conceptorSynthesisExpt_kick.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": [
    "%pylab inline\n",
    "from IPython.display import Audio\n",
    "import librosa\n",
    "import scipy as sp\n",
    "from numba import jit\n",
    "from scipy.spatial import distance\n",
    "\n",
    "figsize(20,6)\n",
    "prefix=\"baseline\"\n",
    "def filepre(nm):\n",
    "    return \"tmp/\"+prefix+\"_\"+nm\n",
    "from IPython.display import clear_output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# %load_ext autoreload\n",
    "# %autoreload 1\n",
    "\n",
    "# %aimport microbial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sr = 22050\n",
    "def nrmse(output,target):\n",
    "    assert(output.shape[0] == target.shape[0])\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": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from datetime import datetime\n",
    "import os\n",
    "resultsFolderName = \"csynthResults/results\" + str(datetime.now()) + \"/\"\n",
    "os.mkdir(resultsFolderName)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['909a_22k.wav']\n"
     ]
    }
   ],
   "source": [
    "# relevant_path = \"audios/ixi\"\n",
    "# ixiFiles = [fn for fn in os.listdir(relevant_path)\n",
    "#               if fn.endswith('wav')]\n",
    "ixiFiles = ['909a_22k.wav']\n",
    "print(ixiFiles)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def ixistr(id):\n",
    "    return str(id) + \"_\" + ixiFiles[id] + \"_\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def compareMFCCs(seq1, seq2):\n",
    "    fftSize=2048\n",
    "    hop=64    \n",
    "    melspec = librosa.feature.melspectrogram(y=seq1, sr=sr, n_fft=fftSize,hop_length=hop)\n",
    "    mfccs = librosa.feature.mfcc(S=melspec,n_mfcc=20)[1:,:]\n",
    "    melspec2 = librosa.feature.melspectrogram(y=seq2, sr=sr, n_fft=fftSize,hop_length=hop)\n",
    "    mfccs2 = librosa.feature.mfcc(S=melspec2,n_mfcc=20)[1:,:]    \n",
    "    return nrmse(mfccs.flatten(), mfccs2.flatten())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def computeConceptor(p, net, i_pattern, alpha):\n",
    "    print('Computing conceptor, alpha: ', alpha)\n",
    "#     Cs = np.zeros((4, 1), dtype=np.object)\n",
    "    R = net['patternRs'][0,i_pattern]\n",
    "    [U,s,V] = svd(R)\n",
    "    S = diag(s)\n",
    "    Snew = (S.dot(linalg.inv(S + pow(alpha, -2) * np.eye(p['N']))))\n",
    "\n",
    "    C =  U.dot(Snew).dot(U.T);\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 C\n",
    "\n",
    "\n",
    "def testConceptor(p, C, net, recallTestLength):\n",
    "    trials = 1\n",
    "    attens = np.zeros(trials)\n",
    "#     LR = np.diag(p['LR'])\n",
    "#     LROneMinus = np.diag(1.0 - p['LR'])\n",
    "    LR = array(p['LR'])\n",
    "    LROneMinus = array(1.0 - p['LR'])\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",
    "        for n in range(recallTestLength + p['washoutLength']):\n",
    "            xOld = x\n",
    "            Wtarget = (net['W'].dot(x))\n",
    "            leakTerm = LROneMinus.dot(xOld)\n",
    "            newX =tanh(Wtarget + net['Wbias'])\n",
    "            newXLeaked = LR.dot(newX)\n",
    "            z = leakTerm + newXLeaked\n",
    "            x = C.dot(z)\n",
    "#             xPrev = x\n",
    "            if (n > p['washoutLength']):\n",
    "                x_CTestPL[:,n-p['washoutLength']] = x.T\n",
    "                z_CTestPL[:,n-p['washoutLength']] = z.T\n",
    "\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"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def makeLoadedNetwork_v2(p):\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",
    "#     LR = np.diag(p['LR'])\n",
    "#     LROneMinus = np.diag(1.0 - p['LR'])\n",
    "    LR = array(p['LR'])\n",
    "    LROneMinus = array(1.0 - p['LR'])\n",
    "#     print(LR)\n",
    "#     print(LROneMinus)\n",
    "    for i_pattern in range(p['patts'].shape[0]):\n",
    "        print(i_pattern, \" \", end='')\n",
    "        patt = p['patts'][i_pattern]\n",
    "        pattLearnLen = patt.size * p['learnLength']\n",
    "        xCollector = np.zeros((p['N'] + 1, pattLearnLen));\n",
    "        xOldCollector = np.zeros((p['N'], pattLearnLen));\n",
    "        WTargetCollector = np.zeros((p['N'], pattLearnLen));\n",
    "        pCollector = np.zeros((1, pattLearnLen));\n",
    "        x = np.zeros((p['N'], 1));\n",
    "\n",
    "        for n in range(p['washoutLength'] + pattLearnLen):\n",
    "            u = patt.take(n, mode='wrap')\n",
    "            xOld = x\n",
    "            Wtarget = (Wstar.dot(x)) + (Win.dot(u))\n",
    "            leakTerm = LROneMinus.dot(xOld)\n",
    "            newX =tanh(Wtarget + Wbias)\n",
    "            newXLeaked = LR.dot(newX)\n",
    "            x = leakTerm + newXLeaked\n",
    "\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",
    "                \n",
    "            uOld = u\n",
    "        \n",
    "        R = xCollector[0:-1].dot(xCollector[0:-1].T) / pattLearnLen\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"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "def render(p, patternCs, bestNet, lrMod=1.0, speed=1.0, xFade=0.05, srMod=1):\n",
    "    audio = np.zeros(0)\n",
    "    x = 0.5 * np.random.randn(p['N'],1)\n",
    "    C = patternCs[0]\n",
    "    \n",
    "    LR = p['LR']\n",
    "    LR = array(LR * lrMod)\n",
    "    LROneMinus = array(1.0 - p['LR'])\n",
    "    Wmod = bestNet['W'] * srMod\n",
    "    \n",
    "    #run to washout\n",
    "    for n in range(p['washoutLength']):\n",
    "        xOld = x\n",
    "        Wtarget = (Wmod.dot(x))\n",
    "        z = (LROneMinus.dot(xOld)) + (LR.dot(tanh(Wtarget + bestNet['Wbias'])))\n",
    "        x = C.dot(z)\n",
    "\n",
    "\n",
    "    for i_patt in range(p['patts'].shape[0]):\n",
    "        xFadeTime=int(p['patts'][i_patt].shape[0] * xFade)\n",
    "        for n in range(int(p['patts'][i_patt].shape[0] * speed)):\n",
    "            C = patternCs[i_patt]\n",
    "            v=int(p['patts'][i_patt].shape[0] * speed)\n",
    "            stepL = min(v - n - 1, xFadeTime)\n",
    "            stepU = min(n, xFadeTime)\n",
    "            m1 = 1.0\n",
    "            if(n > v-xFadeTime-1 and i_patt < p['patts'].shape[0]-1):\n",
    "                m1 = (stepL + stepU) / (2*xFadeTime)\n",
    "                nextC = patternCs[i_patt+1]\n",
    "                C = (m1 * C) + ((1.0-m1) * nextC)\n",
    "            else:\n",
    "                if (n < xFadeTime and i_patt > 0):\n",
    "                    m1 = 0.5 - (n / (2*xFadeTime))\n",
    "                    prevC = patternCs[i_patt-1]\n",
    "                    C = (m1 * prevC) + ((1.0-m1) * C)\n",
    "#                 else:\n",
    "#                     C = cNet['Cs'][0,0]\n",
    "            \n",
    "            xOld = x\n",
    "            Wtarget = (Wmod.dot(x))\n",
    "            z = (LROneMinus.dot(xOld)) + (LR.dot(tanh(Wtarget + bestNet['Wbias'])))\n",
    "            x = C.dot(z)\n",
    "\n",
    "            newSample = bestNet['Wout'].dot(np.concatenate((x[:,0], np.array([1]))))\n",
    "            audio = np.concatenate((audio, newSample))\n",
    "    return audio\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def evalModel(genome, patterns, patternLengths, orgAudio, N=900):\n",
    "    LR =genome[0]\n",
    "    modelParams = {'N':N, 'NetSR':1.5, 'NetinpScaling':1.2,'BiasScaling':0.3, 'TychonovAlpha':0.0001,\n",
    "             'washoutLength':50, 'learnLength':4, 'TychonovAlphaReadout':0.0001,\n",
    "              'LR': LR,\n",
    "              'patts':patterns\n",
    "             }   \n",
    "    newNetwork  = makeLoadedNetwork_v2(modelParams)\n",
    "    \n",
    "    import scipy\n",
    "    def fitnessf(aperture, *args):\n",
    "        print('Pattern: ', args[0])\n",
    "        params = args[1]\n",
    "        net = args[2]\n",
    "        try:\n",
    "            C = computeConceptor(params, net, args[0], aperture)\n",
    "        except:\n",
    "            return 999\n",
    "        atten = testConceptor(params, C, net, params['patts'][args[0]].size * params['learnLength'])\n",
    "        return atten\n",
    "\n",
    "    apertures = [scipy.optimize.fminbound(fitnessf, 0,1000,  disp=2, xtol=15, args = (x,modelParams, newNetwork)) \n",
    "                 for x in np.arange(modelParams['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",
    "        patternCs[i_patt] = computeConceptor(modelParams, newNetwork, i_patt, apertures[i_patt])\n",
    "        \n",
    "#     figsize(20,3)\n",
    "    audio = render(modelParams, patternCs, newNetwork, 1.0,1.0, 0.05)\n",
    "    error = compareMFCCs(audio, orgAudio)\n",
    "\n",
    "#     plot(audio)\n",
    "    return {\"error\":error, \"waveform\":audio, 'apertures':apertures, 'net':newNetwork}\n",
    "#     plt.savefig(filepre('TwoRecon.pdf'))\n",
    "#     librosa.output.write_wav(filepre(\"twoRecon.wav\"), librosa.resample(audio, sr, 44100), 44100)\n",
    "#     Audio(data=audio,rate=sr)        \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def evalFitness(genome, data):\n",
    "\n",
    "    modelData = evalModel(genome, data['patterns'], data['patternLengths'], data['orgAudio'], data['N'])\n",
    "    if (\"winner\" in data):\n",
    "        if modelData['error'] < data['winner']['error']:\n",
    "            data['winner'] = modelData\n",
    "    else:\n",
    "        data['winner'] = modelData\n",
    "    return modelData['error']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0:36:46.691288\n",
      "Runtime: None\n",
      "0 / 1  :  909a_22k.wav\n",
      "[0.63502185 0.34583375 0.11817947 0.30028384 0.4683209  0.11732662\n",
      " 0.2487437  0.1632119  0.20836699 0.        ]\n",
      "hi res trial 9\n",
      "Loading patterns: 0  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  \n",
      "NRMSE readout:  0.0124745745973169  :: absSize readout:  0.02928236091128867\n",
      "NRMSE W:  0.005084365444310596  :: absSize W:  0.030111976402458968\n",
      "Pattern:  0\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  0\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  0\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  0\n",
      "Computing conceptor, alpha:  427.8957449168695\n",
      "Pattern:  0\n",
      "Computing conceptor, alpha:  376.9660055846336\n",
      "Pattern:  0\n",
      "Computing conceptor, alpha:  323.1477478040605\n",
      "Pattern:  0\n",
      "Computing conceptor, alpha:  350.7330062624684\n",
      "Pattern:  0\n",
      "Computing conceptor, alpha:  366.94589147041944\n",
      "Pattern:  0\n",
      "Computing conceptor, alpha:  371.96599999332415\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  1\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  1\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  1\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  1\n",
      "Computing conceptor, alpha:  145.89803375031545\n",
      "Pattern:  1\n",
      "Computing conceptor, alpha:  214.52210404657677\n",
      "Pattern:  1\n",
      "Computing conceptor, alpha:  291.7960675006309\n",
      "Pattern:  1\n",
      "Computing conceptor, alpha:  257.35421375199786\n",
      "Pattern:  1\n",
      "Computing conceptor, alpha:  227.83818615796642\n",
      "Pattern:  1\n",
      "Computing conceptor, alpha:  244.19859625557302\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  2\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  2\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  2\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  2\n",
      "Computing conceptor, alpha:  145.89803375031545\n",
      "Pattern:  2\n",
      "Computing conceptor, alpha:  90.16994374947424\n",
      "Pattern:  2\n",
      "Computing conceptor, alpha:  130.08984128190014\n",
      "Pattern:  2\n",
      "Computing conceptor, alpha:  180.33988749894849\n",
      "Pattern:  2\n",
      "Computing conceptor, alpha:  159.0536512467403\n",
      "Pattern:  2\n",
      "Computing conceptor, alpha:  140.8980315862979\n",
      "Pattern:  2\n",
      "Computing conceptor, alpha:  150.92303249095693\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  3\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  3\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  3\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  3\n",
      "Computing conceptor, alpha:  145.89803375031545\n",
      "Pattern:  3\n",
      "Computing conceptor, alpha:  202.96930078170206\n",
      "Pattern:  3\n",
      "Computing conceptor, alpha:  207.96930379222331\n",
      "Pattern:  3\n",
      "Computing conceptor, alpha:  181.17001655673369\n",
      "Pattern:  3\n",
      "Computing conceptor, alpha:  194.64271513818355\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  4\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  4\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  4\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  4\n",
