  :: absSize W:  0.1582600948975851\n",
      "NRMSEs: best:  0.009111550605637554 avg:  0.009111550605637554\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.313936        initial\n",
      "    2        618.034      0.32491        golden\n",
      "    3        236.068     0.284322        golden\n",
      "    4        145.898   0.00319455        golden\n",
      "    5        90.1699   0.00321331        golden\n",
      "    6        118.042   0.00321868        parabolic\n",
      "    7         180.34     0.247116        golden\n",
      "    8        159.054     0.155783        golden\n",
      "    9        135.258     0.003205        golden\n",
      "   10        150.923   0.00318943        golden\n",
      "   11        154.029   0.00318616        golden\n",
      "   12        155.948   0.00318433        golden\n",
      "   13        157.134   0.00318289        golden\n",
      "   14        157.867   0.00318229        golden\n",
      "   15        158.321   0.00318199        golden\n",
      "   16        158.654   0.00318112        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.28414095 1.28445444 1.28408437 1.28445445 1.28443081]\n",
      "Trial error:  1.2844308094586918\n",
      "LR:  0.57\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00020261096714028985  :: absSize readout:  0.007093648645804491\n",
      "NRMSE W:  0.008821742561652869  :: absSize W:  0.1385469458172836\n",
      "NRMSEs: best:  0.008821742561652869 avg:  0.008821742561652869\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.313146        initial\n",
      "    2        618.034     0.326235        golden\n",
      "    3        236.068     0.279284        golden\n",
      "    4        145.898     0.107316        golden\n",
      "    5        90.1699   0.00123014        golden\n",
      "    6        55.7281   0.00142011        golden\n",
      "    7        73.0793   0.00139261        parabolic\n",
      "    8        111.456   0.00135204        golden\n",
      "    9        93.5994   0.00141695        parabolic\n",
      "   10        89.8366  0.000874833        parabolic\n",
      "   11        83.4359    0.0012446        golden\n",
      "   12        86.8093   0.00150192        parabolic\n",
      "   13        89.5033   0.00106756        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.63224135 0.8825098  0.87570623 0.7654543  0.56604169]\n",
      "Trial error:  0.7654543035685965\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00025395944106917874  :: absSize readout:  0.007898158702698001\n",
      "NRMSE W:  0.007677108800065889  :: absSize W:  0.15605615123075287\n",
      "NRMSEs: best:  0.007677108800065889 avg:  0.007677108800065889\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.274616        initial\n",
      "    2        618.034     0.283001        golden\n",
      "    3        236.068      0.21298        golden\n",
      "    4        145.898   0.00448222        golden\n",
      "    5        90.1699   0.00193068        golden\n",
      "    6         116.56   0.00272609        parabolic\n",
      "    7        75.1064   0.00119477        parabolic\n",
      "    8        46.4183   0.00239438        golden\n",
      "    9        70.8512   0.00135012        parabolic\n",
      "   10        77.1099   0.00130662        parabolic\n",
      "   11        74.2161   0.00135441        parabolic\n",
      "   12        75.8717   0.00121896        golden\n",
      "   13         74.773   0.00131496        parabolic\n",
      "   14        75.4397   0.00141931        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.7348121  0.73925109 0.79794301 0.76834342 0.73685015]\n",
      "Trial error:  0.7392510947966487\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00022414243382555402  :: absSize readout:  0.006874868460583727\n",
      "NRMSE W:  0.0075421354281508465  :: absSize W:  0.1591336395753947\n",
      "NRMSEs: best:  0.0075421354281508465 avg:  0.0075421354281508465\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.273277        initial\n",
      "    2        618.034     0.283425        golden\n",
      "    3        236.068     0.230597        golden\n",
      "    4        145.898  0.000576488        golden\n",
      "    5        90.1699  0.000842442        golden\n",
      "    6         118.17  0.000590599        parabolic\n",
      "    7        133.705  0.000608673        parabolic\n",
      "    8         180.34    0.0158597        golden\n",
      "    9        159.054   0.00194798        golden\n",
      "   10        150.923   0.00158662        golden\n",
      "   11        141.241  0.000625933        golden\n",
      "   12        147.817  0.000549656        golden\n",
      "   13        149.004   0.00054037        golden\n",
      "   14        149.337  0.000562516        parabolic\n",
      "   15        148.491  0.000535834        parabolic\n",
      "   16        148.157  0.000539419        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.40111223 1.38375602 1.37386196 1.40572954 1.40548522]\n",
      "Trial error:  1.4011122264173221\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0002197420133319723  :: absSize readout:  0.007400269679238897\n",
      "NRMSE W:  0.008670611845077782  :: absSize W:  0.13791686237104334\n",
      "NRMSEs: best:  0.008670611845077782 avg:  0.008670611845077782\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.273091        initial\n",
      "    2        618.034     0.286234        golden\n",
      "    3        236.068     0.236068        golden\n",
      "    4        145.898    0.0161139        golden\n",
      "    5        90.1699    0.0142526        golden\n",
      "    6        117.021    0.0152551        parabolic\n",
      "    7        55.7281    0.0118881        golden\n",
      "    8        34.4419   0.00868359        golden\n",
      "    9        21.2862  0.000921065        golden\n",
      "   10        13.1556  0.000868094        golden\n",
      "   11        17.1021  0.000836037        parabolic\n",
      "   12        16.2898  0.000789728        parabolic\n",
      "   13        15.3243  0.000787814        parabolic\n",
      "   14        14.4959  0.000777785        golden\n",
      "   15         13.984  0.000900479        golden\n",
      "   16        14.8293  0.000786308        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.24515974 1.20348484 1.23754171 1.24405838 1.20224363]\n",
      "Trial error:  1.2375417100765191\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00029166981228184464  :: absSize readout:  0.007518337198513867\n",
      "NRMSE W:  0.007773967408423064  :: absSize W:  0.16209365028137881\n",
      "NRMSEs: best:  0.007773967408423064 avg:  0.007773967408423064\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.250576        initial\n",
      "    2        618.034     0.246642        golden\n",
      "    3        763.932     0.242827        golden\n",
      "    4        854.102     0.240496        golden\n",
      "    5         909.83     0.239075        golden\n",
      "    6        944.272     0.238226        golden\n",
      "    7        965.558     0.237702        golden\n",
      "    8        978.714     0.237373        golden\n",
      "    9        986.844     0.237184        golden\n",
      "   10        991.869     0.237054        golden\n",
      "   11        994.975     0.236987        golden\n",
      "   12        996.894     0.236941        golden\n",
      "   13        998.081     0.236904        golden\n",
      "   14        998.814     0.236887        golden\n",
      "   15        999.267     0.236883        golden\n",
      "   16          999.6     0.236875        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43653261 1.43653261 1.43653261 1.43653261 1.43653261]\n",
      "Trial error:  1.4365326108003578\n",
      "LR:  0.6399999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00010135569007678216  :: absSize readout:  0.005434393452789239\n",
      "NRMSE W:  0.007403257296570203  :: absSize W:  0.09966964382125122\n",
      "NRMSEs: best:  0.007403257296570203 avg:  0.007403257296570203\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.347011        initial\n",
      "    2        618.034     0.359873        golden\n",
      "    3        236.068     0.324197        golden\n",
      "    4        145.898     0.220262        golden\n",
      "    5        90.1699   0.00122507        golden\n",
      "    6        55.7281  0.000506817        golden\n",
      "    7        72.7085   0.00089465        parabolic\n",
      "    8        34.4419  0.000294989        golden\n",
      "    9        30.3119  0.000267148        parabolic\n",
      "   10        18.7338   0.00026063        golden\n",
      "   11        23.8072  0.000247631        parabolic\n",
      "   12        24.1405  0.000247938        parabolic\n",
      "   13        23.2599  0.000247309        parabolic\n",
      "   14         21.531  0.000249256        golden\n",
      "   15        22.7488  0.000247456        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.42952569 0.42284219 0.42926647 0.4215331  0.4219001 ]\n",
      "Trial error:  0.4228421915624885\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001427208174728346  :: absSize readout:  0.005527582051157604\n",
      "NRMSE W:  0.0074157864534840166  :: absSize W:  0.09478876162193257\n",
      "NRMSEs: best:  0.0074157864534840166 avg:  0.0074157864534840166\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966      0.27806        initial\n",
      "    2        618.034     0.293916        golden\n",
      "    3        236.068     0.207405        golden\n",
      "    4        145.898   0.00244001        golden\n",
      "    5        90.1699  8.38041e-05        golden\n",
      "    6        116.651  0.000781041        parabolic\n",
      "    7        79.2732  7.13316e-05        parabolic\n",
      "    8        83.8722  7.57752e-05        parabolic\n",
      "    9        48.9935  0.000117382        golden\n",
      "   10        64.5187  8.21942e-05        parabolic\n",
      "   11        76.0807  7.19495e-05        parabolic\n",
      "   12        78.3271  7.23451e-05        parabolic\n",
      "   13        81.0299  7.47699e-05        golden\n",
      "   14        79.9442  7.41469e-05        golden\n",
      "   15        78.9118  7.28832e-05        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   16        79.6066   7.1863e-05        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.37817124 0.4016116  0.29957385 0.37336858 0.39226716]\n",
      "Trial error:  0.3781712362862613\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00015062592226128407  :: absSize readout:  0.005772983218538807\n",
      "NRMSE W:  0.007656712942339716  :: absSize W:  0.1136044831507527\n",
      "NRMSEs: best:  0.007656712942339716 avg:  0.007656712942339716\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.321834        initial\n",
      "    2        618.034     0.329054        golden\n",
      "    3        236.068     0.305568        golden\n",
      "    4        145.898    0.0161384        golden\n",
      "    5        90.1699   0.00278576        golden\n",
      "    6        112.149   0.00263324        parabolic\n",
      "    7        101.635   0.00233184        parabolic\n",
      "    8        102.277   0.00236192        parabolic\n",
      "    9        97.2555   0.00202315        golden\n",
      "   10         94.549    0.0022522        golden\n",
      "   11        97.8351   0.00208152        parabolic\n",
      "   12        96.6525   0.00194871        parabolic\n",
      "   13        95.8491    0.0017965        golden\n",
      "   14        95.3525   0.00192794        golden\n",
      "   15        96.1824   0.00187254        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.28783339 1.28789706 1.28786656 1.28786633 1.28786028]\n",
      "Trial error:  1.2878663347592239\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0002004228113553985  :: absSize readout:  0.005531529862549834\n",
      "NRMSE W:  0.005699225274255669  :: absSize W:  0.15435714224940944\n",
      "NRMSEs: best:  0.005699225274255669 avg:  0.005699225274255669\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.220696        initial\n",
      "    2        618.034      0.23153        golden\n",
      "    3        236.068     0.190207        golden\n",
      "    4        145.898   0.00950127        golden\n",
      "    5        90.1699   0.00167673        golden\n",
      "    6        112.538  0.000392937        parabolic\n",
      "    7        106.194  0.000327771        parabolic\n",
      "    8         108.15  0.000414736        parabolic\n",
      "    9        100.073   0.00235546        golden\n",
      "   10        103.856   0.00029038        golden\n",
      "   11        102.411  0.000367245        golden\n",
      "   12        104.588  0.000348609        parabolic\n",
      "   13        103.304  0.000315883        golden\n",
      "   14        104.189  0.000403486        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.35716122 0.33242798 0.38194092 0.35361697 0.27976935]\n",
      "Trial error:  0.3536169673212825\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00018093530382326815  :: absSize readout:  0.005999987527735038\n",
      "NRMSE W:  0.00751471359239397  :: absSize W:  0.10559952867113555\n",
      "NRMSEs: best:  0.00751471359239397 avg:  0.00751471359239397\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.263417        initial\n",
      "    2        618.034     0.272476        golden\n",
      "    3        236.068       0.2387        golden\n",
      "    4        145.898     0.137755        golden\n",
      "    5        90.1699  0.000218874        golden\n",
      "    6        55.7281  6.80513e-05        golden\n",
      "    7        72.8689  7.97402e-05        parabolic\n",
      "    8        62.7029  6.94936e-05        parabolic\n",
      "    9        57.0034  6.79652e-05        parabolic\n",
      "   10        57.3368  6.79746e-05        parabolic\n",
      "   11        56.6701  6.79601e-05        parabolic\n",
      "   12        56.3103  6.79948e-05        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.25979353 0.24743801 0.23913924 0.24883826 0.25025777]\n",
      "Trial error:  0.24883825748459726\n",
      "LR:  0.71\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001211966084821614  :: absSize readout:  0.0046476806020137856\n",
      "NRMSE W:  0.007388646238764141  :: absSize W:  0.071982744013581\n",
      "NRMSEs: best:  0.007388646238764141 avg:  0.007388646238764141\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.171519        initial\n",
      "    2        618.034     0.187102        golden\n",
      "    3        236.068     0.131104        golden\n",
      "    4        145.898    0.0277372        golden\n",
      "    5        90.1699    0.0182267        golden\n",
      "    6        105.275    0.0214831        parabolic\n",
      "    7        55.7281   0.00758175        golden\n",
      "    8        34.4419  0.000886867        golden\n",
      "    9        21.2862   0.00124253        golden\n",
      "   10        29.2271   0.00100042        parabolic\n",
      "   11        42.5725  0.000989423        golden\n",
      "   12        36.0597  0.000926824        parabolic\n",
      "   13          32.45  0.000917085        golden\n",
      "   14        34.1085   0.00165995        parabolic\n",
      "   15        34.7752   0.00181403        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.6788278  0.47275608 0.52066424 0.76747852 0.71419611]\n",
      "Trial error:  0.6788277961046049\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00015193941533088887  :: absSize readout:  0.005025521847678984\n",
      "NRMSE W:  0.00832785000754695  :: absSize W:  0.12018367313071045\n",
      "NRMSEs: best:  0.00832785000754695 avg:  0.00832785000754695\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.228965        initial\n",
      "    2        618.034     0.242065        golden\n",
      "    3        236.068     0.194815        golden\n",
      "    4        145.898  0.000429107        golden\n",
      "    5        90.1699   0.00104514        golden\n",
      "    6        118.406   0.00083725        parabolic\n",
      "    7         180.34     0.145702        golden\n",
      "    8        159.054    0.0958997        golden\n",
      "    9        135.397  0.000651657        golden\n",
      "   10        150.923   0.00116111        golden\n",
      "   11        141.887  0.000373545        golden\n",
      "   12         142.61    0.0003277        parabolic\n",
      "   13        143.598  0.000368778        parabolic\n",
      "   14        142.943  0.000344695        parabolic\n",
      "   15        142.277  0.000383374        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.33926286 1.33926681 1.33926851 1.33925957 1.33926801]\n",
      "Trial error:  1.3392668117480282\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00010454178284621754  :: absSize readout:  0.0046633868033623465\n",
      "NRMSE W:  0.00813101406640658  :: absSize W:  0.10077485339774885\n",
      "NRMSEs: best:  0.00813101406640658 avg:  0.00813101406640658\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.275414        initial\n",
      "    2        618.034     0.288207        golden\n",
      "    3        236.068     0.253533        golden\n",
      "    4        145.898    0.0092075        golden\n",
      "    5        90.1699   0.00204539        golden\n",
      "    6        114.402   0.00249923        parabolic\n",
      "    7        99.5993   0.00385522        parabolic\n",
      "    8        55.7281   0.00283041        golden\n",
      "    9        89.0529   0.00163975        parabolic\n",
      "   10         76.324  0.000570728        golden\n",
      "   11        80.6057   0.00113227        parabolic\n",
      "   12         68.457   0.00250993        golden\n",
      "   13        73.3191  0.000415736        golden\n",
      "   14        72.4597  0.000406534        parabolic\n",
      "   15        72.1264  0.000403093        parabolic\n",
      "   16        70.7248  0.000378808        golden\n",
      "   17        69.8586  0.000388563        golden\n",
      "   18        71.0582  0.000388463        parabolic\n",
      "   19        70.3915  0.000377049        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.61769563 0.43042935 0.43510668 0.42769992 1.3455409 ]\n",
      "Trial error:  0.4351066755557167\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00016141732210753597  :: absSize readout:  0.0048470719613282745\n",
      "NRMSE W:  0.007988988203294547  :: absSize W:  0.12078818917337135\n",
      "NRMSEs: best:  0.007988988203294547 avg:  0.007988988203294547\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.225225        initial\n",
      "    2        618.034      0.22154        golden\n",
      "    3        763.932     0.218412        golden\n",
      "    4        854.102      0.21649        golden\n",
      "    5         909.83     0.215331        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    6        944.272     0.214621        golden\n",
      "    7        965.558       0.2142        golden\n",
      "    8        978.714     0.213936        golden\n",
      "    9        986.844     0.213775        golden\n",
      "   10        991.869     0.213668        golden\n",
      "   11        994.975     0.213606        golden\n",
      "   12        996.894     0.213568        golden\n",
      "   13        998.081     0.213545        golden\n",
      "   14        998.814      0.21353        golden\n",
      "   15        999.267     0.213521        golden\n",
      "   16          999.6     0.213515        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43612635 1.43612635 1.43612635 1.43612635 1.43612635]\n",
      "Trial error:  1.4361263489690463\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00011022890563605853  :: absSize readout:  0.004488242060446978\n",
      "NRMSE W:  0.00765390407357326  :: absSize W:  0.10714289446727947\n",
      "NRMSEs: best:  0.00765390407357326 avg:  0.00765390407357326\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.260255        initial\n",
      "    2        618.034     0.278661        golden\n",
      "    3        236.068   0.00634862        golden\n",
      "    4        145.898   0.00616157        golden\n",
      "    5        190.842   0.00649462        parabolic\n",
      "    6        90.1699   0.00286672        golden\n",
      "    7        55.7281   0.00174653        golden\n",
      "    8        17.8219  0.000418293        parabolic\n",
      "    9        32.3008   0.00175048        golden\n",
      "   10        11.0146   0.00235647        golden\n",
      "   11        23.3524   0.00236683        golden\n",
      "   12        15.2218  0.000987815        golden\n",
      "   13        19.9344  0.000561192        golden\n",
      "   14        18.3222  0.000441794        parabolic\n",
      "   15        16.8288  0.000340208        golden\n",
      "   16        16.2149   0.00142524        golden\n",
      "   17        17.2081  0.000378661        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.31844309 0.31581275 0.51401218 0.96807994 0.31914426]\n",
      "Trial error:  0.3191442608023201\n",
      "LR:  0.7799999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00015698233855637668  :: absSize readout:  0.004375459937722729\n",
      "NRMSE W:  0.007104577933202686  :: absSize W:  0.10450721738956639\n",
      "NRMSEs: best:  0.007104577933202686 avg:  0.007104577933202686\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.110919        initial\n",
      "    2        618.034     0.121194        golden\n",
      "    3        236.068    0.0845234        golden\n",
      "    4        145.898   0.00246936        golden\n",
      "    5        90.1699   0.00387355        golden\n",
      "    6        119.999   0.00360339        parabolic\n",
      "    7         180.34   0.00246279        golden\n",
      "    8        163.251    0.0623417        parabolic\n",
      "    9        201.626    0.0772773        golden\n",
      "   10        188.471    0.0735042        golden\n",
      "   11        173.813    0.0680051        golden\n",
      "   12        183.446    0.0718146        golden\n",
      "   13        177.847    0.0697077        golden\n",
      "   14        181.526    0.0711219        golden\n",
      "   15        179.388    0.0703148        golden\n",
      "   16        180.793    0.0708496        golden\n",
      "   17        179.976    0.0705409        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.28798821 1.28798821 1.28798821 1.3031691  1.28798821]\n",
      "Trial error:  1.2879882136835057\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00012155497694627977  :: absSize readout:  0.004141251436377242\n",
      "NRMSE W:  0.006896496091688648  :: absSize W:  0.08169406869083666\n",
      "NRMSEs: best:  0.006896496091688648 avg:  0.006896496091688648\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0518281        initial\n",
      "    2        618.034    0.0452718        golden\n",
      "    3        763.932    0.0399791        golden\n",
      "    4        854.102    0.0370665        golden\n",
      "    5         909.83    0.0354675        golden\n",
      "    6        944.272    0.0345567        golden\n",
      "    7        965.558    0.0340223        golden\n",
      "    8        978.714    0.0337028        golden\n",
      "    9        986.844    0.0335096        golden\n",
      "   10        991.869    0.0333915        golden\n",
      "   11        994.975    0.0333191        golden\n",
      "   12        996.894    0.0332746        golden\n",
      "   13        998.081    0.0332469        golden\n",
      "   14        998.814      0.03323        golden\n",
      "   15        999.267    0.0332197        golden\n",
      "   16          999.6    0.0332121        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.30874672 1.30874672 1.30874672 1.30874672 1.30874672]\n",
      "Trial error:  1.3087467168805005\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00018888579707601385  :: absSize readout:  0.004476517318281055\n",
      "NRMSE W:  0.007580270571861436  :: absSize W:  0.11405299577138817\n",
      "NRMSEs: best:  0.007580270571861436 avg:  0.007580270571861436\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.163225        initial\n",
      "    2        618.034     0.165464        golden\n",
      "    3        236.068      0.12965        golden\n",
      "    4        145.898    0.0842488        golden\n",
      "    5        90.1699   0.00610622        golden\n",
      "    6        55.7281  0.000871177        golden\n",
      "    7        67.4678   0.00167459        parabolic\n",
      "    8        52.3014   0.00121205        parabolic\n",
      "    9        58.5073   0.00136739        parabolic\n",
      "   10         55.125    0.0017159        parabolic\n",
      "   11        56.0614   0.00107574        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.8819192  0.9582818  0.90787997 1.08997299 0.80499582]\n",
      "Trial error:  0.907879969737752\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00018926586175181803  :: absSize readout:  0.004419706919598728\n",
      "NRMSE W:  0.006993305001409668  :: absSize W:  0.08568235681185182\n",
      "NRMSEs: best:  0.006993305001409668 avg:  0.006993305001409668\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.029849        initial\n",
      "    2        618.034    0.0269743        golden\n",
      "    3        763.932     0.024031        golden\n",
      "    4        854.102    0.0225812        golden\n",
      "    5         909.83    0.0219268        golden\n",
      "    6        944.272    0.0216495        golden\n",
      "    7        965.558    0.0215395        golden\n",
      "    8        978.714    0.0215145        golden\n",
      "    9        982.157    0.0215133        parabolic\n",
      "   10        982.491    0.0215123        parabolic\n",
      "   11        989.179    0.0215264        golden\n",
      "   12        985.045    0.0215154        golden\n",
      "   13        983.466    0.0215155        golden\n",
      "   14        982.863    0.0215126        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.32256869 1.32256486 1.32256402 1.32257048 1.32256303]\n",
      "Trial error:  1.3225648550687266\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00012427720579897213  :: absSize readout:  0.004216964446551923\n",
      "NRMSE W:  0.0074363745373352665  :: absSize W:  0.09178941961104684\n",
      "NRMSEs: best:  0.0074363745373352665 avg:  0.0074363745373352665\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.217557        initial\n",
      "    2        618.034     0.241614        golden\n",
      "    3        236.068     0.145173        golden\n",
      "    4        145.898    0.0691204        golden\n",
      "    5        90.1699   0.00231768        golden\n",
      "    6        55.7281  0.000839038        golden\n",
      "    7        71.2744   0.00131409        parabolic\n",
      "    8        40.1701     0.001165        parabolic\n",
      "    9         54.275  0.000932047        parabolic\n",
      "   10        61.6662  0.000973314        golden\n",
      "   11        57.7324  0.000880322        parabolic\n",
      "   12        56.3094   0.00084785        parabolic\n",
      "   13        55.1731  0.000903554        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.39806367 0.39297471 0.24605598 0.37239414 0.35814994]\n",
      "Trial error:  0.37239413825864054\n",
      "LR:  0.8499999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00010807803916550351  :: absSize readout:  0.0036404097285908958\n",
      "NRMSE W:  0.00696483912316207  :: absSize W:  0.07549966401558676\n",
      "NRMSEs: best:  0.00696483912316207 avg:  0.00696483912316207\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00192699        initial\n",
      "    2        618.034   0.00352055        golden\n",
      "    3        236.068   0.00282937        golden\n",
      "    4        400.334     0.178744        parabolic\n",
      "    5        326.238     0.163992        golden\n",
      "    6         360.68   0.00362866        golden\n",
      "    7        373.835   0.00262364        golden\n",
      "    8        388.982     0.177491        golden\n",
      "    9         378.86     0.176178        golden\n",
      "   10        384.646   0.00303323        golden\n",
      "   11         380.78     0.176447        golden\n",
      "   12         382.99   0.00309434        golden\n",
      "   13        381.513   0.00355314        golden\n",
      "   14        382.357    0.0030386        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.41838267 0.76509642 1.40132204 0.49485292 1.40132204]\n",
      "Trial error:  0.7650964177070954\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001060940446579158  :: absSize readout:  0.0037918235388222457\n",
      "NRMSE W:  0.007170001740334696  :: absSize W:  0.08418017154832234\n",
      "NRMSEs: best:  0.007170001740334696 avg:  0.007170001740334696\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.174434        initial\n",
      "    2        618.034   0.00860299        golden\n",
      "    3        763.932     0.224548        golden\n",
      "    4        561.469     0.202887        parabolic\n",
      "    5        673.762     0.221673        golden\n",
      "    6        596.428     0.206587        golden\n",
      "    7         639.32     0.217064        golden\n",
      "    8        609.781     0.208606        golden\n",
      "    9        626.165     0.212714        golden\n",
      "   10        614.882     0.209605        golden\n",
      "   11         621.14     0.211148        golden\n",
      "   12         616.83     0.210043        golden\n",
      "   13         619.22      0.21063        golden\n",
      "   14        617.574      0.21022        golden\n",
      "   15        618.487     0.210444        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.28457521 1.43379365 1.43379365 1.43379365 1.43379365]\n",
      "Trial error:  1.4337936545133554\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0004824240939713252  :: absSize readout:  0.0053282237878827365\n",
      "NRMSE W:  0.005701542193333902  :: absSize W:  0.08940120580129865\n",
      "NRMSEs: best:  0.005701542193333902 avg:  0.005701542193333902\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0455876        initial\n",
      "    2        618.034    0.0284009        golden\n",
      "    3        763.932    0.0199638        golden\n",
      "    4        854.102    0.0155086        golden\n",
      "    5         909.83    0.0138127        golden\n",
      "    6        944.272     0.013014        golden\n",
      "    7        965.558    0.0125832        golden\n",
      "    8        978.714    0.0123359        golden\n",
      "    9        986.844    0.0121901        golden\n",
      "   10        991.869    0.0121018        golden\n",
      "   11        994.975    0.0120483        golden\n",
      "   12        996.894    0.0120155        golden\n",
      "   13        998.081    0.0119956        golden\n",
      "   14        998.814    0.0119829        golden\n",
      "   15        999.267    0.0119755        golden\n",
      "   16          999.6    0.0119697        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.3662877 1.3662877 1.3662877 1.3662877 1.3662877]\n",
      "Trial error:  1.3662876978659737\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  6.357640723607406e-05  :: absSize readout:  0.0033678021278991887\n",
      "NRMSE W:  0.0068742478398206635  :: absSize W:  0.07068575337306918\n",
      "NRMSEs: best:  0.0068742478398206635 avg:  0.0068742478398206635\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0993267        initial\n",
      "    2        618.034    0.0023681        golden\n",
      "    3        763.932     0.113506        golden\n",
      "    4        566.902    0.0555322        parabolic\n",
      "    5        673.762    0.0216485        golden\n",
      "    6        632.558   0.00235682        parabolic\n",
      "    7        625.342   0.00989287        parabolic\n",
      "    8        648.297    0.0424943        golden\n",
      "    9         638.57    0.0402591        golden\n",
      "   10        629.802   0.00235913        golden\n",
      "   11        634.854    0.0156258        golden\n",
      "   12        631.505    0.0290898        golden\n",
      "   13        633.435   0.00235639        golden\n",
      "   14        633.977   0.00235594        golden\n",
      "   15        634.312    0.0121486        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.4342198  1.04234789 1.43398058 1.43365564 1.13703414]\n",
      "Trial error:  1.4336556409814065\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00011963448399293052  :: absSize readout:  0.0039650132760655765\n",
      "NRMSE W:  0.006971051906420633  :: absSize W:  0.06598735133024487\n",
      "NRMSEs: best:  0.006971051906420633 avg:  0.006971051906420633\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966      0.14038        initial\n",
      "    2        618.034      0.13992        golden\n",
      "    3        763.932     0.138861        golden\n",
      "    4        854.102     0.138781        golden\n",
      "    5         825.48     0.138739        parabolic\n",
      "    6        820.768     0.138725        parabolic\n",
      "    7        799.059     0.138765        golden\n",
      "    8        814.859     0.138743        parabolic\n",
      "    9        820.435     0.138725        parabolic\n",
      "   10        821.102      0.13874        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.40736991 1.40736991 1.40736991 1.40736991 1.40736991]\n",
      "Trial error:  1.4073699079361925\n",
      "LR:  0.9199999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  6.238836353002215e-05  :: absSize readout:  0.0030820633083665527\n",
      "NRMSE W:  0.008681839128314443  :: absSize W:  0.08437397365714944\n",
      "NRMSEs: best:  0.008681839128314443 avg:  0.008681839128314443\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.238563        initial\n",
      "    2        618.034     0.247216        golden\n",
      "    3        236.068      0.17472        golden\n",
      "    4        145.898     0.117079        golden\n",
      "    5        90.1699  0.000925821        golden\n",
      "    6        55.7281   0.00211984        golden\n",
      "    7        73.6866   0.00103831        parabolic\n",
      "    8        84.1292  0.000942432        parabolic\n",
      "    9         90.673  0.000937215        parabolic\n",
      "   10        87.5038  0.000914608        parabolic\n",
      "   11        86.2148  0.000906417        golden\n",
      "   12        85.4182  0.000939782        golden\n",
      "   13        86.7071  0.000900404        golden\n",
      "   14        87.0405  0.000905964        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.78490687 0.78941056 0.81044023 0.79053659 0.7874433 ]\n",
      "Trial error:  0.789410559457394\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00016805339193028083  :: absSize readout:  0.004125668592411652\n",
      "NRMSE W:  0.0068473144138096725  :: absSize W:  0.08538586003884463\n",
      "NRMSEs: best:  0.0068473144138096725 avg:  0.0068473144138096725\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00258149        initial\n",
      "    2        618.034     0.110436        golden\n",
      "    3        236.068   0.00244602        golden\n",
      "    4        308.628   0.00270679        parabolic\n",
      "    5        145.898   0.00554655        golden\n",
      "    6        201.626    0.0032849        golden\n",
      "    7        263.783   0.00371044        golden\n",
      "    8        222.912    0.0025731        golden\n",
      "    9        246.654   0.00313433        golden\n",
      "   10        231.043   0.00316585        golden\n",
      "   11        240.112   0.00278409        golden\n",
      "   12        234.149    0.0032234        golden\n",
      "   13        237.613   0.00277707        golden\n",
      "   14        235.335   0.00271038        golden\n",
      "   15        236.658   0.00394552        golden\n",
      "   16        235.735   0.00383382        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.16887036 1.10931515 1.06087339 1.0386446  1.02928141]\n",
      "Trial error:  1.060873387909779\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00016990816425177914  :: absSize readout:  0.0036384134633769315\n",
      "NRMSE W:  0.002517624921876042  :: absSize W:  0.07109511276491717\n",
      "NRMSEs: best:  0.002517624921876042 avg:  0.002517624921876042\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0167357        initial\n",
      "    2        618.034    0.0147932        golden\n",
      "    3        763.932       0.0128        golden\n",
      "    4        854.102    0.0115735        golden\n",
      "    5         909.83    0.0107713        golden\n",
      "    6        944.272    0.0103121        golden\n",
      "    7        965.558    0.0100381        golden\n",
      "    8        978.714   0.00987653        golden\n",
      "    9        986.844    0.0112517        golden\n",
      "   10        973.689   0.00993968        golden\n",
      "   11        981.819   0.00984264        golden\n",
      "   12        983.739   0.00982233        golden\n",
      "   13        984.925    0.0106197        golden\n",
      "   14        983.006    0.0098305        golden\n",
      "   15        984.192    0.0104217        golden\n",
      "   16        983.405   0.00982636        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.28215683 1.28228811 1.18755128 1.28186052 1.23971131]\n",
      "Trial error:  1.2818605167826465\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001224268117636352  :: absSize readout:  0.003624047704644462\n",
      "NRMSE W:  0.007229378486182466  :: absSize W:  0.09161472988725441\n",
      "NRMSEs: best:  0.007229378486182466 avg:  0.007229378486182466\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.201336        initial\n",
      "    2        618.034     0.202435        golden\n",
      "    3        236.068   0.00606171        golden\n",
      "    4        145.898   0.00576661        golden\n",
      "    5        190.694   0.00607494        parabolic\n",
      "    6        90.1699   0.00429807        golden\n",
      "    7        55.7281    0.0060743        golden\n",
      "    8        111.456   0.00509836        golden\n",
      "    9        77.0143    0.0035809        golden\n",
      "   10        68.8837    0.0069384        golden\n",
      "   11        82.0393   0.00387552        golden\n",
      "   12        73.9087   0.00721112        golden\n",
      "   13        78.9337   0.00531237        golden\n",
      "   14        75.8281   0.00601751        golden\n",
      "   15        77.7475   0.00701273        golden\n",
      "   16        76.5612   0.00670403        golden\n",
      "   17        77.3477    0.0075105        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.75714885 0.92532966 0.79742822 0.98181375 1.06765486]\n",
      "Trial error:  0.9253296566737892\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00013375526991748669  :: absSize readout:  0.0035802532565342208\n",
      "NRMSE W:  0.006784477070466907  :: absSize W:  0.06998717469831528\n",
      "NRMSEs: best:  0.006784477070466907 avg:  0.006784477070466907\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0602115        initial\n",
      "    2        618.034    0.0961788        golden\n",
      "    3        236.068    0.0391093        golden\n",
      "    4        145.898    0.0284887        golden\n",
      "    5        90.1699   0.00230693        golden\n",
      "    6        55.7281   0.00242667        golden\n",
      "    7        73.2802   0.00214155        parabolic\n",
      "    8        75.2485   0.00206313        parabolic\n",
      "    9        80.2531   0.00189923        parabolic\n",
      "   10         84.041   0.00226674        golden\n",
      "   11        78.3416    0.0019616        golden\n",
      "   12           81.7   0.00288197        golden\n",
      "   13         79.523   0.00208866        golden\n",
      "   14        80.8058   0.00219746        golden\n",
      "   15        79.9198   0.00228681        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.6162434  0.68638488 0.68112859 0.74704024 0.72555018]\n",
      "Trial error:  0.6863848776673459\n",
      "LR:  0.99\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00269694951817925  :: absSize readout:  0.006450068131345827\n",
      "NRMSE W:  0.008781391881379661  :: absSize W:  0.05104089603547186\n",
      "NRMSEs: best:  0.008781391881379661 avg:  0.008781391881379661\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00211558        initial\n",
      "    2        618.034   0.00064165        golden\n",
      "    3        763.932   0.00035068        golden\n",
      "    4        780.616  0.000328926        parabolic\n",
      "    5        864.413  0.000241711        golden\n",
      "    6        916.203  0.000201888        golden\n",
      "    7        948.211  0.000181284        golden\n",
      "    8        967.992  0.000169837        golden\n",
      "    9        980.218  0.000163206        golden\n",
      "   10        987.774  0.000159266        golden\n",
      "   11        992.444   0.00015689        golden\n",
      "   12         995.33  0.000155443        golden\n",
      "   13        997.114  0.000154557        golden\n",
      "   14        998.216  0.000154013        golden\n",
      "   15        998.898  0.000153677        golden\n",
      "   16        999.319  0.000153471        golden\n",
      "   17        999.652  0.000153307        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.35356085 1.35356085 1.35356085 1.35356085 1.35356085]\n",
      "Trial error:  1.353560845663198\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0009348075538222329  :: absSize readout:  0.00618528756516711\n",
      "NRMSE W:  0.010031430367429154  :: absSize W:  0.11897972728397399\n",
      "NRMSEs: best:  0.010031430367429154 avg:  0.010031430367429154\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0284617        initial\n",
      "    2        618.034    0.0167806        golden\n",
      "    3        763.932    0.0127029        golden\n",
      "    4         938.87   0.00937164        parabolic\n",
      "    5         872.05    0.0104912        golden\n",
      "    6        913.347   0.00978157        golden\n",
      "    7         962.22   0.00901282        golden\n",
      "    8         976.65   0.00880108        golden\n",
      "    9        985.569    0.0086733        golden\n",
      "   10        991.081   0.00859788        golden\n",
      "   11        994.488    0.0085502        golden\n",
      "   12        996.593    0.0085209        golden\n",
      "   13        997.895   0.00850048        golden\n",
      "   14        998.699   0.00849173        golden\n",
      "   15        999.196   0.00848486        golden\n",
      "   16        999.529   0.00847788        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43389284 1.43389284 1.43389284 1.43389284 1.43389284]\n",
      "Trial error:  1.433892836438769\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0010512463405722777  :: absSize readout:  0.005693819532190709\n",
      "NRMSE W:  0.008819328691455013  :: absSize W:  0.09668831802646587\n",
      "NRMSEs: best:  0.008819328691455013 avg:  0.008819328691455013\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0134891        initial\n",
      "    2        618.034    0.0114146        golden\n",
      "    3        763.932   0.00985578        golden\n",
      "    4        854.102   0.00873955        golden\n",
      "    5         909.83   0.00817812        golden\n",
      "    6        944.272   0.00790596        golden\n",
      "    7        965.558   0.00776837        golden\n",
      "    8        978.714   0.00769422        golden\n",
      "    9        986.844   0.00765728        golden\n",
      "   10        991.869   0.00763453        golden\n",
      "   11        994.975   0.00762246        golden\n",
      "   12        996.894   0.00761493        golden\n",
      "   13        998.081   0.00761158        golden\n",
      "   14        998.814   0.00760749        golden\n",
      "   15        999.267   0.00760604        golden\n",
      "   16          999.6   0.00760563        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.35860265 1.35861275 1.35860093 1.35860084 1.35860147]\n",
      "Trial error:  1.35860146989556\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001552335226367232  :: absSize readout:  0.003488165105407548\n",
      "NRMSE W:  0.006585183806914937  :: absSize W:  0.049385092368510596\n",
      "NRMSEs: best:  0.006585183806914937 avg:  0.006585183806914937\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00702022        initial\n",
      "    2        618.034   0.00629588        golden\n",
      "    3        763.932    0.0057162        golden\n",
      "    4        854.102   0.00539238        golden\n",
      "    5         909.83   0.00520674        golden\n",
      "    6        944.272    0.0050972        golden\n",
      "    7        965.558   0.00503136        golden\n",
      "    8        978.714   0.00499136        golden\n",
      "    9        986.844   0.00496689        golden\n",
      "   10        991.869   0.00495186        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   11        994.975   0.00494261        golden\n",
      "   12        996.894   0.00493691        golden\n",
      "   13        998.081   0.00493339        golden\n",
      "   14        998.814   0.00493121        golden\n",
      "   15        999.267   0.00492987        golden\n",
      "   16          999.6   0.00492888        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.27895313 1.27895313 1.27895313 1.27895313 1.27895313]\n",
      "Trial error:  1.2789531316165001\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001938863679992289  :: absSize readout:  0.0037929257860598795\n",
      "NRMSE W:  0.0021193902341672638  :: absSize W:  0.06869163008043752\n",
      "NRMSEs: best:  0.0021193902341672638 avg:  0.0021193902341672638\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0141502        initial\n",
      "    2        618.034    0.0334584        golden\n",
      "    3        236.068    0.0992711        golden\n",
      "    4        476.518    0.0145903        parabolic\n",
      "    5        424.973     0.014325        parabolic\n",
      "    6        326.238    0.0135189        golden\n",
      "    7        291.796     0.013021        golden\n",
      "    8         270.51    0.0125828        golden\n",
      "    9        257.354    0.0995545        golden\n",
      "   10         278.64    0.0127666        golden\n",
      "   11        265.485     0.012454        golden\n",
      "   12        262.379    0.0123712        golden\n",
      "   13         260.46    0.0123201        golden\n",
      "   14        259.274    0.0995248        golden\n",
      "   15        261.193     0.012342        golden\n",
      "   16        260.007    0.0123039        golden\n",
      "   17        259.673    0.0122953        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.41886941 1.36816984 1.36819914 1.36823282 1.3682136 ]\n",
      "Trial error:  1.3682136027912946\n",
      "Best error:  0.23426551902471393  best LR:  0.15\n",
      "0.15\n",
      "testing at  20.0  Hz\n",
      "Error:  0.22121764579704145\n",
      "testing at  21.189261887185907  Hz\n",
      "Error:  0.29232345444381963\n",
      "testing at  22.44924096618746  Hz\n",
      "Error:  0.1739971077897236\n",
      "testing at  23.78414230005442  Hz\n",
      "Error:  0.21736399039618798\n",
      "testing at  25.198420997897465  Hz\n",
      "Error:  0.2766129975145485\n",
      "testing at  26.69679708340069  Hz\n",
      "Error:  0.2387543414463976\n",
      "testing at  28.284271247461902  Hz\n",
      "Error:  0.21919011957902304\n",
      "testing at  29.96614153753363  Hz\n",
      "Error:  0.2353238871035458\n",
      "testing at  31.748021039363987  Hz\n",
      "Error:  0.26443695117954524\n",
      "testing at  33.63585661014858  Hz\n",
      "Error:  0.25774564407407163\n",
      "testing at  35.635948725613574  Hz\n",
      "Error:  0.2777470877552481\n",
      "testing at  37.75497250726774  Hz\n",
      "Error:  0.21346619723747307\n",
      "testing at  40.0  Hz\n",
      "Error:  0.21673468653677166\n",
      "testing at  42.378523774371814  Hz\n",
      "Error:  0.23970291951524017\n",
      "testing at  44.89848193237492  Hz\n",
      "Error:  0.17226821890567853\n",
      "testing at  47.56828460010884  Hz\n",
      "Error:  0.1882409051923474\n",
      "testing at  50.39684199579493  Hz\n",
      "Error:  0.18435283774990632\n",
      "testing at  53.39359416680138  Hz\n",
      "Error:  0.10285944635275968\n",
      "testing at  56.568542494923804  Hz\n",
      "Error:  0.24388871664829764\n",
      "testing at  59.93228307506726  Hz\n",
      "Error:  0.20256012120036826\n",
      "testing at  63.496042078727974  Hz\n",
      "Error:  0.248277381177713\n",
      "testing at  67.27171322029716  Hz\n",
      "Error:  0.19816030580213861\n",
      "testing at  71.27189745122715  Hz\n",
      "Error:  0.2327750295572135\n",
      "testing at  75.50994501453548  Hz\n",
      "Error:  0.25558243935278874\n",
      "testing at  80.0  Hz\n",
      "Error:  0.22752959436630318\n",
      "====== running test:  80 <function bl_sawtooth at 0x7fc012258e18>\n",
      "LR:  0.01\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.44839694893480525  :: absSize readout:  1.4753666364911564\n",
      "NRMSE W:  0.07438814577513589  :: absSize W:  0.740490685331863\n",
      "NRMSEs: best:  0.07438814577513589 avg:  0.07438814577513589\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  1.35708e-07        initial\n",
      "    2        618.034  1.13413e-07        golden\n",
      "    3        763.932  1.33157e-07        golden\n",
      "    4          578.5  1.45153e-07        parabolic\n",
      "    5        618.367  1.05797e-07        parabolic\n",
      "    6        673.968  1.22623e-07        golden\n",
      "    7        645.802  1.14184e-07        parabolic\n",
      "    8        631.901  9.90614e-08        parabolic\n",
      "    9        625.289    1.566e-07        parabolic\n",
      "   10        632.235  1.29042e-07        parabolic\n",
      "   11        629.376  1.54679e-07        golden\n",
      "   12        630.937  1.34499e-07        golden\n",
      "   13        631.533  1.21713e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43972131 1.43344257 1.43981144 1.43666696 1.43804058]\n",
      "Trial error:  1.4380405760819166\n",
      "Best so far ***************** \n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.454781339167017  :: absSize readout:  1.473750328966449\n",
      "NRMSE W:  0.07977979877858421  :: absSize W:  0.7555661405041225\n",
      "NRMSEs: best:  0.07977979877858421 avg:  0.07977979877858421\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  6.77299e-05        initial\n",
      "    2        618.034  6.09457e-08        golden\n",
      "    3        763.932  5.97694e-08        golden\n",
      "    4        690.988   6.4583e-08        parabolic\n",
      "    5        854.102  6.75836e-08        golden\n",
      "    6        798.374  6.01624e-08        golden\n",
      "    7        728.314  6.85967e-08        parabolic\n",
      "    8        750.327   5.9941e-08        golden\n",
      "    9        769.741  5.99029e-08        parabolic\n",
      "   10        760.569  6.40884e-08        parabolic\n",
      "   11        766.151  5.97473e-08        golden\n",
      "   12        765.584  6.19332e-08        parabolic\n",
      "   13        767.522  6.19889e-08        golden\n",
      "   14        766.675  6.26223e-08        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43029038 1.43584106 1.43136049 1.43170399 1.43662989]\n",
      "Trial error:  1.431703985315576\n",
      "Best so far ***************** \n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.44093108162635003  :: absSize readout:  1.6071662736024128\n",
      "NRMSE W:  0.08480380237744929  :: absSize W:  0.840661977525743\n",
      "NRMSEs: best:  0.08480380237744929 avg:  0.08480380237744929\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  1.31962e-06        initial\n",
      "    2        618.034  1.13809e-06        golden\n",
      "    3        763.932  1.36416e-06        golden\n",
      "    4        563.344   1.1452e-06        parabolic\n",
      "    5        609.137  1.24464e-06        parabolic\n",
      "    6        673.762  1.20713e-06        golden\n",
      "    7         639.32  1.22283e-06        golden\n",
      "    8        626.165  1.25433e-06        golden\n",
      "    9        614.636  1.13717e-06        golden\n",
      "   10        612.535  1.54852e-06        golden\n",
      "   11        615.934  1.34178e-06        golden\n",
      "   12        613.833  1.29966e-06        golden\n",
      "   13        615.131  1.38014e-06        golden\n",
      "   14        614.302  1.30312e-06        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.95926908 1.18364208 0.8280661  1.3363107  0.67677291]\n",
      "Trial error:  0.9592690802218186\n",
      "Best so far ***************** \n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.44249994435543294  :: absSize readout:  1.57617336809886\n",
      "NRMSE W:  0.08366720942268978  :: absSize W:  0.8352046581827417\n",
      "NRMSEs: best:  0.08366720942268978 avg:  0.08366720942268978\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   7.3254e-07        initial\n",
      "    2        618.034  7.85631e-07        golden\n",
      "    3        236.068  1.18338e-06        golden\n",
      "    4        487.043  4.58048e-05        parabolic\n",
      "    5        326.238  4.63799e-05        golden\n",
      "    6        422.102  5.57055e-05        golden\n",
      "    7         360.68  5.35566e-05        golden\n",
      "    8        397.297  5.68755e-05        golden\n",
      "    9        373.835  5.45762e-05        golden\n",
      "   10        387.822  5.53779e-05        golden\n",
      "   11         378.86  4.48753e-05        golden\n",
      "   12        384.203  5.61603e-05        golden\n",
      "   13         380.78  5.50592e-05        golden\n",
      "   14         382.82  5.56811e-05        golden\n",
      "   15        381.513  5.35183e-05        golden\n",
      "   16        382.299  5.74591e-05        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1.41892185 1.41919155 1.41924626 1.42066952 1.41966368]\n",
      "Trial error:  1.419246264127903\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.4306093512242736  :: absSize readout:  1.5649949913404395\n",
      "NRMSE W:  0.08453758886892893  :: absSize W:  0.8101886980309934\n",
      "NRMSEs: best:  0.08453758886892893 avg:  0.08453758886892893\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  4.58685e-05        initial\n",
      "    2        618.034  1.29969e-07        golden\n",
      "    3        763.932  1.16451e-07        golden\n",
      "    4        691.074  1.09137e-07        parabolic\n",
      "    5         708.51  1.11209e-07        parabolic\n",
      "    6        663.175  1.17763e-07        golden\n",
      "    7        680.418  1.10092e-07        golden\n",
      "    8        691.784  1.30202e-07        parabolic\n",
      "    9        690.741  1.15343e-07        parabolic\n",
      "   10        691.408  1.23577e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.40909028 1.43545403 1.43065915 1.21337453 1.35992379]\n",
      "Trial error:  1.4090902849896434\n",
      "LR:  0.07999999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.11042698257029636  :: absSize readout:  0.44493499107161694\n",
      "NRMSE W:  0.03698066004222711  :: absSize W:  0.35044348414396476\n",
      "NRMSEs: best:  0.03698066004222711 avg:  0.03698066004222711\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00112817        initial\n",
      "    2        618.034   0.00215618        golden\n",
      "    3        236.068  0.000574209        golden\n",
      "    4        145.898  0.000232208        golden\n",
      "    5        90.1699  4.50004e-06        golden\n",
      "    6        55.7281  4.54493e-06        golden\n",
      "    7        72.9634  4.52099e-06        parabolic\n",
      "    8        111.456  4.45592e-06        golden\n",
      "    9        124.612  6.56902e-05        golden\n",
      "   10        103.326   4.4711e-06        golden\n",
      "   11        116.481  9.17597e-06        golden\n",
      "   12        108.351  4.45755e-06        golden\n",
      "   13        110.886  4.45046e-06        parabolic\n",
      "   14        109.969  4.45561e-06        parabolic\n",
      "   15        110.553  4.45409e-06        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43080052 1.43428295 1.43537563 1.43377468 1.43528635]\n",
      "Trial error:  1.4342829497753105\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.10523139815668876  :: absSize readout:  0.4433295234620468\n",
      "NRMSE W:  0.0345768911223287  :: absSize W:  0.3341174742928348\n",
      "NRMSEs: best:  0.0345768911223287 avg:  0.0345768911223287\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000850127        initial\n",
      "    2        618.034  0.000927629        golden\n",
      "    3        236.068  0.000869086        golden\n",
      "    4        363.175  0.000704092        parabolic\n",
      "    5        310.062   0.00062074        parabolic\n",
      "    6        281.799   0.00100744        golden\n",
      "    7        327.522  0.000610411        parabolic\n",
      "    8        323.672  0.000647567        parabolic\n",
      "    9         341.14  0.000631859        golden\n",
      "   10        332.724  0.000678598        golden\n",
      "   11        329.509  0.000671824        golden\n",
      "   12        326.051  0.000665363        golden\n",
      "   13        328.281  0.000659163        golden\n",
      "   14         326.96  0.000658352        golden\n",
      "   15        327.855  0.000654411        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.44627347 1.44761168 1.44905755 1.44773497 1.44873924]\n",
      "Trial error:  1.447734966997424\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.11279097352909158  :: absSize readout:  0.47746382055539655\n",
      "NRMSE W:  0.035149399803050166  :: absSize W:  0.3503393675157868\n",
      "NRMSEs: best:  0.035149399803050166 avg:  0.035149399803050166\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00121166        initial\n",
      "    2        618.034   0.00180035        golden\n",
      "    3        236.068  0.000430522        golden\n",
      "    4        145.898  4.65391e-06        golden\n",
      "    5        90.1699  4.37629e-06        golden\n",
      "    6        117.957  4.50026e-06        parabolic\n",
      "    7        55.7281  5.31889e-06        golden\n",
      "    8        77.0143  4.88853e-06        golden\n",
      "    9        100.784  4.61253e-06        golden\n",
      "   10        85.1449  4.28307e-06        golden\n",
      "   11        82.0393  4.74162e-06        golden\n",
      "   12        87.0643  3.89571e-06        golden\n",
      "   13        87.5267  3.83445e-06        parabolic\n",
      "   14        88.5363  5.16375e-06        golden\n",
      "   15        87.9124  4.98701e-06        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.77592214 0.66513399 0.97898118 0.71879759 0.74306074]\n",
      "Trial error:  0.7430607385305997\n",
      "Best so far ***************** \n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.11149428767494761  :: absSize readout:  0.4316320299641575\n",
      "NRMSE W:  0.03724118661929589  :: absSize W:  0.3529657935607466\n",
      "NRMSEs: best:  0.03724118661929589 avg:  0.03724118661929589\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00176667        initial\n",
      "    2        618.034   0.00261839        golden\n",
      "    3        236.068   7.4419e-05        golden\n",
      "    4        145.898  1.49096e-05        golden\n",
      "    5        183.862  1.47407e-05        parabolic\n",
      "    6        165.055  1.49394e-05        parabolic\n",
      "    7        203.803  1.44365e-05        golden\n",
      "    8        216.127  1.44679e-05        golden\n",
      "    9        207.656  1.44337e-05        parabolic\n",
      "   10        206.669  1.43752e-05        parabolic\n",
      "   11        205.747    1.443e-05        parabolic\n",
      "   12        207.003  1.44389e-05        parabolic\n",
      "   13        206.336  1.44223e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.38870672 1.38447767 1.38692941 1.38710784 1.38176679]\n",
      "Trial error:  1.386929412418421\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.1094067780783772  :: absSize readout:  0.4249719111586648\n",
      "NRMSE W:  0.034397419079263566  :: absSize W:  0.3645663596654933\n",
      "NRMSEs: best:  0.034397419079263566 avg:  0.034397419079263566\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000651178        initial\n",
      "    2        618.034   0.00148641        golden\n",
      "    3        236.068   9.2411e-05        golden\n",
      "    4        145.898  3.64772e-05        golden\n",
      "    5         168.17  4.01636e-05        parabolic\n",
      "    6        144.679  1.19008e-05        parabolic\n",
      "    7        89.4166   7.9701e-06        golden\n",
      "    8        116.948  7.73261e-06        parabolic\n",
      "    9        104.682  7.90912e-06        parabolic\n",
      "   10         127.54  7.74482e-06        golden\n",
      "   11        121.396  7.86462e-06        parabolic\n",
      "   12        112.263  7.80812e-06        golden\n",
      "   13        115.158  7.93871e-06        golden\n",
      "   14        118.647  7.57575e-06        golden\n",
      "   15        119.697   7.8036e-06        golden\n",
      "   16        117.998  7.83633e-06        golden\n",
      "   17        119.048  7.67778e-06        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43766756 1.43927597 1.43915396 1.43860477 1.43899248]\n",
      "Trial error:  1.4389924831657837\n",
      "LR:  0.15\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.04400208961343244  :: absSize readout:  0.26054662392133615\n",
      "NRMSE W:  0.02336952733235609  :: absSize W:  0.3115939623922393\n",
      "NRMSEs: best:  0.02336952733235609 avg:  0.02336952733235609\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00352479        initial\n",
      "    2        618.034   0.00330384        golden\n",
      "    3        763.932    0.0048802        golden\n",
      "    4        515.225   0.00342331        parabolic\n",
      "    5        673.762   0.00363387        golden\n",
      "    6        578.765   0.00330103        golden\n",
      "    7        596.558   0.00328086        parabolic\n",
      "    8        597.762   0.00328222        parabolic\n",
      "    9        592.429    0.0033121        parabolic\n",
      "   10        594.981   0.00328214        golden\n",
      "   11        596.224   0.00326133        parabolic\n",
      "   12        595.778   0.00327328        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1.43136495 1.43517033 1.43332138 1.43199855 1.4329247 ]\n",
      "Trial error:  1.4329247020560139\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0470489723469373  :: absSize readout:  0.2859871424524547\n",
      "NRMSE W:  0.026272658569096347  :: absSize W:  0.32128989636504934\n",
      "NRMSEs: best:  0.026272658569096347 avg:  0.026272658569096347\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00284171        initial\n",
      "    2        618.034   0.00265165        golden\n",
      "    3        763.932    0.0028185        golden\n",
      "    4        578.904    0.0026683        parabolic\n",
      "    5         623.56    0.0026612        parabolic\n",
      "    6        602.879   0.00272701        parabolic\n",
      "    7        612.245   0.00267556        golden\n",
      "    8        615.823   0.00260299        golden\n",
      "    9        614.456   0.00265524        golden\n",
      "   10        616.275   0.00254653        parabolic\n",
      "   11        616.797   0.00261629        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.44313673 1.44266403 1.43721563 1.43858345 1.44474479]\n",
      "Trial error:  1.4426640309780132\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.04416153957341053  :: absSize readout:  0.2693927274866099\n",
      "NRMSE W:  0.024585186269304858  :: absSize W:  0.3082020648271543\n",
      "NRMSEs: best:  0.024585186269304858 avg:  0.024585186269304858\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00329107        initial\n",
      "    2        618.034   0.00302827        golden\n",
      "    3        763.932    0.0029537        golden\n",
      "    4        853.113   0.00306737        parabolic\n",
      "    5        724.627   0.00293952        parabolic\n",
      "    6        722.222   0.00302783        parabolic\n",
      "    7        744.077   0.00294754        parabolic\n",
      "    8        732.056   0.00294455        golden\n",
      "    9        727.465   0.00292856        golden\n",
      "   10        727.999   0.00293277        parabolic\n",
      "   11          726.6   0.00302883        parabolic\n",
      "   12        727.131   0.00293802        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.67380676 1.43061297 1.42857948 1.42865058 1.42945425]\n",
      "Trial error:  1.4286505788122301\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.04364015281210676  :: absSize readout:  0.24589374547959642\n",
      "NRMSE W:  0.02260541663422896  :: absSize W:  0.33771082185567947\n",
      "NRMSEs: best:  0.02260541663422896 avg:  0.02260541663422896\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000602608        initial\n",
      "    2        618.034   0.00208095        golden\n",
      "    3        236.068  0.000968501        golden\n",
      "    4        363.629  0.000777638        parabolic\n",
      "    5        472.136  0.000671011        golden\n",
      "    6        423.057   0.00055343        parabolic\n",
      "    7        417.531  0.000549321        parabolic\n",
      "    8        413.479  0.000549038        parabolic\n",
      "    9        415.008  0.000547335        parabolic\n",
      "   10         415.43  0.000547325        parabolic\n",
      "   11        415.764  0.000555733        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.20517605 1.31375641 1.29942393 1.31377606 1.31242823]\n",
      "Trial error:  1.312428228334414\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.04542389844120544  :: absSize readout:  0.2582796833060384\n",
      "NRMSE W:  0.0246908674855586  :: absSize W:  0.3386025938288988\n",
      "NRMSEs: best:  0.0246908674855586 avg:  0.0246908674855586\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00262642        initial\n",
      "    2        618.034    0.0982112        golden\n",
      "    3        236.068   0.00200465        golden\n",
      "    4        306.986   0.00240678        parabolic\n",
      "    5        145.898  0.000677071        golden\n",
      "    6        90.1699  0.000199961        golden\n",
      "    7        55.7281  6.78182e-05        golden\n",
      "    8         36.338  6.84346e-05        parabolic\n",
      "    9        46.2542  7.09809e-05        parabolic\n",
      "   10        68.8837  6.79709e-05        golden\n",
      "   11        57.9529  7.02288e-05        parabolic\n",
      "   12        55.3948  6.79946e-05        parabolic\n",
      "   13        56.5779  7.08336e-05        golden\n",
      "   14        56.0614  6.88722e-05        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.39284477 1.43247318 1.42623212 1.40232577 1.38219871]\n",
      "Trial error:  1.4023257664319004\n",
      "LR:  0.21999999999999997\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.019716669528328963  :: absSize readout:  0.15697540062100931\n",
      "NRMSE W:  0.020277158467253167  :: absSize W:  0.32355377128865664\n",
      "NRMSEs: best:  0.020277158467253167 avg:  0.020277158467253167\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00429766        initial\n",
      "    2        618.034     0.224219        golden\n",
      "    3        236.068   0.00421168        golden\n",
      "    4        308.896   0.00434084        parabolic\n",
      "    5        145.898   0.00328289        golden\n",
      "    6        90.1699    0.0019354        golden\n",
      "    7        55.7281  0.000788572        golden\n",
      "    8        34.4419  0.000752608        golden\n",
      "    9        43.5955  0.000771257        parabolic\n",
      "   10        21.2862  0.000690495        golden\n",
      "   11        13.1556  0.000594543        golden\n",
      "   12        8.13062  0.000282477        golden\n",
      "   13          5.025  0.000209718        golden\n",
      "   14        4.11512  0.000228937        parabolic\n",
      "   15        5.52198  0.000209618        parabolic\n",
      "   16        5.85531  0.000214217        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.76206399 0.76344556 0.76225092 0.76560136 0.76232886]\n",
      "Trial error:  0.7623288588939863\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.015871507585536748  :: absSize readout:  0.12893140622404625\n",
      "NRMSE W:  0.015421002842943407  :: absSize W:  0.2711471330369593\n",
      "NRMSEs: best:  0.015421002842943407 avg:  0.015421002842943407\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000856197        initial\n",
      "    2        618.034     0.001917        golden\n",
      "    3        236.068  0.000409478        golden\n",
      "    4        145.898  0.000124786        golden\n",
      "    5        90.1699  1.69371e-05        golden\n",
      "    6        55.7281  1.71666e-05        golden\n",
      "    7        73.1037  1.44672e-05        parabolic\n",
      "    8        73.4371  1.44062e-05        parabolic\n",
      "    9        77.9405  1.42735e-05        parabolic\n",
      "   10        82.6117  1.46588e-05        golden\n",
      "   11        76.1537   1.4629e-05        parabolic\n",
      "   12        79.7247  1.43401e-05        golden\n",
      "   13         77.076  1.52119e-05        parabolic\n",
      "   14        78.2738   1.5098e-05        parabolic\n",
      "   15        77.6071  1.44341e-05        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.17465412 0.18138197 0.17642468 0.13834236 0.3243561 ]\n",
      "Trial error:  0.1764246771209863\n",
      "Best so far ***************** \n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.020516720626069664  :: absSize readout:  0.16352206702226807\n",
      "NRMSE W:  0.018052865009430513  :: absSize W:  0.27998625970368596\n",
      "NRMSEs: best:  0.018052865009430513 avg:  0.018052865009430513\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000732672        initial\n",
      "    2        618.034      0.19673        golden\n",
      "    3        236.068  0.000595371        golden\n",
      "    4          308.8  0.000713377        parabolic\n",
      "    5        145.898  0.000461944        golden\n",
      "    6        90.1699  0.000269212        golden\n",
      "    7        55.7281  0.000133501        golden\n",
      "    8        34.4419  9.39497e-05        golden\n",
      "    9        20.2213  7.89918e-05        parabolic\n",
      "   10        12.4975  7.68406e-05        golden\n",
      "   11        12.1641  7.85595e-05        parabolic\n",
      "   12        16.1529   7.3607e-05        parabolic\n",
      "   13        17.7069  7.55773e-05        golden\n",
      "   14        15.3956  7.18086e-05        parabolic\n",
      "   15        14.2886  7.37832e-05        golden\n",
      "   16        15.0623   7.1491e-05        parabolic\n",
      "   17         14.729  7.23551e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.2488505  0.40320982 0.25483124 0.38746522 0.35512895]\n",
      "Trial error:  0.35512894549968144\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.018086724481945496  :: absSize readout:  0.1441389690677803\n",
      "NRMSE W:  0.01838550106330578  :: absSize W:  0.289410001190287\n",
      "NRMSEs: best:  0.01838550106330578 avg:  0.01838550106330578\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0023221        initial\n",
      "    2        618.034     0.242854        golden\n",
      "    3        236.068   0.00212832        golden\n",
      "    4        308.768   0.00225909        parabolic\n",
      "    5        145.898   0.00181264        golden\n",
      "    6        90.1699   0.00357033        golden\n",
      "    7         180.34   0.00195771        golden\n",
      "    8        124.612   0.00279046        golden\n",
      "    9        159.054   0.00187222        golden\n",
      "   10        137.767   0.00177266        golden\n",
      "   11        132.742   0.00287685        golden\n",
      "   12        140.873   0.00178825        golden\n",
      "   13        135.848   0.00176281        golden\n",
      "   14        134.662   0.00288298        golden\n",
      "   15        136.581   0.00176659        golden\n",
      "   16        135.395   0.00263774        golden\n",
      "   17        136.181   0.00176454        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42160823 1.41567225 1.42008257 1.42008256 1.42110416]\n",
      "Trial error:  1.4200825674743005\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.017443910753719344  :: absSize readout:  0.1396326776413759\n",
      "NRMSE W:  0.018561154154154262  :: absSize W:  0.32424221288070626\n",
      "NRMSEs: best:  0.018561154154154262 avg:  0.018561154154154262\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.178228        initial\n",
      "    2        618.034     0.375023        golden\n",
      "    3        236.068   0.00183136        golden\n",
      "    4        145.898   0.00176805        golden\n",
      "    5        190.914   0.00203641        parabolic\n",
      "    6        90.1699    0.0015606        golden\n",
      "    7        55.7281  0.000704734        golden\n",
      "    8        34.4419  0.000249973        golden\n",
      "    9        21.2862  0.000235819        golden\n",
      "   10        26.9508  0.000239046        parabolic\n",
      "   11        19.9041  0.000236233        parabolic\n",
      "   12        21.8096  0.000235809        parabolic\n",
      "   13        22.1429  0.000235841        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43957467 1.43956721 1.43956746 1.43957429 1.43956722]\n",
      "Trial error:  1.4395674625374943\n",
      "LR:  0.29\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.008084041503398531  :: absSize readout:  0.07177822936632293\n",
      "NRMSE W:  0.012588673094054306  :: absSize W:  0.23616986148276486\n",
      "NRMSEs: best:  0.012588673094054306 avg:  0.012588673094054306\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00195466        initial\n",
      "    2        618.034     0.140255        golden\n",
      "    3        236.068   0.00214936        golden\n",