      "Computing conceptor, alpha:  485.8448768413173\n",
      "Pattern:  4\n",
      "Computing conceptor, alpha:  326.23792124926393\n",
      "Pattern:  4\n",
      "Computing conceptor, alpha:  423.65881214630974\n",
      "Pattern:  4\n",
      "Computing conceptor, alpha:  360.67977499789697\n",
      "Pattern:  4\n",
      "Computing conceptor, alpha:  355.24410634608404\n",
      "Pattern:  4\n",
      "Computing conceptor, alpha:  365.6797803476426\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  5\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  5\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  5\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  5\n",
      "Computing conceptor, alpha:  433.26238253559694\n",
      "Pattern:  5\n",
      "Computing conceptor, alpha:  376.9660055846336\n",
      "Pattern:  5\n",
      "Computing conceptor, alpha:  401.55948158162886\n",
      "Pattern:  5\n",
      "Computing conceptor, alpha:  386.96601691557663\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  6\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  6\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  6\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  6\n",
      "Computing conceptor, alpha:  96.77597273506467\n",
      "Pattern:  6\n",
      "Computing conceptor, alpha:  167.3081255372062\n",
      "Pattern:  6\n",
      "Computing conceptor, alpha:  291.7960675006309\n",
      "Pattern:  6\n",
      "Computing conceptor, alpha:  177.51362512216917\n",
      "Pattern:  6\n",
      "Computing conceptor, alpha:  182.51362775512172\n",
      "Pattern:  6\n",
      "Computing conceptor, alpha:  199.73952578739497\n",
      "Pattern:  6\n",
      "Computing conceptor, alpha:  213.61575958287042\n",
      "Pattern:  6\n",
      "Computing conceptor, alpha:  204.73952875001092\n",
      "Pattern:  6\n",
      "Computing conceptor, alpha:  193.15981822580653\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  7\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  7\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  7\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  7\n",
      "Computing conceptor, alpha:  145.89803375031545\n",
      "Pattern:  7\n",
      "Computing conceptor, alpha:  189.178687217225\n",
      "Pattern:  7\n",
      "Computing conceptor, alpha:  90.16994374947424\n",
      "Pattern:  7\n",
      "Computing conceptor, alpha:  114.09028500251137\n",
      "Pattern:  7\n",
      "Computing conceptor, alpha:  55.72809000084122\n",
      "Pattern:  7\n",
      "Computing conceptor, alpha:  34.44185374863304\n",
      "Pattern:  7\n",
      "Computing conceptor, alpha:  68.88370749726606\n",
      "Pattern:  7\n",
      "Computing conceptor, alpha:  47.597471245057875\n",
      "Pattern:  7\n",
      "Computing conceptor, alpha:  60.75308874148271\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  8\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  8\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  8\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  8\n",
      "Computing conceptor, alpha:  313.2930956948079\n",
      "Pattern:  8\n",
      "Computing conceptor, alpha:  472.1359549995793\n",
      "Pattern:  8\n",
      "Computing conceptor, alpha:  416.4078649987381\n",
      "Pattern:  8\n",
      "Computing conceptor, alpha:  355.7352916145329\n",
      "Pattern:  8\n",
      "Computing conceptor, alpha:  339.52381533038005\n",
      "Pattern:  8\n",
      "Computing conceptor, alpha:  329.50457197896077\n",
      "Pattern:  8\n",
      "Computing conceptor, alpha:  345.7160482631136\n",
      "Pattern:  8\n",
      "Computing conceptor, alpha:  334.523810294428\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  9\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  9\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  9\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  9\n",
      "Computing conceptor, alpha:  328.90425494580126\n",
      "Pattern:  9\n",
      "Computing conceptor, alpha:  348.954530411434\n",
      "Pattern:  9\n",
      "Computing conceptor, alpha:  293.4439523504401\n",
      "Pattern:  9\n",
      "Computing conceptor, alpha:  271.5282800951508\n",
      "Pattern:  9\n",
      "Computing conceptor, alpha:  306.988582690512\n",
      "Pattern:  9\n",
      "Computing conceptor, alpha:  285.07291043522264\n",
      "Pattern:  9\n",
      "Computing conceptor, alpha:  298.6175407752945\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  10\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  10\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  10\n",
      "Computing conceptor, alpha:  763.9320225002102\n",
      "Pattern:  10\n",
      "Computing conceptor, alpha:  788.5235933794428\n",
      "Pattern:  10\n",
      "Computing conceptor, alpha:  704.6083997091602\n",
      "Pattern:  10\n",
      "Computing conceptor, alpha:  741.272414929807\n",
      "Pattern:  10\n",
      "Computing conceptor, alpha:  773.3251667393249\n",
      "Pattern:  10\n",
      "Computing conceptor, alpha:  755.2768225800506\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  11\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  11\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  11\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  11\n",
      "Computing conceptor, alpha:  352.9338315535362\n",
      "Pattern:  11\n",
      "Computing conceptor, alpha:  386.96601691557663\n",
      "Pattern:  11\n",
      "Computing conceptor, alpha:  370.8767053735104\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pattern:  11\n",
      "Computing conceptor, alpha:  376.9660055846336\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  12\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  12\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  12\n",
      "Computing conceptor, alpha:  763.9320225002102\n",
      "Pattern:  12\n",
      "Computing conceptor, alpha:  854.1019662496844\n",
      "Pattern:  12\n",
      "Computing conceptor, alpha:  909.8300562505257\n",
      "Pattern:  12\n",
      "Computing conceptor, alpha:  849.101953581305\n",
      "Pattern:  12\n",
      "Computing conceptor, alpha:  859.1019789180639\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  13\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  13\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  13\n",
      "Computing conceptor, alpha:  763.9320225002102\n",
      "Pattern:  13\n",
      "Computing conceptor, alpha:  583.582228968205\n",
      "Pattern:  13\n",
      "Computing conceptor, alpha:  623.0339979168202\n",
      "Pattern:  13\n",
      "Computing conceptor, alpha:  604.8745874855359\n",
      "Pattern:  13\n",
      "Computing conceptor, alpha:  613.007544738508\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  14\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  14\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  14\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  14\n",
      "Computing conceptor, alpha:  481.3294292165482\n",
      "Pattern:  14\n",
      "Computing conceptor, alpha:  496.99046454183\n",
      "Pattern:  14\n",
      "Computing conceptor, alpha:  440.00304666960926\n",
      "Pattern:  14\n",
      "Computing conceptor, alpha:  417.8348717456402\n",
      "Pattern:  14\n",
      "Computing conceptor, alpha:  448.1723050511703\n",
      "Pattern:  14\n",
      "Computing conceptor, alpha:  432.4953013884988\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  15\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  15\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  15\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  15\n",
      "Computing conceptor, alpha:  361.6423134780021\n",
      "Pattern:  15\n",
      "Computing conceptor, alpha:  334.26729411075513\n",
      "Pattern:  15\n",
      "Computing conceptor, alpha:  351.1859865224004\n",
      "Pattern:  15\n",
      "Computing conceptor, alpha:  369.4052752498649\n",
      "Pattern:  15\n",
      "Computing conceptor, alpha:  356.64230811397977\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  16\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  16\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  16\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  16\n",
      "Computing conceptor, alpha:  409.8675582112209\n",
      "Pattern:  16\n",
      "Computing conceptor, alpha:  489.38005936025024\n",
      "Pattern:  16\n",
      "Computing conceptor, alpha:  429.2977406820931\n",
      "Pattern:  16\n",
      "Computing conceptor, alpha:  399.21011561077603\n",
      "Pattern:  16\n",
      "Computing conceptor, alpha:  417.28922750748166\n",
      "Pattern:  16\n",
      "Computing conceptor, alpha:  404.8675521319026\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  17\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  17\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  17\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Pattern:  17\n",
      "Computing conceptor, alpha:  461.70810805220094\n",
      "Pattern:  17\n",
      "Computing conceptor, alpha:  498.88743758157653\n",
      "Pattern:  17\n",
      "Computing conceptor, alpha:  450.1040097292956\n",
      "Pattern:  17\n",
      "Computing conceptor, alpha:  466.7081149004389\n",
      "Pattern:  17\n",
      "Computing conceptor, alpha:  456.708101203963\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Pattern:  18\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Pattern:  18\n",
      "Computing conceptor, alpha:  618.0339887498948\n",
      "Pattern:  18\n",
      "Computing conceptor, alpha:  763.9320225002102\n",
      "Pattern:  18\n",
      "Computing conceptor, alpha:  555.5325171185367\n",
      "Pattern:  18\n",
      "Computing conceptor, alpha:  673.762078750736\n",
      "Pattern:  18\n",
      "Computing conceptor, alpha:  594.1605509336033\n",
      "Pattern:  18\n",
      "Computing conceptor, alpha:  602.8712532647166\n",
      "Pattern:  18\n",
      "Computing conceptor, alpha:  579.4059549348282\n",
      "Pattern:  18\n",
      "Computing conceptor, alpha:  588.5247967523444\n",
      "\n",
      "Optimization terminated successfully;\n",
      "The returned value satisfies the termination criteria\n",
      "(using xtol =  15 )\n",
      "Computing conceptor, alpha:  371.96599999332415\n",
      "Computing conceptor, alpha:  236.0679774997897\n",
      "Computing conceptor, alpha:  145.89803375031545\n",
      "Computing conceptor, alpha:  202.96930078170206\n",
      "Computing conceptor, alpha:  360.67977499789697\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Computing conceptor, alpha:  199.73952578739497\n",
      "Computing conceptor, alpha:  60.75308874148271\n",
      "Computing conceptor, alpha:  339.52381533038005\n",
      "Computing conceptor, alpha:  293.4439523504401\n",
      "Computing conceptor, alpha:  763.9320225002102\n",
      "Computing conceptor, alpha:  381.9660112501051\n",
      "Computing conceptor, alpha:  854.1019662496844\n",
      "Computing conceptor, alpha:  613.007544738508\n",
      "Computing conceptor, alpha:  440.00304666960926\n",
      "Computing conceptor, alpha:  361.6423134780021\n",
      "Computing conceptor, alpha:  409.8675582112209\n",
      "Computing conceptor, alpha:  461.70810805220094\n",
      "Computing conceptor, alpha:  594.1605509336033\n",
      "[0.63502185 0.34583375 0.11817947 0.30028384 0.4683209  0.11732662\n",
      " 0.2487437  0.1632119  0.20836699 0.29482693]\n"
     ]
    },
    {
     "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": [
    "startTS = datetime.now()\n",
    "testing = False\n",
    "learningRates = linspace(0.05,0.95,10)\n",
    "# for currentIxi in range(len(ixiFiles)):\n",
    "for currentIxi in range(1) if testing else range(len(ixiFiles)):\n",
    "    def log(msg):\n",
    "        f = open(resultsFolderName + ixistr(currentIxi) + \"searchLog.txt\", \"a\")\n",
    "        f.write(str(datetime.now()) + \":\")\n",
    "        f.write(msg)\n",
    "        f.write('\\r\\n')\n",
    "        f.close()  \n",
    "        print(msg)\n",
    "    clear_output()\n",
    "    print(\"loading: \", ixiFiles[currentIxi])\n",
    "    y, sr = librosa.load(\"audios/\" + ixiFiles[currentIxi], sr=22050)\n",
    "    y = y[:5000] / np.max(y) * 0.5\n",
    "    print(sr)\n",
    "\n",
    "    #divide out windows\n",
    "    patterns = []\n",
    "\n",
    "    minPatternSize = 9\n",
    "    lastCrossing=0\n",
    "    for i in range(y.shape[0]-1):\n",
    "        if (i-lastCrossing) > minPatternSize and y[i] >=0 and y[i+1] < 0:\n",
    "            print(i)\n",
    "            segment = y[lastCrossing:i]\n",
    "            patterns.append(segment)\n",