      "    4        309.451   0.00205151        parabolic\n",
      "    5        472.136    0.0720807        golden\n",
      "    6        416.408    0.0389161        golden\n",
      "    7        354.268   0.00199317        golden\n",
      "    8        395.122   0.00193457        golden\n",
      "    9        403.252    0.0285661        golden\n",
      "   10        390.097   0.00194244        golden\n",
      "   11        398.227   0.00192956        golden\n",
      "   12        400.147    0.0019264        golden\n",
      "   13        401.333   0.00192442        golden\n",
      "   14        402.066    0.0275768        golden\n",
      "   15         400.88    0.0265593        golden\n",
      "   16        401.666    0.0272371        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43604831 1.43604831 1.43604831 1.43604831 1.4223725 ]\n",
      "Trial error:  1.436048309301213\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00698114291680605  :: absSize readout:  0.07265349545992002\n",
      "NRMSE W:  0.013290586101003872  :: absSize W:  0.2379948766844074\n",
      "NRMSEs: best:  0.013290586101003872 avg:  0.013290586101003872\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.243675        initial\n",
      "    2        618.034     0.331658        golden\n",
      "    3        236.068   0.00296531        golden\n",
      "    4        145.898  0.000144482        golden\n",
      "    5        188.702  0.000454446        parabolic\n",
      "    6        160.166  0.000187938        parabolic\n",
      "    7        142.676  3.96777e-05        parabolic\n",
      "    8        88.1786  2.23868e-05        golden\n",
      "    9        115.143  1.94946e-05        parabolic\n",
      "   10        105.139  2.49391e-05        parabolic\n",
      "   11         125.66   3.9644e-05        golden\n",
      "   12         119.16  2.75763e-05        golden\n",
      "   13        111.322   2.4466e-05        golden\n",
      "   14        116.677  1.94924e-05        golden\n",
      "   15        116.104  1.85424e-05        parabolic\n",
      "   16         115.77   1.8554e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.18823448 0.20417448 0.19772996 0.20468145 0.19644532]\n",
      "Trial error:  0.19772996302761534\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.007718415934050759  :: absSize readout:  0.07063137441919629\n",
      "NRMSE W:  0.011654654257387797  :: absSize W:  0.2848178063305524\n",
      "NRMSEs: best:  0.011654654257387797 avg:  0.011654654257387797\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.463869        initial\n",
      "    2        618.034     0.476319        golden\n",
      "    3        236.068     0.413101        golden\n",
      "    4        145.898   0.00104035        golden\n",
      "    5        90.1699  0.000603347        golden\n",
      "    6        117.909  0.000992965        parabolic\n",
      "    7        55.7281  0.000313268        golden\n",
      "    8        26.3873  0.000349997        parabolic\n",
      "    9        45.1844  0.000485238        parabolic\n",
      "   10        68.8837   0.00036664        golden\n",
      "   11        46.0681  0.000465645        parabolic\n",
      "   12        60.7531  0.000246082        golden\n",
      "   13        63.8587  0.000338553        golden\n",
      "   14        59.5004  0.000278788        parabolic\n",
      "   15        61.9393   0.00024317        golden\n",
      "   16        61.4728  0.000173546        parabolic\n",
      "   17        61.1394  0.000250159        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.95514253 1.20452431 1.15220367 1.07504251 1.15657975]\n",
      "Trial error:  1.1522036733058156\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.007352041941748149  :: absSize readout:  0.06646602848630273\n",
      "NRMSE W:  0.011841300935970932  :: absSize W:  0.25706122826849304\n",
      "NRMSEs: best:  0.011841300935970932 avg:  0.011841300935970932\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966      0.40509        initial\n",
      "    2        618.034     0.429793        golden\n",
      "    3        236.068     0.278597        golden\n",
      "    4        145.898   0.00135115        golden\n",
      "    5        90.1699   0.00183236        golden\n",
      "    6        118.238   0.00174692        parabolic\n",
      "    7         180.34   0.00190238        golden\n",
      "    8        159.054   0.00127019        golden\n",
      "    9        167.875    0.0014241        parabolic\n",
      "   10        155.341   0.00104093        parabolic\n",
      "   11        152.904   0.00107527        parabolic\n",
      "   12        154.694  0.000840349        parabolic\n",
      "   13        154.161   0.00080836        parabolic\n",
      "   14        153.828  0.000770503        parabolic\n",
      "   15        153.475   0.00110495        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.28598034 1.19535531 1.23140743 1.23532009 1.30697376]\n",
      "Trial error:  1.2353200878939476\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.009859590144288376  :: absSize readout:  0.08891264140840945\n",
      "NRMSE W:  0.016088305139371012  :: absSize W:  0.3042229695401171\n",
      "NRMSEs: best:  0.016088305139371012 avg:  0.016088305139371012\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.176082        initial\n",
      "    2        618.034     0.244219        golden\n",
      "    3        236.068   0.00523816        golden\n",
      "    4        145.898   0.00542676        golden\n",
      "    5        191.193   0.00554435        parabolic\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    6        291.796   0.00464209        golden\n",
      "    7        326.238   0.00460845        golden\n",
      "    8        313.547    0.0832929        parabolic\n",
      "    9        326.571   0.00441005        parabolic\n",
      "   10         347.73     0.145326        golden\n",
      "   11        334.653     0.127845        golden\n",
      "   12        329.658     0.119781        golden\n",
      "   13         327.75     0.116425        golden\n",
      "   14        327.022   0.00444668        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43216716 1.43216716 1.43216716 1.43216716 1.31758122]\n",
      "Trial error:  1.4321671579706465\n",
      "LR:  0.36\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0027568918588291897  :: absSize readout:  0.035639113304049175\n",
      "NRMSE W:  0.011153372251739537  :: absSize W:  0.2230610586637657\n",
      "NRMSEs: best:  0.011153372251739537 avg:  0.011153372251739537\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.322329        initial\n",
      "    2        618.034     0.355456        golden\n",
      "    3        236.068   0.00841401        golden\n",
      "    4        145.898   0.00524906        golden\n",
      "    5        189.026   0.00615273        parabolic\n",
      "    6        132.622   0.00540717        parabolic\n",
      "    7        149.481   0.00527594        parabolic\n",
      "    8        144.431    0.0051996        parabolic\n",
      "    9        139.921   0.00533315        golden\n",
      "   10        142.708   0.00542378        golden\n",
      "   11        143.773   0.00528668        golden\n",
      "   12        144.992    0.0076078        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42164788 1.42164788 1.34552951 1.33795676 1.36764937]\n",
      "Trial error:  1.36764937266557\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0031810375645605856  :: absSize readout:  0.040605531846954125\n",
      "NRMSE W:  0.009879008598435891  :: absSize W:  0.22103164533039085\n",
      "NRMSEs: best:  0.009879008598435891 avg:  0.009879008598435891\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.416069        initial\n",
      "    2        618.034     0.436179        golden\n",
      "    3        236.068     0.335654        golden\n",
      "    4        145.898   0.00671682        golden\n",
      "    5        90.1699   0.00604504        golden\n",
      "    6        117.792   0.00641159        parabolic\n",
      "    7        55.7281   0.00542609        golden\n",
      "    8        34.4419   0.00450031        golden\n",
      "    9        21.2862   0.00192768        golden\n",
      "   10        13.1556   0.00044866        golden\n",
      "   11        8.13062  0.000468384        golden\n",
      "   12        10.7821  0.000455173        parabolic\n",
      "   13        15.0481  0.000441124        parabolic\n",
      "   14        17.4309  0.000445556        golden\n",
      "   15        15.5589  0.000441887        parabolic\n",
      "   16        14.3253  0.000445267        golden\n",
      "   17        14.7148  0.000445621        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.37589825 1.37496616 1.37497728 1.37522701 1.37587576]\n",
      "Trial error:  1.375227008412022\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.002889262552123355  :: absSize readout:  0.03605396410455181\n",
      "NRMSE W:  0.011307879367484452  :: absSize W:  0.24546719830403568\n",
      "NRMSEs: best:  0.011307879367484452 avg:  0.011307879367484452\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.408151        initial\n",
      "    2        618.034     0.421662        golden\n",
      "    3        236.068      0.36295        golden\n",
      "    4        145.898   0.00544254        golden\n",
      "    5        90.1699   0.00339271        golden\n",
      "    6        117.351   0.00461115        parabolic\n",
      "    7        55.7281   0.00150426        golden\n",
      "    8        34.4419   0.00099842        golden\n",
      "    9        23.7709   0.00104847        parabolic\n",
      "   10        31.7403   0.00101315        parabolic\n",
      "   11        42.5725  0.000764593        golden\n",
      "   12        47.5975  0.000722855        golden\n",
      "   13        47.9308  0.000768766        parabolic\n",
      "   14        45.2374  0.000442517        parabolic\n",
      "   15        44.2195  0.000941089        golden\n",
      "   16         44.904  0.000410331        parabolic\n",
      "   17        44.5707  0.000371107        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.95187879 0.92824347 0.95263031 0.94812146 0.95141675]\n",
      "Trial error:  0.9514167470674298\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.003348539768036352  :: absSize readout:  0.04115960732035177\n",
      "NRMSE W:  0.010557770242845526  :: absSize W:  0.2168303387416447\n",
      "NRMSEs: best:  0.010557770242845526 avg:  0.010557770242845526\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.116476        initial\n",
      "    2        618.034     0.326706        golden\n",
      "    3        236.068   0.00273182        golden\n",
      "    4        145.898    0.0028519        golden\n",
      "    5        191.184   0.00277901        parabolic\n",
      "    6        291.796    0.0267499        golden\n",
      "    7        257.354   0.00390529        golden\n",
      "    8        218.924   0.00179989        golden\n",
      "    9        213.889   0.00207832        parabolic\n",
      "   10        221.999   0.00179508        parabolic\n",
      "   11         220.58   0.00171071        parabolic\n",
      "   12        220.246   0.00216341        parabolic\n",
      "   13        220.913   0.00206093        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.94707479 0.95360712 1.08013877 0.87900207 0.93210505]\n",
      "Trial error:  0.9470747856275632\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.002776689814350124  :: absSize readout:  0.03702987387735377\n",
      "NRMSE W:  0.009158654128999712  :: absSize W:  0.18370916042908914\n",
      "NRMSEs: best:  0.009158654128999712 avg:  0.009158654128999712\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.375735        initial\n",
      "    2        618.034     0.403424        golden\n",
      "    3        236.068     0.268508        golden\n",
      "    4        145.898  0.000428719        golden\n",
      "    5        90.1699   0.00044016        golden\n",
      "    6        118.039  0.000435011        parabolic\n",
      "    7         180.34  6.96023e-05        golden\n",
      "    8        201.626    7.169e-05        golden\n",
      "    9        190.723  6.08105e-05        parabolic\n",
      "   10         190.39  0.000433295        parabolic\n",
      "   11        191.057   6.5089e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.32762132 0.63078137 0.61232161 1.21836076 1.25820248]\n",
      "Trial error:  1.2183607601595534\n",
      "LR:  0.42999999999999994\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.001259761641913882  :: absSize readout:  0.020279897582206895\n",
      "NRMSE W:  0.0071631749404607  :: absSize W:  0.14636156224979766\n",
      "NRMSEs: best:  0.0071631749404607 avg:  0.0071631749404607\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.345008        initial\n",
      "    2        618.034     0.360703        golden\n",
      "    3        236.068     0.292576        golden\n",
      "    4        145.898   0.00273444        golden\n",
      "    5        90.1699   0.00108558        golden\n",
      "    6        117.356   0.00125968        parabolic\n",
      "    7        99.8211   0.00082587        parabolic\n",
      "    8        102.078  0.000862975        parabolic\n",
      "    9        98.6912  0.000825219        parabolic\n",
      "   10        98.3579  0.000881102        parabolic\n",
      "   11        99.0245  0.000834392        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.1785491  1.01464433 0.70168486 1.00833801 1.03494185]\n",
      "Trial error:  1.0146443253618844\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0010859353249513116  :: absSize readout:  0.01869075620604332\n",
      "NRMSE W:  0.008193003531376863  :: absSize W:  0.17902337569634266\n",
      "NRMSEs: best:  0.008193003531376863 avg:  0.008193003531376863\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.367207        initial\n",
      "    2        618.034      0.38215        golden\n",
      "    3        236.068     0.324668        golden\n",
      "    4        145.898  3.76596e-05        golden\n",
      "    5        90.1699  2.81118e-05        golden\n",
      "    6        118.031  1.46325e-05        parabolic\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    7         114.39  1.52413e-05        parabolic\n",
      "    8        122.606  1.51563e-05        parabolic\n",
      "    9        118.649  1.56817e-05        parabolic\n",
      "   10         116.64   1.4227e-05        golden\n",
      "   11        115.781  1.41159e-05        golden\n",
      "   12        115.315  1.37522e-05        parabolic\n",
      "   13        114.962  1.44207e-05        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.19787298 0.15215047 0.15632906 0.15523416 0.16502099]\n",
      "Trial error:  0.15632906341454575\n",
      "Best so far ***************** \n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0011418796318062939  :: absSize readout:  0.021588226351985904\n",
      "NRMSE W:  0.007966419725701784  :: absSize W:  0.17568708767472793\n",
      "NRMSEs: best:  0.007966419725701784 avg:  0.007966419725701784\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.301923        initial\n",
      "    2        618.034     0.336551        golden\n",
      "    3        236.068     0.180389        golden\n",
      "    4        145.898  0.000131563        golden\n",
      "    5        90.1699  4.33391e-05        golden\n",
      "    6        117.976  6.49377e-05        parabolic\n",
      "    7        90.6246   0.00185698        parabolic\n",
      "    8        89.8366  4.33607e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.27258224 0.22728669 0.21685766 0.22731034 0.2234848 ]\n",
      "Trial error:  0.22728669144064365\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0010912090722969992  :: absSize readout:  0.019953924726196927\n",
      "NRMSE W:  0.00872081010765099  :: absSize W:  0.22100489806416945\n",
      "NRMSEs: best:  0.00872081010765099 avg:  0.00872081010765099\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.384678        initial\n",
      "    2        618.034     0.390035        golden\n",
      "    3        236.068     0.358872        golden\n",
      "    4        145.898     0.209495        golden\n",
      "    5        90.1699   0.00546593        golden\n",
      "    6        55.7281   0.00254946        golden\n",
      "    7        71.8816   0.00464886        parabolic\n",
      "    8        34.4419   0.00183192        golden\n",
      "    9        38.5293   0.00167751        parabolic\n",
      "   10        41.0299   0.00219327        parabolic\n",
      "   11        36.9955   0.00169965        parabolic\n",
      "   12        38.8626    0.0018185        parabolic\n",
      "   13        37.7932   0.00189518        parabolic\n",
      "   14        38.1959   0.00198861        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.2829876  1.1238632  1.24098125 1.18847444 1.26556294]\n",
      "Trial error:  1.240981249585707\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0014752007584456822  :: absSize readout:  0.024149009133834946\n",
      "NRMSE W:  0.007989500868955076  :: absSize W:  0.17243256153551173\n",
      "NRMSEs: best:  0.007989500868955076 avg:  0.007989500868955076\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.330237        initial\n",
      "    2        618.034     0.355545        golden\n",
      "    3        236.068      0.23971        golden\n",
      "    4        145.898  4.82755e-06        golden\n",
      "    5        90.1699  8.71274e-06        golden\n",
      "    6        118.036  6.18453e-06        parabolic\n",
      "    7        164.259   0.00168258        parabolic\n",
      "    8        135.256  5.18094e-06        golden\n",
      "    9        159.025   0.00167443        parabolic\n",
      "   10        150.912  4.68053e-06        golden\n",
      "   11        154.011  4.76772e-06        golden\n",
      "   12        150.475  4.80765e-06        parabolic\n",
      "   13        152.306  4.74017e-06        parabolic\n",
      "   14        151.444  4.78529e-06        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.21777688 1.41174925 0.24221098 1.40902044 0.29439833]\n",
      "Trial error:  0.29439833182248937\n",
      "LR:  0.49999999999999994\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0006052196455604346  :: absSize readout:  0.012139352640524403\n",
      "NRMSE W:  0.007961025568204506  :: absSize W:  0.19086019670547455\n",
      "NRMSEs: best:  0.007961025568204506 avg:  0.007961025568204506\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.249767        initial\n",
      "    2        618.034     0.249258        golden\n",
      "    3        763.932     0.245885        golden\n",
      "    4        854.102     0.243542        golden\n",
      "    5         909.83     0.242073        golden\n",
      "    6        944.272      0.24116        golden\n",
      "    7        965.558     0.240602        golden\n",
      "    8        978.714     0.240257        golden\n",
      "    9        986.844     0.240041        golden\n",
      "   10        991.869     0.239909        golden\n",
      "   11        994.975     0.239831        golden\n",
      "   12        996.894     0.239778        golden\n",
      "   13        998.081      0.23975        golden\n",
      "   14        998.814     0.239728        golden\n",
      "   15        999.267     0.239716        golden\n",
      "   16          999.6      0.23971        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43789605 1.43789605 1.43789605 1.43789605 1.43789605]\n",
      "Trial error:  1.4378960506567309\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00041822703072833363  :: absSize readout:  0.010258816996369536\n",
      "NRMSE W:  0.007355060699503558  :: absSize W:  0.1687890438525011\n",
      "NRMSEs: best:  0.007355060699503558 avg:  0.007355060699503558\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.406042        initial\n",
      "    2        618.034     0.410379        golden\n",
      "    3        236.068     0.390451        golden\n",
      "    4        145.898     0.341937        golden\n",
      "    5        90.1699   0.00531231        golden\n",
      "    6        55.7281   0.00473487        golden\n",
      "    7        72.8235   0.00519425        parabolic\n",
      "    8        34.4419   0.00042507        golden\n",
      "    9        21.2862  0.000429956        golden\n",
      "   10        27.8956  0.000430599        parabolic\n",
      "   11        42.5725  0.000452435        golden\n",
      "   12        37.5475  0.000424712        golden\n",
      "   13        39.6479  0.000450994        parabolic\n",
      "   14        36.3612  0.000438466        golden\n",
      "   15        38.3498   0.00044887        golden\n",
      "   16        37.0944  0.000420476        golden\n",
      "   17         36.761  0.000422263        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.20389199 1.23249207 1.21928183 1.21601166 1.21105335]\n",
      "Trial error:  1.2160116612861098\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0005144469958591936  :: absSize readout:  0.012042153766756575\n",
      "NRMSE W:  0.008559909430821046  :: absSize W:  0.18416683152343957\n",
      "NRMSEs: best:  0.008559909430821046 avg:  0.008559909430821046\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.330685        initial\n",
      "    2        618.034     0.338578        golden\n",
      "    3        236.068     0.305278        golden\n",
      "    4        145.898      0.22293        golden\n",
      "    5        90.1699   0.00201243        golden\n",
      "    6        55.7281    0.0018279        golden\n",
      "    7         72.888   0.00183486        parabolic\n",
      "    8        63.6007   0.00178406        parabolic\n",
      "    9         63.993   0.00118571        parabolic\n",
      "   10        67.3906   0.00183944        golden\n",
      "   11        65.2908   0.00189379        golden\n",
      "   12        64.4887    0.0022146        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.75016355 0.99971293 0.8636244  0.91376538 1.02789513]\n",
      "Trial error:  0.9137653799881824\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0004181031199987773  :: absSize readout:  0.01071750559737617\n",
      "NRMSE W:  0.006666474631660087  :: absSize W:  0.1400479308924311\n",
      "NRMSEs: best:  0.006666474631660087 avg:  0.006666474631660087\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.296529        initial\n",
      "    2        618.034     0.312326        golden\n",
      "    3        236.068     0.239971        golden\n",
      "    4        145.898    0.0170132        golden\n",
      "    5        90.1699    0.0122702        golden\n",
      "    6        115.434    0.0143475        parabolic\n",
      "    7        55.7281   0.00864136        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    8        34.4419   0.00465381        golden\n",
      "    9        21.2862   0.00421839        golden\n",
      "   10        24.1685   0.00439197        parabolic\n",
      "   11        13.1556    0.0035429        golden\n",
      "   12        8.13062   0.00399144        golden\n",
      "   13          14.05     0.003501        parabolic\n",
      "   14         16.814   0.00374519        golden\n",
      "   15        14.3833   0.00348747        parabolic\n",
      "   16        15.3118   0.00350685        golden\n",
      "   17        14.7167   0.00347755        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43628528 1.4363237  1.43671301 1.43672937 1.4366125 ]\n",
      "Trial error:  1.4366124987952342\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0006408750677976749  :: absSize readout:  0.014154169313828126\n",
      "NRMSE W:  0.0071996447577767474  :: absSize W:  0.1450289843449084\n",
      "NRMSEs: best:  0.0071996447577767474 avg:  0.0071996447577767474\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.208545        initial\n",
      "    2        618.034      0.23473        golden\n",
      "    3        236.068   0.00418655        golden\n",
      "    4        145.898   1.4379e-05        golden\n",
      "    5        186.951  2.72016e-05        parabolic\n",
      "    6        166.257  2.63316e-05        parabolic\n",
      "    7        90.1699  1.48551e-05        golden\n",
      "    8         118.58  1.03715e-05        parabolic\n",
      "    9        118.913  1.01969e-05        parabolic\n",
      "   10        129.287   8.8786e-06        parabolic\n",
      "   11        135.632  8.48305e-06        golden\n",
      "   12        140.512  2.03225e-05        parabolic\n",
      "   13        133.209  1.50868e-05        golden\n",
      "   14        137.496  2.53006e-05        golden\n",
      "   15        134.706  9.91221e-06        golden\n",
      "   16        136.344  3.53795e-05        golden\n",
      "   17        135.279  8.77042e-06        golden\n",
      "   18        135.965  1.99156e-05        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.25677411 0.26993837 0.30701272 0.21475249 0.20052013]\n",
      "Trial error:  0.2567741106999083\n",
      "LR:  0.57\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0003884916605218138  :: absSize readout:  0.008176844185478852\n",
      "NRMSE W:  0.006462679394841733  :: absSize W:  0.13498819093055756\n",
      "NRMSEs: best:  0.006462679394841733 avg:  0.006462679394841733\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966      0.21249        initial\n",
      "    2        618.034     0.208115        golden\n",
      "    3        763.932     0.203547        golden\n",
      "    4        854.102     0.200654        golden\n",
      "    5         909.83      0.19888        golden\n",
      "    6        944.272     0.197814        golden\n",
      "    7        965.558     0.197155        golden\n",
      "    8        978.714     0.196742        golden\n",
      "    9        986.844     0.196493        golden\n",
      "   10        991.869      0.19634        golden\n",
      "   11        994.975     0.196255        golden\n",
      "   12        996.894     0.196197        golden\n",
      "   13        998.081     0.196151        golden\n",
      "   14        998.814     0.196129        golden\n",
      "   15        999.267     0.196115        golden\n",
      "   16          999.6     0.196105        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43756975 1.43756975 1.43756975 1.43756975 1.43756975]\n",
      "Trial error:  1.4375697453066028\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0002511654044817266  :: absSize readout:  0.00728851626370815\n",
      "NRMSE W:  0.006889903984936194  :: absSize W:  0.12197850746412855\n",
      "NRMSEs: best:  0.006889903984936194 avg:  0.006889903984936194\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.232293        initial\n",
      "    2        618.034     0.238098        golden\n",
      "    3        236.068     0.208301        golden\n",
      "    4        145.898     0.108313        golden\n",
      "    5        90.1699    0.0019852        golden\n",
      "    6        55.7281   0.00109675        golden\n",
      "    7        72.3311   0.00147396        parabolic\n",
      "    8        34.4419  6.87692e-05        golden\n",
      "    9        21.2862  0.000120282        golden\n",
      "   10        29.1556  6.87813e-05        parabolic\n",
      "   11         31.801  6.71917e-05        parabolic\n",
      "   12        32.1344  6.71857e-05        parabolic\n",
      "   13        32.4677   6.7346e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.09305574 0.26568109 0.28431876 0.19470104 0.20503485]\n",
      "Trial error:  0.20503485210244374\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00027711392020354005  :: absSize readout:  0.007167551050471943\n",
      "NRMSE W:  0.007449678917452617  :: absSize W:  0.1300482602009436\n",
      "NRMSEs: best:  0.007449678917452617 avg:  0.007449678917452617\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0107939        initial\n",
      "    2        618.034    0.0463935        golden\n",
      "    3        236.068   0.00947284        golden\n",
      "    4        296.817    0.0103117        parabolic\n",
      "    5        145.898   0.00587768        golden\n",
      "    6        90.1699   0.00521083        golden\n",
      "    7        86.7522   0.00550172        parabolic\n",
      "    8        114.389     0.002425        parabolic\n",
      "    9        126.424   0.00245164        golden\n",
      "   10        120.064   0.00243623        parabolic\n",
      "   11        105.138   0.00242609        golden\n",
      "   12        110.183   0.00575962        parabolic\n",
      "   13        114.722   0.00604326        parabolic\n",
      "   14        112.782   0.00242416        golden\n",
      "   15        111.789   0.00596253        golden\n",
      "   16        113.396   0.00538661        golden\n",
      "   17        112.403   0.00242405        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42113992 1.39772089 1.39186737 1.3963967  1.41471057]\n",
      "Trial error:  1.3977208940578898\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00023190837489359375  :: absSize readout:  0.007777035996494751\n",
      "NRMSE W:  0.007213203107816131  :: absSize W:  0.17203639592646217\n",
      "NRMSEs: best:  0.007213203107816131 avg:  0.007213203107816131\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.376189        initial\n",
      "    2        618.034     0.379878        golden\n",
      "    3        236.068     0.359137        golden\n",
      "    4        145.898     0.307064        golden\n",
      "    5        90.1699    0.0277764        golden\n",
      "    6        55.7281    0.0191074        golden\n",
      "    7         70.565    0.0232139        parabolic\n",
      "    8        34.4419    0.0158224        golden\n",
      "    9         22.324    0.0133233        parabolic\n",
      "   10         13.797  0.000344441        golden\n",
      "   11        8.52699  0.000412019        golden\n",
      "   12        11.2196   0.00037427        parabolic\n",
      "   13         17.054    0.0113743        golden\n",
      "   14         15.041  0.000331878        golden\n",
      "   15        15.8099    0.0107193        golden\n",
      "   16        14.5658   0.00992532        golden\n",
      "   17        15.3744    0.0104602        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.4193855  1.43633301 1.4193855  1.4193855  1.4193855 ]\n",
      "Trial error:  1.4193854991328474\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0002855029463300036  :: absSize readout:  0.007596231415990639\n",
      "NRMSE W:  0.006953824369306978  :: absSize W:  0.1446405662827648\n",
      "NRMSEs: best:  0.006953824369306978 avg:  0.006953824369306978\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.301219        initial\n",
      "    2        618.034     0.314428        golden\n",
      "    3        236.068     0.266275        golden\n",
      "    4        145.898    0.0138336        golden\n",
      "    5        90.1699   0.00544883        golden\n",
      "    6        113.891   0.00657674        parabolic\n",
      "    7        94.6359   0.00512135        parabolic\n",
      "    8        98.2433   0.00609978        parabolic\n",
      "    9         93.262   0.00628462        parabolic\n",
      "   10        96.0138   0.00606612        golden\n",
      "   11        95.1622   0.00563866        golden\n",
      "   12        94.1111   0.00621896        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.24095402 1.25981193 1.2730751  1.21137465 1.24486035]\n",
      "Trial error:  1.2448603475511106\n",
      "LR:  0.6399999999999999\n",
      "Trial:  0\n",
      "0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001543207416903176  :: absSize readout:  0.005707728724434046\n",
      "NRMSE W:  0.005878120587693599  :: absSize W:  0.09872063229304649\n",
      "NRMSEs: best:  0.005878120587693599 avg:  0.005878120587693599\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.260427        initial\n",
      "    2        618.034     0.272069        golden\n",
      "    3        236.068     0.235117        golden\n",
      "    4        145.898     0.180193        golden\n",
      "    5        90.1699    0.0268116        golden\n",
      "    6        55.7281    0.0211859        golden\n",
      "    7        70.1046   0.00203665        parabolic\n",
      "    8        71.8531    0.0249169        parabolic\n",
      "    9        63.6612    0.0233657        parabolic\n",
      "   10        67.6434    0.0241923        golden\n",
      "   11        69.1645    0.0244704        golden\n",
      "   12        70.7725     0.024744        golden\n",
      "   13        69.7455    0.0245716        golden\n",
      "   14        70.4379    0.0246887        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.40932419 1.40932419 1.40932419 1.40932419 1.40932419]\n",
      "Trial error:  1.4093241934501404\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00017968623235431103  :: absSize readout:  0.00561214005596518\n",
      "NRMSE W:  0.005748831755880474  :: absSize W:  0.10288144978697485\n",
      "NRMSEs: best:  0.005748831755880474 avg:  0.005748831755880474\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.256755        initial\n",
      "    2        618.034     0.273546        golden\n",
      "    3        236.068     0.221301        golden\n",
      "    4        145.898    0.0318492        golden\n",
      "    5        90.1699    0.0163204        golden\n",
      "    6         106.88    0.0226273        parabolic\n",
      "    7        55.7281  0.000819775        golden\n",
      "    8        34.4419   0.00155202        golden\n",
      "    9        47.0635   0.00110419        parabolic\n",
      "   10        68.8837  3.48483e-05        golden\n",
      "   11        77.0143   0.00624662        golden\n",
      "   12        63.8587  3.99222e-05        golden\n",
      "   13        66.4412  3.71141e-05        parabolic\n",
      "   14        71.9893  2.98962e-05        golden\n",
      "   15        73.9087  3.36554e-05        golden\n",
      "   16        71.5641  3.05738e-05        parabolic\n",
      "   17        72.3227  2.90855e-05        parabolic\n",
      "   18        72.9285  3.72598e-05        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.29179784 0.26103429 0.16273208 0.27069081 0.19265851]\n",
      "Trial error:  0.26103429279669915\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00021957898256950206  :: absSize readout:  0.005716897669875502\n",
      "NRMSE W:  0.006281955160890241  :: absSize W:  0.10879310218866255\n",
      "NRMSEs: best:  0.006281955160890241 avg:  0.006281955160890241\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.263741        initial\n",
      "    2        618.034     0.265916        golden\n",
      "    3        236.068     0.250501        golden\n",
      "    4        145.898     0.204605        golden\n",
      "    5        90.1699   0.00373548        golden\n",
      "    6        55.7281  0.000874467        golden\n",
      "    7        71.8855   0.00322134        parabolic\n",
      "    8        34.4419  0.000108226        golden\n",
      "    9        38.9165  0.000134893        parabolic\n",
      "   10        35.0114  0.000103714        parabolic\n",
      "   11        35.8409  0.000102511        parabolic\n",
      "   12        36.1743  0.000104223        parabolic\n",
      "   13        35.5076  0.000102953        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.23876436 0.07991052 0.28537781 0.128539   0.19850218]\n",
      "Trial error:  0.19850218341734313\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00017274854820544674  :: absSize readout:  0.00547959098382608\n",
      "NRMSE W:  0.006384097843064716  :: absSize W:  0.10539595635223142\n",
      "NRMSEs: best:  0.006384097843064716 avg:  0.006384097843064716\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.209588        initial\n",
      "    2        618.034     0.233253        golden\n",
      "    3        236.068    0.0382249        golden\n",
      "    4        145.898   0.00616507        golden\n",
      "    5        139.741   0.00636994        parabolic\n",
      "    6        146.941    0.0372741        parabolic\n",
      "    7        142.823    0.0368668        parabolic\n",
      "    8        144.724    0.0062002        golden\n",
      "    9        145.449     0.037135        golden\n",
      "   10        146.296   0.00615607        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43876947 1.43876483 1.43876828 1.43860335 1.43795134]\n",
      "Trial error:  1.438764826657654\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00015850879503543006  :: absSize readout:  0.0051401200520415306\n",
      "NRMSE W:  0.006328376685604923  :: absSize W:  0.12595934408995496\n",
      "NRMSEs: best:  0.006328376685604923 avg:  0.006328376685604923\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.288428        initial\n",
      "    2        618.034     0.297187        golden\n",
      "    3        236.068      0.26715        golden\n",
      "    4        145.898     0.193989        golden\n",
      "    5        90.1699    0.0237804        golden\n",
      "    6        55.7281    0.0187535        golden\n",
      "    7        70.6864    0.0222325        parabolic\n",
      "    8        34.4419  0.000570538        golden\n",
      "    9        21.2862  0.000561827        golden\n",
      "   10        27.8507  0.000718118        parabolic\n",