    "            lastCrossing = i\n",
    "            \n",
    "    #convert to numpy\n",
    "    patterns = np.array(patterns, dtype=np.object)\n",
    "    maxPatterns = 150\n",
    "    patterns = patterns[:maxPatterns]\n",
    "    patternLengths = [x.shape[0] for x in patterns]\n",
    "    y = y[:sum(patternLengths)]\n",
    "    maxPatternLen = np.max(patternLengths)\n",
    "    minPatternLen = np.min(patternLengths)\n",
    "    print(\"max length: \", maxPatternLen)\n",
    "    print(\"min length: \", minPatternLen)\n",
    "    # for p in patterns:\n",
    "    #     plot(p)\n",
    "    print(patterns.shape)\n",
    "    \n",
    "    \n",
    "    data = {'patterns':patterns, 'patternLengths':patternLengths, 'orgAudio':y}\n",
    "    def onEpochStart():\n",
    "        clear_output()\n",
    "        print('Runtime:', print(datetime.now() - startTS))\n",
    "        print(currentIxi, '/', len(ixiFiles), ' : ', ixiFiles[currentIxi])\n",
    "        log(str(scores))\n",
    "\n",
    "#     brute force search of learning rates\n",
    "#     do the search at low res with smaller N\n",
    "    data['N'] = 600\n",
    "    scores = zeros_like(learningRates)\n",
    "    for i,l in enumerate(learningRates):\n",
    "        trials = 5\n",
    "        trialScores = zeros(trials)\n",
    "        for trial in range(trials):\n",
    "            onEpochStart()\n",
    "            log(\"lo res trial\" + str(trial))\n",
    "            log(str(scores))\n",
    "            log(str(trialScores))\n",
    "            trialScores[trial] = evalFitness(array([l]), data)\n",
    "        scores[i] = np.median(trialScores)\n",
    "        log(str(trialScores))\n",
    "    \n",
    "    winningScore = np.min(scores)\n",
    "    log(\"Winning score: \" + str(winningScore))\n",
    "    bestLR = learningRates[np.argmin(scores)]\n",
    "#     now generate the best of x at high res to find a good network\n",
    "    data['N'] = 900\n",
    "    trials = 10\n",
    "    scores = zeros(trials)\n",
    "    del data['winner']\n",
    "    for trial in range(trials):\n",
    "        onEpochStart()\n",
    "        log(\"hi res trial \" + str(trial))\n",
    "        scores[trial] = evalFitness(array([bestLR]), data)\n",
    "        log(str(scores))\n",
    "        \n",
    "        \n",
    "    \n",
    "#     microbial.evolve(pop, evalFitness, microbial.criterionMin, data, 3, 1 if testing else 15, 0.0, 1.0, 0.5 , onEpochStart)\n",
    "    \n",
    "    plt.close()\n",
    "    figsize(20,4)\n",
    "    plt.xlabel(\"Time (samples)\", fontsize=20)\n",
    "    plt.ylabel(\"Amplitude\", fontsize=20)\n",
    "    plt.xticks(fontsize=18, rotation=0)\n",
    "    plt.yticks(fontsize=18, rotation=0) \n",
    "    plot(y, label='Original', alpha=0.6)\n",
    "    plot(data['winner']['waveform'], alpha=1.0, label='Reconstruction')\n",
    "    plt.legend(fontsize=20)\n",
    "    plt.savefig(resultsFolderName + ixistr(currentIxi) + \"compare.pdf\", bbox_inches='tight')\n",
    "\n",
    "    librosa.output.write_wav(resultsFolderName + ixistr(currentIxi) + \"org.wav\",y, sr)\n",
    "    librosa.output.write_wav(resultsFolderName + ixistr(currentIxi) + \"recon.wav\", data['winner']['waveform'], sr)\n",
    "\n",
    "    import dill as pickle\n",
    "    with open(resultsFolderName + ixistr(currentIxi) + r\"model.dill.pickled\", \"wb\") as output_file:    \n",
    "        pickle.dump({'winner':data['winner'], 'original':y, 'patterns':patterns}, output_file, protocol=0)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'BiasScaling': 0.3,\n",
       " 'LR': 0.15,\n",
       " 'N': 900,\n",
       " 'NetSR': 1.5,\n",
       " 'NetinpScaling': 1.2,\n",
       " 'TychonovAlpha': 0.0001,\n",
       " 'TychonovAlphaReadout': 0.0001,\n",
       " 'learnLength': 4,\n",
       " 'patts': array([array([0.0000000e+00, 0.0000000e+00, 0.0000000e+00, 0.0000000e+00,\n",
       "        0.0000000e+00, 2.8227856e-05, 2.8227856e-05, 2.8227856e-05,\n",
       "        0.0000000e+00, 2.8227856e-05, 0.0000000e+00, 0.0000000e+00,\n",
       "        2.8227856e-05, 0.0000000e+00, 2.8227856e-05, 0.0000000e+00],\n",
       "       dtype=float32),\n",
       "        array([ 5.64557122e-05, -5.64557122e-05,  8.46835683e-05, -8.46835683e-05,\n",
       "         1.41139273e-04, -1.69367137e-04,  2.82278546e-04, -4.23417834e-04,\n",
       "         1.72189914e-03,  1.27307624e-02,  2.37396266e-02,  3.28007676e-02,\n",
       "         4.23135534e-02,  4.86365929e-02,  5.37176095e-02,  5.73872291e-02,\n",
       "         5.97583689e-02,  5.97301424e-02,  5.85727990e-02,  5.78388758e-02,\n",
       "         5.43950759e-02,  5.33224195e-02,  5.23908995e-02,  5.23062162e-02,\n",
       "         4.89470996e-02,  3.92367207e-02,  2.83125397e-02,  2.45017782e-02,\n",
       "         1.87150687e-02,  2.08321568e-02,  2.82843113e-02,  3.06554511e-02,\n",
       "         2.56591197e-02,  1.61745604e-02,  1.22508891e-02,  1.78400036e-02,\n",
       "         1.22791175e-02,  3.35911475e-03], dtype=float32),\n",
       "        array([ 0.00214532, -0.00203241, -0.00270987, -0.01098064, -0.01346469,\n",
       "        -0.01809406, -0.0207757 , -0.02204596, -0.01998532, -0.01953368,\n",
       "        -0.01964659, -0.03782533, -0.06300458, -0.09416813, -0.12654547,\n",
       "        -0.1561565 , -0.18379156, -0.21481398, -0.24343702, -0.45949304,\n",
       "        -0.7292102 , -0.65691864, -0.52729636, -0.5391803 , -0.5870829 ,\n",
       "        -0.58530456, -0.5494834 , -0.54047877, -0.5396037 , -0.5407328 ,\n",
       "        -0.5387004 , -0.5428499 , -0.55182636, -0.5515441 , -0.5539999 ,\n",
       "        -0.5526167 , -0.5466324 , -0.53836167, -0.53378874, -0.52927226,\n",
       "        -0.5192514 , -0.5120815 , -0.5036696 , -0.49223733, -0.48407948,\n",
       "        -0.47431266, -0.4593519 , -0.43468073, -0.40255743, -0.38192287,\n",
       "        -0.3432225 , -0.29896122, -0.25190538, -0.20617625, -0.15937448,\n",
       "        -0.11971433, -0.08171964, -0.04383786, -0.00937165,  0.04047874,\n",
       "         0.08262293,  0.11979902,  0.16168916,  0.2069384 ,  0.25038108,\n",
       "         0.2864845 ,  0.3180997 ,  0.3535821 ,  0.3740191 ,  0.3877096 ,\n",
       "         0.40388414,  0.41511884,  0.43214023,  0.45204088,  0.46601367,\n",
       "         0.47518772,  0.47834924,  0.48021227,  0.4875233 ,  0.48794672,\n",
       "         0.48597077,  0.4782081 ,  0.47267544,  0.46581608,  0.46254164,\n",
       "         0.46033987,  0.45026252,  0.4453509 ,  0.43620506,  0.42330492,\n",
       "         0.41195732,  0.39598036,  0.36922035,  0.33975047,  0.30844578,\n",
       "         0.28236324,  0.25331676,  0.22347993,  0.1947722 ,  0.16016485,\n",
       "         0.12600915,  0.0838085 ,  0.04567267], dtype=float32),\n",
       "        array([ 0.01002089, -0.02738102, -0.06565799, -0.1047818 , -0.14413142,\n",
       "        -0.18926777, -0.23694462, -0.28481907, -0.32896742, -0.37441427,\n",
       "        -0.41331226, -0.45060125, -0.48800316, -0.5199853 , -0.54299104,\n",
       "        -0.5575001 , -0.57505786, -0.5899622 , -0.6033704 , -0.61892396,\n",
       "        -0.6348445 , -0.64418787, -0.64980525, -0.66197145, -0.6705809 ,\n",
       "        -0.67506915, -0.6743917 , -0.672698  , -0.66998816, -0.65993905,\n",
       "        -0.6525433 , -0.64068764, -0.62694067, -0.6138994 , -0.5961441 ,\n",
       "        -0.5769209 , -0.55388695, -0.5332524 , -0.5128155 , -0.48952746,\n",
       "        -0.46053746, -0.42688987, -0.39070174, -0.3469768 , -0.3016711 ,\n",
       "        -0.25613955, -0.21777791, -0.18297295, -0.14170383, -0.09840231,\n",
       "        -0.05854457, -0.01865861,  0.01953368,  0.05439508,  0.08702648,\n",
       "         0.11627053,  0.15370066,  0.18977587,  0.22607689,  0.26059955,\n",
       "         0.29342857,  0.32495907,  0.34875515,  0.3716197 ,  0.3927624 ,\n",
       "         0.41238073,  0.42522442,  0.43451136,  0.44501215,  0.45232314,\n",
       "         0.4608762 ,  0.469401  ,  0.47662735,  0.48148254,  0.48303506,\n",
       "         0.48512393,  0.4917857 ,  0.4956529 ,  0.49155986,  0.4886524 ,\n",
       "         0.48678938,  0.48091796,  0.47137696,  0.46319088,  0.45661378,\n",
       "         0.44947213,  0.43767288,  0.43143454,  0.42381302,  0.4092192 ,\n",
       "         0.39129454,  0.36868402,  0.34071022,  0.31505108,  0.28831932,\n",
       "         0.25884944,  0.23355727,  0.20047423,  0.16970587,  0.13967143,\n",
       "         0.11006041,  0.07677977,  0.04420482], dtype=float32),\n",
       "        array([ 0.01301304, -0.02052165, -0.05230622, -0.08874838, -0.12631965,\n",
       "        -0.16369334, -0.20518827, -0.24758652, -0.28761363, -0.3275278 ,\n",
       "        -0.36433694, -0.4030373 , -0.4406086 , -0.47380456, -0.5019195 ,\n",
       "        -0.5240501 , -0.5433015 , -0.5597019 , -0.57440865, -0.58824027,\n",
       "        -0.59998304, -0.61375827, -0.6232146 , -0.6317112 , -0.64308697,\n",
       "        -0.65056735, -0.65914863, -0.66428614, -0.6705809 , -0.6750974 ,\n",
       "        -0.67091966, -0.6640603 , -0.6564106 , -0.6498617 , -0.64373624,\n",
       "        -0.634167  , -0.6243155 , -0.6130808 , -0.60105574, -0.5892283 ,\n",
       "        -0.5791791 , -0.56684357, -0.55072546, -0.5340146 , -0.51109356,\n",
       "        -0.48362786, -0.45658556, -0.42432112, -0.3877943 , -0.3504206 ,\n",
       "        -0.30951843, -0.27124146, -0.23477107, -0.19996613, -0.16428612,\n",
       "        -0.12911421, -0.09374471, -0.05702027, -0.02159431,  0.0123638 ,\n",
       "         0.04505166,  0.07793711,  0.11014509,  0.14009485,  0.16815333,\n",
       "         0.19784904,  0.2248631 ,  0.25196183,  0.27869362,  0.3032801 ,\n",
       "         0.3287134 ,  0.35214248,  0.37204313,  0.38745555,  0.40320668,\n",
       "         0.41822392,  0.4299667 ,  0.43835035,  0.44614124,  0.45348048,\n",
       "         0.46124315,  0.46722746,  0.46945745,  0.47400215,  0.47947836,\n",
       "         0.48424885,  0.48927343,  0.49455202,  0.49661267,  0.49331   ,\n",
       "         0.48924518,  0.48534974,  0.48009935,  0.47360694,  0.46827188,\n",
       "         0.46149722,  0.4564444 ,  0.44907695,  0.4401852 ,  0.43177328,\n",
       "         0.42242986,  0.41288882,  0.39970642,  0.38426578,  0.36509907,\n",
       "         0.34381527,  0.321233  ,  0.29771918,  0.2722012 ,  0.2456388 ,\n",
       "         0.21870942,  0.19175182,  0.1645684 ,  0.13656637,  0.1085361 ,\n",
       "         0.07855812,  0.04889065], dtype=float32),\n",
       "        array([ 0.01953368, -0.01024671, -0.03988596, -0.07102128, -0.10156383,\n",
       "        -0.1334613 , -0.16665725, -0.20016372, -0.23341614, -0.26734602,\n",
       "        -0.3009654 , -0.3340202 , -0.3662282 , -0.39812568, -0.4299949 ,\n",
       "        -0.45983177, -0.48721278, -0.5113476 , -0.53260314, -0.54979396,\n",
       "        -0.5647547 , -0.57757014, -0.5896799 , -0.6003783 , -0.6117823 ,\n",
       "        -0.6216621 , -0.630328  , -0.6377237 , -0.64384913, -0.65045446,\n",
       "        -0.6550556 , -0.6614351 , -0.6664879 , -0.6719923 , -0.6730085 ,\n",
       "        -0.6680122 , -0.66290295, -0.6572291 , -0.65248686, -0.64678484,\n",
       "        -0.64068764, -0.6329532 , -0.62510586, -0.6158471 , -0.60622144,\n",
       "        -0.5961158 , -0.58445776, -0.5718117 , -0.5574155 , -0.54172075,\n",
       "        -0.52354205, -0.5033309 , -0.47964773, -0.4538757 , -0.42519617,\n",
       "        -0.39465365, -0.36250213, -0.33020946, -0.2976063 , -0.26452324,\n",
       "        -0.23211765, -0.20002258, -0.16877435, -0.13727206, -0.10735054,\n",
       "        -0.07720318, -0.04807204, -0.01984418,  0.00869418,  0.03680912,\n",
       "         0.06500875,  0.09238977,  0.11931914,  0.14520408,  0.17049624,\n",
       "         0.19584486,  0.2207136 ,  0.24459437,  0.26822108,  0.29100096,\n",
       "         0.3125388 ,  0.33297578,  0.35236832,  0.37001073,  0.38601592,\n",
       "         0.40015808,  0.41249365,  0.42386946,  0.43357986,  0.44207644,\n",
       "         0.44885114,  0.4551177 ,  0.46053746,  0.46559024,  0.47055835,\n",
       "         0.47482076,  0.47868797,  0.4823576 ,  0.4857167 ,  0.48859593,\n",
       "         0.4918986 ,  0.49390277,  0.49175745,  0.48837012,  0.4854062 ,\n",
       "         0.48168012,  0.47778466,  0.47397393,  0.4697115 ,  0.46479985,\n",
       "         0.4596624 ,  0.45424265,  0.44794783,  0.4411449 ,  0.4336363 ,\n",
       "         0.42465985,  0.41492125,  0.40334782,  0.39019364,  0.3752611 ,\n",
       "         0.3587196 ,  0.3408231 ,  0.32131767,  0.3010783 ,  0.28004855,\n",
       "         0.25825664,  0.23573081,  0.2129227 ,  0.18952182,  0.16575396,\n",
       "         0.14147802,  0.11711738,  0.09219217,  0.06673065,  0.04152317],\n",
       "       dtype=float32),\n",
       "        array([ 0.01549709, -0.01041608, -0.03711963, -0.06404901, -0.09140179,\n",
       "        -0.11881104, -0.14658725, -0.17484334, -0.20346639, -0.23245639,\n",
       "        -0.26141816, -0.29122677, -0.32018855, -0.34937617, -0.37805566,\n",
       "        -0.40622705, -0.43324113, -0.45870265, -0.48252696, -0.5036132 ,\n",
       "        -0.5225823 , -0.5388133 , -0.5540846 , -0.5676339 , -0.5789251 ,\n",
       "        -0.58934116, -0.5989104 , -0.6073788 , -0.61508495, -0.62253714,\n",
       "        -0.6292271 , -0.635409  , -0.64150625, -0.6471236 , -0.6518094 ,\n",
       "        -0.65652347, -0.6619432 , -0.666629  , -0.6708632 , -0.6728109 ,\n",
       "        -0.66936713, -0.6649918 , -0.66016483, -0.655846  , -0.6509061 ,\n",
       "        -0.64483714, -0.6390504 , -0.63247335, -0.62623495, -0.619178  ,\n",
       "        -0.6116976 , -0.6033986 , -0.59442216, -0.5848247 , -0.5740981 ,\n",
       "        -0.5628352 , -0.5493141 , -0.5346638 , -0.5190538 , -0.5006492 ,\n",
       "        -0.48049456, -0.45830745, -0.4338339 , -0.4079772 , -0.38127363,\n",
       "        -0.35372326, -0.32572123, -0.29771918, -0.2697736 , -0.24174336,\n",
       "        -0.21444702, -0.18743296, -0.1604189 , -0.13422345, -0.10836674,\n",
       "        -0.08282053, -0.05721786, -0.03198216, -0.00748038,  0.01671089,\n",
       "         0.04101507,  0.06489584,  0.08787331,  0.11113307,  0.1334613 ,\n",
       "         0.1559589 ,  0.17763789,  0.19878055,  0.22048777,  0.24120702,\n",
       "         0.26141816,  0.28086716,  0.29980806,  0.31832552,  0.33605263,\n",
       "         0.35265058,  0.36792186,  0.38243097,  0.3953029 ,  0.40648112,\n",
       "         0.4166996 ,  0.42607126,  0.43400326,  0.44103202,  0.44806075,\n",
       "         0.45342404,  0.4585615 ,  0.4636425 ,  0.46782026,  0.47196975,\n",
       "         0.4752724 ,  0.47857505,  0.48179305,  0.48444647,  0.48715633,\n",
       "         0.48980975,  0.49212444,  0.49421328,  0.49319708,  0.49062836,\n",
       "         0.48851126,  0.48566025,  0.48280925,  0.47942188,  0.476091  ,\n",
       "         0.47264722,  0.46914697,  0.46499747,  0.46031162,  0.45565403,\n",
       "         0.45057303,  0.44453228,  0.43835035,  0.43177328,  0.42353073,\n",
       "         0.41463897,  0.40470275,  0.39352453,  0.38138655,  0.3673573 ,\n",
       "         0.35248122,  0.33661717,  0.31996274,  0.302405  ,  0.28419805,\n",
       "         0.26596284,  0.24699374,  0.22743183,  0.20744652,  0.18732005,\n",
       "         0.16725004,  0.14698245,  0.12609383,  0.10489471,  0.08352622,\n",
       "         0.06210128,  0.04030938], dtype=float32),\n",
       "        array([ 0.01848925, -0.00420595, -0.0269576 , -0.04985039, -0.07274318,\n",
       "        -0.09600294, -0.1201942 , -0.14365156, -0.16744764, -0.1920341 ,\n",
       "        -0.21673347, -0.24126348, -0.2660193 , -0.29128325, -0.31603906,\n",
       "        -0.3408796 , -0.36608705, -0.39089933, -0.4150624 , -0.4375882 ,\n",
       "        -0.45827922, -0.4784339 , -0.4976571 , -0.51425505, -0.5293852 ,\n",
       "        -0.5428499 , -0.55507255, -0.5657709 , -0.5766951 , -0.5856998 ,\n",
       "        -0.5937165 , -0.60170496, -0.6088466 , -0.61545193, -0.6218314 ,\n",
       "        -0.62826735, -0.6330097 , -0.63820356, -0.6434257 , -0.648168  ,\n",
       "        -0.6519788 , -0.6562129 , -0.66078585, -0.6639474 , -0.66733474,\n",
       "        -0.6714278 , -0.6717947 , -0.6678428 , -0.6642579 , -0.6603907 ,\n",
       "        -0.65590245, -0.6516118 , -0.64751875, -0.64308697, -0.63775194,\n",
       "        -0.6326145 , -0.62722296, -0.62115395, -0.6149156 , -0.60816914,\n",
       "        -0.6012815 , -0.59385765, -0.5852481 , -0.5762152 , -0.5664766 ,\n",
       "        -0.5558347 , -0.5449105 , -0.5322362 , -0.5174166 , -0.5016655 ,\n",
       "        -0.485011  , -0.46711454, -0.44735506, -0.4258454 , -0.4035454 ,\n",
       "        -0.38045505, -0.35733643, -0.33362502, -0.30954665, -0.28577882,\n",
       "        -0.26198274, -0.23866652, -0.21515271, -0.19197765, -0.16866143,\n",
       "        -0.14588155, -0.12406142, -0.10164851, -0.07996951, -0.05809293,\n",
       "        -0.03661153, -0.01575114,  0.00519393,  0.02585671,  0.046209  ,\n",
       "         0.06605318,  0.08564331,  0.10512053,  0.12488003,  0.1439903 ,\n",
       "         0.16233839,  0.18096878,  0.19965562,  0.21743917,  0.23525095,\n",
       "         0.2528369 ,  0.26983008,  0.28631514,  0.3026873 ,  0.31872073,\n",
       "         0.33342743,  0.34745666,  0.3612601 ,  0.37449896,  0.38626996,\n",
       "         0.39708123,  0.40698922,  0.41588098,  0.42372835,  0.43103936,\n",
       "         0.43770114,  0.44357252,  0.4490205 ,  0.4541015 ,  0.4585615 ,\n",
       "         0.46256986,  0.46669114,  0.4703043 ,  0.47346583,  0.4764862 ,\n",
       "         0.47959125,  0.4822729 ,  0.48458758,  0.48746684,  0.48952746,\n",
       "         0.49144697,  0.49328178,  0.49548355,  0.49599165,  0.49393103,\n",
       "         0.49192682,  0.48949924,  0.4870434 ,  0.4847005 ,  0.4822729 ,\n",
       "         0.47928074,  0.47685316,  0.47366342,  0.47027606,  0.46691695,\n",
       "         0.46327555,  0.4590414 ,  0.45483544,  0.4505448 ,  0.44495568,\n",
       "         0.43973354,  0.43386215,  0.42717212,  0.4198611 ,  0.4121267 ,\n",
       "         0.40317845,  0.39360923,  0.383052  ,  0.37130922,  0.3592559 ,\n",
       "         0.3464969 ,  0.332383  ,  0.3175069 ,  0.30263084,  0.28735957,\n",
       "         0.27186248,  0.2556597 ,  0.2391464 ,  0.22237904,  0.20563993,\n",
       "         0.1882798 ,  0.17131485,  0.15449105,  0.1360865 ,  0.11835939,\n",
       "         0.1010275 ,  0.08262293,  0.06444419,  0.04651951,  0.02783266],\n",
       "       dtype=float32),\n",
       "        array([ 0.00844013, -0.01067013, -0.02952634, -0.04886242, -0.06825496,\n",
       "        -0.08750635, -0.10785864, -0.12843674, -0.14864789, -0.1690284 ,\n",
       "        -0.18921131, -0.2105798 , -0.23180714, -0.25238526, -0.27338678,\n",
       "        -0.2949811 , -0.3161802 , -0.33689946, -0.3580139 , -0.37808388,\n",
       "        -0.39854908, -0.41856265, -0.43801162, -0.45633152, -0.47298595,\n",
       "        -0.48961216, -0.5043471 , -0.51730365, -0.52983683, -0.54189014,\n",
       "        -0.55202395, -0.56105685, -0.57000506, -0.5778807 , -0.58493763,\n",
       "        -0.5927285 , -0.59955966, -0.6052899 , -0.61090726, -0.61646813,\n",
       "        -0.6216621 , -0.62631965, -0.6307232 , -0.63515496, -0.63978434,\n",
       "        -0.643708  , -0.6468131 , -0.64991814, -0.6538418 , -0.6571445 ,\n",
       "        -0.6599955 , -0.66270536, -0.6661774 , -0.6686332 , -0.6667137 ,\n",
       "        -0.66403204, -0.6613222 , -0.6576808 , -0.65398294, -0.6507932 ,\n",
       "        -0.64641786, -0.64274824, -0.6395303 , -0.6348162 , -0.6297917 ,\n",
       "        -0.6248518 , -0.6204765 , -0.6154802 , -0.61017334, -0.6044995 ,\n",
       "        -0.59857166, -0.592051  , -0.5858974 , -0.57875574, -0.5709366 ,\n",
       "        -0.5625812 , -0.5533789 , -0.54412013, -0.5335347 , -0.52213067,\n",
       "        -0.50999266, -0.49649975, -0.48218822, -0.46708632, -0.45034721,\n",
       "        -0.43307176, -0.41486478, -0.39586744, -0.37624907, -0.35623553,\n",
       "        -0.33627844, -0.3163778 , -0.29647717, -0.2758426 , -0.2558008 ,\n",
       "        -0.23581551, -0.21602778, -0.19640942, -0.17698865, -0.15762435,\n",
       "        -0.13859877, -0.12013775, -0.1008299 , -0.08203015, -0.06413369,\n",
       "        -0.04646305, -0.02831254, -0.01016203,  0.00753684,  0.02489697,\n",
       "         0.04194659,  0.05896799,  0.07629989,  0.09301078,  0.1090442 ,\n",
       "         0.12527522,  0.14139333,  0.15756789,  0.17320612,  0.18844916,\n",
       "         0.20391802,  0.21924575,  0.23434766,  0.24882855,  0.26364818,\n",
       "         0.27781856,  0.2918478 ,  0.30539715,  0.31843844,  0.33164907,\n",
       "         0.3440129 ,  0.3558968 ,  0.36741376,  0.37814036,  0.38799188,\n",
       "         0.39730707,  0.4059448 ,  0.41407442,  0.42121607,  0.4277367 ,\n",
       "         0.4338339 ,  0.43925366,  0.44430643,  0.44896403,  0.45331112,\n",
       "         0.45729125,  0.461102  ,  0.4644329 ,  0.4676509 ,  0.4706995 ,\n",
       "         0.4737763 ,  0.4764015 ,  0.47846216,  0.4813696 ,  0.48326087,\n",
       "         0.48537797,  0.48786202,  0.4898662 ,  0.4918986 ,  0.49370518,\n",
       "         0.49554   ,  0.49712077,  0.49841923,  0.4976571 ,  0.49559647,\n",
       "         0.49376166,  0.49184215,  0.48978153,  0.4876362 ,  0.4854344 ,\n",
       "         0.48320442,  0.48077682,  0.47863153,  0.4758652 ,  0.4729577 ,\n",
       "         0.4701914 ,  0.4671428 ,  0.46417886,  0.46053746,  0.45712188,\n",
       "         0.45311353,  0.44868177,  0.44444758,  0.4396206 ,  0.43425733,\n",
       "         0.42883757,  0.4227686 ,  0.41588098,  0.40888047,  0.4014001 ,\n",
       "         0.39279062,  0.38324958,  0.37379324,  0.3638006 ,  0.3527635 ,\n",
       "         0.341303  ,  0.3294473 ,  0.31753513,  0.30488905,  0.29190424,\n",
       "         0.2789759 ,  0.26545474,  0.25182068,  0.2380455 ,  0.22449613,\n",
       "         0.21009992,  0.1957884 ,  0.18156156,  0.16634674,  0.1518941 ,\n",
       "         0.13713092,  0.12214193,  0.10737876,  0.09207927,  0.07652572,\n",
       "         0.06066166,  0.04505166,  0.02946988], dtype=float32),\n",
       "        array([ 0.01391633, -0.00220177, -0.01826342, -0.03446621, -0.05095128,\n",
       "        -0.06763394, -0.08434483, -0.1011404 , -0.1181618 , -0.13504206,\n",
       "        -0.1525151 , -0.16976233, -0.18712245, -0.20448259, -0.22201209,\n",
       "        -0.23956981, -0.2574945 , -0.27541918, -0.2930898 , -0.3112968 ,\n",
       "        -0.32879806, -0.34706149, -0.36509907, -0.3822616 , -0.39914188,\n",
       "        -0.41582453, -0.43236607, -0.44803253, -0.46285215, -0.47654265,\n",
       "        -0.4894428 , -0.5020324 , -0.51343644, -0.5243042 , -0.5341275 ,\n",
       "        -0.54355556, -0.55202395, -0.56040764, -0.56785977, -0.5746627 ,\n",
       "        -0.58143735, -0.5873652 , -0.5932931 , -0.59792244, -0.60317284,\n",
       "        -0.60867727, -0.6135324 , -0.6176537 , -0.6216903 , -0.6256986 ,\n",
       "        -0.62953764, -0.6331508 , -0.636397  , -0.63967144, -0.6432846 ,\n",
       "        -0.6465308 , -0.64901483, -0.65245867, -0.6556766 , -0.6581889 ,\n",
       "        -0.6608141 , -0.6638627 , -0.6665161 , -0.66775817, -0.66583866,\n",
       "        -0.6636087 , -0.6613786 , -0.65785015, -0.6549427 , -0.6512448 ,\n",
       "        -0.64833736, -0.6452606 , -0.64181674, -0.6387964 , -0.6356066 ,\n",
       "        -0.63165474, -0.628098  , -0.62420255, -0.62005305, -0.61610115,\n",
       "        -0.6113307 , -0.60704005, -0.6020155 , -0.5970756 , -0.5919099 ,\n",
       "        -0.5856998 , -0.57957435, -0.57322305, -0.56684357, -0.55950433,\n",
       "        -0.55193925, -0.54372495, -0.53477675, -0.52585673, -0.5163157 ,\n",
       "        -0.5057303 , -0.4941004 , -0.48278102, -0.46985266, -0.45568228,\n",
       "        -0.44156834, -0.42680517, -0.4109129 , -0.3945125 , -0.3778863 ,\n",
       "        -0.36111894, -0.34412578, -0.32738668, -0.31061932, -0.29340032,\n",
       "        -0.2765483 , -0.2594987 , -0.24275956, -0.22630271, -0.20964828,\n",
       "        -0.19279625, -0.17605713, -0.15982611, -0.14351042, -0.12719472,\n",
       "        -0.11104839, -0.09555129, -0.07946141, -0.06359736, -0.04889065,\n",
       "        -0.03367583, -0.01863038, -0.00366962,  0.01137583,  0.0259414 ,\n",
       "         0.04045052,  0.05521369,  0.06944052,  0.08361091,  0.09792243,\n",
       "         0.11175408,  0.12538813,  0.13936092,  0.15291029,  0.16626206,\n",
       "         0.17961384,  0.19251397,  0.2054141 ,  0.21763676,  0.231045  ,\n",
       "         0.24374753,  0.2556597 ,  0.26791057,  0.28004855,  0.29187602,\n",
       "         0.3037317 ,  0.31530514,  0.32574946,  0.33687124,  0.34776717,\n",
       "         0.35781628,  0.36741376,  0.37684187,  0.3857901 ,  0.39383504,\n",
       "         0.40154123,  0.40834415,  0.41494948,  0.42144188,  0.4269463 ,\n",
       "         0.43197086,  0.43708012,  0.44199175,  0.4463953 ,  0.45006493,\n",
       "         0.453791  ,  0.45748886,  0.46084797,  0.4638119 ,  0.4667758 ,\n",
       "         0.46982443,  0.4722238 ,  0.47470784,  0.47724834,  0.47964773,\n",
       "         0.48142606,  0.4834867 ,  0.48563203,  0.48726922,  0.4886242 ,\n",
       "         0.49051544,  0.49223733,  0.49370518,  0.49556822,  0.4970643 ,\n",
       "         0.4981934 ,  0.49960482,  0.5       ,  0.4986733 ,  0.49734658,\n",