      "   11        13.1556  0.000738303        golden\n",
      "   12        18.1806  0.000642533        golden\n",
      "   13        23.7936  0.000559364        golden\n",
      "   14        25.7205  0.000718587        parabolic\n",
      "   15        22.8359  0.000556857        golden\n",
      "   16        22.5026  0.000563677        parabolic\n",
      "   17        23.1692  0.000630938        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.64649891 0.56721372 0.62823769 0.64414329 0.64573871]\n",
      "Trial error:  0.6441432862830563\n",
      "LR:  0.71\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00033433379067414397  :: absSize readout:  0.005753056957547778\n",
      "NRMSE W:  0.00533431191245699  :: absSize W:  0.09102139548570343\n",
      "NRMSEs: best:  0.00533431191245699 avg:  0.00533431191245699\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0252737        initial\n",
      "    2        618.034    0.0219964        golden\n",
      "    3        763.932    0.0208506        golden\n",
      "    4        939.746    0.0198439        parabolic\n",
      "    5        872.591    0.0201947        golden\n",
      "    6        914.095    0.0199738        golden\n",
      "    7        962.761    0.0197312        golden\n",
      "    8        976.985    0.0196632        golden\n",
      "    9        985.776    0.0196219        golden\n",
      "   10        991.209    0.0195965        golden\n",
      "   11        994.567     0.019581        golden\n",
      "   12        996.642    0.0195714        golden\n",
      "   13        997.925    0.0195655        golden\n",
      "   14        998.717    0.0195618        golden\n",
      "   15        999.207    0.0195596        golden\n",
      "   16        999.541     0.019558        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.41004264 1.41004264 1.41004264 1.41004264 1.41004264]\n",
      "Trial error:  1.410042638194659\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001238523125478699  :: absSize readout:  0.004620298726899452\n",
      "NRMSE W:  0.00562119683724259  :: absSize W:  0.10621440788233297\n",
      "NRMSEs: best:  0.00562119683724259 avg:  0.00562119683724259\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.294267        initial\n",
      "    2        618.034      0.29654        golden\n",
      "    3        236.068     0.284247        golden\n",
      "    4        145.898   0.00437197        golden\n",
      "    5        90.1699    0.0043114        golden\n",
      "    6        118.008   0.00236861        parabolic\n",
      "    7        117.675   0.00253359        parabolic\n",
      "    8        128.661   0.00224316        golden\n",
      "    9        123.469   0.00230642        parabolic\n",
      "   10        135.245   0.00215059        golden\n",
      "   11        139.314   0.00402088        golden\n",
      "   12         132.73   0.00218906        golden\n",
      "   13        136.799   0.00258841        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   14        134.285   0.00216618        golden\n",
      "   15        135.839   0.00213977        golden\n",
      "   16        136.206   0.00213206        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42491354 1.42491711 1.42491354 1.42491355 1.42491352]\n",
      "Trial error:  1.4249135381981144\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00017532481263965054  :: absSize readout:  0.005054131143415497\n",
      "NRMSE W:  0.005108501252253612  :: absSize W:  0.07324304630726924\n",
      "NRMSEs: best:  0.005108501252253612 avg:  0.005108501252253612\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.168299        initial\n",
      "    2        618.034      0.17056        golden\n",
      "    3        236.068    0.0320061        golden\n",
      "    4        145.898    0.0112639        golden\n",
      "    5        152.422   0.00888582        parabolic\n",
      "    6        174.801    0.0142746        parabolic\n",
      "    7         160.97    0.0162732        golden\n",
      "    8        157.863    0.0144417        parabolic\n",
      "    9         149.93   0.00927021        golden\n",
      "   10        152.756    0.0171717        parabolic\n",
      "   11        152.089    0.0110541        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.09847076 1.08180404 0.90192028 0.98329815 1.02263816]\n",
      "Trial error:  1.0226381603141113\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00010975532986288121  :: absSize readout:  0.0048792832280807885\n",
      "NRMSE W:  0.005203747331021586  :: absSize W:  0.06975454693137582\n",
      "NRMSEs: best:  0.005203747331021586 avg:  0.005203747331021586\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00892806        initial\n",
      "    2        618.034     0.195553        golden\n",
      "    3        236.068   0.00540297        golden\n",
      "    4        302.996    0.0119014        parabolic\n",
      "    5        145.898   0.00171053        golden\n",
      "    6        90.1699  0.000525914        golden\n",
      "    7          39.29  0.000198351        parabolic\n",
      "    8        41.5728  0.000198086        parabolic\n",
      "    9        40.8611  0.000199201        parabolic\n",
      "   10        60.1352  0.000295095        golden\n",
      "   11         48.663  0.000227592        golden\n",
      "   12         44.281  0.000203099        golden\n",
      "   13        42.6072    0.0001989        golden\n",
      "   14        41.9679  0.000204971        golden\n",
      "   15        41.2395  0.000198804        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.76749147 0.74143326 0.74242397 0.75114514 0.64634428]\n",
      "Trial error:  0.742423969946129\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0004889176749909924  :: absSize readout:  0.006272389035113995\n",
      "NRMSE W:  0.005331983346376519  :: absSize W:  0.11648641254594302\n",
      "NRMSEs: best:  0.005331983346376519 avg:  0.005331983346376519\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0147589        initial\n",
      "    2        618.034    0.0133415        golden\n",
      "    3        763.932    0.0129888        golden\n",
      "    4        819.704    0.0124371        parabolic\n",
      "    5        888.571    0.0114651        golden\n",
      "    6        931.133    0.0120236        golden\n",
      "    7         883.01    0.0122015        parabolic\n",
      "    8        907.683    0.0116573        parabolic\n",
      "    9        895.871    0.0119755        golden\n",
      "   10        891.359    0.0121679        golden\n",
      "   11        886.447    0.0123543        golden\n",
      "   12        889.636    0.0119042        golden\n",
      "   13         887.76    0.0123831        golden\n",
      "   14        888.978    0.0119655        golden\n",
      "   15        888.238    0.0124069        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.45281622 1.44546974 1.47011303 1.45547044 1.46636451]\n",
      "Trial error:  1.4554704428439327\n",
      "LR:  0.7799999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00023655496882277218  :: absSize readout:  0.004681277793111594\n",
      "NRMSE W:  0.005657210632691725  :: absSize W:  0.10400793438015\n",
      "NRMSEs: best:  0.005657210632691725 avg:  0.005657210632691725\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0783697        initial\n",
      "    2        618.034    0.0888907        golden\n",
      "    3        236.068   0.00525918        golden\n",
      "    4        145.898   0.00405682        golden\n",
      "    5        187.756   0.00419674        parabolic\n",
      "    6        158.746   0.00409892        parabolic\n",
      "    7        90.1699   0.00344064        golden\n",
      "    8        55.7281   0.00284604        golden\n",
      "    9        34.4419   0.00615406        golden\n",
      "   10        68.8837   0.00584977        golden\n",
      "   11        47.5975    0.0061739        golden\n",
      "   12        60.7531   0.00560548        golden\n",
      "   13        52.6225   0.00432066        golden\n",
      "   14        57.6475   0.00422857        golden\n",
      "   15        54.5418   0.00387296        golden\n",
      "   16        56.4612   0.00269645        golden\n",
      "   17        56.9143   0.00485793        golden\n",
      "   18        56.1279   0.00441711        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.09991178 1.27810681 1.19030731 1.01500604 1.24013647]\n",
      "Trial error:  1.1903073133217226\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00010649515734406524  :: absSize readout:  0.004122270354599346\n",
      "NRMSE W:  0.004852458635595412  :: absSize W:  0.05925103834437372\n",
      "NRMSEs: best:  0.004852458635595412 avg:  0.004852458635595412\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00718344        initial\n",
      "    2        618.034   0.00788691        golden\n",
      "    3        236.068   0.00363187        golden\n",
      "    4        145.898   0.00153386        golden\n",
      "    5        90.1699  6.11808e-05        golden\n",
      "    6        55.7281   0.00769086        golden\n",
      "    7        111.456  6.25505e-05        golden\n",
      "    8        100.771  8.00962e-05        parabolic\n",
      "    9        77.0143  7.88022e-05        golden\n",
      "   10        100.024  4.56689e-05        parabolic\n",
      "   11        99.6904  0.000133427        parabolic\n",
      "   12        100.357  4.19552e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.30728018 0.15792481 0.1822916  0.2263032  0.16596001]\n",
      "Trial error:  0.18229159604383247\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00010081841973939889  :: absSize readout:  0.004111224227690851\n",
      "NRMSE W:  0.00508335042758998  :: absSize W:  0.07684475974940064\n",
      "NRMSEs: best:  0.00508335042758998 avg:  0.00508335042758998\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.151629        initial\n",
      "    2        618.034      0.16839        golden\n",
      "    3        236.068     0.123591        golden\n",
      "    4        145.898    0.0596554        golden\n",
      "    5        90.1699   0.00494813        golden\n",
      "    6        55.7281   0.00546389        golden\n",
      "    7        73.6264   0.00196329        parabolic\n",
      "    8        73.9598    0.0069317        parabolic\n",
      "    9        73.2931  0.000781702        parabolic\n",
      "   10        66.5839   0.00558146        golden\n",
      "   11        70.7304   0.00504489        golden\n",
      "   12        72.3142   0.00094104        golden\n",
      "   13        72.9598   0.00634076        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.94197479 0.15369574 0.9858265  0.17885094 0.79618862]\n",
      "Trial error:  0.7961886231361182\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  8.463348657837208e-05  :: absSize readout:  0.0038620828992753497\n",
      "NRMSE W:  0.005479189296979251  :: absSize W:  0.07531576027119322\n",
      "NRMSEs: best:  0.005479189296979251 avg:  0.005479189296979251\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.228077        initial\n",
      "    2        618.034     0.240842        golden\n",
      "    3        236.068     0.198559        golden\n",
      "    4        145.898   0.00536883        golden\n",
      "    5        90.1699   0.00496147        golden\n",
      "    6        117.784    0.0053376        parabolic\n",
      "    7        55.7281   0.00353763        golden\n",
      "    8        34.4419    0.0018159        golden\n",
      "    9        21.2862  0.000341472        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   10        13.1556   0.00106653        golden\n",
      "   11         21.937  0.000315406        parabolic\n",
      "   12         25.192    0.0002105        parabolic\n",
      "   13        28.7251   0.00120268        golden\n",
      "   14        26.5415  0.000199884        golden\n",
      "   15        26.8749   0.00021015        parabolic\n",
      "   16         26.038  0.000209181        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.32627843 0.33138193 0.36919661 0.32321852 0.34364024]\n",
      "Trial error:  0.33138192980333186\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00017497622716709836  :: absSize readout:  0.004296611348986786\n",
      "NRMSE W:  0.0055909885681050125  :: absSize W:  0.0827066023602396\n",
      "NRMSEs: best:  0.0055909885681050125 avg:  0.0055909885681050125\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00606595        initial\n",
      "    2        618.034   0.00506349        golden\n",
      "    3        763.932   0.00351036        golden\n",
      "    4        854.102   0.00417958        golden\n",
      "    5         760.53   0.00631591        parabolic\n",
      "    6        798.374   0.00171862        golden\n",
      "    7        805.538   0.00399763        parabolic\n",
      "    8        784.077   0.00362345        parabolic\n",
      "    9        792.913     0.003657        golden\n",
      "   10         801.11   0.00274219        golden\n",
      "   11        796.288   0.00374253        golden\n",
      "   12        799.419   0.00362217        golden\n",
      "   13        797.577   0.00426176        golden\n",
      "   14        798.773   0.00243602        golden\n",
      "   15        798.041   0.00403012        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.33520454 1.40321862 1.38133136 1.38816207 1.35277472]\n",
      "Trial error:  1.3813313579780462\n",
      "LR:  0.8499999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00016001648202985098  :: absSize readout:  0.003997905875808507\n",
      "NRMSE W:  0.004934236853567073  :: absSize W:  0.04253134315597852\n",
      "NRMSEs: best:  0.004934236853567073 avg:  0.004934236853567073\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0408481        initial\n",
      "    2        618.034    0.0362751        golden\n",
      "    3        763.932     0.034213        golden\n",
      "    4        854.102    0.0332446        golden\n",
      "    5         909.83    0.0327525        golden\n",
      "    6        944.272    0.0324798        golden\n",
      "    7        965.558    0.0323186        golden\n",
      "    8        978.714    0.0322313        golden\n",
      "    9        986.844     0.032176        golden\n",
      "   10        991.869    0.0197201        golden\n",
      "   11        994.975    0.0321219        golden\n",
      "   12        990.914    0.0321439        parabolic\n",
      "   13        992.945    0.0321304        parabolic\n",
      "   14         992.28    0.0197397        golden\n",
      "   15        991.536    0.0321398        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.40857766 1.40857766 1.40857766 1.40857766 1.40857766]\n",
      "Trial error:  1.4085776587896377\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0005722607262022412  :: absSize readout:  0.005931718875786061\n",
      "NRMSE W:  0.007444367502095617  :: absSize W:  0.12564824467700345\n",
      "NRMSEs: best:  0.007444367502095617 avg:  0.007444367502095617\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0242088        initial\n",
      "    2        618.034    0.0552325        golden\n",
      "    3        236.068    0.0192164        golden\n",
      "    4        241.783    0.0189254        parabolic\n",
      "    5        280.846    0.0172634        parabolic\n",
      "    6         319.47    0.0193612        golden\n",
      "    7        265.925    0.0178005        golden\n",
      "    8        295.599    0.0191299        golden\n",
      "    9        275.146    0.0174501        golden\n",
      "   10        286.481    0.0170856        golden\n",
      "   11        289.964    0.0191974        golden\n",
      "   12        284.328    0.0171643        golden\n",
      "   13        287.811    0.0170574        golden\n",
      "   14        288.633    0.0170417        golden\n",
      "   15        289.141    0.0170335        golden\n",
      "   16        289.475    0.0170272        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.40859525 1.40761485 1.40859027 1.40648028 1.40859321]\n",
      "Trial error:  1.4085902746881866\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001263779386342949  :: absSize readout:  0.003828044291696322\n",
      "NRMSE W:  0.005578938910379606  :: absSize W:  0.08485124588401473\n",
      "NRMSEs: best:  0.005578938910379606 avg:  0.005578938910379606\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.137479        initial\n",
      "    2        618.034     0.149451        golden\n",
      "    3        236.068    0.0588085        golden\n",
      "    4        145.898    0.0461769        golden\n",
      "    5        149.556    0.0464526        parabolic\n",
      "    6        90.1699     0.035407        golden\n",
      "    7        55.7281   0.00623942        golden\n",
      "    8        34.4419   0.00552526        golden\n",
      "    9        43.9355   0.00437118        parabolic\n",
      "   10        43.6022   0.00532699        parabolic\n",
      "   11        48.4399   0.00550755        golden\n",
      "   12        45.9921   0.00588353        parabolic\n",
      "   13        44.7211   0.00619299        golden\n",
      "   14        44.2689   0.00521626        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.93536436 0.6743295  0.692973   0.81844575 0.77795192]\n",
      "Trial error:  0.777951920029861\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00012535234445727557  :: absSize readout:  0.003867257501428658\n",
      "NRMSE W:  0.005058904503003586  :: absSize W:  0.07246911899845797\n",
      "NRMSEs: best:  0.005058904503003586 avg:  0.005058904503003586\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0903896        initial\n",
      "    2        618.034     0.076251        golden\n",
      "    3        763.932    0.0758845        golden\n",
      "    4        699.345    0.0679093        parabolic\n",
      "    5        691.794    0.0642518        parabolic\n",
      "    6         663.62    0.0726406        golden\n",
      "    7        684.507    0.0725674        parabolic\n",
      "    8        693.359    0.0720624        parabolic\n",
      "    9        689.011    0.0722784        golden\n",
      "   10        690.731     0.070987        golden\n",
      "   11        692.392    0.0664036        golden\n",
      "   12        691.388    0.0651528        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.46062657 1.45905402 1.47458387 1.46564253 1.46797039]\n",
      "Trial error:  1.4656425264429147\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001259372148911831  :: absSize readout:  0.0040173695957584025\n",
      "NRMSE W:  0.0052866408564290835  :: absSize W:  0.08780722452615894\n",
      "NRMSEs: best:  0.0052866408564290835 avg:  0.0052866408564290835\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.221211        initial\n",
      "    2        618.034     0.224844        golden\n",
      "    3        236.068     0.198408        golden\n",
      "    4        145.898    0.0154028        golden\n",
      "    5        90.1699    0.0102481        golden\n",
      "    6        114.551    0.0112272        parabolic\n",
      "    7        90.5033    0.0102745        parabolic\n",
      "    8        55.7281   0.00617193        golden\n",
      "    9        34.4419   0.00367491        golden\n",
      "   10        21.2862   0.00305582        golden\n",
      "   11         16.328   0.00308147        parabolic\n",
      "   12         19.704   0.00338173        parabolic\n",
      "   13        26.3112   0.00295986        golden\n",
      "   14        29.4169    0.0035186        golden\n",
      "   15        24.3919   0.00342582        golden\n",
      "   16        27.4975   0.00339068        golden\n",
      "   17        25.5781   0.00364466        golden\n",
      "   18        26.7643   0.00393912        golden\n",
      "   19        25.9779   0.00407989        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.04437897 1.11632266 0.98538127 0.91859187 1.05213512]\n",
      "Trial error:  1.0443789749700008\n",
      "LR:  0.9199999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0006149940668885177  :: absSize readout:  0.005364568953701745\n",
      "NRMSE W:  0.004440869530396523  :: absSize W:  0.09130947381634971\n",
      "NRMSEs: best:  0.004440869530396523 avg:  0.004440869530396523\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0424788        initial\n",
      "    2        618.034    0.0318553        golden\n",
      "    3        763.932    0.0271913        golden\n",
      "    4        854.102    0.0248062        golden\n",
      "    5         909.83    0.0234837        golden\n",
      "    6        944.272    0.0227172        golden\n",
      "    7        965.558    0.0222615        golden\n",
      "    8        978.714    0.0219865        golden\n",
      "    9        986.844     0.021819        golden\n",
      "   10        991.869    0.0217173        golden\n",
      "   11        994.975    0.0216542        golden\n",
      "   12        996.894    0.0216144        golden\n",
      "   13        998.081    0.0215914        golden\n",
      "   14        998.814    0.0215766        golden\n",
      "   15        999.267    0.0215666        golden\n",
      "   16          999.6    0.0215608        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43548114 1.43548114 1.43548114 1.43548114 1.43548114]\n",
      "Trial error:  1.435481140636053\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00012518550106525817  :: absSize readout:  0.0038727080615101927\n",
      "NRMSE W:  0.0047530031478256966  :: absSize W:  0.05309684659249464\n",
      "NRMSEs: best:  0.0047530031478256966 avg:  0.0047530031478256966\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0125956        initial\n",
      "    2        618.034    0.0751929        golden\n",
      "    3        236.068   0.00526048        golden\n",
      "    4        264.335   0.00512034        parabolic\n",
      "    5         255.48   0.00828632        parabolic\n",
      "    6        309.266   0.00712744        golden\n",
      "    7        281.497   0.00698983        golden\n",
      "    8         270.89   0.00544184        golden\n",
      "    9        260.953   0.00701834        golden\n",
      "   10        266.838   0.00761504        golden\n",
      "   11        263.043   0.00848387        golden\n",
      "   12        265.291    0.0063143        golden\n",
      "   13        263.841   0.00651987        golden\n",
      "   14          264.7   0.00638012        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.6097151  0.89591876 0.73565713 0.97918811 0.78238367]\n",
      "Trial error:  0.7823836739086093\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00013995707623195623  :: absSize readout:  0.0035520288660872944\n",
      "NRMSE W:  0.005181315043939378  :: absSize W:  0.05915562121877104\n",
      "NRMSEs: best:  0.005181315043939378 avg:  0.005181315043939378\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00836035        initial\n",
      "    2        618.034   0.00966376        golden\n",
      "    3        236.068     0.007368        golden\n",
      "    4        145.898   0.00333581        golden\n",
      "    5        90.1699   0.00382025        golden\n",
      "    6        129.907   0.00373298        parabolic\n",
      "    7         180.34   0.00401502        golden\n",
      "    8        159.054   0.00297893        golden\n",
      "    9        167.184   0.00335533        golden\n",
      "   10        156.408   0.00218539        parabolic\n",
      "   11        152.912   0.00285196        parabolic\n",
      "   12        155.853   0.00187683        parabolic\n",
      "   13        155.035   0.00300025        parabolic\n",
      "   14         155.52   0.00285446        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.77480743 0.84680683 0.71542011 0.14841782 0.69220442]\n",
      "Trial error:  0.7154201118588682\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  8.284109550414982e-05  :: absSize readout:  0.0036080514957564534\n",
      "NRMSE W:  0.0047891881305249925  :: absSize W:  0.04495984739400139\n",
      "NRMSEs: best:  0.0047891881305249925 avg:  0.0047891881305249925\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0745401        initial\n",
      "    2        618.034    0.0695191        golden\n",
      "    3        763.932    0.0668958        golden\n",
      "    4        854.102    0.0658804        golden\n",
      "    5         909.83    0.0653116        golden\n",
      "    6        944.272    0.0649885        golden\n",
      "    7        965.558    0.0647826        golden\n",
      "    8        978.714    0.0646631        golden\n",
      "    9        986.844    0.0645902        golden\n",
      "   10        991.869    0.0645455        golden\n",
      "   11        994.975    0.0645269        golden\n",
      "   12        996.894    0.0645011        golden\n",
      "   13        998.081    0.0644906        golden\n",
      "   14        998.814    0.0644842        golden\n",
      "   15        999.267    0.0644802        golden\n",
      "   16          999.6    0.0644773        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.41063829 1.41063829 1.41063829 1.41063829 1.41063829]\n",
      "Trial error:  1.41063829236894\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0006487075667532134  :: absSize readout:  0.005770362650140861\n",
      "NRMSE W:  0.00576844269350334  :: absSize W:  0.0762054246360173\n",
      "NRMSEs: best:  0.00576844269350334 avg:  0.00576844269350334\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00188397        initial\n",
      "    2        618.034   0.00623681        golden\n",
      "    3        236.068    0.0047917        golden\n",
      "    4        408.219   0.00271023        parabolic\n",
      "    5         342.39  0.000991855        parabolic\n",
      "    6        301.779   0.00308369        golden\n",
      "    7        326.878  0.000999759        golden\n",
      "    8        335.243   0.00103892        parabolic\n",
      "    9        357.507   0.00148672        golden\n",
      "   10        348.164   0.00152392        golden\n",
      "   11         339.66   0.00141821        golden\n",
      "   12        344.596   0.00150742        golden\n",
      "   13        341.347   0.00159068        golden\n",
      "   14        343.233   0.00152743        golden\n",
      "   15        341.992  0.000952426        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.25710133 1.28801063 1.29890598 1.23494041 1.32496182]\n",
      "Trial error:  1.2880106292215416\n",
      "LR:  0.99\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00014171613414422457  :: absSize readout:  0.0037063887855361215\n",
      "NRMSE W:  0.004942143817593569  :: absSize W:  0.0382345222840573\n",
      "NRMSEs: best:  0.004942143817593569 avg:  0.004942143817593569\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0137663        initial\n",
      "    2        618.034   0.00205507        golden\n",
      "    3        763.932   0.00139089        golden\n",
      "    4        710.279   0.00157901        parabolic\n",
      "    5        854.102   0.00143126        golden\n",
      "    6        800.876   0.00126735        parabolic\n",
      "    7        805.875   0.00131654        parabolic\n",
      "    8        787.723   0.00134373        parabolic\n",
      "    9        795.852   0.00133341        golden\n",
      "   10         801.23    0.0013169        parabolic\n",
      "   11        798.957    0.0013187        golden\n",
      "   12        800.143   0.00131778        golden\n",
      "   13        800.542    0.0011991        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.39848012 1.39971376 1.39845289 1.40389017 1.39852282]\n",
      "Trial error:  1.3985228206658977\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0030607045852839806  :: absSize readout:  0.006745500154753945\n",
      "NRMSE W:  0.012046190293511555  :: absSize W:  0.06299520175332451\n",
      "NRMSEs: best:  0.012046190293511555 avg:  0.012046190293511555\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00136647        initial\n",
      "    2        618.034  0.000496921        golden\n",
      "    3        763.932  0.000268278        golden\n",
      "    4        832.405  0.000207723        parabolic\n",
      "    5        896.421  0.000164471        golden\n",
      "    6        935.984  0.000142877        golden\n",
      "    7        960.436  0.000131157        golden\n",
      "    8        975.548  0.000124506        golden\n",
      "    9        984.888  0.000120584        golden\n",
      "   10         990.66  0.000118215        golden\n",
      "   11        994.228  0.000116804        golden\n",
      "   12        996.433  0.000115917        golden\n",
      "   13        997.795  0.000115398        golden\n",
      "   14        998.637  0.000115072        golden\n",
      "   15        999.158  0.000114866        golden\n",
      "   16        999.491  0.000114737        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1.41555121 1.41558801 1.41558801 1.41557257 1.41558785]\n",
      "Trial error:  1.4155878484637903\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0006385697819591926  :: absSize readout:  0.00527867322832582\n",
      "NRMSE W:  0.004037906547886235  :: absSize W:  0.06010756866547085\n",
      "NRMSEs: best:  0.004037906547886235 avg:  0.004037906547886235\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.007597        initial\n",
      "    2        618.034    0.0132906        golden\n",
      "    3        236.068    0.0051166        golden\n",
      "    4        145.898   0.00506934        golden\n",
      "    5        187.228   0.00517549        parabolic\n",
      "    6        90.1699   0.00347539        golden\n",
      "    7        55.7281   0.00202017        golden\n",
      "    8        34.4419    0.0024495        golden\n",
      "    9        54.0884   0.00220888        parabolic\n",
      "   10        68.8837   0.00239582        golden\n",
      "   11        60.8354    0.0021934        parabolic\n",
      "   12        57.5138    0.0022131        parabolic\n",
      "   13        56.4102    0.0021081        golden\n",
      "   14        55.1018   0.00241523        golden\n",
      "   15        56.0614    0.0022847        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.49663037 0.87996764 0.9495564  0.73666164 0.57376161]\n",
      "Trial error:  0.7366616437942306\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.008777035578983856  :: absSize readout:  0.00639696569753535\n",
      "NRMSE W:  0.026466501772677165  :: absSize W:  0.04059202313269578\n",
      "NRMSEs: best:  0.026466501772677165 avg:  0.026466501772677165\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  2.20838e-05        initial\n",
      "    2        618.034  3.70539e-06        golden\n",
      "    3        763.932   1.6412e-06        golden\n",
      "    4        733.399  1.92099e-06        parabolic\n",
      "    5        854.102   1.0639e-06        golden\n",
      "    6        948.958  7.05046e-07        parabolic\n",
      "    7        912.726  8.21274e-07        golden\n",
      "    8        968.454  6.51256e-07        golden\n",
      "    9        980.504  6.20468e-07        golden\n",
      "   10        987.951  6.02228e-07        golden\n",
      "   11        992.553  5.91337e-07        golden\n",
      "   12        995.398  5.84654e-07        golden\n",
      "   13        997.156  5.80705e-07        golden\n",
      "   14        998.242  5.78346e-07        golden\n",
      "   15        998.914  5.76714e-07        golden\n",
      "   16        999.329  5.75797e-07        golden\n",
      "   17        999.662  5.74965e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.40968497 1.40969458 1.40969857 1.40969476 1.40969   ]\n",
      "Trial error:  1.4096945763562354\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.002987652213817784  :: absSize readout:  0.006216129593861718\n",
      "NRMSE W:  0.009024349975411287  :: absSize W:  0.03941150391840377\n",
      "NRMSEs: best:  0.009024349975411287 avg:  0.009024349975411287\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000311654        initial\n",
      "    2        618.034  6.80276e-05        golden\n",
      "    3        763.932  6.43876e-05        golden\n",
      "    4        695.714  3.78849e-05        parabolic\n",
      "    5        693.327  8.54205e-05        parabolic\n",
      "    6        721.771  4.43762e-05        golden\n",
      "    7        705.667  8.12237e-05        golden\n",
      "    8        699.516   6.0817e-05        golden\n",
      "    9        697.166  8.41267e-05        golden\n",
      "   10        694.802  6.17738e-05        golden\n",
      "   11        696.269  8.42971e-05        golden\n",
      "   12        695.366  8.46681e-05        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.41646527 1.27174729 1.38179417 1.41648914 1.21780464]\n",
      "Trial error:  1.3817941669418996\n",
      "Best error:  0.15632906341454575  best LR:  0.42999999999999994\n",
      "0.42999999999999994\n",
      "testing at  40.0  Hz\n",
      "Error:  0.4353125530772639\n",
      "testing at  42.378523774371814  Hz\n",
      "Error:  0.4456994308226285\n",
      "testing at  44.89848193237492  Hz\n",
      "Error:  0.39360984457109355\n",
      "testing at  47.56828460010884  Hz\n",
      "Error:  0.38777487983292847\n",
      "testing at  50.39684199579493  Hz\n",
      "Error:  0.33764093918078514\n",
      "testing at  53.39359416680138  Hz\n",
      "Error:  0.2476335170001\n",
      "testing at  56.568542494923804  Hz\n",
      "Error:  0.44841559321310115\n",
      "testing at  59.93228307506726  Hz\n",
      "Error:  0.20492100112254688\n",
      "testing at  63.496042078727974  Hz\n",
      "Error:  0.2517932788112723\n",
      "testing at  67.27171322029716  Hz\n",
      "Error:  0.11696490817429046\n",
      "testing at  71.27189745122715  Hz\n",
      "Error:  0.18682035515479653\n",
      "testing at  75.50994501453548  Hz\n",
      "Error:  0.2158284252169886\n",
      "testing at  80.0  Hz\n",
      "Error:  0.15617327477090404\n",
      "testing at  84.75704754874363  Hz\n",
      "Error:  0.12614169582247917\n",
      "testing at  89.79696386474984  Hz\n",
      "Error:  0.5374172436490241\n",
      "testing at  95.13656920021768  Hz\n",
      "Error:  0.7984972546514334\n",
      "testing at  100.79368399158986  Hz\n",
      "Error:  0.9336974973162374\n",
      "testing at  106.78718833360276  Hz\n",
      "Error:  1.0063028080783298\n",
      "testing at  113.13708498984761  Hz\n",
      "Error:  1.302719754328094\n",
      "testing at  119.86456615013452  Hz\n",
      "Error:  1.3944327511855072\n",
      "testing at  126.99208415745595  Hz\n",
      "Error:  1.4147472080012813\n",
      "testing at  134.54342644059432  Hz\n",
      "Error:  1.4177659593326704\n",
      "testing at  142.5437949024543  Hz\n",
      "Error:  1.4173945699038228\n",
      "testing at  151.01989002907095  Hz\n",
      "Error:  1.416515207902504\n",
      "testing at  160.0  Hz\n",
      "Error:  1.4171143744732597\n",
      "====== running test:  160 <function bl_sawtooth at 0x7fc012258e18>\n",
      "LR:  0.01\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.538400022747049  :: absSize readout:  1.1657555923276588\n",
      "NRMSE W:  0.08516624428091016  :: absSize W:  0.6118990035696952\n",
      "NRMSEs: best:  0.08516624428091016 avg:  0.08516624428091016\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000120745        initial\n",
      "    2        618.034  0.000230016        golden\n",
      "    3        236.068  1.32995e-07        golden\n",
      "    4        145.898  1.11517e-07        golden\n",
      "    5        190.949  1.14568e-07        parabolic\n",
      "    6        159.462  1.29236e-07        parabolic\n",
      "    7        90.1699  1.19181e-07        golden\n",
      "    8        151.798  1.26249e-07        parabolic\n",
      "    9        145.565  1.18754e-07        parabolic\n",
      "   10        148.152  1.11727e-07        golden\n",
      "   11        146.759  1.09657e-07        golden\n",
      "   12        147.092  1.06359e-07        parabolic\n",
      "   13        147.497  1.23548e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.44264494 1.4426478  1.4426277  1.44261626 1.44265793]\n",
      "Trial error:  1.4426449443769318\n",
      "Best so far ***************** \n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.5400108236755422  :: absSize readout:  1.227078711244919\n",
      "NRMSE W:  0.0990663088263279  :: absSize W:  0.6431696776547271\n",
      "NRMSEs: best:  0.0990663088263279 avg:  0.0990663088263279\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  9.32501e-05        initial\n",
      "    2        618.034  0.000142547        golden\n",
      "    3        236.068  3.12749e-07        golden\n",
      "    4        145.898  2.75949e-07        golden\n",
      "    5        190.907  2.49944e-07        parabolic\n",
      "    6        181.635   2.3375e-07        parabolic\n",
      "    7        172.844  2.35346e-07        parabolic\n",
      "    8         178.09  2.77961e-07        parabolic\n",
      "    9        181.969  2.94819e-07        parabolic\n",
      "   10        180.281   2.6619e-07        golden\n",
      "   11        181.118  2.68027e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.44264526 1.44266102 1.44266504 1.44266118 1.44266227]\n",