       "         0.49590695,  0.49407214,  0.4922938 ,  0.4906848 ,  0.48893467,\n",
       "         0.48707163,  0.4854062 ,  0.48320442,  0.48108733,  0.4788009 ,\n",
       "         0.47645798,  0.4742562 ,  0.47146162,  0.46855417,  0.4659572 ,\n",
       "         0.46304974,  0.46022695,  0.4571501 ,  0.45333937,  0.449952  ,\n",
       "         0.44633883,  0.44247162,  0.43770114,  0.4333258 ,  0.42883757,\n",
       "         0.42324847,  0.41757467,  0.41167504,  0.4055496 ,  0.39809746,\n",
       "         0.39067352,  0.38319314,  0.37438604,  0.3653249 ,  0.35651782,\n",
       "         0.34694856,  0.33687124,  0.32679388,  0.31640604,  0.30545363,\n",
       "         0.29472703,  0.2838593 ,  0.27245525,  0.26107943,  0.24956247,\n",
       "         0.2380455 ,  0.22635917,  0.21424942,  0.20194207,  0.18977587,\n",
       "         0.17752498,  0.16510473,  0.15237397,  0.14001016,  0.12725118,\n",
       "         0.11392763,  0.10116863,  0.08866369,  0.07576356,  0.06266584,\n",
       "         0.04934229,  0.03615988,  0.02269519], dtype=float32),\n",
       "        array([ 0.00928696, -0.00457291, -0.01829165, -0.03158697, -0.04553153,\n",
       "        -0.05922204, -0.0734771 , -0.08747812, -0.10170496, -0.11587535,\n",
       "        -0.13027155, -0.1451194 , -0.15960029, -0.17444815, -0.18884435,\n",
       "        -0.20394625, -0.2186812 , -0.23330322, -0.24823576, -0.26330942,\n",
       "        -0.27875006, -0.29371083, -0.30895388, -0.3232654 , -0.33833906,\n",
       "        -0.35372326, -0.36820415, -0.38274148, -0.39741996, -0.4109129 ,\n",
       "        -0.42454696, -0.43784225, -0.45062947, -0.46324733, -0.4740586 ,\n",
       "        -0.48506746, -0.49528596, -0.50491166, -0.5144809 , -0.5230057 ,\n",
       "        -0.530966  , -0.5382205 , -0.544967  , -0.5517699 , -0.55809295,\n",
       "        -0.5637103 , -0.5690171 , -0.5744933 , -0.5796308 , -0.5843166 ,\n",
       "        -0.5895952 , -0.5936318 , -0.59735787, -0.6014791 , -0.6050923 ,\n",
       "        -0.60890305, -0.6127138 , -0.6160165 , -0.61821824, -0.6213798 ,\n",
       "        -0.62437195, -0.62702537, -0.6297634 , -0.6324451 , -0.63470334,\n",
       "        -0.63676393, -0.6393327 , -0.64142156, -0.64365155, -0.6455146 ,\n",
       "        -0.6476599 , -0.65000284, -0.6514989 , -0.65358776, -0.6538418 ,\n",
       "        -0.6510755 , -0.64870435, -0.6465308 , -0.64362335, -0.6407441 ,\n",
       "        -0.6380342 , -0.6354937 , -0.6319652 , -0.6292271 , -0.6266584 ,\n",
       "        -0.6231299 , -0.6199684 , -0.61646813, -0.6127138 , -0.60907245,\n",
       "        -0.60582626, -0.6023542 , -0.5992774 , -0.59498674, -0.5902162 ,\n",
       "        -0.58688533, -0.5827923 , -0.57757014, -0.57268673, -0.5680856 ,\n",
       "        -0.5627505 , -0.55679446, -0.5514312 , -0.54567266, -0.53875685,\n",
       "        -0.5328008 , -0.5258285 , -0.5182352 , -0.5105854 , -0.502851  ,\n",
       "        -0.49446735, -0.48458758, -0.4758652 , -0.4653362 , -0.4543838 ,\n",
       "        -0.44362897, -0.43174505, -0.41895783, -0.4064529 , -0.39355275,\n",
       "        -0.37960818, -0.36600238, -0.3520296 , -0.3382544 , -0.32382995,\n",
       "        -0.30974427, -0.2954045 , -0.28134704, -0.26745892, -0.2532603 ,\n",
       "        -0.23937221, -0.22545588, -0.21165246, -0.19784904, -0.18367866,\n",
       "        -0.16959296, -0.15655169, -0.14300232, -0.12973522, -0.1162423 ,\n",
       "        -0.10308813, -0.08993395, -0.07666685, -0.06404901, -0.05174166,\n",
       "        -0.03889798, -0.02619545, -0.01425507, -0.00177835,  0.0103314 ,\n",
       "         0.02207418,  0.03381697,  0.04615254,  0.05775419,  0.06904534,\n",
       "         0.08109863,  0.09258737,  0.10342686,  0.11483092,  0.12631965,\n",
       "         0.13713092,  0.14828092,  0.15914865,  0.16995992,  0.1806865 ,\n",
       "         0.19149777,  0.2017727 ,  0.21204765,  0.22288714,  0.23282336,\n",
       "         0.24298538,  0.25286514,  0.26268843,  0.2722012 ,  0.28157285,\n",
       "         0.29097274,  0.30020323,  0.3095749 ,  0.31784564,  0.3263987 ,\n",
       "         0.33472592,  0.34285554,  0.35061818,  0.35778806,  0.3645063 ,\n",
       "         0.37125275,  0.37754756,  0.3831649 ,  0.38906452,  0.39462543,\n",
       "         0.39922655,  0.40315023,  0.40786427,  0.41190088,  0.41590923,\n",
       "         0.4193248 ,  0.42231694,  0.42547846,  0.4286682 ,  0.43134984,\n",
       "         0.43369275,  0.43637443,  0.43877378,  0.44080618,  0.44241518,\n",
       "         0.4446734 ,  0.4463953 ,  0.44806075,  0.45020607,  0.45141986,\n",
       "         0.45274657,  0.45401683,  0.45534354,  0.4569525 ,  0.45816633,\n",
       "         0.4592672 ,  0.46033987,  0.46141252,  0.46254164,  0.46336025,\n",
       "         0.46460226,  0.46573138,  0.46657822,  0.46753797,  0.46711454,\n",
       "         0.4657596 ,  0.46420708,  0.46293682,  0.46144074,  0.45963416,\n",
       "         0.4583357 ,  0.45681137,  0.4549201 ,  0.45333937,  0.45178682,\n",
       "         0.45000848,  0.44825834,  0.44653645,  0.44430643,  0.44233048,\n",
       "         0.44024163,  0.43784225,  0.43572518,  0.43369275,  0.43143454,\n",
       "         0.42880934,  0.42609948,  0.4235025 ,  0.42067972,  0.41748998,\n",
       "         0.41455427,  0.41136453,  0.407582  ,  0.40388414,  0.4004121 ,\n",
       "         0.39640376,  0.39171794,  0.38739908,  0.38271326,  0.37749112,\n",
       "         0.3719302 ,  0.36667985,  0.36058262,  0.3540902 ,  0.34782362,\n",
       "         0.34099248,  0.3338226 ,  0.32648337,  0.31874895,  0.31044996,\n",
       "         0.3027155 ,  0.29433185,  0.2858635 ,  0.2773669 ,  0.26836222,\n",
       "         0.2594422 ,  0.2506916 ,  0.24137639,  0.23220234,  0.22328234,\n",
       "         0.21354373,  0.20417207,  0.19463105,  0.18542878,  0.17540789,\n",
       "         0.16575396,  0.15652345,  0.14667194,  0.13662282,  0.12648901,\n",
       "         0.11686332,  0.10692712,  0.09716028,  0.08665951,  0.07644103,\n",
       "         0.06622255,  0.05594761,  0.04572913,  0.03522836,  0.02481228,\n",
       "         0.01405747], dtype=float32),\n",
       "        array([ 0.00355671, -0.00683114, -0.01772709, -0.0288771 , -0.03934963,\n",
       "        -0.05047141, -0.06139559, -0.0719528 , -0.08344154, -0.09433749,\n",
       "        -0.10517699, -0.1166375 , -0.12719472, -0.1384294 , -0.15011573,\n",
       "        -0.16123751, -0.17312144, -0.1841303 , -0.19553435, -0.20727713,\n",
       "        -0.21859652, -0.23028284, -0.24185626, -0.25342968, -0.26522893,\n",
       "        -0.27711284, -0.28854513, -0.29977983, -0.31143793, -0.32337832,\n",
       "        -0.33447185, -0.34508553, -0.35615084, -0.36670807, -0.37698302,\n",
       "        -0.3868063 , -0.3968272 , -0.40611416, -0.41472363, -0.42316377,\n",
       "        -0.43072885, -0.4386891 , -0.44614124, -0.45263365, -0.45892847,\n",
       "        -0.46578783, -0.4709253 , -0.4759781 , -0.48210353, -0.48678938,\n",
       "        -0.49091062, -0.4953424 , -0.4994919 , -0.50279456, -0.50702876,\n",
       "        -0.511517  , -0.514001  , -0.5166262 , -0.5199006 , -0.5229775 ,\n",
       "        -0.5256309 , -0.52814317, -0.5306837 , -0.53297013, -0.5353695 ,\n",
       "        -0.53782535, -0.5401682 , -0.5422571 , -0.5445153 , -0.5467453 ,\n",
       "        -0.5486366 , -0.5504714 , -0.552278  , -0.5541975 , -0.555383  ,\n",
       "        -0.55727434, -0.55893975, -0.56004065, -0.5617343 , -0.56300455,\n",
       "        -0.5645289 , -0.56585556, -0.5669    , -0.5683961 , -0.5687066 ,\n",
       "        -0.56658953, -0.5643595 , -0.56244004, -0.56077456, -0.5586293 ,\n",
       "        -0.55617344, -0.554508  , -0.5523909 , -0.5498504 , -0.5476486 ,\n",
       "        -0.5457856 , -0.5430757 , -0.54098684, -0.5392085 , -0.5361317 ,\n",
       "        -0.5330548 , -0.5304014 , -0.5275786 , -0.52458644, -0.5218484 ,\n",
       "        -0.51868683, -0.5152148 , -0.5118557 , -0.5087506 , -0.50519395,\n",
       "        -0.5020324 , -0.4981934 , -0.49443913, -0.49108   , -0.4868176 ,\n",
       "        -0.4826681 , -0.4782081 , -0.47388923, -0.46886468, -0.46358606,\n",
       "        -0.45822278, -0.45317   , -0.44766554, -0.44139898, -0.43482187,\n",
       "        -0.42776492, -0.42090556, -0.41356632, -0.40602946, -0.39798453,\n",
       "        -0.38965732, -0.3809349 , -0.37178907, -0.3627844 , -0.35281995,\n",
       "        -0.34296843, -0.33294755, -0.32233387, -0.3115226 , -0.30082425,\n",
       "        -0.2905211 , -0.27914527, -0.26841867, -0.2570993 , -0.2456388 ,\n",
       "        -0.23536386, -0.22410095, -0.213205  , -0.20253485, -0.19178005,\n",
       "        -0.18105346, -0.17007282, -0.15943092, -0.14881726, -0.13817535,\n",
       "        -0.12801333, -0.11787952, -0.10757636, -0.09710382, -0.08728053,\n",
       "        -0.07723141, -0.06690001, -0.0570485 , -0.04725343, -0.03748659,\n",
       "        -0.02749393, -0.01823519, -0.00835545,  0.00115734,  0.01027494,\n",
       "         0.0195619 ,  0.02884887,  0.0383052 ,  0.04730989,  0.05639926,\n",
       "         0.06506521,  0.07401343,  0.08298989,  0.09165584,  0.10023712,\n",
       "         0.10895952,  0.11771016,  0.12603737,  0.13436459,  0.14274827,\n",
       "         0.15118839,  0.1590075 ,  0.16722181,  0.17552081,  0.1836222 ,\n",
       "         0.19149777,  0.19948626,  0.20716423,  0.21489866,  0.22240727,\n",
       "         0.22988765,  0.23784791,  0.24510247,  0.2522441 ,  0.259414  ,\n",
       "         0.26664034,  0.2732174 ,  0.27996388,  0.28673854,  0.29289222,\n",
       "         0.29921526,  0.3051149 ,  0.31098628,  0.31674477,  0.32213628,\n",
       "         0.32727376,  0.332383  ,  0.33684298,  0.3411901 ,  0.34590414,\n",
       "         0.3497149 ,  0.35344097,  0.35688478,  0.36018744,  0.36357477,\n",
       "         0.3667363 ,  0.36933327,  0.371902  ,  0.3746401 ,  0.37672895,\n",
       "         0.3789025 ,  0.3809349 ,  0.38310844,  0.38477388,  0.3864111 ,\n",
       "         0.38816124,  0.38945973,  0.39129454,  0.3926777 ,  0.39397618,\n",
       "         0.39550048,  0.39696833,  0.3979563 ,  0.39902896,  0.40021452,\n",
       "         0.4010896 ,  0.40213403,  0.40340427,  0.40422288,  0.40498504,\n",
       "         0.40614238,  0.406961  ,  0.4076949 ,  0.40837237,  0.4093039 ,\n",
       "         0.40984023,  0.41054592,  0.41122338,  0.4118444 ,  0.41221136,\n",
       "         0.4118444 ,  0.41068706,  0.40955794,  0.40882403,  0.40755376,\n",
       "         0.4062553 ,  0.40498504,  0.40388414,  0.40269858,  0.4014848 ,\n",
       "         0.4001863 ,  0.398662  ,  0.39727885,  0.39600858,  0.39476657,\n",
       "         0.39301643,  0.3915768 ,  0.39008072,  0.3883306 ,  0.3866934 ,\n",
       "         0.3849715 ,  0.383052  ,  0.3813301 ,  0.37960818,  0.37726527,\n",
       "         0.37528932,  0.37317225,  0.3711116 ,  0.3687969 ,  0.366454  ,\n",
       "         0.36388528,  0.36117542,  0.3587196 ,  0.35572743,  0.3525659 ,\n",
       "         0.34943262,  0.34610173,  0.34268618,  0.33901653,  0.33503643,\n",
       "         0.33094338,  0.3266245 ,  0.32222098,  0.31745046,  0.31259528,\n",
       "         0.30751425,  0.30215096,  0.29675946,  0.29091626,  0.2850731 ,\n",
       "         0.27903235,  0.27279398,  0.26661208,  0.2602326 ,  0.2538531 ,\n",
       "         0.24741715,  0.24050133,  0.23367019,  0.22681081,  0.21992321,\n",
       "         0.21317676,  0.2059222 ,  0.19886523,  0.1918365 ,  0.18435612,\n",
       "         0.1769322 ,  0.16936713,  0.16197143,  0.15480156,  0.14681308,\n",
       "         0.13924801,  0.131937  ,  0.12406142,  0.11646813,  0.10895952,\n",
       "         0.101338  ,  0.09329306,  0.08578445,  0.0783323 ,  0.07020268,\n",
       "         0.06215774,  0.05442331,  0.04635014,  0.03793824,  0.03026026,\n",
       "         0.02170722,  0.01354937], dtype=float32),\n",
       "        array([ 0.00544798, -0.0027381 , -0.01115   , -0.01916671, -0.02740925,\n",
       "        -0.03632925, -0.0444871 , -0.05312482, -0.06139559, -0.06989217,\n",
       "        -0.0783323 , -0.08697002, -0.09597471, -0.10464066, -0.11333484,\n",
       "        -0.12208547, -0.13100548, -0.1396432 , -0.14864789, -0.15801953,\n",
       "        -0.16682662, -0.17537966, -0.18469486, -0.19381246, -0.20256309,\n",
       "        -0.21165246, -0.2200926 , -0.22904082, -0.23815842, -0.24753006,\n",
       "        -0.25656298, -0.26531363, -0.27409247, -0.28267375, -0.29153728,\n",
       "        -0.3002597 , -0.3086716 , -0.31674477, -0.32476148, -0.33235475,\n",
       "        -0.33952466, -0.34638402, -0.35344097, -0.36030033, -0.3663129 ,\n",
       "        -0.37215605, -0.37763226, -0.3830802 , -0.38847175, -0.39335516,\n",
       "        -0.39767402, -0.4021058 , -0.40628353, -0.409812  , -0.4136792 ,\n",
       "        -0.4172077 , -0.42048213, -0.42403883, -0.4273697 , -0.43033364,\n",
       "        -0.4332129 , -0.43606392, -0.43905607, -0.44137076, -0.44331846,\n",
       "        -0.44611302, -0.4482019 , -0.4502343 , -0.4525772 , -0.45477897,\n",
       "        -0.45596454, -0.45729125, -0.4595777 , -0.4613843 , -0.462711  ,\n",
       "        -0.4642353 , -0.4660419 , -0.46739683, -0.46886468, -0.4708124 ,\n",
       "        -0.4722238 , -0.47340935, -0.47490543, -0.47603455, -0.47733304,\n",
       "        -0.4785186 , -0.47956303, -0.48066393, -0.48142606, -0.4826681 ,\n",
       "        -0.4835996 , -0.48458758, -0.4854344 , -0.48546267, -0.48424885,\n",
       "        -0.48218822, -0.48108733, -0.47939366, -0.4780105 , -0.47642973,\n",
       "        -0.47431266, -0.4723085 , -0.47103822, -0.46942922, -0.46739683,\n",
       "        -0.46530798, -0.46316266, -0.46163833, -0.45940834, -0.45737594,\n",
       "        -0.45525885, -0.45331112, -0.45108113, -0.44904873, -0.44673404,\n",
       "        -0.44379833, -0.4417377 , -0.43956417, -0.4369672 , -0.43408796,\n",
       "        -0.43168858, -0.4289505 , -0.42607126, -0.42322025, -0.42070794,\n",
       "        -0.41785693, -0.41455427, -0.41096935, -0.40735617, -0.4044205 ,\n",
       "        -0.40055326, -0.39679897, -0.39242363, -0.38841528, -0.38423756,\n",
       "        -0.37955174, -0.37478122, -0.36955908, -0.36487326, -0.35965112,\n",
       "        -0.35392085, -0.34813413, -0.34226274, -0.33560097, -0.32930616,\n",
       "        -0.32247502, -0.31578502, -0.30833286, -0.3003726 , -0.29297692,\n",
       "        -0.2848473 , -0.27671766, -0.26858804, -0.26068425, -0.25179246,\n",
       "        -0.24354993, -0.23527917, -0.22607689, -0.21755208, -0.20922486,\n",
       "        -0.2006436 , -0.19189297, -0.18336815, -0.17467397, -0.16620561,\n",
       "        -0.1576808 , -0.14932536, -0.14119573, -0.13269915, -0.12423079,\n",
       "        -0.11635522, -0.10808446, -0.09987015, -0.09182521, -0.08389319,\n",
       "        -0.07601761, -0.06777508, -0.06018179, -0.05208039, -0.04431773,\n",
       "        -0.03655507, -0.02890532, -0.02145317, -0.01397279, -0.00666177,\n",
       "         0.00124203,  0.00818608,  0.01544064,  0.02263874,  0.02963925,\n",
       "         0.03714786,  0.04434596,  0.05126179,  0.05834698,  0.06548862,\n",
       "         0.07220685,  0.07917913,  0.08592559,  0.09264382,  0.09933382,\n",
       "         0.10624965,  0.11274205,  0.11895218,  0.12578332,  0.13219105,\n",
       "         0.13854231,  0.14528877,  0.15180941,  0.15790662,  0.1640603 ,\n",
       "         0.1700446 ,  0.17636764,  0.18238017,  0.18836448,  0.19429232,\n",
       "         0.20022018,  0.20603512,  0.21176536,  0.2175803 ,  0.22300005,\n",
       "         0.2287303 ,  0.23406537,  0.23962626,  0.24467905,  0.2499012 ,\n",
       "         0.25509512,  0.26009145,  0.2648055 ,  0.26971716,  0.2745159 ,\n",
       "         0.2784678 ,  0.2828149 ,  0.2869079 ,  0.290634  ,  0.29410604,\n",
       "         0.29788855,  0.30124766,  0.30418336,  0.30711907,  0.31002653,\n",
       "         0.31276464,  0.31513578,  0.3177045 ,  0.31999096,  0.32196692,\n",
       "         0.324084  ,  0.32605997,  0.3277254 ,  0.32947552,  0.33142325,\n",
       "         0.33306047,  0.33435893,  0.335855  ,  0.33709705,  0.33850843,\n",
       "         0.33966577,  0.3407949 ,  0.34217805,  0.34350476,  0.3444645 ,\n",
       "         0.34522668,  0.34621465,  0.347372  ,  0.34785187,  0.34875515,\n",
       "         0.3496302 ,  0.35025123,  0.35106984,  0.3517473 ,  0.35239655,\n",
       "         0.35301757,  0.35392085,  0.35442895,  0.35496527,  0.35555807,\n",
       "         0.3561791 ,  0.35651782,  0.35708237,  0.35787275,  0.3581268 ,\n",
       "         0.35840908,  0.35829616,  0.35744932,  0.35651782,  0.35578388,\n",
       "         0.35502174,  0.354062  ,  0.3530458 ,  0.35214248,  0.35109806,\n",
       "         0.35019478,  0.34948906,  0.34807768,  0.34692034,  0.34601706,\n",
       "         0.3449444 ,  0.3438435 ,  0.34248856,  0.3414159 ,  0.3400045 ,\n",
       "         0.33873427,  0.33763337,  0.336222  ,  0.33503643,  0.33365324,\n",
       "         0.33204427,  0.3306893 ,  0.3292215 ,  0.32738668,  0.32569298,\n",
       "         0.3241969 ,  0.32230565,  0.32047084,  0.3188054 ,  0.31691414,\n",
       "         0.31462768,  0.3123412 ,  0.31039348,  0.3078812 ,  0.30548185,\n",
       "         0.30305424,  0.3001468 ,  0.29715464,  0.29413426,  0.29131147,\n",
       "         0.28775474,  0.28442386,  0.28109297,  0.27722576,  0.27344322,\n",
       "         0.26971716,  0.26556766,  0.26102298,  0.256817  ,  0.25241348,\n",
       "         0.24772766,  0.24307007,  0.23796082,  0.23310563,  0.22808108,\n",
       "         0.22277424,  0.2175803 ,  0.2121888 ,  0.20702308,  0.20154689,\n",
       "         0.19576018,  0.19048157,  0.18509005,  0.17941625,  0.17374244,\n",
       "         0.16787106,  0.16208434,  0.15655169,  0.15076497,  0.14466776,\n",
       "         0.13916333,  0.1332637 ,  0.12719472,  0.12129509,  0.11536724,\n",
       "         0.10938294,  0.10331395,  0.09735787,  0.09117597,  0.08499407,\n",
       "         0.07889686,  0.07257382,  0.06610963,  0.06026647,  0.05371761,\n",
       "         0.0474228 ,  0.04138204,  0.03503077,  0.02834077,  0.02207418,\n",
       "         0.01606165,  0.00937165], dtype=float32),\n",
       "        array([ 0.00276633, -0.00355671, -0.0101338 , -0.01696494, -0.0235138 ,\n",
       "        -0.02992153, -0.03675267, -0.04355558, -0.05061254, -0.05775419,\n",
       "        -0.06413369, -0.07071078, -0.07751369, -0.08426015, -0.09123243,\n",
       "        -0.09848699, -0.10520522, -0.11226218, -0.11940382, -0.12612206,\n",
       "        -0.13334839, -0.14068763, -0.14771637, -0.15449105, -0.16154802,\n",
       "        -0.16897194, -0.1760289 , -0.18308587, -0.19025575, -0.1975103 ,\n",
       "        -0.20453905, -0.21139841, -0.21825777, -0.22525829, -0.23228702,\n",
       "        -0.23892057, -0.24580817, -0.25266755, -0.25907525, -0.26508778,\n",
       "        -0.27155197, -0.2777621 , -0.283323  , -0.2889968 , -0.29382375,\n",
       "        -0.2988483 , -0.3042398 , -0.30903855, -0.31347033, -0.31807148,\n",
       "        -0.32210806, -0.32591882, -0.32987073, -0.33356857, -0.33712527,\n",
       "        -0.34025857, -0.34305313, -0.34601706, -0.34906566, -0.35169086,\n",
       "        -0.35423136, -0.356913  , -0.3592559 , -0.3612601 , -0.36354655,\n",
       "        -0.36569187, -0.3675549 , -0.36941794, -0.37159148, -0.373511  ,\n",
       "        -0.3751482 , -0.37672895, -0.37822503, -0.38008806, -0.38166884,\n",
       "        -0.3830802 , -0.3846892 , -0.38607237, -0.38734263, -0.3886129 ,\n",
       "        -0.38971376, -0.39104047, -0.39211315, -0.39307287, -0.39437136,\n",
       "        -0.39516175, -0.39620617, -0.3973635 , -0.3985773 , -0.3994806 ,\n",
       "        -0.40029922, -0.40145656, -0.40187997, -0.40269858, -0.40394062,\n",
       "        -0.40458986, -0.40512618, -0.4057472 , -0.40619883, -0.40701744,\n",
       "        -0.4066787 , -0.40498504, -0.40385592, -0.40289617, -0.40165415,\n",
       "        -0.40024275, -0.39891604, -0.3973353 , -0.3958392 , -0.3947101 ,\n",
       "        -0.39329872, -0.39202845, -0.39016542, -0.3886129 , -0.38737085,\n",
       "        -0.38533846, -0.383673  , -0.38223338, -0.38070908, -0.3787049 ,\n",
       "        -0.37715238, -0.3752611 , -0.37314403, -0.37159148, -0.3697849 ,\n",
       "        -0.36795008, -0.36608705, -0.36408287, -0.36205047, -0.35965112,\n",
       "        -0.35739288, -0.35527578, -0.3528482 , -0.3504206 , -0.34802124,\n",
       "        -0.34565008, -0.34310958, -0.34017387, -0.3372382 , -0.3346977 ,\n",
       "        -0.33190313, -0.3282335 , -0.32515666, -0.32179755, -0.3180997 ,\n",
       "        -0.31448653, -0.31076047, -0.30700615, -0.30246148, -0.29831198,\n",
       "        -0.29385197, -0.28905323, -0.28448033, -0.27956867, -0.2743183 ,\n",
       "        -0.26929373, -0.26333767, -0.25712752, -0.25179246, -0.24561056,\n",
       "        -0.23917462, -0.23251285, -0.22616157, -0.21952803, -0.21255575,\n",
       "        -0.20583752, -0.19911928, -0.19240107, -0.18551347, -0.1788517 ,\n",
       "        -0.17202055, -0.16544347, -0.15855587, -0.15180941, -0.14509118,\n",
       "        -0.13845763, -0.13173941, -0.12476712, -0.11807712, -0.11130244,\n",
       "        -0.10495117, -0.09823294, -0.09159939, -0.08493762, -0.07844521,\n",
       "        -0.07217862, -0.06565799, -0.05930672, -0.05284255, -0.04657596,\n",
       "        -0.04030938, -0.03421216, -0.02819963, -0.02215887, -0.01594874,\n",
       "        -0.01002089, -0.00398013,  0.00180658,  0.00767798,  0.01374697,\n",
       "         0.01939254,  0.02506633,  0.03088127,  0.03675267,  0.04203128,\n",
       "         0.04778976,  0.05357647,  0.05910913,  0.0647547 ,  0.07028736,\n",
       "         0.07559419,  0.0808728 ,  0.08651838,  0.09176876,  0.09713205,\n",
       "         0.10246711,  0.10763281,  0.11265737,  0.11793598,  0.12307345,\n",
       "         0.12801333,  0.13298142,  0.13800599,  0.14305878,  0.14777282,\n",
       "         0.15282561,  0.15770903,  0.16259244,  0.16713713,  0.17199232,\n",
       "         0.17664991,  0.1811946 ,  0.18602157,  0.19005814,  0.19460283,\n",
       "         0.1990346 ,  0.20363575,  0.20806752,  0.21213233,  0.21611246,\n",
       "         0.22034664,  0.22441144,  0.2281093 ,  0.2318636 ,  0.23564613,\n",
       "         0.23923108,  0.24256197,  0.2459493 ,  0.24899791,  0.25196183,\n",
       "         0.25492576,  0.25774854,  0.2601479 ,  0.26263195,  0.265116  ,\n",
       "         0.26745892,  0.26968893,  0.27149552,  0.27344322,  0.27547565,\n",
       "         0.2771411 ,  0.27877828,  0.2805002 ,  0.28196806,  0.2832665 ,\n",
       "         0.28450856,  0.2859764 ,  0.28716198,  0.28834754,  0.2895331 ,\n",
       "         0.29057753,  0.2915655 ,  0.29255348,  0.29362616,  0.2943883 ,\n",
       "         0.29526338,  0.2960255 ,  0.29675946,  0.29757804,  0.29828376,\n",
       "         0.29887652,  0.2996387 ,  0.30028793,  0.30085248,  0.30152994,\n",
       "         0.30203804,  0.30260262,  0.30316716,  0.30387285,  0.30426806,\n",
       "         0.3048326 ,  0.30545363,  0.30565122,  0.3059335 ,  0.30652627,\n",
       "         0.30692148,  0.30711907,  0.3078812 ,  0.3081635 ,  0.30813527,\n",
       "         0.30790946,  0.30726022,  0.30658275,  0.3059335 ,  0.30525604,\n",
       "         0.30443743,  0.30364704,  0.30291313,  0.30215096,  0.30136058,\n",
       "         0.30054197,  0.29980806,  0.29887652,  0.29794502,  0.297324  ,\n",
       "         0.2962231 ,  0.29517868,  0.29461414,  0.2934568 ,  0.2924406 ,\n",
       "         0.29142436,  0.29037994,  0.2895331 ,  0.2884322 ,  0.28713375,\n",
       "         0.28569412,  0.28459322,  0.28357702,  0.28225031,  0.2808107 ,\n",
       "         0.27956867,  0.27821374,  0.27680236,  0.27536273,  0.2738102 ,\n",
       "         0.2721165 ,  0.27047932,  0.26878563,  0.26683792,  0.265116  ,\n",
       "         0.26311183,  0.26102298,  0.25882122,  0.25664765,  0.2543612 ,\n",
       "         0.25176424,  0.24933665,  0.24659854,  0.243804  ,  0.24095297,\n",
       "         0.237735  ,  0.23443234,  0.23132728,  0.2279964 ,  0.22449613,\n",
       "         0.22085474,  0.21721335,  0.21360017,  0.20978943,  0.2059222 ,\n",
       "         0.20208322,  0.19818777,  0.19384068,  0.1894936 ,  0.18528764,\n",
       "         0.18094055,  0.17659347,  0.17255688,  0.16798396,  0.1635522 ,\n",
       "         0.15903574,  0.15443459,  0.15020041,  0.14557105,  0.14099814,\n",
       "         0.13650991,  0.13173941,  0.12719472,  0.12250889,  0.11785129,\n",
       "         0.11316547,  0.1082256 ,  0.10353977,  0.09857167,  0.09349066,\n",
       "         0.08874838,  0.0838085 ,  0.07878394,  0.07421103,  0.06927115,\n",
       "         0.06421837,  0.05916559,  0.05428217,  0.04925761,  0.04412014,\n",
       "         0.03929317,  0.03418393,  0.02899001,  0.02390899,  0.01868684,\n",
       "         0.01323886,  0.00798848], dtype=float32),\n",
       "        array([ 0.00304861, -0.00228646, -0.00781912, -0.01292836, -0.0180376 ,\n",
       "        -0.02373963, -0.02893355, -0.03443798, -0.03988596, -0.04505166,\n",
       "        -0.05055609, -0.05606052, -0.06148027, -0.06678711, -0.07243267,\n",
       "        -0.07807825, -0.08383673, -0.08925648, -0.09507142, -0.10066053,\n",
       "        -0.10613674, -0.11200813, -0.1176537 , -0.12344041, -0.1290013 ,\n",
       "        -0.13453396, -0.14051826, -0.14607915, -0.15158358, -0.15753967,\n",