      "Trial error:  1.4426611833802774\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.5582235132978131  :: absSize readout:  1.128877437747788\n",
      "NRMSE W:  0.13670754459327458  :: absSize W:  0.6200201298314082\n",
      "NRMSEs: best:  0.13670754459327458 avg:  0.13670754459327458\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  7.63072e-05        initial\n",
      "    2        618.034  8.57039e-05        golden\n",
      "    3        236.068  4.47459e-08        golden\n",
      "    4        145.898  4.35057e-08        golden\n",
      "    5         190.98  4.44599e-08        parabolic\n",
      "    6        90.1699  4.34481e-08        golden\n",
      "    7        115.445   4.3767e-08        parabolic\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    8        55.7281  4.25746e-08        golden\n",
      "    9        34.4419  4.50101e-08        golden\n",
      "   10        68.8837  4.35736e-08        golden\n",
      "   11        47.5975  4.36478e-08        golden\n",
      "   12        60.7531  4.26936e-08        golden\n",
      "   13        52.6225  4.28286e-08        golden\n",
      "   14        57.3276  4.22305e-08        parabolic\n",
      "   15        58.0707  4.32927e-08        parabolic\n",
      "   16        56.7167  4.28243e-08        golden\n",
      "   17         57.661  4.32987e-08        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.44268346 1.4426719  1.44269053 1.44268432 1.44268646]\n",
      "Trial error:  1.4426843213842393\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.5387965482825227  :: absSize readout:  1.2018231902360328\n",
      "NRMSE W:  0.09679609573996759  :: absSize W:  0.6331501854410962\n",
      "NRMSEs: best:  0.09679609573996759 avg:  0.09679609573996759\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  6.76851e-05        initial\n",
      "    2        618.034  8.97565e-05        golden\n",
      "    3        236.068  5.34172e-07        golden\n",
      "    4        145.898  4.49331e-07        golden\n",
      "    5        190.741  4.11546e-07        parabolic\n",
      "    6        179.027  5.23994e-07        parabolic\n",
      "    7        208.054  4.22149e-07        golden\n",
      "    8        186.317  3.44833e-07        parabolic\n",
      "    9        183.532  5.40642e-07        golden\n",
      "   10        188.007  4.80051e-07        golden\n",
      "   11        185.253  5.52516e-07        golden\n",
      "   12        186.962  3.49799e-07        golden\n",
      "   13        185.911   4.4608e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.44264701 1.44264277 1.44266967 1.44264614 1.44264278]\n",
      "Trial error:  1.442646140847769\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.5460517649979599  :: absSize readout:  1.1862251910597283\n",
      "NRMSE W:  0.10246985685772488  :: absSize W:  0.6158757076716331\n",
      "NRMSEs: best:  0.10246985685772488 avg:  0.10246985685772488\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  9.24913e-05        initial\n",
      "    2        618.034  0.000117902        golden\n",
      "    3        236.068  9.44883e-08        golden\n",
      "    4        145.898  1.15689e-07        golden\n",
      "    5        191.027  9.11825e-08        parabolic\n",
      "    6        208.179  1.03302e-07        parabolic\n",
      "    7        173.789  8.74709e-08        golden\n",
      "    8        163.136  1.02287e-07        golden\n",
      "    9        180.373  8.80256e-08        golden\n",
      "   10        173.456  1.06837e-07        parabolic\n",
      "   11        174.122   1.0021e-07        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.44260828 1.44261265 1.44263255 1.44263223 1.44262414]\n",
      "Trial error:  1.4426241361872376\n",
      "Best so far ***************** \n",
      "LR:  0.07999999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.14783391031299342  :: absSize readout:  0.5752374074410499\n",
      "NRMSE W:  0.041861711074113704  :: absSize W:  0.42233018307448217\n",
      "NRMSEs: best:  0.041861711074113704 avg:  0.041861711074113704\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000238137        initial\n",
      "    2        618.034  0.000124561        golden\n",
      "    3        763.932  0.000137928        golden\n",
      "    4        660.432  0.000133184        parabolic\n",
      "    5        527.864  0.000109483        golden\n",
      "    6        472.136  0.000134112        golden\n",
      "    7        562.306   0.00010808        golden\n",
      "    8        550.545  0.000113397        parabolic\n",
      "    9        583.592  0.000115219        golden\n",
      "   10        570.437  0.000126259        golden\n",
      "   11        557.813  0.000114074        golden\n",
      "   12        565.412  0.000115057        golden\n",
      "   13         560.59  0.000125062        golden\n",
      "   14        563.492  0.000114929        golden\n",
      "   15         561.65  0.000117516        golden\n",
      "   16        562.759  0.000114392        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.48197177 1.47770178 1.476949   1.50518665 1.47841882]\n",
      "Trial error:  1.4784188165530123\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.14328654637148056  :: absSize readout:  0.5678396423569888\n",
      "NRMSE W:  0.04146690627611337  :: absSize W:  0.43706191277785644\n",
      "NRMSEs: best:  0.04146690627611337 avg:  0.04146690627611337\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00883268        initial\n",
      "    2        618.034   0.00849603        golden\n",
      "    3        763.932   0.00732032        golden\n",
      "    4        854.102   0.00637914        golden\n",
      "    5         909.83   0.00613367        golden\n",
      "    6        935.228   0.00598055        parabolic\n",
      "    7        959.968   0.00576243        golden\n",
      "    8        975.259   0.00564858        golden\n",
      "    9        984.709   0.00565418        golden\n",
      "   10        979.072   0.00566193        parabolic\n",
      "   11        969.419   0.00571507        golden\n",
      "   12        973.028   0.00566824        golden\n",
      "   13        976.716   0.00561594        golden\n",
      "   14        977.616   0.00562247        golden\n",
      "   15        977.049   0.00562528        parabolic\n",
      "   16        976.159   0.00568566        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42673938 1.42526069 1.42441038 1.42677023 1.42576096]\n",
      "Trial error:  1.4257609593868654\n",
      "Best so far ***************** \n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.14394802830500003  :: absSize readout:  0.6076383936261405\n",
      "NRMSE W:  0.03961442457082741  :: absSize W:  0.4276376353346739\n",
      "NRMSEs: best:  0.03961442457082741 avg:  0.03961442457082741\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00197246        initial\n",
      "    2        618.034   0.00265016        golden\n",
      "    3        236.068  0.000833184        golden\n",
      "    4        145.898   0.00100872        golden\n",
      "    5        214.537   0.00087439        parabolic\n",
      "    6        291.796  0.000732497        golden\n",
      "    7        326.238  0.000675525        golden\n",
      "    8        347.524  0.000642451        golden\n",
      "    9         360.68   0.00169143        golden\n",
      "   10        339.394  0.000654892        golden\n",
      "   11        352.549   0.00063488        golden\n",
      "   12        355.655   0.00164068        golden\n",
      "   13         350.63    0.0015358        golden\n",
      "   14        353.735    0.0015828        golden\n",
      "   15        351.816    0.0015491        golden\n",
      "   16        353.002   0.00158661        golden\n",
      "   17        352.216    0.0015552        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42576011 1.4254402  1.42699138 1.42809169 1.41640446]\n",
      "Trial error:  1.4257601077738495\n",
      "Best so far ***************** \n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.145448530978122  :: absSize readout:  0.581128239755107\n",
      "NRMSE W:  0.04477739842222408  :: absSize W:  0.44706547146484354\n",
      "NRMSEs: best:  0.04477739842222408 avg:  0.04477739842222408\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000342921        initial\n",
      "    2        618.034  0.000341001        golden\n",
      "    3        763.932  0.000399998        golden\n",
      "    4        503.765  0.000359067        parabolic\n",
      "    5        673.762  0.000338378        golden\n",
      "    6        708.204  0.000395153        golden\n",
      "    7        652.476  0.000329131        golden\n",
      "    8        647.582  0.000329325        parabolic\n",
      "    9        651.124  0.000377943        parabolic\n",
      "   10        652.809  0.000323639        parabolic\n",
      "   11        660.812  0.000327051        golden\n",
      "   12        656.706  0.000328305        parabolic\n",
      "   13        654.298  0.000319533        golden\n",
      "   14        655.217  0.000323468        golden\n",
      "   15        653.964  0.000383774        parabolic\n",
      "   16        654.631  0.000315506        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.50247526 1.49412728 1.49612045 1.49436135 1.49686226]\n",
      "Trial error:  1.4961204482637451\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.14594702455636013  :: absSize readout:  0.5855222998975508\n",
      "NRMSE W:  0.04224787353584113  :: absSize W:  0.41085799260806155\n",
      "NRMSEs: best:  0.04224787353584113 avg:  0.04224787353584113\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000506242        initial\n",
      "    2        618.034  0.000747592        golden\n",
      "    3        236.068   0.00036388        golden\n",
      "    4        145.898  0.000704691        golden\n",
      "    5        284.798  0.000187609        parabolic\n",
      "    6        298.696  0.000275911        parabolic\n",
      "    7        271.793  0.000192176        parabolic\n",
      "    8            279  0.000190417        parabolic\n",
      "    9        290.106  0.000232893        golden\n",
      "   10        282.583  0.000207474        golden\n",
      "   11        286.825  0.000205232        golden\n",
      "   12        283.952  0.000185532        golden\n",
      "   13        283.429  0.000191532        golden\n",
      "   14        284.285  0.000204521        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.4347331  1.43329135 1.42966307 1.42535104 1.43469195]\n",
      "Trial error:  1.4332913511988699\n",
      "LR:  0.15\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.04863243535488976  :: absSize readout:  0.33353889286895416\n",
      "NRMSE W:  0.02798443758233813  :: absSize W:  0.3820841495957131\n",
      "NRMSEs: best:  0.02798443758233813 avg:  0.02798443758233813\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000446206        initial\n",
      "    2        618.034     0.325758        golden\n",
      "    3        236.068   0.00113004        golden\n",
      "    4        309.664  0.000818275        parabolic\n",
      "    5        472.136     0.239374        golden\n",
      "    6        416.408     0.161976        golden\n",
      "    7        354.349  0.000585703        golden\n",
      "    8        395.122      0.11604        golden\n",
      "    9        371.417  0.000520765        golden\n",
      "   10        386.991    0.0937337        golden\n",
      "   11        377.937    0.0593713        golden\n",
      "   12        383.885    0.0838887        golden\n",
      "   13        380.427    0.0711862        golden\n",
      "   14        382.699  0.000407511        golden\n",
      "   15        383.152    0.0813812        golden\n",
      "   16        382.366    0.0786114        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.41709264 1.41709264 1.41709264 1.41709264 1.41709264]\n",
      "Trial error:  1.4170926363967362\n",
      "Best so far ***************** \n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.04958686973040161  :: absSize readout:  0.3347565721765647\n",
      "NRMSE W:  0.025922901845973113  :: absSize W:  0.37758661149875933\n",
      "NRMSEs: best:  0.025922901845973113 avg:  0.025922901845973113\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.120321        initial\n",
      "    2        618.034     0.396849        golden\n",
      "    3        236.068   0.00220551        golden\n",
      "    4        145.898    0.0025303        golden\n",
      "    5        191.506   0.00261822        parabolic\n",
      "    6        291.796   0.00144373        golden\n",
      "    7        326.238  0.000849773        golden\n",
      "    8        347.524  0.000891613        golden\n",
      "    9         334.03  0.000717673        parabolic\n",
      "   10         336.18  0.000736505        parabolic\n",
      "   11        333.412   0.00076311        parabolic\n",
      "   12        334.958  0.000750303        parabolic\n",
      "   13        334.384  0.000767495        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.4655301  1.45447197 1.47104609 1.45093729 1.46795793]\n",
      "Trial error:  1.4655300989528603\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.04920164364558263  :: absSize readout:  0.33852074342231564\n",
      "NRMSE W:  0.031234222153424702  :: absSize W:  0.3943183049532847\n",
      "NRMSEs: best:  0.031234222153424702 avg:  0.031234222153424702\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000931633        initial\n",
      "    2        618.034     0.295997        golden\n",
      "    3        236.068   0.00140339        golden\n",
      "    4         309.51   0.00121611        parabolic\n",
      "    5        472.136     0.162148        golden\n",
      "    6        416.408  0.000835288        golden\n",
      "    7        437.694    0.0813228        golden\n",
      "    8        403.252  0.000831419        golden\n",
      "    9        408.818  0.000826895        parabolic\n",
      "   10        409.151  0.000837503        parabolic\n",
      "   11        408.484  0.000845029        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.46144749 1.46418307 1.46334824 1.45542462 1.45691451]\n",
      "Trial error:  1.4614474854365025\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.04931414309800781  :: absSize readout:  0.3344253003674011\n",
      "NRMSE W:  0.030747672756666764  :: absSize W:  0.41580252348561914\n",
      "NRMSEs: best:  0.030747672756666764 avg:  0.030747672756666764\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  0.000615125        initial\n",
      "    2        618.034      0.25517        golden\n",
      "    3        236.068   0.00213605        golden\n",
      "    4        310.846   0.00117374        parabolic\n",
      "    5        472.136      0.05396        golden\n",
      "    6        460.673    0.0401537        parabolic\n",
      "    7        412.029   0.00060174        golden\n",
      "    8        400.038  0.000683447        parabolic\n",
      "    9         430.61  0.000524588        golden\n",
      "   10        442.093  0.000452255        golden\n",
      "   11         449.19     0.024893        golden\n",
      "   12        437.707  0.000454097        golden\n",
      "   13        440.153  0.000505415        parabolic\n",
      "   14        444.804  0.000473519        golden\n",
      "   15        443.128  0.000539504        golden\n",
      "   16        441.352  0.000533673        golden\n",
      "   17        442.488    0.0004234        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.47268725 1.46193975 1.50957618 1.46643624 1.50287377]\n",
      "Trial error:  1.472687251361168\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.048387078680351775  :: absSize readout:  0.3165650202355027\n",
      "NRMSE W:  0.027952123074573103  :: absSize W:  0.3873107958259514\n",
      "NRMSEs: best:  0.027952123074573103 avg:  0.027952123074573103\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00171642        initial\n",
      "    2        618.034     0.287485        golden\n",
      "    3        236.068   0.00325978        golden\n",
      "    4        310.671   0.00245243        parabolic\n",
      "    5        472.136     0.168434        golden\n",
      "    6        416.408    0.0544814        golden\n",
      "    7        354.734   0.00202482        golden\n",
      "    8        395.122   0.00157014        golden\n",
      "    9        403.252    0.0015436        golden\n",
      "   10        403.609   0.00149303        parabolic\n",
      "   11        408.498   0.00155978        golden\n",
      "   12        405.823   0.00147872        parabolic\n",
      "   13        404.777   0.00148254        parabolic\n",
      "   14        406.156   0.00151808        parabolic\n",
      "   15         405.49   0.00149011        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.4205368  1.4210547  1.41707866 1.41999064 1.41880441]\n",
      "Trial error:  1.4199906373805296\n",
      "LR:  0.21999999999999997\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.015973913355824258  :: absSize readout:  0.14785488346015593\n",
      "NRMSE W:  0.019367755694436033  :: absSize W:  0.30900341919298\n",
      "NRMSEs: best:  0.019367755694436033 avg:  0.019367755694436033\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.367326        initial\n",
      "    2        618.034     0.424101        golden\n",
      "    3        236.068   0.00180213        golden\n",
      "    4        145.898   0.00146929        golden\n",
      "    5        190.809   0.00167722        parabolic\n",
      "    6        90.1699  0.000894425        golden\n",
      "    7        55.7281  0.000131281        golden\n",
      "    8        34.4419  0.000138247        golden\n",
      "    9        45.4905  0.000134532        parabolic\n",
      "   10        68.8837  3.79327e-05        golden\n",
      "   11        77.0143  0.000151587        golden\n",
      "   12        63.8587  0.000129018        golden\n",
      "   13        71.9893   5.8569e-05        golden\n",
      "   14         69.346  0.000127522        parabolic\n",
      "   15        66.9643  0.000128054        golden\n",
      "   16        68.1506  0.000127873        golden\n",
      "   17        68.5504  0.000127811        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1.36370318 1.10212168 1.09823689 1.38826577 1.41607119]\n",
      "Trial error:  1.3637031808103357\n",
      "Best so far ***************** \n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.01994341184356617  :: absSize readout:  0.15686754467967326\n",
      "NRMSE W:  0.01804826340978001  :: absSize W:  0.31351108501250474\n",
      "NRMSEs: best:  0.01804826340978001 avg:  0.01804826340978001\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.446434        initial\n",
      "    2        618.034     0.486321        golden\n",
      "    3        236.068  0.000847976        golden\n",
      "    4        145.898   0.00093079        golden\n",
      "    5        191.018   0.00104023        parabolic\n",
      "    6        291.796     0.363566        golden\n",
      "    7        257.354     0.258795        golden\n",
      "    8        218.861  0.000932858        golden\n",
      "    9        244.199     0.169488        golden\n",
      "   10        229.495  0.000879319        golden\n",
      "   11        239.174  0.000833893        golden\n",
      "   12        241.093  0.000823948        golden\n",
      "   13        242.279     0.148865        golden\n",
      "   14         240.36     0.121327        golden\n",
      "   15        241.546     0.139602        golden\n",
      "   16         240.76     0.128143        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42736458 1.39586721 1.42736458 1.42736458 1.39546457]\n",
      "Trial error:  1.4273645766338237\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.019152121610204714  :: absSize readout:  0.15680607911074235\n",
      "NRMSE W:  0.019866591506334578  :: absSize W:  0.3821091772356736\n",
      "NRMSEs: best:  0.019866591506334578 avg:  0.019866591506334578\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.472093        initial\n",
      "    2        618.034     0.493403        golden\n",
      "    3        236.068     0.315674        golden\n",
      "    4        145.898   0.00162659        golden\n",
      "    5        90.1699   0.00277328        golden\n",
      "    6        118.462   0.00183507        parabolic\n",
      "    7        140.463   0.00166263        parabolic\n",
      "    8         180.34    0.0014366        golden\n",
      "    9        201.626   0.00134287        golden\n",
      "   10        214.782     0.184631        golden\n",
      "   11        193.496     0.001377        golden\n",
      "   12        206.651    0.0013227        golden\n",
      "   13        209.757     0.123013        golden\n",
      "   14        204.732   0.00133032        golden\n",
      "   15        207.837   0.00131803        golden\n",
      "   16        208.571    0.0996472        golden\n",
      "   17        207.384    0.0013198        golden\n",
      "   18        208.171   0.00131673        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42465488 1.42465495 1.42874338 1.42465489 1.42465491]\n",
      "Trial error:  1.424654910907465\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.018349353614062033  :: absSize readout:  0.13944889836612173\n",
      "NRMSE W:  0.018816141011846282  :: absSize W:  0.3234475663966952\n",
      "NRMSEs: best:  0.018816141011846282 avg:  0.018816141011846282\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.413982        initial\n",
      "    2        618.034     0.464732        golden\n",
      "    3        236.068   0.00201071        golden\n",
      "    4        145.898   0.00115784        golden\n",
      "    5        190.586   0.00168968        parabolic\n",
      "    6        90.1699   0.00108814        golden\n",
      "    7        112.138  0.000897495        parabolic\n",
      "    8        115.908  0.000745487        parabolic\n",
      "    9        127.363  0.000601412        golden\n",
      "   10        125.086  0.000653216        parabolic\n",
      "   11        134.443    0.0009882        golden\n",
      "   12        130.067  0.000682235        golden\n",
      "   13         127.03  0.000662031        parabolic\n",
      "   14        128.396  0.000633006        golden\n",
      "   15        127.757  0.000686785        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.35840024 1.37251167 1.3805555  1.3507643  1.29648891]\n",
      "Trial error:  1.358400238018976\n",
      "Best so far ***************** \n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.018348664472614965  :: absSize readout:  0.15875238145175205\n",
      "NRMSE W:  0.019995555670878733  :: absSize W:  0.33787215003126925\n",
      "NRMSEs: best:  0.019995555670878733 avg:  0.019995555670878733\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.403157        initial\n",
      "    2        618.034     0.457992        golden\n",
      "    3        236.068   0.00293276        golden\n",
      "    4        145.898   0.00344325        golden\n",
      "    5        191.226   0.00317714        parabolic\n",
      "    6        291.796     0.277521        golden\n",
      "    7        257.354   0.00282446        golden\n",
      "    8         270.51     0.189817        golden\n",
      "    9        249.224   0.00286521        golden\n",
      "   10        262.379     0.128563        golden\n",
      "   11        254.249   0.00283993        golden\n",
      "   12        259.274   0.00281496        golden\n",
      "   13         260.46     0.098183        golden\n",
      "   14         258.54   0.00281858        golden\n",
      "   15        259.727   0.00281272        golden\n",
      "   16         260.06   0.00281107        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42204905 1.42204876 1.42204885 1.42204916 1.42204918]\n",
      "Trial error:  1.4220490518975857\n",
      "LR:  0.29\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0066023591349593045  :: absSize readout:  0.07295709322457425\n",
      "NRMSE W:  0.01580640964339739  :: absSize W:  0.2934640114657549\n",
      "NRMSEs: best:  0.01580640964339739 avg:  0.01580640964339739\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.460445        initial\n",
      "    2        618.034     0.491123        golden\n",
      "    3        236.068     0.314614        golden\n",
      "    4        145.898   0.00378734        golden\n",
      "    5        90.1699   0.00415126        golden\n",
      "    6        118.172    0.0042644        parabolic\n",
      "    7         180.34   0.00313713        golden\n",
      "    8        201.626   0.00294702        golden\n",
      "    9        215.998     0.167039        parabolic\n",
      "   10        193.496   0.00304218        golden\n",
      "   11        207.116   0.00281071        golden\n",
      "   12        210.508   0.00276499        golden\n",
      "   13        214.085   0.00254784        parabolic\n",
      "   14        212.719   0.00271424        golden\n",
      "   15        214.816   0.00261741        golden\n",
      "   16        213.751   0.00249625        parabolic\n",
      "   17        213.357   0.00271499        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.48988771 1.47877538 1.50399633 1.49311269 1.49428229]\n",
      "Trial error:  1.4931126946907862\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.007498772468584633  :: absSize readout:  0.07856772680441698\n",
      "NRMSE W:  0.01507191391882994  :: absSize W:  0.24361712643805564\n",
      "NRMSEs: best:  0.01507191391882994 avg:  0.01507191391882994\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.448759        initial\n",
      "    2        618.034     0.460581        golden\n",
      "    3        236.068     0.404943        golden\n",
      "    4        145.898  0.000608903        golden\n",
      "    5        90.1699   0.00204187        golden\n",
      "    6         118.45   0.00150391        parabolic\n",
      "    7         180.34     0.332641        golden\n",
      "    8        159.054  0.000508946        golden\n",
      "    9        158.644  0.000519239        parabolic\n",
      "   10        167.184     0.276941        golden\n",
      "   11        162.159     0.230791        golden\n",
      "   12         160.24     0.196642        golden\n",
      "   13        159.507  0.000524257        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.34239374 1.35929961 1.34245313 1.33910011 1.35172167]\n",
      "Trial error:  1.3424531347750654\n",
      "Best so far ***************** \n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.006441357751775714  :: absSize readout:  0.07469699313864538\n",
      "NRMSE W:  0.017392765423882534  :: absSize W:  0.28716215175396104\n",
      "NRMSEs: best:  0.017392765423882534 avg:  0.017392765423882534\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.335677        initial\n",
      "    2        618.034     0.365916        golden\n",
      "    3        236.068     0.228969        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    4        145.898   0.00118588        golden\n",
      "    5        90.1699  0.000384012        golden\n",
      "    6        117.616  0.000516834        parabolic\n",
      "    7        96.7268  0.000526808        parabolic\n",
      "    8        55.7281  6.93296e-05        golden\n",
      "    9        34.4419  0.000152461        golden\n",
      "   10        53.4288   7.0152e-05        parabolic\n",
      "   11        55.3948  6.82927e-05        parabolic\n",
      "   12        54.6798   6.9997e-05        parabolic\n",
      "   13        55.0614   7.0122e-05        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.52261871 0.54052778 0.97053995 1.04092995 1.09664164]\n",
      "Trial error:  0.9705399484751636\n",
      "Best so far ***************** \n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.006781495257780319  :: absSize readout:  0.07164623200946918\n",
      "NRMSE W:  0.015697035544009783  :: absSize W:  0.24943479008375577\n",
      "NRMSEs: best:  0.015697035544009783 avg:  0.015697035544009783\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.399468        initial\n",
      "    2        618.034     0.413206        golden\n",
      "    3        236.068     0.338661        golden\n",
      "    4        145.898  0.000482131        golden\n",
      "    5        90.1699  0.000264407        golden\n",
      "    6        117.958  0.000500001        parabolic\n",
      "    7        55.7281  3.24086e-05        golden\n",
      "    8        34.4419  1.88962e-05        golden\n",
      "    9        42.1859  1.95508e-05        parabolic\n",
      "   10        37.2737  1.87331e-05        parabolic\n",
      "   11        36.8531  1.87583e-05        parabolic\n",
      "   12          39.15   1.8879e-05        golden\n",
      "   13        37.6071  1.92626e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.14675959 0.1835493  0.33963824 0.33986817 0.26129729]\n",
      "Trial error:  0.26129728615425746\n",
      "Best so far ***************** \n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.008595877806347407  :: absSize readout:  0.0846705486227899\n",
      "NRMSE W:  0.0175506412197795  :: absSize W:  0.3262430840451864\n",
      "NRMSEs: best:  0.0175506412197795 avg:  0.0175506412197795\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.347021        initial\n",
      "    2        618.034     0.377684        golden\n",
      "    3        236.068     0.220047        golden\n",
      "    4        145.898   0.00261378        golden\n",
      "    5        90.1699   0.00385556        golden\n",
      "    6        118.702   0.00337352        parabolic\n",
      "    7         180.34   0.00184408        golden\n",
      "    8        201.626   0.00148881        golden\n",
      "    9        214.782     0.120728        golden\n",
      "   10        193.496   0.00162509        golden\n",
      "   11        206.651   0.00142299        golden\n",
      "   12        209.757   0.00136709        golden\n",
      "   13        211.676   0.00134113        golden\n",
      "   14        212.862   0.00131043        golden\n",
      "   15        213.595     0.104016        golden\n",
      "   16        212.409   0.00132166        golden\n",
      "   17        213.196    0.0935453        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42881749 1.42900433 1.4285841  1.42913093 1.42922484]\n",
      "Trial error:  1.429004328937722\n",
      "LR:  0.36\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0042990948529282105  :: absSize readout:  0.05267015030704316\n",
      "NRMSE W:  0.014214650706815995  :: absSize W:  0.3252076250920439\n",
      "NRMSEs: best:  0.014214650706815995 avg:  0.014214650706815995\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.271689        initial\n",
      "    2        618.034     0.277957        golden\n",
      "    3        236.068     0.242367        golden\n",
      "    4        145.898  0.000646573        golden\n",
      "    5        90.1699    0.0014146        golden\n",
      "    6        118.407  0.000374364        parabolic\n",
      "    7         126.25   0.00129941        parabolic\n",
      "    8        107.621  0.000920005        golden\n",
      "    9        114.287   0.00112035        golden\n",
      "   10        121.403   0.00120807        golden\n",
      "   11        116.834  0.000690792        golden\n",
      "   12        119.551  0.000298862        golden\n",
      "   13        120.259  0.000874253        golden\n",
      "   14        119.114  0.000485012        golden\n",
      "   15        119.885  0.000546208        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.7652723  1.13333447 1.06320774 1.28581643 1.2323563 ]\n",
      "Trial error:  1.1333344741714435\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.003114729220933338  :: absSize readout:  0.03970017234965187\n",
      "NRMSE W:  0.01412869634037634  :: absSize W:  0.27218617297352954\n",
      "NRMSEs: best:  0.01412869634037634 avg:  0.01412869634037634\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.443064        initial\n",
      "    2        618.034      0.44889        golden\n",
      "    3        236.068     0.408303        golden\n",
      "    4        145.898     0.209839        golden\n",
      "    5        90.1699   0.00189156        golden\n",
      "    6        55.7281  0.000327652        golden\n",
      "    7        72.3936   0.00169387        parabolic\n",
      "    8        34.4419  0.000328497        golden\n",
      "    9        45.0942   0.00032746        parabolic\n",
      "   10         48.743   0.00032742        parabolic\n",
      "   11        48.2262  0.000327418        parabolic\n",
      "   12        47.8929  0.000327418        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.44260044 1.44260044 1.44260044 1.44260044 1.44260044]\n",
      "Trial error:  1.4426004440230775\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.002661014498475305  :: absSize readout:  0.03775196112680126\n",
      "NRMSE W:  0.014945005414411293  :: absSize W:  0.22055514256403796\n",
      "NRMSEs: best:  0.014945005414411293 avg:  0.014945005414411293\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.302826        initial\n",
      "    2        618.034     0.318339        golden\n",
      "    3        236.068     0.246686        golden\n",
      "    4        145.898   0.00261889        golden\n",
      "    5        90.1699   0.00253858        golden\n",
      "    6        117.995   0.00253799        parabolic\n",
      "    7        104.285   0.00251699        parabolic\n",
      "    8        103.951   0.00251675        parabolic\n",
      "    9        98.0558   0.00251991        parabolic\n",
      "   10        101.699   0.00251715        golden\n",
      "   11        103.082   0.00251706        parabolic\n",
      "   12        103.618   0.00251674        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.41587698 1.41434891 1.42125561 1.41777303 1.42122488]\n",
      "Trial error:  1.4177730251028873\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0024654334134537057  :: absSize readout:  0.03406523240925669\n",
      "NRMSE W:  0.014184630594247155  :: absSize W:  0.22793848870810166\n",
      "NRMSEs: best:  0.014184630594247155 avg:  0.014184630594247155\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966      0.36803        initial\n",
      "    2        618.034      0.38589        golden\n",
      "    3        236.068     0.315995        golden\n",
      "    4        145.898   0.00119769        golden\n",
      "    5        90.1699   0.00282666        golden\n",
      "    6         118.64   0.00199533        parabolic\n",
      "    7         180.34  0.000615191        golden\n",
      "    8        205.369     0.279652        parabolic\n",
      "    9        167.184  0.000910882        golden\n",
      "   10          189.9     0.234918        golden\n",
      "   11        175.315  0.000737035        golden\n",
      "   12        183.992     0.195707        golden\n",
      "   13        178.421  0.000695824        golden\n",
      "   14        181.735     0.159225        golden\n",
      "   15        179.607  0.000617909        golden\n",
      "   16        180.673  0.000617738        parabolic\n",
      "   17        180.007  0.000807882        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.20713873 1.17353301 1.19891458 1.16460267 1.22599392]\n",
      "Trial error:  1.1989145766031226\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.003066749331288791  :: absSize readout:  0.042860893245757516\n",
      "NRMSE W:  0.014034668265867224  :: absSize W:  0.25337441017769285\n",
      "NRMSEs: best:  0.014034668265867224 avg:  0.014034668265867224\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.428119        initial\n",
      "    2        618.034     0.438945        golden\n",
      "    3        236.068     0.390694        golden\n",
      "    4        145.898    0.0051281        golden\n",
      "    5        90.1699   0.00544795        golden\n",
      "    6        118.132   0.00526469        parabolic\n",
      "    7         180.34     0.330854        golden\n",
      "    8        159.054     0.217357        golden\n",
      "    9        135.292   0.00517499        golden\n",
      "   10        150.923   0.00510798        golden\n",
      "   11        154.029   0.00509618        golden\n",
      "   12        155.948   0.00508913        golden\n",
      "   13        157.134   0.00508486        golden\n",
      "   14        157.867     0.188235        golden\n",
      "   15        156.681   0.00508649        golden\n",
      "   16        157.468   0.00508368        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43319087 1.43319087 1.43319087 1.43319087 1.43319087]\n",
      "Trial error:  1.4331908652547984\n",
      "LR:  0.42999999999999994\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.000974125008802465  :: absSize readout:  0.019662833414019597\n",
      "NRMSE W:  0.012349283927924637  :: absSize W:  0.17342118524906097\n",
      "NRMSEs: best:  0.012349283927924637 avg:  0.012349283927924637\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.352173        initial\n",
      "    2        618.034     0.371659        golden\n",
      "    3        236.068      0.30392        golden\n",
      "    4        145.898   0.00189995        golden\n",
      "    5        90.1699   0.00152207        golden\n",
      "    6        117.886   0.00212455        parabolic\n",
      "    7        55.7281   0.00173757        golden\n",
      "    8        94.5859   0.00148106        parabolic\n",
      "    9        103.486   0.00239946        golden\n",
      "   10        97.9853   0.00144494        golden\n",
      "   11        100.086   0.00142006        golden\n",
      "   12        101.385    0.0014035        golden\n",
      "   13        102.187   0.00139275        golden\n",
      "   14        102.683   0.00138589        golden\n",
      "   15        103.017   0.00138118        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43775692 1.43775692 1.43775692 1.43775692 1.43775692]\n",
      "Trial error:  1.4377569177918714\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.001240759968035327  :: absSize readout:  0.020849795546053537\n",
      "NRMSE W:  0.012796376233978765  :: absSize W:  0.2004984798131509\n",
      "NRMSEs: best:  0.012796376233978765 avg:  0.012796376233978765\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.264083        initial\n",
      "    2        618.034      0.30562        golden\n",
      "    3        236.068   0.00134312        golden\n",
      "    4        145.898   0.00165981        golden\n",
      "    5        191.213   0.00160539        parabolic\n",
      "    6        291.796    0.0018016        golden\n",
      "    7        257.354   0.00156873        golden\n",
      "    8        225.398  0.000802497        parabolic\n",
      "    9         212.34  0.000930189        golden\n",
      "   10        220.788   0.00125368        parabolic\n",
      "   11        229.474   0.00193446        golden\n",
      "   12        223.637   0.00169421        golden\n",
      "   13        226.955   0.00151973        golden\n",
      "   14        224.725   0.00110192        golden\n",
      "   15        225.993   0.00174352        golden\n",
      "   16        225.065   0.00128591        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.14861347 0.89736375 0.94500858 0.89703512 0.95806992]\n",
      "Trial error:  0.9450085811934889\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0012634125650779714  :: absSize readout:  0.0208117658667892\n",
      "NRMSE W:  0.011769549471737853  :: absSize W:  0.16303743155512845\n",
      "NRMSEs: best:  0.011769549471737853 avg:  0.011769549471737853\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.249208        initial\n",
      "    2        618.034     0.263972        golden\n",
      "    3        236.068   0.00627788        golden\n",
      "    4        145.898   0.00134356        golden\n",
      "    5        186.972   0.00546741        parabolic\n",
      "    6        90.1699  0.000962347        golden\n",
      "    7        114.495  0.000961465        parabolic\n",
      "    8        102.415    0.0013203        parabolic\n",
      "    9         126.49    0.0012143        golden\n",
      "   10        109.881   0.00120934        golden\n",
      "   11        119.077   0.00109873        golden\n",
      "   12        112.733   0.00122719        golden\n",
      "   13        116.245   0.00169918        golden\n",
      "   14        113.822   0.00115916        golden\n",
      "   15        115.164   0.00150215        golden\n",
      "   16        114.162   0.00132743        golden\n",
      "   17        114.828   0.00121268        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.99410478 1.00647791 0.5202448  0.96202296 1.11059234]\n",
      "Trial error:  0.9941047805706212\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0010414740283018456  :: absSize readout:  0.019608930725920706\n",
      "NRMSE W:  0.011735692349839033  :: absSize W:  0.1538723166647264\n",
      "NRMSEs: best:  0.011735692349839033 avg:  0.011735692349839033\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.310121        initial\n",
      "    2        618.034     0.332689        golden\n",
      "    3        236.068     0.247018        golden\n",
      "    4        145.898   0.00313878        golden\n",
      "    5        90.1699     0.002114        golden\n",
      "    6        117.535   0.00280615        parabolic\n",
      "    7        55.7281   0.00106764        golden\n",
      "    8        34.4419  0.000384505        golden\n",
      "    9        21.2862  6.52397e-05        golden\n",
      "   10        13.1556  9.30314e-05        golden\n",
      "   11        18.5349  7.02103e-05        parabolic\n",
      "   12        22.9258  6.43735e-05        parabolic\n",
      "   13        23.2591  6.42213e-05        parabolic\n",
      "   14        27.5306  7.03418e-05        golden\n",
      "   15        24.8907  0.000458164        golden\n",
      "   16        23.8823  0.000471736        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.23601107 0.17560168 1.38923236 0.12210755 0.17634397]\n",
      "Trial error:  0.1763439653070522\n",
      "Best so far ***************** \n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0010522497736084517  :: absSize readout:  0.019549999995406464\n",
      "NRMSE W:  0.012824091831625023  :: absSize W:  0.17439648830982116\n",
      "NRMSEs: best:  0.012824091831625023 avg:  0.012824091831625023\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.272893        initial\n",
      "    2        618.034     0.307487        golden\n",
      "    3        236.068     0.143749        golden\n",
      "    4        145.898  0.000811633        golden\n",
      "    5        90.1699  0.000385156        golden\n",
      "    6         117.68  0.000458867        parabolic\n",
      "    7        96.3238  0.000647361        parabolic\n",
      "    8        55.7281  1.74628e-05        golden\n",
      "    9        34.4419     2.66e-05        golden\n",
      "   10         46.162   0.00391174        parabolic\n",
      "   11        68.8837  1.38883e-05        golden\n",
      "   12        77.0143   0.00395988        golden\n",
      "   13        63.8587   0.00401959        golden\n",
      "   14        71.9893   0.00399028        golden\n",
      "   15        66.9643  1.43509e-05        golden\n",
      "   16        70.0699   0.00399972        golden\n",
      "   17        68.1506  1.36946e-05        golden\n",
      "   18        68.4839   0.00400646        parabolic\n",
      "   19        67.8172  1.40603e-05        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.27630688 0.09484397 0.10508874 0.08400251 0.08730826]\n",
      "Trial error:  0.09484397311932018\n",
      "Best so far ***************** \n",
      "LR:  0.49999999999999994\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0005512828652505453  :: absSize readout:  0.012166791867643391\n",
      "NRMSE W:  0.011504065865467357  :: absSize W:  0.1586323277033636\n",
      "NRMSEs: best:  0.011504065865467357 avg:  0.011504065865467357\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.296289        initial\n",
      "    2        618.034     0.312979        golden\n",
      "    3        236.068      0.22636        golden\n",
      "    4        145.898   0.00110887        golden\n",
      "    5        90.1699  0.000791808        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    6        117.867  0.000585724        parabolic\n",
      "    7        111.961  0.000718243        parabolic\n",
      "    8        128.574   0.00105964        golden\n",
      "    9        121.957  0.000857298        golden\n",
      "   10        115.611  0.000430876        golden\n",
      "   11        115.278  0.000882344        parabolic\n",
      "   12        115.945  0.000762079        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.77116311 0.29317252 0.75582236 1.05398564 0.56720688]\n",
      "Trial error:  0.7558223636680551\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0009018058160882241  :: absSize readout:  0.013457245906262077\n",
      "NRMSE W:  0.009005993532385532  :: absSize W:  0.1937320165466111\n",
      "NRMSEs: best:  0.009005993532385532 avg:  0.009005993532385532\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.208382        initial\n",
      "    2        618.034     0.199475        golden\n",
      "    3        763.932     0.193008        golden\n",
      "    4        854.102     0.189178        golden\n",
      "    5         909.83     0.186907        golden\n",
      "    6        944.272     0.185529        golden\n",
      "    7        965.558     0.184694        golden\n",
      "    8        978.714     0.184182        golden\n",
      "    9        986.844     0.183868        golden\n",
      "   10        991.869      0.18368        golden\n",
      "   11        994.975     0.183561        golden\n",
      "   12        996.894     0.183482        golden\n",
      "   13        998.081     0.183442        golden\n",
      "   14        998.814     0.183414        golden\n",
      "   15        999.267     0.183396        golden\n",
      "   16          999.6     0.183378        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43996919 1.43996919 1.43996919 1.43996919 1.43996919]\n",
      "Trial error:  1.4399691914004262\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00042558432109888864  :: absSize readout:  0.011121710798105429\n",
      "NRMSE W:  0.010307745505723965  :: absSize W:  0.15883239387937353\n",
      "NRMSEs: best:  0.010307745505723965 avg:  0.010307745505723965\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.350838        initial\n",
      "    2        618.034     0.362871        golden\n",
      "    3        236.068     0.305552        golden\n",
      "    4        145.898   0.00233289        golden\n",
      "    5        90.1699   0.00240835        golden\n",
      "    6        118.063   0.00236513        parabolic\n",
      "    7         180.34   0.00235682        golden\n",
      "    8        151.443   0.00231832        parabolic\n",
      "    9        160.096    0.0023218        parabolic\n",
      "   10        154.748   0.00230636        golden\n",
      "   11        155.502   0.00230736        parabolic\n",
      "   12        154.415   0.00230661        parabolic\n",
      "   13        155.082   0.00231678        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.34535756 1.34437926 1.34580683 1.34266557 1.34400694]\n",
      "Trial error:  1.344379257215028\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0004341993558139791  :: absSize readout:  0.011412640113144181\n",
      "NRMSE W:  0.011073577216013946  :: absSize W:  0.15761705924534022\n",
      "NRMSEs: best:  0.011073577216013946 avg:  0.011073577216013946\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966      0.35551        initial\n",
      "    2        618.034      0.36513        golden\n",
      "    3        236.068     0.327606        golden\n",
      "    4        145.898     0.235091        golden\n",
      "    5        90.1699   0.00714465        golden\n",
      "    6        55.7281   0.00663212        golden\n",
      "    7        72.7844   0.00683787        parabolic\n",
      "    8        34.4419   0.00668019        golden\n",
      "    9        48.1079   0.00661028        parabolic\n",
      "   10        47.7746   0.00661055        parabolic\n",
      "   11         48.809   0.00661005        parabolic\n",
      "   12        51.4519   0.00661333        golden\n",
      "   13        48.4757    0.0066101        parabolic\n",
      "   14        49.1424   0.00661011        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.4315659 1.4315659 1.4315659 1.4315659 1.4315659]\n",
      "Trial error:  1.4315659018694282\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0004667187874911767  :: absSize readout:  0.011709097891987033\n",
      "NRMSE W:  0.011208760020744977  :: absSize W:  0.187279613137202\n",
      "NRMSEs: best:  0.011208760020744977 avg:  0.011208760020744977\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.396332        initial\n",
      "    2        618.034     0.400165        golden\n",
      "    3        236.068     0.378829        golden\n",
      "    4        145.898     0.318403        golden\n",
      "    5        90.1699    0.0173646        golden\n",
      "    6        55.7281    0.0200334        golden\n",
      "    7        73.5866    0.0186065        parabolic\n",
      "    8        111.456     0.164361        golden\n",
      "    9        98.3006    0.0167945        golden\n",
      "   10        103.326    0.0164535        golden\n",
      "   11        106.431    0.0162468        golden\n",
      "   12        108.351     0.125696        golden\n",
      "   13        105.245    0.0163254        golden\n",
      "   14        107.164    0.0161984        golden\n",
      "   15        107.617    0.0161686        golden\n",
      "   16        107.951     0.115867        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.4313086  1.43130858 1.43130857 1.43130857 1.43130854]\n",
      "Trial error:  1.4313085746687617\n",
      "LR:  0.57\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00016494157442045512  :: absSize readout:  0.007176982652060537\n",
      "NRMSE W:  0.010748185734826709  :: absSize W:  0.15867487011006523\n",
      "NRMSEs: best:  0.010748185734826709 avg:  0.010748185734826709\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.351723        initial\n",
      "    2        618.034     0.355917        golden\n",
      "    3        236.068     0.339327        golden\n",
      "    4        145.898      0.31056        golden\n",
      "    5        90.1699    0.0174794        golden\n",
      "    6        55.7281   0.00828394        golden\n",
      "    7        70.5378    0.0100268        parabolic\n",
      "    8        55.3948   0.00825355        parabolic\n",
      "    9        34.2358   0.00600666        golden\n",
      "   10        21.1589    0.0008373        golden\n",
      "   11        13.0769   0.00195688        golden\n",
      "   12        19.8633  0.000566973        parabolic\n",
      "   13        18.4719  0.000272565        parabolic\n",
      "   14        16.4112  0.000526976        golden\n",
      "   15        18.0775   0.00023842        parabolic\n",
      "   16        17.7442  0.000279473        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.21500956 0.21919453 0.24641501 0.30370631 0.30718285]\n",
      "Trial error:  0.2464150108404217\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.000290175246408821  :: absSize readout:  0.007421528931459343\n",
      "NRMSE W:  0.010919657288431374  :: absSize W:  0.1373381863548721\n",
      "NRMSEs: best:  0.010919657288431374 avg:  0.010919657288431374\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.191754        initial\n",
      "    2        618.034     0.211353        golden\n",
      "    3        236.068     0.104545        golden\n",
      "    4        145.898    0.0098339        golden\n",
      "    5        90.1699   0.00829575        golden\n",
      "    6        116.065   0.00911392        parabolic\n",
      "    7        55.7281    0.0061612        golden\n",
      "    8        34.4419   0.00329743        golden\n",
      "    9        21.2862   0.00332832        golden\n",
      "   10        28.1595   0.00326965        parabolic\n",
      "   11        29.0561   0.00327058        parabolic\n",
      "   12        27.7842    0.0032697        parabolic\n",
      "   13        28.4928   0.00326983        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.44261042 1.44152935 1.44252123 1.4426104  1.44261052]\n",
      "Trial error:  1.4426103985706455\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001767491662359591  :: absSize readout:  0.0071175171325008665\n",
      "NRMSE W:  0.010360879057727078  :: absSize W:  0.151839786345246\n",
      "NRMSEs: best:  0.010360879057727078 avg:  0.010360879057727078\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.329521        initial\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    2        618.034     0.338327        golden\n",
      "    3        236.068     0.306267        golden\n",
      "    4        145.898     0.245212        golden\n",
      "    5        90.1699   0.00306108        golden\n",
      "    6        55.7281   0.00336988        golden\n",
      "    7        73.0419   0.00186549        parabolic\n",
      "    8        73.9344   0.00303395        parabolic\n",
      "    9        66.4286   0.00276232        golden\n",
      "   10        70.0875    0.0033029        parabolic\n",
      "   11        72.7085   0.00297157        parabolic\n",
      "   12        73.3828   0.00276556        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.36006013 1.33135991 1.3162352  1.36665611 1.3497312 ]\n",
      "Trial error:  1.3497311959128697\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00026921856029260674  :: absSize readout:  0.007455033431874814\n",
      "NRMSE W:  0.010137108106275926  :: absSize W:  0.1197897576332555\n",
      "NRMSEs: best:  0.010137108106275926 avg:  0.010137108106275926\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.223158        initial\n",
      "    2        618.034     0.226609        golden\n",
      "    3        236.068      0.20412        golden\n",
      "    4        145.898     0.104549        golden\n",
      "    5        90.1699   0.00445299        golden\n",
      "    6        55.7281   0.00633947        golden\n",
      "    7        74.2832   0.00504914        parabolic\n",
      "    8         102.41   0.00383982        parabolic\n",
      "    9        119.021   0.00475738        golden\n",
      "   10        103.151   0.00400416        parabolic\n",
      "   11        97.4851   0.00398137        parabolic\n",
      "   12        100.529   0.00482414        golden\n",
      "   13        101.692   0.00412322        golden\n",
      "   14        102.744   0.00493282        golden\n",
      "   15        102.077   0.00452611        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.15635766 1.20828146 1.19211707 1.21408836 1.12880542]\n",
      "Trial error:  1.1921170681059212\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00020133098680317915  :: absSize readout:  0.007342678990441329\n",
      "NRMSE W:  0.005264911556896557  :: absSize W:  0.14458916456760737\n",
      "NRMSEs: best:  0.005264911556896557 avg:  0.005264911556896557\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.333304        initial\n",
      "    2        618.034     0.340504        golden\n",
      "    3        236.068     0.316154        golden\n",
      "    4        145.898     0.279334        golden\n",
      "    5        90.1699   0.00523632        golden\n",
      "    6        55.7281   0.00386239        golden\n",
      "    7        72.5804   0.00483384        parabolic\n",
      "    8        34.4419   0.00308609        golden\n",
      "    9        12.2418  0.000254277        parabolic\n",
      "   10        20.7215  0.000851132        golden\n",
      "   11        8.03582   0.00044346        parabolic\n",
      "   12        12.6118  0.000238578        parabolic\n",
      "   13        15.7094  0.000143681        golden\n",
      "   14        18.6637  0.000149869        parabolic\n",
      "   15        16.9929  0.000122959        parabolic\n",
      "   16        17.3262  0.000117682        parabolic\n",
      "   17        17.8371  0.000108798        golden\n",
      "   18        18.1704  0.000112302        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.07587512 0.1347742  0.32740914 0.33302035 0.13178414]\n",
      "Trial error:  0.13477420391377842\n",
      "LR:  0.6399999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00018679595388224578  :: absSize readout:  0.0055749973314713035\n",
      "NRMSE W:  0.009272495708927981  :: absSize W:  0.14168840494926885\n",
      "NRMSEs: best:  0.009272495708927981 avg:  0.009272495708927981\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.100899        initial\n",
      "    2        618.034     0.170258        golden\n",
      "    3        236.068    0.0319238        golden\n",
      "    4        145.898    0.0106793        golden\n",
      "    5        73.7203   0.00710618        parabolic\n",
      "    6        88.2161   0.00736201        parabolic\n",
      "    7        64.9906   0.00462016        parabolic\n",
      "    8        40.1664   0.00268671        golden\n",
      "    9        46.2627   0.00402745        parabolic\n",
      "   10        24.8242   0.00320251        golden\n",
      "   11        33.9166   0.00364758        parabolic\n",
      "   12        37.7792   0.00277285        golden\n",
      "   13         42.495   0.00359023        golden\n",
      "   14        39.2546   0.00302515        golden\n",
      "   15        41.0558   0.00343866        golden\n",
      "   16        39.8181   0.00340587        golden\n",
      "   17        40.5061   0.00344234        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.02727729 1.38131826 1.01859206 1.02015822 1.09051219]\n",
      "Trial error:  1.0272772922039972\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001625361640979151  :: absSize readout:  0.005197375134623445\n",
      "NRMSE W:  0.00935568430159486  :: absSize W:  0.12899358021770638\n",
      "NRMSEs: best:  0.00935568430159486 avg:  0.00935568430159486\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.322357        initial\n",
      "    2        618.034     0.331742        golden\n",
      "    3        236.068     0.300113        golden\n",
      "    4        145.898     0.011027        golden\n",
      "    5        90.1699   0.00157567        golden\n",
      "    6        113.959   0.00945738        parabolic\n",
      "    7        55.7281   0.00214294        golden\n",
      "    8        74.3279   0.00172415        parabolic\n",
      "    9        94.5297   0.00635313        parabolic\n",
      "   10        89.8366   0.00157512        parabolic\n",
      "   11        88.8503   0.00158093        parabolic\n",
      "   12        89.5033   0.00566688        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43587203 1.43528902 1.4279533  1.43063795 1.43005819]\n",
      "Trial error:  1.4306379533423157\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00014400750599646367  :: absSize readout:  0.005057419285843986\n",
      "NRMSE W:  0.00971343771694812  :: absSize W:  0.13154044943688878\n",
      "NRMSEs: best:  0.00971343771694812 avg:  0.00971343771694812\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.145872        initial\n",
      "    2        618.034     0.236059        golden\n",
      "    3        236.068    0.0227938        golden\n",
      "    4        145.898   0.00880148        golden\n",
      "    5        164.377    0.0101767        parabolic\n",
      "    6        122.102    0.0074962        parabolic\n",
      "    7         75.463   0.00398285        golden\n",
      "    8        46.6387   0.00416985        golden\n",
      "    9        64.0426   0.00311814        parabolic\n",
      "   10        61.7376   0.00402967        parabolic\n",
      "   11          68.65   0.00401954        parabolic\n",
      "   12        65.8025   0.00403608        golden\n",
      "   13        63.1622   0.00344773        golden\n",
      "   14        64.7148    0.0040319        golden\n",
      "   15        63.7063   0.00404857        golden\n",
      "   16         64.376   0.00404457        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43797972 1.35021801 1.29400221 1.44222344 1.20033165]\n",
      "Trial error:  1.3502180130311259\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0002226918208236265  :: absSize readout:  0.005662592356603391\n",
      "NRMSE W:  0.009380871812896362  :: absSize W:  0.11739646238301919\n",
      "NRMSEs: best:  0.009380871812896362 avg:  0.009380871812896362\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0263905        initial\n",
      "    2        618.034    0.0139112        golden\n",
      "    3        763.932   0.00787462        golden\n",
      "    4        854.102   0.00889992        golden\n",
      "    5        783.572    0.0074401        parabolic\n",
      "    6        797.046   0.00890195        parabolic\n",
      "    7         776.07   0.00733326        golden\n",
      "    8        777.444   0.00834586        parabolic\n",
      "    9        771.434   0.00758214        golden\n",
      "   10        774.299   0.00751814        golden\n",
      "   11        775.394   0.00783354        golden\n",
      "   12        776.595   0.00807673        golden\n",
      "   13        775.737   0.00764043        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.82036653 0.65428967 0.68583362 0.81970803 0.8201286 ]\n",
      "Trial error:  0.8197080263423281\n",
      "Trial:  4\n",
      "0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00016735230713895918  :: absSize readout:  0.0051583776425290565\n",
      "NRMSE W:  0.00957661441959672  :: absSize W:  0.1256464152317106\n",
      "NRMSEs: best:  0.00957661441959672 avg:  0.00957661441959672\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0126999        initial\n",
      "    2        618.034   0.00991286        golden\n",
      "    3        763.932    0.0105704        golden\n",
      "    4        638.219   0.00821786        parabolic\n",
      "    5        687.781    0.0110266        parabolic\n",
      "    6         657.15   0.00915355        golden\n",
      "    7        640.438   0.00914617        parabolic\n",
      "    8        630.509    0.0110749        golden\n",
      "    9        635.274   0.00920591        golden\n",
      "   10        637.094   0.00984139        golden\n",
      "   11        639.066    0.0100931        golden\n",
      "   12        637.789   0.00887855        golden\n",
      "   13        638.552   0.00798964        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.23914903 1.37066132 1.1996996  1.15066627 1.32944657]\n",
      "Trial error:  1.2391490324764842\n",
      "LR:  0.71\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00011519708857982791  :: absSize readout:  0.004512266603036878\n",
      "NRMSE W:  0.008874303281280334  :: absSize W:  0.12069748201314631\n",
      "NRMSEs: best:  0.008874303281280334 avg:  0.008874303281280334\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00987374        initial\n",
      "    2        618.034   0.00830992        golden\n",
      "    3        763.932   0.00963656        golden\n",
      "    4        580.494    0.0121974        parabolic\n",
      "    5        618.367    0.0149467        parabolic\n",
      "    6        603.695   0.00775956        golden\n",
      "    7        594.833   0.00785233        golden\n",
      "    8         601.75    0.0104728        parabolic\n",
      "    9        604.028   0.00704775        parabolic\n",
      "   10        609.378   0.00743916        golden\n",
      "   11        606.609   0.00866588        parabolic\n",
      "   12        605.014   0.00661034        golden\n",
      "   13        605.623    0.0111879        golden\n",
      "   14        604.638   0.00974917        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.33853197 1.35651544 1.34926404 1.35469349 1.35858278]\n",
      "Trial error:  1.3546934884174187\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00028135916168243053  :: absSize readout:  0.005598579790056487\n",
      "NRMSE W:  0.008350564132193991  :: absSize W:  0.11689779930466897\n",
      "NRMSEs: best:  0.008350564132193991 avg:  0.008350564132193991\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.137511        initial\n",
      "    2        618.034     0.130596        golden\n",
      "    3        763.932     0.126336        golden\n",
      "    4        854.102     0.123817        golden\n",
      "    5         909.83     0.122315        golden\n",
      "    6        944.272     0.121398        golden\n",
      "    7        965.558     0.120827        golden\n",
      "    8        978.714     0.120483        golden\n",
      "    9        986.844     0.120283        golden\n",
      "   10        991.869     0.120153        golden\n",
      "   11        994.975     0.120072        golden\n",
      "   12        996.894     0.120022        golden\n",
      "   13        998.081     0.119981        golden\n",
      "   14        998.814     0.119973        golden\n",
      "   15        999.147     0.119953        parabolic\n",
      "   16         999.48     0.119945        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43961946 1.43961946 1.43961946 1.43961946 1.43961946]\n",
      "Trial error:  1.4396194631793071\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0002406864930147713  :: absSize readout:  0.005274744780499505\n",
      "NRMSE W:  0.009674393326899313  :: absSize W:  0.12291573665125241\n",
      "NRMSEs: best:  0.009674393326899313 avg:  0.009674393326899313\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0118205        initial\n",
      "    2        618.034    0.0671597        golden\n",
      "    3        236.068    0.0100489        golden\n",
      "    4        298.584    0.0134425        parabolic\n",
      "    5        145.898   0.00893968        golden\n",
      "    6        90.1699   0.00918635        golden\n",
      "    7        137.337    0.0082662        parabolic\n",
      "    8         119.29     0.007954        parabolic\n",
      "    9        108.167   0.00899463        golden\n",
      "   10        124.563   0.00856845        parabolic\n",
      "   11        115.042   0.00871607        golden\n",
      "   12        121.304   0.00949617        golden\n",
      "   13        117.667   0.00925606        golden\n",
      "   14         120.06   0.00780407        golden\n",
      "   15        120.535   0.00969763        golden\n",
      "   16        119.726   0.00781778        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.27124758 1.2810928  1.29230242 1.34899898 1.38319258]\n",
      "Trial error:  1.2923024205538336\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001217413945513585  :: absSize readout:  0.004514533153706663\n",
      "NRMSE W:  0.00869418550359465  :: absSize W:  0.11568670147232595\n",
      "NRMSEs: best:  0.00869418550359465 avg:  0.00869418550359465\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.274262        initial\n",
      "    2        618.034     0.273365        golden\n",
      "    3        763.932     0.271525        golden\n",
      "    4        854.102     0.270204        golden\n",
      "    5         909.83     0.269344        golden\n",
      "    6        944.272     0.268803        golden\n",
      "    7        965.558      0.26847        golden\n",
      "    8        978.714     0.268261        golden\n",
      "    9        986.844     0.268127        golden\n",
      "   10        991.869     0.268047        golden\n",
      "   11        994.975     0.267997        golden\n",
      "   12        996.894     0.267971        golden\n",
      "   13        998.081     0.267952        golden\n",
      "   14        998.814      0.26794        golden\n",
      "   15        999.267     0.267933        golden\n",
      "   16          999.6     0.267927        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43932241 1.43932241 1.43932241 1.43932241 1.43932241]\n",
      "Trial error:  1.4393224085629204\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00012176471039989074  :: absSize readout:  0.0044983292533184675\n",
      "NRMSE W:  0.008209432051604968  :: absSize W:  0.1286776316058358\n",
      "NRMSEs: best:  0.008209432051604968 avg:  0.008209432051604968\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0105533        initial\n",
      "    2        618.034    0.0351575        golden\n",
      "    3        236.068    0.0114141        golden\n",
      "    4        319.248   0.00878509        parabolic\n",
      "    5        316.214    0.0114595        parabolic\n",
      "    6        343.204    0.0121065        golden\n",
      "    7        328.399    0.0129246        golden\n",
      "    8        322.743    0.0104591        golden\n",
      "    9        320.583   0.00988398        golden\n",
      "   10        318.089    0.0108169        golden\n",
      "   11        319.758    0.0135453        golden\n",
      "   12        318.806   0.00900984        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.39399183 1.23540963 1.3884784  1.37924236 1.31193339]\n",
      "Trial error:  1.3792423635394504\n",
      "LR:  0.7799999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00014026892103613858  :: absSize readout:  0.004166565641040514\n",
      "NRMSE W:  0.008603626188857218  :: absSize W:  0.11939778459903319\n",
      "NRMSEs: best:  0.008603626188857218 avg:  0.008603626188857218\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00576903        initial\n",
      "    2        618.034   0.00742504        golden\n",
      "    3        236.068   0.00539404        golden\n",
      "    4         198.58   0.00272459        parabolic\n",
      "    5        122.729   0.00272872        golden\n",
      "    6        160.698   0.00329113        parabolic\n",
      "    7        184.111   0.00379656        golden\n",
      "    8        212.899   0.00311508        golden\n",
      "    9        193.054    0.0041027        golden\n",
      "   10         204.05   0.00326163        golden\n",
      "   11        196.469   0.00445996        golden\n",
      "   12         200.67   0.00344313        golden\n",
      "   13        197.774   0.00285911        golden\n",
      "   14        199.378   0.00334332        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   15        198.247   0.00238629        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.34821017 0.67998837 0.51600202 0.39568011 0.70054742]\n",
      "Trial error:  0.5160020163845417\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00015101209443348345  :: absSize readout:  0.004020109057604318\n",
      "NRMSE W:  0.008945669278045246  :: absSize W:  0.12036003349877823\n",
      "NRMSEs: best:  0.008945669278045246 avg:  0.008945669278045246\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00326273        initial\n",
      "    2        618.034   0.00197308        golden\n",
      "    3        763.932    0.0021956        golden\n",
      "    4        649.302   0.00185543        parabolic\n",
      "    5        674.452   0.00184498        parabolic\n",
      "    6        665.381   0.00208658        parabolic\n",
      "    7        674.785   0.00194049        parabolic\n",
      "    8        670.987   0.00195471        golden\n",
      "    9        673.128   0.00194666        golden\n",
      "   10        673.946   0.00193328        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.39705763 1.39287814 1.39665705 1.41026087 1.39862471]\n",
      "Trial error:  1.3970576267121735\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00012008470500110696  :: absSize readout:  0.0038940857581458888\n",
      "NRMSE W:  0.008173670281692084  :: absSize W:  0.13389409428938717\n",
      "NRMSEs: best:  0.008173670281692084 avg:  0.008173670281692084\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.130704        initial\n",
      "    2        618.034     0.154116        golden\n",
      "    3        236.068    0.0177501        golden\n",
      "    4        145.898   0.00520163        golden\n",
      "    5        165.116   0.00569614        parabolic\n",
      "    6         147.46   0.00553369        parabolic\n",
      "    7        90.1699   0.00393685        golden\n",
      "    8         114.61   0.00502962        parabolic\n",
      "    9        55.7281   0.00385127        golden\n",
      "   10        71.2166   0.00350468        parabolic\n",
      "   11        72.0019   0.00165754        parabolic\n",
      "   12        78.9415   0.00368024        golden\n",
      "   13        75.0458   0.00402664        parabolic\n",
      "   14        72.3352   0.00399891        parabolic\n",
      "   15        71.6685   0.00282423        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.16427598 0.67594193 0.64741143 0.74690802 0.94206194]\n",
      "Trial error:  0.7469080177916322\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00010091128000526435  :: absSize readout:  0.004095213731040206\n",
      "NRMSE W:  0.008451999533139692  :: absSize W:  0.11778882258549395\n",
      "NRMSEs: best:  0.008451999533139692 avg:  0.008451999533139692\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0477505        initial\n",
      "    2        618.034    0.0442547        golden\n",
      "    3        763.932   0.00320794        golden\n",
      "    4        854.102    0.0532165        golden\n",
      "    5        708.204    0.0431198        golden\n",
      "    6        798.374   0.00316758        golden\n",
      "    7        781.227    0.0453546        parabolic\n",
      "    8         819.66   0.00314199        golden\n",
      "    9        832.816   0.00312598        golden\n",
      "   10        840.946    0.0522187        golden\n",
      "   11        827.791    0.0488581        golden\n",
      "   12        835.921    0.0494936        golden\n",
      "   13        830.896    0.0485233        golden\n",
      "   14        834.002    0.0485871        golden\n",
      "   15        832.083     0.048728        golden\n",
      "   16        833.269    0.0485092        golden\n",
      "   17        832.482   0.00312638        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.42180847 1.42246896 1.42190777 1.42606003 1.42256177]\n",
      "Trial error:  1.4224689593976756\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0002429298957019462  :: absSize readout:  0.004768232725216895\n",
      "NRMSE W:  0.008932988067089667  :: absSize W:  0.11713891937161922\n",
      "NRMSEs: best:  0.008932988067089667 avg:  0.008932988067089667\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00653063        initial\n",
      "    2        618.034    0.0143104        golden\n",