       "        -0.16321346, -0.16894372, -0.17467397, -0.1804889 , -0.18616271,\n",
       "        -0.19155422, -0.19714335, -0.20270424, -0.2078417 , -0.21309207,\n",
       "        -0.21831423, -0.22302829, -0.22788347, -0.23234348, -0.23688816,\n",
       "        -0.24140462, -0.24552588, -0.24956247, -0.25300628, -0.256817  ,\n",
       "        -0.26048666, -0.2637893 , -0.26706374, -0.27042285, -0.27355614,\n",
       "        -0.27660474, -0.2790888 , -0.28177044, -0.2843674 , -0.28631514,\n",
       "        -0.28877097, -0.29094452, -0.29300514, -0.29475525, -0.2968159 ,\n",
       "        -0.2988483 , -0.30042908, -0.30192515, -0.3037882 , -0.30514312,\n",
       "        -0.30641338, -0.30790946, -0.3092644 , -0.31084514, -0.31186134,\n",
       "        -0.3131316 , -0.3146559 , -0.31589794, -0.3168859 , -0.31781742,\n",
       "        -0.31908768, -0.32010388, -0.3211483 , -0.32227743, -0.32303956,\n",
       "        -0.32385817, -0.32464856, -0.3256083 , -0.32637045, -0.32744312,\n",
       "        -0.3283464 , -0.32876983, -0.32950374, -0.33040705, -0.330774  ,\n",
       "        -0.3311692 , -0.3319878 , -0.3327782 , -0.3333992 , -0.33368146,\n",
       "        -0.33444363, -0.33466944, -0.33509287, -0.3360244 , -0.335855  ,\n",
       "        -0.33486706, -0.3338226 , -0.33317336, -0.33215716, -0.33080223,\n",
       "        -0.32978603, -0.3287416 , -0.32755604, -0.32622933, -0.32515666,\n",
       "        -0.32414046, -0.32278553, -0.3216564 , -0.32075313, -0.31936994,\n",
       "        -0.3179868 , -0.316773  , -0.3154745 , -0.31400666, -0.31276464,\n",
       "        -0.31149438, -0.30994186, -0.30850223, -0.30680856, -0.30525604,\n",
       "        -0.30367526, -0.30232033, -0.30082425, -0.29887652, -0.29735222,\n",
       "        -0.29563034, -0.29348502, -0.29173487, -0.290013  , -0.28817818,\n",
       "        -0.286174  , -0.2843674 , -0.28210917, -0.27976626, -0.27767742,\n",
       "        -0.27505222, -0.2726246 , -0.27039462, -0.26785412, -0.26491842,\n",
       "        -0.2621521 , -0.25938576, -0.25599843, -0.25286514, -0.24978828,\n",
       "        -0.24640095, -0.24275956, -0.23917462, -0.23536386, -0.23129904,\n",
       "        -0.22726247, -0.22302829, -0.21859652, -0.2142212 , -0.20959182,\n",
       "        -0.20453905, -0.19948626, -0.19434878, -0.18983233, -0.18441258,\n",
       "        -0.17873877, -0.17354485, -0.1678146 , -0.16267712, -0.1571727 ,\n",
       "        -0.15164004, -0.14602269, -0.1404618 , -0.13498561, -0.1295094 ,\n",
       "        -0.12417433, -0.11864167, -0.11316547, -0.10760459, -0.10232598,\n",
       "        -0.09670863, -0.09168407, -0.08637724, -0.08078812, -0.07562242,\n",
       "        -0.07020268, -0.06509344, -0.05989951, -0.05459267, -0.04939875,\n",
       "        -0.04417659, -0.03906735, -0.03404279, -0.02907469, -0.02413482,\n",
       "        -0.01896912, -0.01411393, -0.00923051, -0.00431886,  0.00059278,\n",
       "         0.00510924,  0.00990798,  0.01490431,  0.01973127,  0.02441709,\n",
       "         0.02884887,  0.03367583,  0.03813583,  0.04268052,  0.0474228 ,\n",
       "         0.05165698,  0.05620166,  0.0608028 ,  0.06512167,  0.0694123 ,\n",
       "         0.07390053,  0.0781347 ,  0.08222774,  0.08663129,  0.09109129,\n",
       "         0.09532547,  0.09933382,  0.10373737,  0.10785864,  0.11169762,\n",
       "         0.1159318 ,  0.1198837 ,  0.12400497,  0.12787218,  0.13168295,\n",
       "         0.1358042 ,  0.1396432 ,  0.14376447,  0.14763169,  0.15118839,\n",
       "         0.15482979,  0.15861233,  0.16225371,  0.16597979,  0.16959296,\n",
       "         0.17303675,  0.17656523,  0.1798679 ,  0.18317056,  0.18661435,\n",
       "         0.18971941,  0.19290917,  0.19612713,  0.19920397,  0.20188563,\n",
       "         0.20487778,  0.2078417 ,  0.21038221,  0.21306385,  0.21571727,\n",
       "         0.21814486,  0.22026196,  0.2224355 ,  0.22460905,  0.22666968,\n",
       "         0.22841981,  0.23019816,  0.23194829,  0.23355727,  0.23519449,\n",
       "         0.23657766,  0.23798905,  0.2391464 ,  0.24044487,  0.24171512,\n",
       "         0.24281602,  0.24405804,  0.24490488,  0.24603398,  0.24690905,\n",
       "         0.24778411,  0.24877208,  0.2497036 ,  0.25052223,  0.25117147,\n",
       "         0.25187716,  0.25266755,  0.25331676,  0.25379664,  0.25458702,\n",
       "         0.25495398,  0.2553492 ,  0.25605488,  0.25653476,  0.25692993,\n",
       "         0.25743803,  0.25794613,  0.25828487,  0.25879297,  0.25927284,\n",
       "         0.25969627,  0.26017615,  0.26048666,  0.2610512 ,  0.26127702,\n",
       "         0.26167223,  0.26209563,  0.26232147,  0.26251906,  0.26263195,\n",
       "         0.26314005,  0.2633659 ,  0.2631683 ,  0.2625755 ,  0.2619545 ,\n",
       "         0.26130524,  0.26079714,  0.2601479 ,  0.2595269 ,  0.2589906 ,\n",
       "         0.2583978 ,  0.25783324,  0.25707108,  0.25642183,  0.2558573 ,\n",
       "         0.25509512,  0.25438944,  0.25362727,  0.252724  ,  0.25199005,\n",
       "         0.251115  ,  0.2504093 ,  0.2495907 ,  0.24871564,  0.24798171,\n",
       "         0.2470502 ,  0.24609044,  0.24521537,  0.24436854,  0.24323943,\n",
       "         0.24247728,  0.24151753,  0.24041665,  0.2392593 ,  0.23810196,\n",
       "         0.23700108,  0.2356179 ,  0.23443234,  0.23313385,  0.23183538,\n",
       "         0.23028284,  0.22889967,  0.2274036 ,  0.22582284,  0.2241574 ,\n",
       "         0.2224355 ,  0.22068538,  0.21865296,  0.21690284,  0.21464461,\n",
       "         0.2126969 ,  0.21052334,  0.2081522 ,  0.20578106,  0.20335346,\n",
       "         0.20081297,  0.19807486,  0.19550613,  0.19245751,  0.18974765,\n",
       "         0.18672726,  0.18339637,  0.180376  ,  0.17746852,  0.17416586,\n",
       "         0.17077853,  0.16744764,  0.16400383,  0.1606165 ,  0.15711625,\n",
       "         0.15381359,  0.15005928,  0.14650257,  0.14300232,  0.13941738,\n",
       "         0.1356913 ,  0.131937  ,  0.12849319,  0.12479535,  0.1210975 ,\n",
       "         0.11723028,  0.11353243,  0.10980636,  0.10596737,  0.10215661,\n",
       "         0.09828939,  0.09459154,  0.09058319,  0.08654661,  0.08267939,\n",
       "         0.07869926,  0.07454976,  0.07056963,  0.06670243,  0.06263761,\n",
       "         0.05843166,  0.05439508,  0.05038672,  0.04623723,  0.04237001,\n",
       "         0.03847457,  0.03409925,  0.0298933 ,  0.02577203,  0.02182013,\n",
       "         0.01736013,  0.01301304,  0.00903291,  0.00502456], dtype=float32),\n",
       "        array([ 4.79873532e-04, -4.00835555e-03, -8.21430609e-03, -1.24202566e-02,\n",
       "        -1.67673454e-02, -2.14249417e-02, -2.56591197e-02, -3.00344378e-02,\n",
       "        -3.46073508e-02, -3.90108973e-02, -4.34144400e-02, -4.80155833e-02,\n",
       "        -5.24191260e-02, -5.65968491e-02, -6.13109022e-02, -6.59684986e-02,\n",
       "        -7.05131814e-02, -7.51990080e-02, -7.98001438e-02, -8.42601508e-02,\n",
       "        -8.90871137e-02, -9.37447101e-02, -9.83176231e-02, -1.02636479e-01,\n",
       "        -1.07717499e-01, -1.12459779e-01, -1.16891548e-01, -1.21746741e-01,\n",
       "        -1.26093835e-01, -1.30807877e-01, -1.35437250e-01, -1.40207753e-01,\n",
       "        -1.44526616e-01, -1.48930162e-01, -1.53587759e-01, -1.58104211e-01,\n",
       "        -1.62479535e-01, -1.66854858e-01, -1.71004340e-01, -1.75153837e-01,\n",
       "        -1.79303333e-01, -1.83368146e-01, -1.86924860e-01, -1.90650940e-01,\n",
       "        -1.94489926e-01, -1.97764352e-01, -2.00841188e-01, -2.04059169e-01,\n",
       "        -2.07446516e-01, -2.10523337e-01, -2.13204995e-01, -2.15999544e-01,\n",
       "        -2.18540058e-01, -2.21165240e-01, -2.23310560e-01, -2.25597024e-01,\n",
       "        -2.27883473e-01, -2.30057016e-01, -2.31863603e-01, -2.33444363e-01,\n",
       "        -2.35448539e-01, -2.36944616e-01, -2.38807648e-01, -2.40473092e-01,\n",
       "        -2.41856262e-01, -2.43154749e-01, -2.44933099e-01, -2.46457398e-01,\n",
       "        -2.47445375e-01, -2.48828545e-01, -2.50268161e-01, -2.51623094e-01,\n",
       "        -2.52695769e-01, -2.53881335e-01, -2.54643470e-01, -2.55687922e-01,\n",
       "        -2.56958157e-01, -2.57917911e-01, -2.58680075e-01, -2.59583354e-01,\n",
       "        -2.60401964e-01, -2.60966510e-01, -2.62067407e-01, -2.62857795e-01,\n",
       "        -2.63676405e-01, -2.64353871e-01, -2.65200704e-01, -2.66047537e-01,\n",
       "        -2.66612083e-01, -2.67261326e-01, -2.67797649e-01, -2.68531591e-01,\n",
       "        -2.68870324e-01, -2.69265503e-01, -2.69576013e-01, -2.70084113e-01,\n",
       "        -2.70507544e-01, -2.70902723e-01, -2.71439046e-01, -2.72229433e-01,\n",
       "        -2.72793978e-01, -2.72822231e-01, -2.73612589e-01, -2.74233609e-01,\n",
       "        -2.74346530e-01, -2.74685264e-01, -2.74346530e-01, -2.73132741e-01,\n",
       "        -2.72285879e-01, -2.71410823e-01, -2.70620435e-01, -2.69688934e-01,\n",
       "        -2.68757403e-01, -2.67797649e-01, -2.66753227e-01, -2.65962839e-01,\n",
       "        -2.65144259e-01, -2.64156282e-01, -2.63196528e-01, -2.62349695e-01,\n",
       "        -2.61277020e-01, -2.60063231e-01, -2.58905888e-01, -2.58341342e-01,\n",
       "        -2.57268667e-01, -2.55913734e-01, -2.54756391e-01, -2.53373235e-01,\n",
       "        -2.52272338e-01, -2.51227915e-01, -2.50014126e-01, -2.48687401e-01,\n",
       "        -2.47473612e-01, -2.46146902e-01, -2.44763732e-01, -2.43691072e-01,\n",
       "        -2.42392600e-01, -2.40670696e-01, -2.39061713e-01, -2.37622082e-01,\n",
       "        -2.35956639e-01, -2.34517023e-01, -2.33105630e-01, -2.31157914e-01,\n",
       "        -2.29548916e-01, -2.27996394e-01, -2.25963980e-01, -2.24496126e-01,\n",
       "        -2.22294360e-01, -2.20064357e-01, -2.18455374e-01, -2.16056004e-01,\n",
       "        -2.13713095e-01, -2.11341947e-01, -2.08914354e-01, -2.06289172e-01,\n",
       "        -2.03805119e-01, -2.01123476e-01, -1.98216006e-01, -1.95393220e-01,\n",
       "        -1.92457512e-01, -1.89239547e-01, -1.85682833e-01, -1.82436630e-01,\n",
       "        -1.78879917e-01, -1.75182074e-01, -1.71597138e-01, -1.67786375e-01,\n",
       "        -1.63778022e-01, -1.59797892e-01, -1.55648395e-01, -1.51555359e-01,\n",
       "        -1.47321180e-01, -1.42720029e-01, -1.38344720e-01, -1.33828267e-01,\n",
       "        -1.29481167e-01, -1.25162303e-01, -1.20589398e-01, -1.16129398e-01,\n",
       "        -1.11443572e-01, -1.07068256e-01, -1.02325976e-01, -9.76683795e-02,\n",
       "        -9.33777466e-02, -8.91153365e-02, -8.47964808e-02, -8.01671073e-02,\n",
       "        -7.58200213e-02, -7.14164749e-02, -6.69564754e-02, -6.24400154e-02,\n",
       "        -5.80364689e-02, -5.38305193e-02, -4.96245697e-02, -4.53339368e-02,\n",
       "        -4.12126705e-02, -3.68373506e-02, -3.26878577e-02, -2.85948180e-02,\n",
       "        -2.45582350e-02, -2.03805119e-02, -1.62027888e-02, -1.21662058e-02,\n",
       "        -8.10139440e-03, -4.06481093e-03, -1.41139273e-04,  3.81076057e-03,\n",
       "         7.96025526e-03,  1.15451925e-02,  1.54688647e-02,  1.95619036e-02,\n",
       "         2.34291200e-02,  2.70705130e-02,  3.07683628e-02,  3.46073508e-02,\n",
       "         3.82487439e-02,  4.20595035e-02,  4.57008965e-02,  4.94552031e-02,\n",
       "         5.29836826e-02,  5.64839393e-02,  6.02100156e-02,  6.38231784e-02,\n",
       "         6.72387481e-02,  7.07672313e-02,  7.45215341e-02,  7.79935643e-02,\n",
       "         8.13526809e-02,  8.48529339e-02,  8.84378701e-02,  9.17123035e-02,\n",
       "         9.50996429e-02,  9.84869897e-02,  1.01761416e-01,  1.04979396e-01,\n",
       "         1.08169138e-01,  1.11471802e-01,  1.14605092e-01,  1.17823064e-01,\n",
       "         1.21125728e-01,  1.24174334e-01,  1.27307624e-01,  1.30525604e-01,\n",
       "         1.33545980e-01,  1.36594594e-01,  1.39784336e-01,  1.42720029e-01,\n",
       "         1.45514593e-01,  1.48337379e-01,  1.51273072e-01,  1.53841808e-01,\n",
       "         1.56608135e-01,  1.59459159e-01,  1.61971435e-01,  1.64568394e-01,\n",
       "         1.67137131e-01,  1.69818774e-01,  1.72077000e-01,  1.74617514e-01,\n",
       "         1.77045107e-01,  1.79021060e-01,  1.81335747e-01,  1.83481053e-01,\n",
       "         1.85287639e-01,  1.87207133e-01,  1.89352453e-01,  1.91046119e-01,\n",
       "         1.92626879e-01,  1.94207639e-01,  1.95731953e-01,  1.97228029e-01,\n",
       "         1.98554739e-01,  1.99824989e-01,  2.01179922e-01,  2.02393726e-01,\n",
       "         2.03466386e-01,  2.04510808e-01,  2.05498785e-01,  2.06514984e-01,\n",
       "         2.07277134e-01,  2.08152205e-01,  2.09083721e-01,  2.09817648e-01,\n",
       "         2.10523337e-01,  2.11229041e-01,  2.11934730e-01,  2.12612212e-01,\n",