      "    3        236.068   0.00743828        golden\n",
      "    4        339.344   0.00678685        parabolic\n",
      "    5        472.136   0.00871778        golden\n",
      "    6        416.408   0.00692629        golden\n",
      "    7        373.892    0.0066075        parabolic\n",
      "    8        395.122   0.00639377        golden\n",
      "    9        403.252   0.00618821        golden\n",
      "   10        408.277   0.00623559        golden\n",
      "   11        403.978   0.00589429        parabolic\n",
      "   12        405.716   0.00655006        parabolic\n",
      "   13        404.642   0.00636002        golden\n",
      "   14        403.645   0.00635973        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43488774 1.43319523 1.43818235 1.39324059 1.4336309 ]\n",
      "Trial error:  1.4336308997822678\n",
      "LR:  0.8499999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  7.096034252104837e-05  :: absSize readout:  0.0034770412279587276\n",
      "NRMSE W:  0.006612145493236742  :: absSize W:  0.12890981062336113\n",
      "NRMSEs: best:  0.006612145493236742 avg:  0.006612145493236742\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00526496        initial\n",
      "    2        618.034    0.0422751        golden\n",
      "    3        236.068   0.00489927        golden\n",
      "    4        305.914   0.00471812        parabolic\n",
      "    5        290.328   0.00618052        parabolic\n",
      "    6        306.247   0.00582682        parabolic\n",
      "    7        299.961   0.00747948        golden\n",
      "    8         303.64   0.00689798        golden\n",
      "    9        305.045   0.00573962        golden\n",
      "   10        305.581   0.00742254        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.60151475 0.60748366 0.27031358 0.38571315 0.48563334]\n",
      "Trial error:  0.48563334136769815\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  9.694890385829292e-05  :: absSize readout:  0.003595017622077955\n",
      "NRMSE W:  0.007872230384258504  :: absSize W:  0.12709344594857547\n",
      "NRMSEs: best:  0.007872230384258504 avg:  0.007872230384258504\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00608811        initial\n",
      "    2        618.034    0.0192811        golden\n",
      "    3        236.068   0.00602958        golden\n",
      "    4        307.636   0.00692218        parabolic\n",
      "    5        145.898   0.00536572        golden\n",
      "    6        90.1699   0.00952244        golden\n",
      "    7         180.34   0.00749077        golden\n",
      "    8        124.612   0.00798712        golden\n",
      "    9        159.054   0.00755151        golden\n",
      "   10        137.767   0.00709506        golden\n",
      "   11        150.923   0.00592903        golden\n",
      "   12        142.792     0.004764        golden\n",
      "   13        140.873   0.00456838        golden\n",
      "   14        139.044   0.00460761        parabolic\n",
      "   15        140.285   0.00559745        parabolic\n",
      "   16        141.606   0.00492019        golden\n",
      "   17        141.206   0.00576858        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.37340785 1.39216    1.38006507 1.39203448 1.39298578]\n",
      "Trial error:  1.392034482675121\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00016757499924504452  :: absSize readout:  0.004089070069439445\n",
      "NRMSE W:  0.00911135028360477  :: absSize W:  0.10745378983416046\n",
      "NRMSEs: best:  0.00911135028360477 avg:  0.00911135028360477\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00585869        initial\n",
      "    2        618.034    0.0219565        golden\n",
      "    3        236.068   0.00448302        golden\n",
      "    4        278.372   0.00437304        parabolic\n",
      "    5        268.414   0.00402708        parabolic\n",
      "    6        258.346   0.00446642        parabolic\n",
      "    7        268.954    0.0042933        parabolic\n",
      "    8        264.568   0.00466683        golden\n",
      "    9        266.945   0.00469313        golden\n",
      "   10        267.853   0.00426005        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1.30088477 1.18392311 1.26440451 0.89555014 1.19624954]\n",
      "Trial error:  1.19624953647514\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  8.056342295856575e-05  :: absSize readout:  0.00346953485292474\n",
      "NRMSE W:  0.007321930954486217  :: absSize W:  0.12710562473511897\n",
      "NRMSEs: best:  0.007321930954486217 avg:  0.007321930954486217\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00844517        initial\n",
      "    2        618.034    0.0704853        golden\n",
      "    3        236.068   0.00504124        golden\n",
      "    4        290.411   0.00565787        parabolic\n",
      "    5        219.893   0.00602366        parabolic\n",
      "    6        257.689   0.00740634        parabolic\n",
      "    7        244.327    0.0062073        golden\n",
      "    8         229.89   0.00511679        golden\n",
      "    9        239.222   0.00646159        golden\n",
      "   10        233.708    0.0059765        golden\n",
      "   11        237.273   0.00508842        golden\n",
      "   12        235.167   0.00423968        golden\n",
      "   13         234.61   0.00458281        golden\n",
      "   14          235.5   0.00709504        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.30685854 1.12095894 1.23423813 1.31665719 1.33626948]\n",
      "Trial error:  1.3068585449301693\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00011834756163759254  :: absSize readout:  0.003903423935438605\n",
      "NRMSE W:  0.0075555613844554715  :: absSize W:  0.10691153841851936\n",
      "NRMSEs: best:  0.0075555613844554715 avg:  0.0075555613844554715\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0315635        initial\n",
      "    2        618.034    0.0345504        golden\n",
      "    3        236.068    0.0301913        golden\n",
      "    4        145.898    0.0239179        golden\n",
      "    5        90.1699    0.0125073        golden\n",
      "    6        55.7281   0.00805737        golden\n",
      "    7        34.4419  0.000160847        golden\n",
      "    8        21.2862  0.000278156        golden\n",
      "    9        28.2683  0.000210503        parabolic\n",
      "   10        42.5725   0.00475499        golden\n",
      "   11        37.5475  0.000145284        golden\n",
      "   12        39.4669    0.0043919        golden\n",
      "   13        36.3612  0.000154362        golden\n",
      "   14        38.2806  0.000148165        golden\n",
      "   15        37.8808  0.000149293        parabolic\n",
      "   16        37.0944  0.000148379        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.34521712 0.35264088 0.1499957  0.16205558 0.27741071]\n",
      "Trial error:  0.27741071340413004\n",
      "LR:  0.9199999999999999\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001252218123733305  :: absSize readout:  0.0038327066858632364\n",
      "NRMSE W:  0.007681552453769267  :: absSize W:  0.1192997782314407\n",
      "NRMSEs: best:  0.007681552453769267 avg:  0.007681552453769267\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0107949        initial\n",
      "    2        618.034    0.0217284        golden\n",
      "    3        236.068  0.000235309        golden\n",
      "    4        145.898  0.000105454        golden\n",
      "    5        188.587   0.00013424        parabolic\n",
      "    6        146.338  7.99029e-05        parabolic\n",
      "    7            167  6.08627e-05        parabolic\n",
      "    8        175.245  0.000105197        golden\n",
      "    9        158.784   5.1554e-05        parabolic\n",
      "   10         159.46  5.81885e-05        parabolic\n",
      "   11         154.03  0.000100434        golden\n",
      "   12        156.968  5.23129e-05        golden\n",
      "   13        157.927  8.24704e-05        parabolic\n",
      "   14         158.45   7.8811e-05        parabolic\n",
      "   15        159.117  0.000112917        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.17075238 0.12491394 0.11488183 0.1163844  0.16104464]\n",
      "Trial error:  0.12491394239412189\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00048290760210006596  :: absSize readout:  0.005151910022686638\n",
      "NRMSE W:  0.009220741585405399  :: absSize W:  0.08903299127688033\n",
      "NRMSEs: best:  0.009220741585405399 avg:  0.009220741585405399\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00490905        initial\n",
      "    2        618.034     0.019563        golden\n",
      "    3        236.068   0.00448296        golden\n",
      "    4        299.588   0.00479791        parabolic\n",
      "    5        145.898   0.00565429        golden\n",
      "    6        201.626    0.0051177        golden\n",
      "    7        260.331    0.0044155        golden\n",
      "    8        255.252   0.00455078        parabolic\n",
      "    9        275.326   0.00418183        golden\n",
      "   10        284.593    0.0046971        golden\n",
      "   11        269.598   0.00432327        golden\n",
      "   12        278.866     0.004365        golden\n",
      "   13        273.959   0.00464274        parabolic\n",
      "   14        275.659   0.00467877        parabolic\n",
      "   15        274.804   0.00451425        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.32359759 1.34896233 1.27119697 1.31381435 1.20998433]\n",
      "Trial error:  1.3138143510872313\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  8.090201957200227e-05  :: absSize readout:  0.0033029224261760352\n",
      "NRMSE W:  0.005811263578312783  :: absSize W:  0.1276463915477651\n",
      "NRMSEs: best:  0.005811263578312783 avg:  0.005811263578312783\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00816026        initial\n",
      "    2        618.034    0.0924006        golden\n",
      "    3        236.068   0.00427092        golden\n",
      "    4        293.598   0.00515316        parabolic\n",
      "    5         204.99   0.00708637        parabolic\n",
      "    6        258.419   0.00645485        parabolic\n",
      "    7        224.197   0.00756005        golden\n",
      "    8        244.605   0.00812295        golden\n",
      "    9        231.534    0.0043618        golden\n",
      "   10        239.329   0.00632409        golden\n",
      "   11        234.336   0.00688407        golden\n",
      "   12        237.314    0.0056823        golden\n",
      "   13        235.406   0.00702775        golden\n",
      "   14        236.544   0.00591685        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[0.53074083 0.48072388 0.58268918 0.3807764  0.42548352]\n",
      "Trial error:  0.48072387895147767\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  6.423863660326216e-05  :: absSize readout:  0.0031008893175098755\n",
      "NRMSE W:  0.00816985629652549  :: absSize W:  0.12184103598121603\n",
      "NRMSEs: best:  0.00816985629652549 avg:  0.00816985629652549\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00740037        initial\n",
      "    2        618.034    0.0904325        golden\n",
      "    3        236.068    0.0111023        golden\n",
      "    4        321.867   0.00798105        parabolic\n",
      "    5        378.299   0.00781896        parabolic\n",
      "    6        472.136   0.00201826        golden\n",
      "    7        478.473   0.00194751        parabolic\n",
      "    8        486.406   0.00191239        parabolic\n",
      "    9        487.128   0.00193589        parabolic\n",
      "   10        482.958   0.00194227        parabolic\n",
      "   11        485.089   0.00196141        golden\n",
      "   12        486.073   0.00195549        parabolic\n",
      "   13         486.74   0.00192913        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.24802708 1.43464033 1.43358472 1.43576821 1.43556458]\n",
      "Trial error:  1.434640331286656\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0009387734854882233  :: absSize readout:  0.005812758664179269\n",
      "NRMSE W:  0.008464863247503838  :: absSize W:  0.08949255461250381\n",
      "NRMSEs: best:  0.008464863247503838 avg:  0.008464863247503838\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966     0.021519        initial\n",
      "    2        618.034    0.0128488        golden\n",
      "    3        763.932    0.0094807        golden\n",
      "    4        854.102   0.00795206        golden\n",
      "    5         909.83   0.00672767        golden\n",
      "    6        944.272   0.00661103        golden\n",
      "    7        935.267   0.00644333        parabolic\n",
      "    8        929.008   0.00651216        parabolic\n",
      "    9        934.933   0.00644618        parabolic\n",
      "   10        938.706   0.00661537        golden\n",
      "   11         936.58   0.00642796        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   12        937.392   0.00641557        golden\n",
      "   13        937.894   0.00641068        golden\n",
      "   14        938.228   0.00647827        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43386607 1.43499507 1.43525563 1.43545628 1.43494089]\n",
      "Trial error:  1.4349950737642483\n",
      "LR:  0.99\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  5.603897223840835e-05  :: absSize readout:  0.0029441624842658993\n",
      "NRMSE W:  0.0070199692557288565  :: absSize W:  0.11618009890029089\n",
      "NRMSEs: best:  0.0070199692557288565 avg:  0.0070199692557288565\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966   0.00708559        initial\n",
      "    2        618.034   0.00165483        golden\n",
      "    3        763.932    0.0193042        golden\n",
      "    4        530.516   0.00168091        parabolic\n",
      "    5         575.25   0.00164409        parabolic\n",
      "    6        586.417   0.00164404        parabolic\n",
      "    7         581.15    0.0016437        parabolic\n",
      "    8        580.816   0.00164379        parabolic\n",
      "    9        583.261   0.00164386        parabolic\n",
      "   10        581.956   0.00164371        golden\n",
      "   11        581.483    0.0016438        parabolic\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.4409584 1.4409584 1.4409584 1.4409584 1.4409584]\n",
      "Trial error:  1.4409584044038504\n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0001456322243603069  :: absSize readout:  0.003814469499826482\n",
      "NRMSE W:  0.005269390495131075  :: absSize W:  0.12213996668957879\n",
      "NRMSEs: best:  0.005269390495131075 avg:  0.005269390495131075\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0316218        initial\n",
      "    2        618.034    0.0605423        golden\n",
      "    3        236.068   0.00525895        golden\n",
      "    4        145.898   0.00475762        golden\n",
      "    5        187.236   0.00494799        parabolic\n",
      "    6        90.1699   0.00533962        golden\n",
      "    7        124.612   0.00532495        golden\n",
      "    8        161.688   0.00503531        golden\n",
      "    9        137.767   0.00457751        golden\n",
      "   10        132.742   0.00548031        golden\n",
      "   11        140.873   0.00439537        golden\n",
      "   12        141.206   0.00473415        parabolic\n",
      "   13        139.414   0.00504187        parabolic\n",
      "   14        140.316   0.00483921        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.10322861 1.02009    1.16165061 1.23987144 1.11248008]\n",
      "Trial error:  1.1124800836087423\n",
      "Trial:  2\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.007214743618723499  :: absSize readout:  0.005524870315682251\n",
      "NRMSE W:  0.01827716622826285  :: absSize W:  0.030121834184850343\n",
      "NRMSEs: best:  0.01827716622826285 avg:  0.01827716622826285\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  1.42175e-05        initial\n",
      "    2        618.034   2.4132e-06        golden\n",
      "    3        763.932  1.07054e-06        golden\n",
      "    4         734.06  1.26228e-06        parabolic\n",
      "    5        854.102  7.22116e-07        golden\n",
      "    6         899.81  5.81859e-07        parabolic\n",
      "    7        938.079  4.83252e-07        golden\n",
      "    8        961.731   4.2107e-07        golden\n",
      "    9        976.348  3.98376e-07        golden\n",
      "   10        985.383  3.87141e-07        golden\n",
      "   11        990.966   3.8932e-07        golden\n",
      "   12        986.428  4.02907e-07        parabolic\n",
      "   13        985.049  3.91065e-07        parabolic\n",
      "   14        985.782  3.65742e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.06657735 1.25732745 1.30212759 1.21211059 1.28555982]\n",
      "Trial error:  1.2573274484959003\n",
      "Trial:  3\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.0003128530914381755  :: absSize readout:  0.004059870249930258\n",
      "NRMSE W:  0.00630171137768689  :: absSize W:  0.08788275149517279\n",
      "NRMSEs: best:  0.00630171137768689 avg:  0.00630171137768689\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0267985        initial\n",
      "    2        618.034    0.0350196        golden\n",
      "    3        236.068   0.00476294        golden\n",
      "    4        145.898   0.00523783        golden\n",
      "    5         194.96    0.0160711        parabolic\n",
      "    6        291.796    0.0135981        golden\n",
      "    7        257.354   0.00466059        golden\n",
      "    8         270.51   0.00467562        golden\n",
      "    9        260.626   0.00463328        parabolic\n",
      "   10        263.337   0.00467616        parabolic\n",
      "   11        259.376   0.00467602        golden\n",
      "   12        261.661   0.00468619        golden\n",
      "   13        260.148   0.00474094        golden\n",
      "   14        261.021   0.00699034        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.33486892 1.36061418 1.35916874 1.32873968 1.33009099]\n",
      "Trial error:  1.3348689219917826\n",
      "Trial:  4\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.00029902656957675916  :: absSize readout:  0.004843746235993506\n",
      "NRMSE W:  0.00597061108610908  :: absSize W:  0.11255752075321046\n",
      "NRMSEs: best:  0.00597061108610908 avg:  0.00597061108610908\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966    0.0482066        initial\n",
      "    2        618.034    0.0619466        golden\n",
      "    3        236.068    0.0049311        golden\n",
      "    4        145.898   0.00562089        golden\n",
      "    5        193.951   0.00514703        parabolic\n",
      "    6        263.829   0.00507021        parabolic\n",
      "    7        232.679   0.00475409        parabolic\n",
      "    8        217.886   0.00492642        golden\n",
      "    9         226.94   0.00469399        parabolic\n",
      "   10        227.666   0.00490073        parabolic\n",
      "   11        223.482   0.00478429        golden\n",
      "   12        225.619   0.00495057        golden\n",
      "   13        226.436   0.00530196        golden\n",
      "   14        227.274   0.00501315        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.29309123 1.28101437 1.32607424 1.06387515 1.29203948]\n",
      "Trial error:  1.2920394821427719\n",
      "Best error:  0.09484397311932018  best LR:  0.42999999999999994\n",
      "0.42999999999999994\n",
      "testing at  80.0  Hz\n",
      "Error:  0.37602095176636074\n",
      "testing at  84.75704754874363  Hz\n",
      "Error:  0.36290325245069516\n",
      "testing at  89.79696386474984  Hz\n",
      "Error:  0.3031714112628761\n",
      "testing at  95.13656920021768  Hz\n",
      "Error:  0.3249902373444835\n",
      "testing at  100.79368399158986  Hz\n",
      "Error:  0.35720411952674563\n",
      "testing at  106.78718833360276  Hz\n",
      "Error:  0.25227559181258086\n",
      "testing at  113.13708498984761  Hz\n",
      "Error:  0.22855752173347493\n",
      "testing at  119.86456615013452  Hz\n",
      "Error:  0.34101203832445687\n",
      "testing at  126.99208415745595  Hz\n",
      "Error:  0.3474762367406137\n",
      "testing at  134.54342644059432  Hz\n",
      "Error:  0.19850777174661266\n",
      "testing at  142.5437949024543  Hz\n",
      "Error:  0.28293091773298307\n",
      "testing at  151.01989002907095  Hz\n",
      "Error:  0.14163058419096358\n",
      "testing at  160.0  Hz\n",
      "Error:  0.08434794076913865\n",
      "testing at  169.51409509748726  Hz\n",
      "Error:  0.2958614630655602\n",
      "testing at  179.59392772949968  Hz\n",
      "Error:  0.3859922835423014\n",
      "testing at  190.27313840043536  Hz\n",
      "Error:  0.6182992054051918\n",
      "testing at  201.58736798317972  Hz\n",
      "Error:  1.2183936643857245\n",
      "testing at  213.5743766672055  Hz\n",
      "Error:  1.222291339539515\n",
      "testing at  226.27416997969522  Hz\n",
      "Error:  1.3339901017354105\n",
      "testing at  239.72913230026904  Hz\n",
      "Error:  1.3898367068560262\n",
      "testing at  253.9841683149119  Hz\n",
      "Error:  1.3593844238506818\n",
      "testing at  269.08685288118863  Hz\n",
      "Error:  1.4109033938862232\n",
      "testing at  285.0875898049086  Hz\n",
      "Error:  1.420017648664127\n",
      "testing at  302.0397800581419  Hz\n",
      "Error:  1.425007669088278\n",
      "testing at  320.0  Hz\n",
      "Error:  1.4298910727413703\n",
      "====== running test:  320 <function bl_sawtooth at 0x7fc012258e18>\n",
      "LR:  0.01\n",
      "Trial:  0\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.5530695074921959  :: absSize readout:  1.4009896370613002\n",
      "NRMSE W:  0.09241929418532077  :: absSize W:  0.7033914376202037\n",
      "NRMSEs: best:  0.09241929418532077 avg:  0.09241929418532077\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  1.83384e-07        initial\n",
      "    2        618.034  2.23299e-07        golden\n",
      "    3        236.068  1.95532e-07        golden\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    4        372.032  2.19766e-07        parabolic\n",
      "    5        472.136  2.72803e-07        golden\n",
      "    6        416.408   2.2627e-07        golden\n",
      "    7        395.122  1.83591e-07        golden\n",
      "    8        375.951  2.54683e-07        parabolic\n",
      "    9        382.299  2.40081e-07        parabolic\n",
      "   10        379.668  2.43141e-07        golden\n",
      "   11        381.088  2.05488e-07        golden\n",
      "   12        381.631  2.68312e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43821233 1.43798277 1.43773536 1.43794535 1.43787436]\n",
      "Trial error:  1.4379453496737788\n",
      "Best so far ***************** \n",
      "Trial:  1\n",
      "0\n",
      "Loading patterns: 0  \n",
      "NRMSE readout:  0.539466250988668  :: absSize readout:  1.39232064259841\n",
      "NRMSE W:  0.10283482247293489  :: absSize W:  0.707347926399247\n",
      "NRMSEs: best:  0.10283482247293489 avg:  0.10283482247293489\n",
      " \n",
      " Func-count     x          f(x)          Procedure\n",
      "    1        381.966  5.62931e-07        initial\n",
      "    2        618.034  4.89843e-07        golden\n",
      "    3        763.932  3.86617e-07        golden\n",
      "    4        854.102  4.82358e-07        golden\n",
      "    5        738.183  5.55815e-07        parabolic\n",
      "    6        764.265  5.51259e-07        parabolic\n",
      "    7        754.097  5.00627e-07        golden\n",
      "    8        760.175  5.80288e-07        golden\n",
      "    9        762.497  5.80591e-07        golden\n",
      "   10        763.384   5.3085e-07        golden\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  1 )\n",
      "[1.43866835 1.43860861 1.43845149 1.438614   1.43877475]\n",
      "Trial error:  1.4386140041020945\n",
      "Trial:  2\n",
      "0\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "sr = 8000\n",
    "testLen = 1024\n",
    "windowSize = testLen\n",
    "\n",
    "\n",
    "def evalModel(modelSize, lr, testPattern):\n",
    "    params = {'N':modelSize, 'NetSR':1.5, 'NetinpScaling':1,'BiasScaling':0.5,'TychonovAlpha':0.001,\n",
    "             'washoutLength':windowSize*1, 'learnLength':windowSize * 1, 'TychonovAlphaReadout':0.001,\n",
    "              'LR': lr,\n",
    "              'patts':testPattern\n",
    "             }\n",
    "\n",
    "    sampleSize = 1\n",
    "    loadedNets = np.zeros(sampleSize, dtype=np.object)\n",
    "    for i_net, net in enumerate(loadedNets):\n",
    "        print(i_net)\n",
    "        loadedNets[i_net] = makeLoadedNetwork_v2(params)\n",
    "    nrmses = [x['NRMSE_W'] for x in loadedNets]\n",
    "    bestNetwork = np.argmin(nrmses)\n",
    "    print(\"NRMSEs: best: \", loadedNets[bestNetwork]['NRMSE_W'], \"avg: \", np.mean(nrmses) )\n",
    "\n",
    "    import scipy\n",
    "    def fitnessf(aperture, *args):\n",
    "#         print('Pattern: ', args[0])\n",
    "        cnet = computeConceptor(params, loadedNets[bestNetwork], args[0], aperture)\n",
    "        atten = testConceptor(params, cnet, windowSize)\n",
    "        return atten\n",
    "\n",
    "    #scipy.optimize.fminbound(fitnessf, 0,1000,  disp=3, xtol=0.1, args = (0,))\n",
    "    apertures = [scipy.optimize.fminbound(fitnessf, 0,1000,  disp=3, xtol=1, args = (x,)) \n",
    "                 for x in np.arange(params['patts'].shape[0])]\n",
    "\n",
    "    #store conceptors with calculated apertures\n",
    "    patternCs = np.zeros(len(apertures), dtype=np.object)\n",
    "    for i_patt in range(patternCs.size):\n",
    "#         print(i_patt, \" \", end='')\n",
    "        patternCs[i_patt] = computeConceptor(params, loadedNets[bestNetwork], i_patt, apertures[i_patt])\n",
    "#     print(\"Done\")\n",
    "    return patternCs\n",
    "    \n",
    "lrList = linspace(0.01,0.99,15)\n",
    "rl = testLen\n",
    "\n",
    "def testModel(modelSize, lr, testPat):\n",
    "    model = evalModel(modelSize, lr, testPat)\n",
    "    srCtl = ones(rl)\n",
    "    lrCtl = ones(rl)\n",
    "    numRenders = 5\n",
    "    renderErrors = zeros(numRenders)\n",
    "    for r in range(numRenders):\n",
    "        audio = render(model, srCtl, lrCtl, rl=rl, oversample=1)\n",
    "        renderErrors[r] = compareMelSpec(audio, testPat[0])\n",
    "    print(renderErrors)\n",
    "    error = median(renderErrors)\n",
    "    return error, model\n",
    "\n",
    "trials=5\n",
    "def searchLRs(modelSize, testPat):\n",
    "    bestError=999\n",
    "    bestLR = 0\n",
    "    bestModel = None\n",
    "    for lr in lrList:\n",
    "        print(\"LR: \", lr)\n",
    "#         lrScores = zeros(trials)\n",
    "        for t in range(trials):\n",
    "            print(\"Trial: \", t)\n",
    "            error, model = testModel(modelSize, lr, testPat)\n",
    "#             lrScores[t] = error\n",
    "            print(\"Trial error: \", error)\n",
    "#         print(\"LR scores: \", lrScores)\n",
    "#         medianError = np.median(lrScores)\n",
    "#         print(\"median error: \", medianError)\n",
    "            if (error < bestError):\n",
    "                print(\"Best so far ***************** \")\n",
    "                bestError = error\n",
    "                bestLR = lr\n",
    "                bestModel = model\n",
    "            \n",
    "    print(\"Best error: \", bestError, \" best LR: \", bestLR)\n",
    "    return bestLR, bestModel, bestError\n",
    "\n",
    "\n",
    "def findBestModel(lr, testPat):\n",
    "    print(\"Finding high res model: \", lr)\n",
    "    bestModel = None\n",
    "    bestError = 999\n",
    "    for t in range(10):\n",
    "        print(\"Trial: \", t)\n",
    "        error, model = testModel(800, lr, testPat)\n",
    "        print(\"Error: \", error)\n",
    "        if (error < bestError):\n",
    "            bestModel = model\n",
    "            bestError = error\n",
    "    print(\"Best high res error: \", bestError)\n",
    "    return bestModel\n",
    "\n",
    "\n",
    "\n",
    "def runOsc(freq, algo):\n",
    "    t = linspace(0, (testLen/sr), num = testLen, endpoint = False)\n",
    "    sig = algo(2 * pi * freq * t) * 0.5\n",
    "    return sig\n",
    "\n",
    "def testPitches(model, algo, testFreq, ):\n",
    "    freqErrors = zeros_like(freqMuls)\n",
    "    print(bestLR)\n",
    "    srCtl = ones(rl)\n",
    "    for i, freqMul in enumerate(freqMuls):\n",
    "        if ((model[0]['net']['p']['LR'] * freqMul) <= 1):\n",
    "            print(\"testing at \", testFreq * freqMul, \" Hz\")\n",
    "            testSig = runOsc(testFreq * freqMul, algo)\n",
    "            lrCtl = ones(rl) * freqMul\n",
    "            numRenders = 5\n",
    "            renderErrors = zeros(numRenders)\n",
    "            for r in range(numRenders):\n",
    "                renderSig = render(model, srCtl, lrCtl, rl=rl, oversample=1)\n",
    "                renderErrors[r] = compareMelSpec(renderSig, testSig)\n",
    "            error = np.min(renderErrors)\n",
    "            print(\"Error: \",error)\n",
    "            freqErrors[i] = error\n",
    "        else:\n",
    "            print(\"can't test, max lr\")\n",
    "            freqErrors[i] = -1\n",
    "    return freqErrors\n",
    "\n",
    "\n",
    "# testAlgos = [bl_sawtooth]\n",
    "# testFreqs = [40]\n",
    "\n",
    "allResults = zeros((len(testAlgos), len(testFreqs), len(freqMuls)))\n",
    "allLRs = zeros((len(testAlgos), len(testFreqs)))\n",
    "\n",
    "\n",
    "for i_algo, algo in enumerate(testAlgos):\n",
    "    for i_testFreq, testFreq in enumerate(testFreqs):\n",
    "        print(\"====== running test: \", testFreq, str(algo))\n",
    "        testPattern = np.array([runOsc(testFreq, algo)])\n",
    "        bestLR, bestModel, bestError = searchLRs(800, testPattern)\n",
    "#         bestLR = 0.2\n",
    "#         bestModel = findBestModel(bestLR, testPattern)\n",
    "        freqErrors = testPitches(bestModel, bl_sawtooth, testFreq)\n",
    "        allResults[i_algo][i_testFreq] = freqErrors\n",
    "        allLRs[i_algo][i_testFreq] = bestLR\n",
    "        import dill as pickle\n",
    "        ts = str(datetime.datetime.timestamp(datetime.datetime.now()))\n",
    "        with open(r\"pitchResults/allres3_\" + ts + \".pickled\", \"wb\") as output_file:    \n",
    "            pickle.dump({\"res\":allResults,\"lrs\":allLRs}, output_file, protocol=0)\n",
    "\n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###### "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(r\"pitchResults/allres3_1558012449.162566.pickled\", \"rb\") as infile:    \n",
    "    data3 = pickle.load(infile)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fbc58d41240>"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "figsize(20,10)\n",
    "for waveform in range(3):\n",
    "    for i, series in enumerate(data3['res'][waveform]):\n",
    "        y = array([q for q in series if q != -1])\n",
    "        x = arange(y.shape[0])\n",
    "        plot(y, label=str(i))\n",
    "        \n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 137,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def convToNan(x):\n",
    "    return NaN if x == -1 else x\n",
    "\n",
    "convToNan(0)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "d = data3['res']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "d = np.vectorize(convToNan)(d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3, 7, 25)"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[0.22121765, 0.29232345, 0.17399711, 0.21736399, 0.276613  ,\n",
       "         0.23875434, 0.21919012, 0.23532389, 0.26443695, 0.25774564,\n",
       "         0.27774709, 0.2134662 , 0.21673469, 0.23970292, 0.17226822,\n",
       "         0.18824091, 0.18435284, 0.10285945, 0.24388872, 0.20256012,\n",
       "         0.24827738, 0.19816031, 0.23277503, 0.25558244, 0.22752959],\n",
       "        [0.43531255, 0.44569943, 0.39360984, 0.38777488, 0.33764094,\n",
       "         0.24763352, 0.44841559, 0.204921  , 0.25179328, 0.11696491,\n",
       "         0.18682036, 0.21582843, 0.15617327, 0.1261417 , 0.53741724,\n",
       "         0.79849725, 0.9336975 , 1.00630281, 1.30271975, 1.39443275,\n",
       "         1.41474721, 1.41776596, 1.41739457, 1.41651521, 1.41711437],\n",
       "        [0.37602095, 0.36290325, 0.30317141, 0.32499024, 0.35720412,\n",
       "         0.25227559, 0.22855752, 0.34101204, 0.34747624, 0.19850777,\n",
       "         0.28293092, 0.14163058, 0.08434794, 0.29586146, 0.38599228,\n",
       "         0.61829921, 1.21839366, 1.22229134, 1.3339901 , 1.38983671,\n",
       "         1.35938442, 1.41090339, 1.42001765, 1.42500767, 1.42989107],\n",
       "        [1.38984019, 1.37955456, 1.3820624 , 1.38256962, 1.39042977,\n",
       "         1.38014661, 1.36964482, 1.37042743, 1.33345912, 1.26841846,\n",
       "         0.7837591 , 0.30618613, 0.07614327,        nan,        nan,\n",
       "                nan,        nan,        nan,        nan,        nan,\n",
       "                nan,        nan,        nan,        nan,        nan],\n",
       "        [1.42427042, 1.42447196, 1.42388507, 1.41804611, 1.41811705,\n",
       "         1.42113856, 1.41103   , 1.40427499, 1.38879638, 1.3377061 ,\n",
       "         1.01501123, 0.59223413, 0.04423231, 0.24186798, 0.30568334,\n",
       "         1.32536456, 1.41782096, 1.42141341, 1.41998537, 1.42043156,\n",
       "         1.4217178 , 1.4200223 , 1.41899268, 1.41864854, 1.41807914],\n",
       "        [1.42787058, 1.42920614, 1.43137937, 1.42686313, 1.42995378,\n",
       "         1.42371429, 1.42561491, 1.4250059 , 1.4250421 , 1.42184163,\n",
       "         1.41344013, 1.13192486, 0.08272889, 1.37932949, 1.3998758 ,\n",
       "         1.41073818, 1.41764845, 1.42085558,        nan,        nan,\n",
       "                nan,        nan,        nan,        nan,        nan],\n",
       "        [1.42541935, 1.42534247, 1.42588501, 1.42698482, 1.42647797,\n",
       "         1.42722836, 1.42610077, 1.4231796 , 1.42032857, 1.41893916,\n",
       "         1.41904289, 1.41905064, 0.09971553, 1.42073042, 1.4208635 ,\n",
       "         1.42391677, 1.42164676, 1.42203673,        nan,        nan,\n",
       "                nan,        nan,        nan,        nan,        nan]],\n",
       "\n",
       "       [[0.40094867, 0.43139725, 0.57398294, 0.73386599, 0.71598565,\n",
       "         0.81713053, 0.90141678, 0.93720186, 0.94486673, 0.90952966,\n",
       "         0.97386912, 0.99783731, 0.95789582, 0.98710609, 1.01437198,\n",
       "         0.9893511 , 0.99541745, 1.00610875, 0.93072688, 0.9920047 ,\n",
       "         1.01381406, 1.00375085, 0.98939258, 0.99577888, 0.99423396],\n",
       "        [0.96625105, 0.99363484, 1.00318483, 1.02545083, 1.01332417,\n",
       "         1.01544517, 0.97321138, 1.02858174, 1.01746227, 1.013884  ,\n",
       "         1.02031494, 1.01258611, 1.01457517, 1.04645006, 1.03612301,\n",
       "         1.03130797, 1.04125521, 1.04664354, 1.05167562, 1.03561826,\n",
       "         1.01771042, 1.00709492, 0.9663428 , 1.02547649, 1.05245497],\n",
       "        [1.00896725, 1.02188096, 1.08490292, 1.04094219, 1.03579544,\n",
       "         1.00798865, 1.0726788 , 1.02212425, 1.00921148, 1.02036895,\n",
       "         0.99403504, 1.01941584, 1.05459972, 1.07448537, 1.04790928,\n",
       "         1.0389457 , 1.03321871, 1.01480872,        nan,        nan,\n",
       "                nan,        nan,        nan,        nan,        nan],\n",
       "        [1.06068395, 1.12913368, 1.1351656 , 1.17355364, 1.23166675,\n",
       "         1.13801346, 1.19129341, 1.17682053, 1.08434636, 1.08857345,\n",
       "         1.1509389 , 1.03710498, 1.05113154, 0.98625555, 0.97848714,\n",
       "         0.91661068, 1.12743024, 1.42148973, 1.40308643, 1.42335164,\n",
       "         1.42599587, 1.42405389, 1.42230793, 1.42065698, 1.41960087],\n",
       "        [1.07877813, 1.08159059, 1.07412331, 1.01274829, 1.08558108,\n",
       "         1.10636054, 1.07663677, 1.10637466, 1.07626577, 1.10288812,\n",
       "         1.08162862, 1.05929808, 1.0260626 ,        nan,        nan,\n",
       "                nan,        nan,        nan,        nan,        nan,\n",
       "                nan,        nan,        nan,        nan,        nan],\n",
       "        [1.42682033, 1.42796904, 1.4301822 , 1.42581817, 1.42968479,\n",
       "         1.42297211, 1.42388157, 1.42475677, 1.42539062, 1.42144569,\n",
       "         1.4207228 , 1.4190517 , 1.04127738, 1.16912926, 1.42273119,\n",
       "         1.42245707, 1.41921311, 1.42061115,        nan,        nan,\n",
       "                nan,        nan,        nan,        nan,        nan],\n",
       "        [1.42536028, 1.42526951, 1.42582841, 1.42691206, 1.42636829,\n",
       "         1.42708831, 1.4258818 , 1.42292568, 1.4198044 , 1.4181785 ,\n",
       "         1.41795571, 1.40807362, 1.01417438, 1.35821927,        nan,\n",
       "                nan,        nan,        nan,        nan,        nan,\n",
       "                nan,        nan,        nan,        nan,        nan]],\n",
       "\n",
       "       [[0.38804569, 0.28775998, 0.33487026, 0.41924878, 0.26465516,\n",
       "         0.42691127, 0.42738776, 0.47326264, 0.51800244, 0.41708484,\n",
       "         0.47470569, 0.46593943, 0.47841408, 0.53270188, 0.53132014,\n",
       "         0.52178002, 0.51377822, 0.53147766, 0.3805077 , 0.72003798,\n",
       "         0.58011485, 0.49054097,        nan,        nan,        nan],\n",
       "        [0.45492392, 0.48430367, 0.53258041, 0.53486582, 0.6067938 ,\n",
       "         0.59814456, 0.39995514, 0.46829381, 0.5104501 , 0.54084199,\n",
       "         0.582375  , 0.45716917, 0.47705907, 0.42259963, 0.65518969,\n",
       "         0.45098857, 0.57755683, 0.44018097, 0.60218659, 0.50247306,\n",
       "         0.5251236 , 0.43974933, 0.45052101, 0.39588781, 0.54558964],\n",
       "        [0.39193598, 0.42337894, 0.49001527, 0.48218909, 0.510779  ,\n",
       "         0.46664388, 0.55631275, 0.46379083, 0.43575433, 0.51408103,\n",
       "         0.48583913, 0.45977947, 0.51799676, 0.52809313, 0.60928211,\n",
       "         0.53166882, 0.41886606, 0.46891627, 0.64018187, 0.58144806,\n",
       "         0.73300813, 0.63406684, 0.74614188, 0.79053999, 0.95249153],\n",
       "        [0.62988218, 0.86169824, 0.61885776, 0.8783529 , 0.88142342,\n",
       "         0.60976513, 0.84153484, 0.67625349, 0.61099323, 0.63816757,\n",
       "         0.53465152, 0.57208244, 0.58675706, 0.56087286, 0.55741082,\n",
       "         0.60838256, 0.61107078, 0.69587329, 0.63781466, 0.59916331,\n",
       "         0.74011998, 0.97916649, 1.43515799, 1.42028404, 1.42400747],\n",
       "        [0.37329548, 0.39885498, 0.31308548, 0.26066533, 0.39685233,\n",
       "         0.42316871, 0.39962058, 0.47807009, 0.42449662, 0.46915612,\n",
       "         0.47985525, 0.44355225, 0.48185912,        nan,        nan,\n",
       "                nan,        nan,        nan,        nan,        nan,\n",
       "                nan,        nan,        nan,        nan,        nan],\n",
       "        [1.4266784 , 1.42840646, 1.42993858, 1.42644017, 1.42911623,\n",
       "         1.42266945, 1.42383108, 1.42349814, 1.42243363, 1.41736043,\n",
       "         1.4076206 , 1.14393678, 0.44629822, 0.91900216, 1.13312823,\n",
       "         1.37481042, 1.40404111, 1.41686347, 1.40428086, 1.40915611,\n",
       "         1.41809412, 1.41732414, 1.41928406, 1.4189168 , 1.41912348],\n",
       "        [1.42541374, 1.42535784, 1.42587335, 1.42699833, 1.42647558,\n",
       "         1.42716105, 1.42605801, 1.42311512, 1.42018925, 1.41868561,\n",
       "         1.41832926, 1.41513678, 0.44478387, 1.38173577, 1.42172122,\n",
       "         1.42063002, 1.42499047, 1.42494334, 1.42794184, 1.42750142,\n",
       "         1.42766361, 1.43028706, 1.42565774, 1.42025987, 1.42930526]]])"
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns; sns.set(color_codes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 506,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fbc2f267278>]"
      ]
     },
     "execution_count": 506,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot(avgAll)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 507,
   "metadata": {},
   "outputs": [],
   "source": [
    "testpitches =arange(-12,13)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 426,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.DataFrame(columns=[\"Waveform\", \"Frequency\", \"Relative Pitch\", \"Reconstruction Error\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 472,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "40\n",
      "80\n",
      "160\n",
      "320\n",
      "640\n",
      "40\n",
      "80\n",
      "160\n",
      "320\n",
      "640\n",
      "40\n",
      "80\n",
      "160\n",
      "320\n",
      "640\n"
     ]
    }
   ],
   "source": [
    "idx=0\n",
    "for i_freq, freq in enumerate(testFreqs[:5]):\n",
    "    print(freq)\n",
    "    for i_note, note in enumerate(arange(-12,13)):\n",
    "        df.loc[idx] = [\"Saw\",freq,note,d[0, i_freq, i_note]]\n",
    "        idx = idx+1\n",
    "for i_freq, freq in enumerate(testFreqs[:5]):\n",
    "    print(freq)\n",
    "    for i_note, note in enumerate(arange(-12,13)):\n",
    "        df.loc[idx] = [\"Square\",freq,note,d[1, i_freq, i_note]]\n",
    "        idx = idx+1\n",
    "for i_freq, freq in enumerate(testFreqs[:5]):\n",
    "    print(freq)\n",
    "    for i_note, note in enumerate(arange(-12,13)):\n",
    "        df.loc[idx] = [\"Triangle\",freq,note,d[2, i_freq, i_note]]\n",
    "        idx = idx+1\n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 525,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Waveform</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>Relative Pitch</th>\n",
       "      <th>Reconstruction Error</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>-12</td>\n",
       "      <td>0.221218</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>-11</td>\n",
       "      <td>0.292323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>-10</td>\n",
       "      <td>0.173997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>-9</td>\n",
       "      <td>0.217364</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>-8</td>\n",
       "      <td>0.276613</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>-7</td>\n",
       "      <td>0.238754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>-6</td>\n",
       "      <td>0.219190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>-5</td>\n",
       "      <td>0.235324</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>-4</td>\n",
       "      <td>0.264437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>-3</td>\n",
       "      <td>0.257746</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>-2</td>\n",
       "      <td>0.277747</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.213466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>0</td>\n",
       "      <td>0.216735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>1</td>\n",
       "      <td>0.239703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>2</td>\n",
       "      <td>0.172268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>3</td>\n",
       "      <td>0.188241</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>4</td>\n",
       "      <td>0.184353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>5</td>\n",
       "      <td>0.102859</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>6</td>\n",
       "      <td>0.243889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>7</td>\n",
       "      <td>0.202560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>8</td>\n",
       "      <td>0.248277</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>9</td>\n",
       "      <td>0.198160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>10</td>\n",
       "      <td>0.232775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>11</td>\n",
       "      <td>0.255582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>Saw</td>\n",
       "      <td>40</td>\n",
       "      <td>12</td>\n",
       "      <td>0.227530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>Saw</td>\n",
       "      <td>80</td>\n",
       "      <td>-12</td>\n",
       "      <td>0.435313</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>Saw</td>\n",
       "      <td>80</td>\n",
       "      <td>-11</td>\n",
       "      <td>0.445699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>Saw</td>\n",
       "      <td>80</td>\n",
       "      <td>-10</td>\n",
       "      <td>0.393610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>Saw</td>\n",
       "      <td>80</td>\n",
       "      <td>-9</td>\n",
       "      <td>0.387775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>Saw</td>\n",
       "      <td>80</td>\n",
       "      <td>-8</td>\n",
       "      <td>0.337641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>345</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>320</td>\n",
       "      <td>8</td>\n",
       "      <td>0.740120</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>346</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>320</td>\n",
       "      <td>9</td>\n",
       "      <td>0.979166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>347</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>320</td>\n",
       "      <td>10</td>\n",
       "      <td>1.435158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>348</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>320</td>\n",
       "      <td>11</td>\n",
       "      <td>1.420284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>349</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>320</td>\n",
       "      <td>12</td>\n",
       "      <td>1.424007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>350</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>-12</td>\n",
       "      <td>0.373295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>351</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>-11</td>\n",
       "      <td>0.398855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>352</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>-10</td>\n",
       "      <td>0.313085</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>353</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>-9</td>\n",
       "      <td>0.260665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>354</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>-8</td>\n",
       "      <td>0.396852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>355</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>-7</td>\n",
       "      <td>0.423169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>356</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>-6</td>\n",
       "      <td>0.399621</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>357</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>-5</td>\n",
       "      <td>0.478070</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>358</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>-4</td>\n",
       "      <td>0.424497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>359</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>-3</td>\n",
       "      <td>0.469156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>360</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>-2</td>\n",
       "      <td>0.479855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>361</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>-1</td>\n",
       "      <td>0.443552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>362</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>0</td>\n",
       "      <td>0.481859</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>363</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>364</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>365</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>366</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>4</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>367</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>5</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>368</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>6</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>369</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>7</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>370</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>8</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>371</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>9</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>372</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>373</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>11</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>374</th>\n",
       "      <td>Triangle</td>\n",
       "      <td>640</td>\n",
       "      <td>12</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>375 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Waveform  Frequency  Relative Pitch  Reconstruction Error\n",
       "0         Saw         40             -12              0.221218\n",
       "1         Saw         40             -11              0.292323\n",
       "2         Saw         40             -10              0.173997\n",
       "3         Saw         40              -9              0.217364\n",
       "4         Saw         40              -8              0.276613\n",
       "5         Saw         40              -7              0.238754\n",
       "6         Saw         40              -6              0.219190\n",
       "7         Saw         40              -5              0.235324\n",
       "8         Saw         40              -4              0.264437\n",
       "9         Saw         40              -3              0.257746\n",
       "10        Saw         40              -2              0.277747\n",
       "11        Saw         40              -1              0.213466\n",
       "12        Saw         40               0              0.216735\n",
       "13        Saw         40               1              0.239703\n",
       "14        Saw         40               2              0.172268\n",
       "15        Saw         40               3              0.188241\n",
       "16        Saw         40               4              0.184353\n",
       "17        Saw         40               5              0.102859\n",
       "18        Saw         40               6              0.243889\n",
       "19        Saw         40               7              0.202560\n",
       "20        Saw         40               8              0.248277\n",
       "21        Saw         40               9              0.198160\n",
       "22        Saw         40              10              0.232775\n",
       "23        Saw         40              11              0.255582\n",
       "24        Saw         40              12              0.227530\n",
       "25        Saw         80             -12              0.435313\n",
       "26        Saw         80             -11              0.445699\n",
       "27        Saw         80             -10              0.393610\n",
       "28        Saw         80              -9              0.387775\n",
       "29        Saw         80              -8              0.337641\n",
       "..        ...        ...             ...                   ...\n",
       "345  Triangle        320               8              0.740120\n",
       "346  Triangle        320               9              0.979166\n",
       "347  Triangle        320              10              1.435158\n",
       "348  Triangle        320              11              1.420284\n",
       "349  Triangle        320              12              1.424007\n",
       "350  Triangle        640             -12              0.373295\n",
       "351  Triangle        640             -11              0.398855\n",
       "352  Triangle        640             -10              0.313085\n",
       "353  Triangle        640              -9              0.260665\n",
       "354  Triangle        640              -8              0.396852\n",
       "355  Triangle        640              -7              0.423169\n",
       "356  Triangle        640              -6              0.399621\n",
       "357  Triangle        640              -5              0.478070\n",
       "358  Triangle        640              -4              0.424497\n",
       "359  Triangle        640              -3              0.469156\n",
       "360  Triangle        640              -2              0.479855\n",
       "361  Triangle        640              -1              0.443552\n",
       "362  Triangle        640               0              0.481859\n",
       "363  Triangle        640               1                   NaN\n",
       "364  Triangle        640               2                   NaN\n",
       "365  Triangle        640               3                   NaN\n",
       "366  Triangle        640               4                   NaN\n",
       "367  Triangle        640               5                   NaN\n",
       "368  Triangle        640               6                   NaN\n",
       "369  Triangle        640               7                   NaN\n",
       "370  Triangle        640               8                   NaN\n",
       "371  Triangle        640               9                   NaN\n",
       "372  Triangle        640              10                   NaN\n",
       "373  Triangle        640              11                   NaN\n",
       "374  Triangle        640              12                   NaN\n",
       "\n",
       "[375 rows x 4 columns]"
      ]
     },
     "execution_count": 525,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 439,
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.set(style='whitegrid', rc={\"grid.linewidth\": 0.1})\n",
    "sns.set_context(\"paper\", font_scale=2)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 480,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figsize(20,5)\n",
    "sns.lineplot(x='Relative Pitch', y='Reconstruction Error', hue=\"Waveform\", data=df, lw=2)\n",
    "xlabel(\"Relative pitch to base frequency (semitones)\", fontsize=20)\n",
    "ylabel(\"Reconstruction error\", fontsize=20)\n",
    "plt.savefig(\"tmp/reconPitchSemis.pdf\", bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 469,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saw\n",
      "40\n",
      "80\n",
      "160\n",
      "320\n",
      "640\n",
      "1280\n",
      "2560\n",
      "excluding freq:  4063.7466930385904\n",
      "excluding freq:  4305.389646099018\n",
      "excluding freq:  4561.401436878537\n",
      "excluding freq:  4832.63648093027\n",
      "excluding freq:  5120.0\n",
      "Square\n",
      "40\n",
      "80\n",
      "160\n",
      "320\n",
      "640\n",
      "1280\n",
      "2560\n",
      "excluding freq:  4063.7466930385904\n",
      "excluding freq:  4305.389646099018\n",
      "excluding freq:  4561.401436878537\n",
      "excluding freq:  4832.63648093027\n",
      "excluding freq:  5120.0\n",
      "Triangle\n",
      "40\n",
      "80\n",
      "160\n",
      "320\n",
      "640\n",
      "1280\n",
      "2560\n",
      "excluding freq:  4063.7466930385904\n",
      "excluding freq:  4305.389646099018\n",
      "excluding freq:  4561.401436878537\n",
      "excluding freq:  4832.63648093027\n",
      "excluding freq:  5120.0\n"
     ]
    }
   ],
   "source": [
    "df2 = pd.DataFrame(columns=[\"Waveform\", \"Base Frequency\", \"note\", \"Reconstruction Error\"])\n",
    "idx=0\n",
    "for i_wave, wave in enumerate([\"Saw\", \"Square\", \"Triangle\"]):\n",
    "    print(wave)\n",
    "    for i_freq, freq in enumerate(testFreqs):\n",
    "        print(freq)\n",
    "        for i_note, note in enumerate(arange(-12,13)):\n",
    "            if (freq * freqMuls[i_note]) < sr/2:\n",
    "                df2.loc[idx] = [wave,freq,note,d[i_wave, i_freq, i_note]]\n",
    "            else:\n",
    "                print(\"excluding freq: \", freq * freqMuls[i_note])\n",
    "            \n",
    "            idx = idx+1\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 481,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figsize(20,4)\n",
    "sns.boxplot(x='Base Frequency', y='Reconstruction Error',data=df2)\n",
    "xlabel(\"Base Frequency (Hz)\", fontsize=20)\n",
    "ylabel(\"Reconstruction Error\", fontsize=20)\n",
    "plt.savefig(\"tmp/pitchReconFreq.pdf\", bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 535,
   "metadata": {},
   "outputs": [],
   "source": [
    "dPitchRecon = df[[\"Relative Pitch\", \"Reconstruction Error\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 595,
   "metadata": {},
   "outputs": [],
   "source": [
    "posPitches = dPitchRecon.loc[dPitchRecon[\"Relative Pitch\"] >= 0]\n",
    "\n",
    "posPitches = posPitches.loc[posPitches[\"Reconstruction Error\"] < 1000000] #remove NaNs\n",
    "\n",
    "posPitches = array(posPitches)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 596,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.        ,  0.21673469],\n",
       "       [ 1.        ,  0.23970292],\n",
       "       [ 2.        ,  0.17226822],\n",
       "       [ 3.        ,  0.18824091],\n",
       "       [ 4.        ,  0.18435284],\n",
       "       [ 5.        ,  0.10285945],\n",
       "       [ 6.        ,  0.24388872],\n",
       "       [ 7.        ,  0.20256012],\n",
       "       [ 8.        ,  0.24827738],\n",
       "       [ 9.        ,  0.19816031],\n",
       "       [10.        ,  0.23277503],\n",
       "       [11.        ,  0.25558244],\n",
       "       [12.        ,  0.22752959],\n",
       "       [ 0.        ,  0.15617327],\n",
       "       [ 1.        ,  0.1261417 ],\n",
       "       [ 2.        ,  0.53741724],\n",
       "       [ 3.        ,  0.79849725],\n",
       "       [ 4.        ,  0.9336975 ],\n",
       "       [ 5.        ,  1.00630281],\n",
       "       [ 6.        ,  1.30271975],\n",
       "       [ 7.        ,  1.39443275],\n",
       "       [ 8.        ,  1.41474721],\n",
       "       [ 9.        ,  1.41776596],\n",
       "       [10.        ,  1.41739457],\n",
       "       [11.        ,  1.41651521],\n",
       "       [12.        ,  1.41711437],\n",
       "       [ 0.        ,  0.08434794],\n",
       "       [ 1.        ,  0.29586146],\n",
       "       [ 2.        ,  0.38599228],\n",
       "       [ 3.        ,  0.61829921],\n",
       "       [ 4.        ,  1.21839366],\n",
       "       [ 5.        ,  1.22229134],\n",
       "       [ 6.        ,  1.3339901 ],\n",
       "       [ 7.        ,  1.38983671],\n",
       "       [ 8.        ,  1.35938442],\n",
       "       [ 9.        ,  1.41090339],\n",
       "       [10.        ,  1.42001765],\n",
       "       [11.        ,  1.42500767],\n",
       "       [12.        ,  1.42989107],\n",
       "       [ 0.        ,  0.07614327],\n",
       "       [ 0.        ,  0.04423231],\n",
       "       [ 1.        ,  0.24186798],\n",
       "       [ 2.        ,  0.30568334],\n",
       "       [ 3.        ,  1.32536456],\n",
       "       [ 4.        ,  1.41782096],\n",
       "       [ 5.        ,  1.42141341],\n",
       "       [ 6.        ,  1.41998537],\n",
       "       [ 7.        ,  1.42043156],\n",
       "       [ 8.        ,  1.4217178 ],\n",
       "       [ 9.        ,  1.4200223 ],\n",
       "       [10.        ,  1.41899268],\n",
       "       [11.        ,  1.41864854],\n",
       "       [12.        ,  1.41807914],\n",
       "       [ 0.        ,  0.95789582],\n",
       "       [ 1.        ,  0.98710609],\n",
       "       [ 2.        ,  1.01437198],\n",
       "       [ 3.        ,  0.9893511 ],\n",
       "       [ 4.        ,  0.99541745],\n",
       "       [ 5.        ,  1.00610875],\n",
       "       [ 6.        ,  0.93072688],\n",
       "       [ 7.        ,  0.9920047 ],\n",
       "       [ 8.        ,  1.01381406],\n",
       "       [ 9.        ,  1.00375085],\n",
       "       [10.        ,  0.98939258],\n",
       "       [11.        ,  0.99577888],\n",
       "       [12.        ,  0.99423396],\n",
       "       [ 0.        ,  1.01457517],\n",
       "       [ 1.        ,  1.04645006],\n",
       "       [ 2.        ,  1.03612301],\n",
       "       [ 3.        ,  1.03130797],\n",
       "       [ 4.        ,  1.04125521],\n",
       "       [ 5.        ,  1.04664354],\n",
       "       [ 6.        ,  1.05167562],\n",
       "       [ 7.        ,  1.03561826],\n",
       "       [ 8.        ,  1.01771042],\n",
       "       [ 9.        ,  1.00709492],\n",
       "       [10.        ,  0.9663428 ],\n",
       "       [11.        ,  1.02547649],\n",
       "       [12.        ,  1.05245497],\n",
       "       [ 0.        ,  1.05459972],\n",
       "       [ 1.        ,  1.07448537],\n",
       "       [ 2.        ,  1.04790928],\n",
       "       [ 3.        ,  1.0389457 ],\n",
       "       [ 4.        ,  1.03321871],\n",
       "       [ 5.        ,  1.01480872],\n",
       "       [ 0.        ,  1.05113154],\n",
       "       [ 1.        ,  0.98625555],\n",
       "       [ 2.        ,  0.97848714],\n",
       "       [ 3.        ,  0.91661068],\n",
       "       [ 4.        ,  1.12743024],\n",
       "       [ 5.        ,  1.42148973],\n",
       "       [ 6.        ,  1.40308643],\n",
       "       [ 7.        ,  1.42335164],\n",
       "       [ 8.        ,  1.42599587],\n",
       "       [ 9.        ,  1.42405389],\n",
       "       [10.        ,  1.42230793],\n",
       "       [11.        ,  1.42065698],\n",
       "       [12.        ,  1.41960087],\n",
       "       [ 0.        ,  1.0260626 ],\n",
       "       [ 0.        ,  0.47841408],\n",
       "       [ 1.        ,  0.53270188],\n",
       "       [ 2.        ,  0.53132014],\n",
       "       [ 3.        ,  0.52178002],\n",
       "       [ 4.        ,  0.51377822],\n",
       "       [ 5.        ,  0.53147766],\n",
       "       [ 6.        ,  0.3805077 ],\n",
       "       [ 7.        ,  0.72003798],\n",
       "       [ 8.        ,  0.58011485],\n",
       "       [ 9.        ,  0.49054097],\n",
       "       [ 0.        ,  0.47705907],\n",
       "       [ 1.        ,  0.42259963],\n",
       "       [ 2.        ,  0.65518969],\n",
       "       [ 3.        ,  0.45098857],\n",
       "       [ 4.        ,  0.57755683],\n",
       "       [ 5.        ,  0.44018097],\n",
       "       [ 6.        ,  0.60218659],\n",
       "       [ 7.        ,  0.50247306],\n",
       "       [ 8.        ,  0.5251236 ],\n",
       "       [ 9.        ,  0.43974933],\n",
       "       [10.        ,  0.45052101],\n",
       "       [11.        ,  0.39588781],\n",
       "       [12.        ,  0.54558964],\n",
       "       [ 0.        ,  0.51799676],\n",
       "       [ 1.        ,  0.52809313],\n",
       "       [ 2.        ,  0.60928211],\n",
       "       [ 3.        ,  0.53166882],\n",
       "       [ 4.        ,  0.41886606],\n",
       "       [ 5.        ,  0.46891627],\n",
       "       [ 6.        ,  0.64018187],\n",
       "       [ 7.        ,  0.58144806],\n",
       "       [ 8.        ,  0.73300813],\n",
       "       [ 9.        ,  0.63406684],\n",
       "       [10.        ,  0.74614188],\n",
       "       [11.        ,  0.79053999],\n",
       "       [12.        ,  0.95249153],\n",
       "       [ 0.        ,  0.58675706],\n",
       "       [ 1.        ,  0.56087286],\n",
       "       [ 2.        ,  0.55741082],\n",
       "       [ 3.        ,  0.60838256],\n",
       "       [ 4.        ,  0.61107078],\n",
       "       [ 5.        ,  0.69587329],\n",
       "       [ 6.        ,  0.63781466],\n",
       "       [ 7.        ,  0.59916331],\n",
       "       [ 8.        ,  0.74011998],\n",
       "       [ 9.        ,  0.97916649],\n",
       "       [10.        ,  1.43515799],\n",
       "       [11.        ,  1.42028404],\n",
       "       [12.        ,  1.42400747],\n",
       "       [ 0.        ,  0.48185912]])"
      ]
     },
     "execution_count": 596,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "posPitches"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 597,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.9247173070907593, 4.6683669552294305e-07)"
      ]
     },
     "execution_count": 597,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scipy.stats.shapiro(posPitches[:,1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 598,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.40299105085353515, 3.4789529018220155e-07)"
      ]
     },
     "execution_count": 598,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scipy.stats.pearsonr(posPitches[:,0], posPitches[:,1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 580,
   "metadata": {},
   "outputs": [],
   "source": [
    "negPitches = dPitchRecon.loc[dPitchRecon[\"Relative Pitch\"] <= 0]\n",
    "\n",
    "negPitches = negPitches.loc[negPitches[\"Reconstruction Error\"] < 1000000] #remove NaNs\n",
    "\n",
    "negPitches = array(negPitches)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 581,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-12.        ,   0.22121765],\n",
       "       [-11.        ,   0.29232345],\n",
       "       [-10.        ,   0.17399711],\n",
       "       [ -9.        ,   0.21736399],\n",
       "       [ -8.        ,   0.276613  ],\n",
       "       [ -7.        ,   0.23875434],\n",
       "       [ -6.        ,   0.21919012],\n",
       "       [ -5.        ,   0.23532389],\n",
       "       [ -4.        ,   0.26443695],\n",
       "       [ -3.        ,   0.25774564],\n",
       "       [ -2.        ,   0.27774709],\n",
       "       [ -1.        ,   0.2134662 ],\n",
       "       [  0.        ,   0.21673469],\n",
       "       [-12.        ,   0.43531255],\n",
       "       [-11.        ,   0.44569943],\n",
       "       [-10.        ,   0.39360984],\n",
       "       [ -9.        ,   0.38777488],\n",
       "       [ -8.        ,   0.33764094],\n",
       "       [ -7.        ,   0.24763352],\n",
       "       [ -6.        ,   0.44841559],\n",
       "       [ -5.        ,   0.204921  ],\n",
       "       [ -4.        ,   0.25179328],\n",
       "       [ -3.        ,   0.11696491],\n",
       "       [ -2.        ,   0.18682036],\n",
       "       [ -1.        ,   0.21582843],\n",
       "       [  0.        ,   0.15617327],\n",
       "       [-12.        ,   0.37602095],\n",
       "       [-11.        ,   0.36290325],\n",
       "       [-10.        ,   0.30317141],\n",
       "       [ -9.        ,   0.32499024],\n",
       "       [ -8.        ,   0.35720412],\n",
       "       [ -7.        ,   0.25227559],\n",
       "       [ -6.        ,   0.22855752],\n",
       "       [ -5.        ,   0.34101204],\n",
       "       [ -4.        ,   0.34747624],\n",
       "       [ -3.        ,   0.19850777],\n",
       "       [ -2.        ,   0.28293092],\n",
       "       [ -1.        ,   0.14163058],\n",
       "       [  0.        ,   0.08434794],\n",
       "       [-12.        ,   1.38984019],\n",
       "       [-11.        ,   1.37955456],\n",
       "       [-10.        ,   1.3820624 ],\n",
       "       [ -9.        ,   1.38256962],\n",
       "       [ -8.        ,   1.39042977],\n",
       "       [ -7.        ,   1.38014661],\n",
       "       [ -6.        ,   1.36964482],\n",
       "       [ -5.        ,   1.37042743],\n",
       "       [ -4.        ,   1.33345912],\n",
       "       [ -3.        ,   1.26841846],\n",
       "       [ -2.        ,   0.7837591 ],\n",
       "       [ -1.        ,   0.30618613],\n",
       "       [  0.        ,   0.07614327],\n",
       "       [-12.        ,   1.42427042],\n",
       "       [-11.        ,   1.42447196],\n",
       "       [-10.        ,   1.42388507],\n",
       "       [ -9.        ,   1.41804611],\n",
       "       [ -8.        ,   1.41811705],\n",
       "       [ -7.        ,   1.42113856],\n",
       "       [ -6.        ,   1.41103   ],\n",
       "       [ -5.        ,   1.40427499],\n",
       "       [ -4.        ,   1.38879638],\n",
       "       [ -3.        ,   1.3377061 ],\n",
       "       [ -2.        ,   1.01501123],\n",
       "       [ -1.        ,   0.59223413],\n",
       "       [  0.        ,   0.04423231],\n",
       "       [-12.        ,   0.40094867],\n",
       "       [-11.        ,   0.43139725],\n",
       "       [-10.        ,   0.57398294],\n",
       "       [ -9.        ,   0.73386599],\n",
       "       [ -8.        ,   0.71598565],\n",
       "       [ -7.        ,   0.81713053],\n",
       "       [ -6.        ,   0.90141678],\n",
       "       [ -5.        ,   0.93720186],\n",
       "       [ -4.        ,   0.94486673],\n",
       "       [ -3.        ,   0.90952966],\n",
       "       [ -2.        ,   0.97386912],\n",
       "       [ -1.        ,   0.99783731],\n",
       "       [  0.        ,   0.95789582],\n",
       "       [-12.        ,   0.96625105],\n",
       "       [-11.        ,   0.99363484],\n",
       "       [-10.        ,   1.00318483],\n",
       "       [ -9.        ,   1.02545083],\n",
       "       [ -8.        ,   1.01332417],\n",
       "       [ -7.        ,   1.01544517],\n",
       "       [ -6.        ,   0.97321138],\n",
       "       [ -5.        ,   1.02858174],\n",
       "       [ -4.        ,   1.01746227],\n",
       "       [ -3.        ,   1.013884  ],\n",
       "       [ -2.        ,   1.02031494],\n",
       "       [ -1.        ,   1.01258611],\n",
       "       [  0.        ,   1.01457517],\n",
       "       [-12.        ,   1.00896725],\n",
       "       [-11.        ,   1.02188096],\n",
       "       [-10.        ,   1.08490292],\n",
       "       [ -9.        ,   1.04094219],\n",
       "       [ -8.        ,   1.03579544],\n",
       "       [ -7.        ,   1.00798865],\n",
       "       [ -6.        ,   1.0726788 ],\n",
       "       [ -5.        ,   1.02212425],\n",
       "       [ -4.        ,   1.00921148],\n",
       "       [ -3.        ,   1.02036895],\n",
       "       [ -2.        ,   0.99403504],\n",
       "       [ -1.        ,   1.01941584],\n",
       "       [  0.        ,   1.05459972],\n",
       "       [-12.        ,   1.06068395],\n",
       "       [-11.        ,   1.12913368],\n",
       "       [-10.        ,   1.1351656 ],\n",
       "       [ -9.        ,   1.17355364],\n",
       "       [ -8.        ,   1.23166675],\n",
       "       [ -7.        ,   1.13801346],\n",
       "       [ -6.        ,   1.19129341],\n",
       "       [ -5.        ,   1.17682053],\n",
       "       [ -4.        ,   1.08434636],\n",
       "       [ -3.        ,   1.08857345],\n",
       "       [ -2.        ,   1.1509389 ],\n",
       "       [ -1.        ,   1.03710498],\n",
       "       [  0.        ,   1.05113154],\n",
       "       [-12.        ,   1.07877813],\n",
       "       [-11.        ,   1.08159059],\n",
       "       [-10.        ,   1.07412331],\n",
       "       [ -9.        ,   1.01274829],\n",
       "       [ -8.        ,   1.08558108],\n",
       "       [ -7.        ,   1.10636054],\n",
       "       [ -6.        ,   1.07663677],\n",
       "       [ -5.        ,   1.10637466],\n",
       "       [ -4.        ,   1.07626577],\n",
       "       [ -3.        ,   1.10288812],\n",
       "       [ -2.        ,   1.08162862],\n",
       "       [ -1.        ,   1.05929808],\n",
       "       [  0.        ,   1.0260626 ],\n",
       "       [-12.        ,   0.38804569],\n",
       "       [-11.        ,   0.28775998],\n",
       "       [-10.        ,   0.33487026],\n",
       "       [ -9.        ,   0.41924878],\n",
       "       [ -8.        ,   0.26465516],\n",
       "       [ -7.        ,   0.42691127],\n",
       "       [ -6.        ,   0.42738776],\n",
       "       [ -5.        ,   0.47326264],\n",
       "       [ -4.        ,   0.51800244],\n",
       "       [ -3.        ,   0.41708484],\n",
       "       [ -2.        ,   0.47470569],\n",
       "       [ -1.        ,   0.46593943],\n",
       "       [  0.        ,   0.47841408],\n",
       "       [-12.        ,   0.45492392],\n",
       "       [-11.        ,   0.48430367],\n",
       "       [-10.        ,   0.53258041],\n",
       "       [ -9.        ,   0.53486582],\n",
       "       [ -8.        ,   0.6067938 ],\n",
       "       [ -7.        ,   0.59814456],\n",
       "       [ -6.        ,   0.39995514],\n",
       "       [ -5.        ,   0.46829381],\n",
       "       [ -4.        ,   0.5104501 ],\n",
       "       [ -3.        ,   0.54084199],\n",
       "       [ -2.        ,   0.582375  ],\n",
       "       [ -1.        ,   0.45716917],\n",
       "       [  0.        ,   0.47705907],\n",
       "       [-12.        ,   0.39193598],\n",
       "       [-11.        ,   0.42337894],\n",
       "       [-10.        ,   0.49001527],\n",
       "       [ -9.        ,   0.48218909],\n",
       "       [ -8.        ,   0.510779  ],\n",
       "       [ -7.        ,   0.46664388],\n",
       "       [ -6.        ,   0.55631275],\n",
       "       [ -5.        ,   0.46379083],\n",
       "       [ -4.        ,   0.43575433],\n",
       "       [ -3.        ,   0.51408103],\n",
       "       [ -2.        ,   0.48583913],\n",
       "       [ -1.        ,   0.45977947],\n",
       "       [  0.        ,   0.51799676],\n",
       "       [-12.        ,   0.62988218],\n",
       "       [-11.        ,   0.86169824],\n",
       "       [-10.        ,   0.61885776],\n",
       "       [ -9.        ,   0.8783529 ],\n",
       "       [ -8.        ,   0.88142342],\n",
       "       [ -7.        ,   0.60976513],\n",
       "       [ -6.        ,   0.84153484],\n",
       "       [ -5.        ,   0.67625349],\n",
       "       [ -4.        ,   0.61099323],\n",
       "       [ -3.        ,   0.63816757],\n",
       "       [ -2.        ,   0.53465152],\n",
       "       [ -1.        ,   0.57208244],\n",
       "       [  0.        ,   0.58675706],\n",
       "       [-12.        ,   0.37329548],\n",
       "       [-11.        ,   0.39885498],\n",
       "       [-10.        ,   0.31308548],\n",
       "       [ -9.        ,   0.26066533],\n",
       "       [ -8.        ,   0.39685233],\n",
       "       [ -7.        ,   0.42316871],\n",
       "       [ -6.        ,   0.39962058],\n",
       "       [ -5.        ,   0.47807009],\n",
       "       [ -4.        ,   0.42449662],\n",
       "       [ -3.        ,   0.46915612],\n",
       "       [ -2.        ,   0.47985525],\n",
       "       [ -1.        ,   0.44355225],\n",
       "       [  0.        ,   0.48185912]])"
      ]
     },
     "execution_count": 581,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "negPitches"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 585,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.09919697879754835, 0.16767597953597702)"
      ]
     },
     "execution_count": 585,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scipy.stats.pearsonr(negPitches[:,0], negPitches[:,1])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}