       "         2.13261440e-01,  2.13910684e-01,  2.14559928e-01,  2.15068027e-01,\n",
       "         2.15576127e-01,  2.16056004e-01,  2.16677010e-01,  2.17326254e-01,\n",
       "         2.17721447e-01,  2.18060181e-01,  2.18370691e-01,  2.18963474e-01,\n",
       "         2.19443351e-01,  2.19782084e-01,  2.20092595e-01,  2.20516011e-01,\n",
       "         2.20741823e-01,  2.21137017e-01,  2.21475750e-01,  2.21673355e-01,\n",
       "         2.22181454e-01,  2.22463727e-01,  2.22661316e-01,  2.22915366e-01,\n",
       "         2.23197654e-01,  2.23479927e-01,  2.23621070e-01,  2.23733976e-01,\n",
       "         2.24157393e-01,  2.24100947e-01,  2.23592848e-01,  2.23169416e-01,\n",
       "         2.22774237e-01,  2.22435504e-01,  2.21758038e-01,  2.21249923e-01,\n",
       "         2.20798284e-01,  2.20290184e-01,  2.19782084e-01,  2.19076380e-01,\n",
       "         2.18624741e-01,  2.18060181e-01,  2.17410937e-01,  2.16931075e-01,\n",
       "         2.16394737e-01,  2.15745494e-01,  2.14955121e-01,  2.14390561e-01,\n",
       "         2.13854238e-01,  2.13176757e-01,  2.12499291e-01,  2.11850047e-01,\n",
       "         2.11200818e-01,  2.10438654e-01,  2.09732965e-01,  2.08970815e-01,\n",
       "         2.08123982e-01,  2.07390055e-01,  2.06684351e-01,  2.05781057e-01,\n",
       "         2.04849541e-01,  2.03918025e-01,  2.02930048e-01,  2.02054992e-01,\n",
       "         2.01123476e-01,  2.00050816e-01,  1.98949918e-01,  1.97849035e-01,\n",
       "         1.96663469e-01,  1.95619032e-01,  1.94489926e-01,  1.92993850e-01,\n",
       "         1.91582456e-01,  1.90340430e-01,  1.88872576e-01,  1.87320054e-01,\n",
       "         1.85880423e-01,  1.84214979e-01,  1.82351947e-01,  1.80545360e-01,\n",
       "         1.78682327e-01,  1.76649913e-01,  1.74730420e-01,  1.72641560e-01,\n",
       "         1.70298651e-01,  1.68322697e-01,  1.66036248e-01,  1.63608655e-01,\n",
       "         1.61378652e-01,  1.58979282e-01,  1.56241179e-01,  1.53757125e-01,\n",
       "         1.51188388e-01,  1.48365602e-01,  1.45768642e-01,  1.43058777e-01,\n",
       "         1.40292436e-01,  1.37356743e-01,  1.34618640e-01,  1.31824091e-01,\n",
       "         1.28775477e-01,  1.25839785e-01,  1.22875854e-01,  1.19911931e-01,\n",
       "         1.16891548e-01,  1.13984078e-01,  1.10991925e-01,  1.07745722e-01,\n",
       "         1.04781799e-01,  1.01817876e-01,  9.85716730e-02,  9.54383761e-02,\n",
       "         9.24744532e-02,  8.92000198e-02,  8.60667303e-02,  8.30463469e-02,\n",
       "         7.99412876e-02,  7.66668543e-02,  7.33077377e-02,  7.00333118e-02,\n",
       "         6.69846982e-02,  6.37384951e-02,  6.04640655e-02,  5.72743192e-02,\n",
       "         5.38869761e-02,  5.05843163e-02,  4.71122898e-02,  4.39225435e-02,\n",
       "         4.05069739e-02,  3.71196307e-02,  3.37322876e-02,  3.03449444e-02,\n",
       "         2.70140581e-02,  2.36549433e-02,  2.02111453e-02,  1.65697504e-02,\n",
       "         1.30977249e-02,  9.62569844e-03,  6.06898870e-03], dtype=float32),\n",
       "        array([ 0.00248405, -0.00110089, -0.00457291, -0.00829899, -0.01168633,\n",
       "        -0.0152995 , -0.01885621, -0.02244115, -0.0259414 , -0.02961102,\n",
       "        -0.03345001, -0.03711963, -0.04078925, -0.04454356, -0.04829786,\n",
       "        -0.05193925, -0.05577824, -0.05956078, -0.06325863, -0.06690001,\n",
       "        -0.07082369, -0.07460622, -0.07847344, -0.08222774, -0.08606673,\n",
       "        -0.090188  , -0.09397053, -0.09780952, -0.10145091, -0.10503585,\n",
       "        -0.10890307, -0.11274205, -0.11652458, -0.12025066, -0.12389205,\n",
       "        -0.12742054, -0.13097724, -0.1344775 , -0.13786484, -0.14130865,\n",
       "        -0.14475244, -0.14794219, -0.15118839, -0.15401118, -0.15700333,\n",
       "        -0.1599108 , -0.16273358, -0.16521764, -0.16747586, -0.16990346,\n",
       "        -0.17218992, -0.17441992, -0.1764241 , -0.17854118, -0.1801784 ,\n",
       "        -0.18198498, -0.18396093, -0.18537232, -0.187066  , -0.18853384,\n",
       "        -0.1900017 , -0.19130017, -0.19262688, -0.19398183, -0.19513917,\n",
       "        -0.19643764, -0.19782081, -0.1990346 , -0.20016372, -0.20134929,\n",
       "        -0.2021679 , -0.20326878, -0.20414385, -0.20516005, -0.2061198 ,\n",
       "        -0.20685372, -0.20747474, -0.20801106, -0.20894258, -0.20984587,\n",
       "        -0.21055157, -0.21097499, -0.21165246, -0.21227348, -0.21289448,\n",
       "        -0.21345904, -0.21393892, -0.21475752, -0.2155479 , -0.21608423,\n",
       "        -0.2164512 , -0.21704398, -0.21743917, -0.21789081, -0.21842714,\n",
       "        -0.21882233, -0.21921752, -0.21975385, -0.22012082, -0.22026196,\n",
       "        -0.22054423, -0.22096765, -0.22136284, -0.22178626, -0.222125  ,\n",
       "        -0.2222379 , -0.22215322, -0.22164512, -0.22077006, -0.22006436,\n",
       "        -0.2191893 , -0.2184836 , -0.21791904, -0.21704398, -0.21622537,\n",
       "        -0.21537854, -0.21475752, -0.21385424, -0.21314853, -0.2124146 ,\n",
       "        -0.21142663, -0.21074916, -0.20990233, -0.20888613, -0.2078417 ,\n",
       "        -0.20716423, -0.20628917, -0.20510359, -0.20425676, -0.2031841 ,\n",
       "        -0.20225258, -0.20140575, -0.20005082, -0.19920397, -0.198216  ,\n",
       "        -0.19708689, -0.19601423, -0.19494157, -0.19367132, -0.19242929,\n",
       "        -0.19130017, -0.18994524, -0.18873145, -0.18720713, -0.18571106,\n",
       "        -0.18415853, -0.18288827, -0.1813922 , -0.17969853, -0.17823067,\n",
       "        -0.17650878, -0.17461751, -0.17289561, -0.17117372, -0.16922599,\n",
       "        -0.16710891, -0.16490713, -0.16301586, -0.16067295, -0.15821713,\n",
       "        -0.1559589 , -0.1535313 , -0.15093434, -0.14802687, -0.14509118,\n",
       "        -0.14232485, -0.1394456 , -0.13628408, -0.13315079, -0.12976345,\n",
       "        -0.12660193, -0.12307345, -0.11960142, -0.11601648, -0.11214927,\n",
       "        -0.1087337 , -0.10506408, -0.10142268, -0.09786598, -0.09439395,\n",
       "        -0.09055496, -0.08694179, -0.08338509, -0.07965901, -0.07598939,\n",
       "        -0.07229154, -0.06859369, -0.0647547 , -0.06111331, -0.05761305,\n",
       "        -0.05408457, -0.05024558, -0.04680178, -0.04318862, -0.03963191,\n",
       "        -0.03610343, -0.03251849, -0.02899001, -0.02563089, -0.02235646,\n",
       "        -0.0187433 , -0.01521481, -0.01191215, -0.00869418, -0.00536329,\n",
       "        -0.00191949,  0.00143962,  0.00457291,  0.00793203,  0.01131937,\n",
       "         0.01433975,  0.0175295 ,  0.02066279,  0.02393722,  0.02709874,\n",
       "         0.03014735,  0.0333371 ,  0.03632925,  0.03929317,  0.04237001,\n",
       "         0.04539039,  0.04818495,  0.05131824,  0.05439508,  0.05733077,\n",
       "         0.06018179,  0.06317394,  0.06602495,  0.0687066 ,  0.07175521,\n",
       "         0.07457799,  0.07728787,  0.08002597,  0.0827923 ,  0.08555863,\n",
       "         0.08824027,  0.09109129,  0.09371648,  0.09620053,  0.09899509,\n",
       "         0.10167673,  0.10416079,  0.10687066,  0.10955231,  0.11189522,\n",
       "         0.11435104,  0.11691978,  0.1193756 ,  0.12188788,  0.12431548,\n",
       "         0.1265737 ,  0.1290013 ,  0.13134421,  0.13371535,  0.13603003,\n",
       "         0.13814713,  0.14032067,  0.14243776,  0.14466776,  0.14661548,\n",
       "         0.14850675,  0.15053914,  0.15243042,  0.15409586,  0.15584598,\n",
       "         0.15751143,  0.15906397,  0.16064472,  0.1621408 ,  0.1635522 ,\n",
       "         0.16485068,  0.16626206,  0.16753232,  0.16863321,  0.16970587,\n",
       "         0.17083497,  0.17179473,  0.1727827 ,  0.17365776,  0.17444815,\n",
       "         0.17529498,  0.17597245,  0.1767346 ,  0.1774403 ,  0.17808954,\n",
       "         0.17879523,  0.17938802,  0.18003726,  0.1804889 ,  0.18096878,\n",
       "         0.18167448,  0.1822108 ,  0.18269068,  0.1831141 ,  0.18350929,\n",
       "         0.18387625,  0.18424322,  0.18469486,  0.18520296,  0.18548524,\n",
       "         0.18579574,  0.1862474 ,  0.18655789,  0.18689662,  0.18712245,\n",
       "         0.18734828,  0.1877717 ,  0.18799752,  0.18825157,  0.18853384,\n",
       "         0.18870321,  0.18890081,  0.18907018,  0.18935245,  0.1896065 ,\n",
       "         0.18980409,  0.19005814,  0.19022752,  0.19036865,  0.19039689,\n",
       "         0.19022752,  0.18977587,  0.18940891,  0.18895726,  0.18844916,\n",
       "         0.18813865,  0.18765879,  0.18723537,  0.18689662,  0.18633208,\n",
       "         0.18582398,  0.18537232,  0.18500537,  0.18452549,  0.18407385,\n",
       "         0.1836222 ,  0.18305764,  0.18243663,  0.18195675,  0.18144865,\n",
       "         0.18074296,  0.18026309,  0.1796703 ,  0.17899282,  0.17828713,\n",
       "         0.17772257,  0.17715801,  0.176537  ,  0.17588776,  0.17512561,\n",
       "         0.17436346,  0.17377068,  0.17309321,  0.17221814,  0.17139953,\n",
       "         0.17052448,  0.16987523,  0.16888726,  0.16787106,  0.16696776,\n",
       "         0.1658951 ,  0.1648789 ,  0.16363688,  0.162536  ,  0.16140687,\n",
       "         0.16016485,  0.15880992,  0.15756789,  0.15618472,  0.15466042,\n",
       "         0.15330549,  0.15164004,  0.15000282,  0.1483656 ,  0.14655903,\n",
       "         0.14472422,  0.14288941,  0.14094168,  0.13896573,  0.13696155,\n",
       "         0.13484447,  0.13281205,  0.13063851,  0.12849319,  0.12617852,\n",
       "         0.12397674,  0.12169028,  0.11929092,  0.11706091,  0.11477446,\n",
       "         0.11231863,  0.10997573,  0.10757636,  0.10520522,  0.10260826,\n",
       "         0.1001242 ,  0.09778129,  0.09518433,  0.0927285 ,  0.09010331,\n",
       "         0.08770394,  0.08507875,  0.08248179,  0.07996951,  0.07740078,\n",
       "         0.0748885 ,  0.07226331,  0.06963812,  0.06701293,  0.06430306,\n",
       "         0.06159318,  0.05902445,  0.05628634,  0.05354824,  0.05092305,\n",
       "         0.04815672,  0.04539039,  0.04259583,  0.03991419,  0.03714786,\n",
       "         0.03426862,  0.03153051,  0.02879241,  0.02605431,  0.02325975,\n",
       "         0.02049342,  0.0175295 ,  0.01456557,  0.01157342,  0.00880709,\n",
       "         0.00584317,  0.00282279], dtype=float32),\n",
       "        array([ 2.54050683e-04, -2.73810187e-03, -5.81493834e-03, -8.80709104e-03,\n",
       "        -1.17427874e-02, -1.47067122e-02, -1.76424086e-02, -2.06345618e-02,\n",
       "        -2.36831699e-02, -2.65906397e-02, -2.96957046e-02, -3.28007676e-02,\n",
       "        -3.58211473e-02, -3.89826670e-02, -4.22006436e-02, -4.50234301e-02,\n",
       "        -4.83260900e-02, -5.14593795e-02, -5.46491295e-02, -5.76977357e-02,\n",
       "        -6.07745722e-02, -6.40207753e-02, -6.70693815e-02, -7.02026784e-02,\n",
       "        -7.32795149e-02, -7.64128044e-02, -7.94331878e-02, -8.27640742e-02,\n",
       "        -8.58691335e-02, -8.90024304e-02, -9.21074897e-02, -9.51561034e-02,\n",
       "        -9.81764793e-02, -1.01112179e-01, -1.04217239e-01, -1.07068256e-01,\n",
       "        -1.09721676e-01, -1.12600915e-01, -1.15310788e-01, -1.17823064e-01,\n",
       "        -1.20448261e-01, -1.22791171e-01, -1.25134081e-01, -1.27561674e-01,\n",
       "        -1.29622310e-01, -1.31570026e-01, -1.33517757e-01, -1.35606617e-01,\n",
       "        -1.37272060e-01, -1.39022186e-01, -1.40800536e-01, -1.42409533e-01,\n",
       "        -1.44131422e-01, -1.45683959e-01, -1.47180036e-01, -1.48563206e-01,\n",
       "        -1.49974599e-01, -1.51075482e-01, -1.52486876e-01, -1.53587759e-01,\n",
       "        -1.54632196e-01, -1.55676618e-01, -1.56608135e-01, -1.57709032e-01,\n",
       "        -1.58386499e-01, -1.59374475e-01, -1.60164848e-01, -1.60955235e-01,\n",
       "        -1.61858529e-01, -1.62535995e-01, -1.63382828e-01, -1.64116755e-01,\n",
       "        -1.64850682e-01, -1.65782198e-01, -1.66459665e-01, -1.66883081e-01,\n",
       "        -1.67588770e-01, -1.68153331e-01, -1.68576747e-01, -1.69197768e-01,\n",
       "        -1.69762328e-01, -1.70355111e-01, -1.70976117e-01, -1.71258405e-01,\n",
       "        -1.71625361e-01, -1.72218144e-01, -1.72613338e-01, -1.73121437e-01,\n",
       "        -1.73516631e-01, -1.73940048e-01, -1.74194098e-01, -1.74561054e-01,\n",
       "        -1.74843341e-01, -1.75097391e-01, -1.75577253e-01, -1.76085368e-01,\n",
       "        -1.76424101e-01, -1.76621690e-01, -1.77016884e-01, -1.77158013e-01,\n",
       "        -1.77553207e-01, -1.77920178e-01, -1.78258911e-01, -1.78541183e-01,\n",
       "        -1.78315356e-01, -1.77835494e-01, -1.77242696e-01, -1.76762834e-01,\n",
       "        -1.76113591e-01, -1.75577253e-01, -1.74984470e-01, -1.74278781e-01,\n",
       "        -1.73657760e-01, -1.72952071e-01, -1.72556877e-01, -1.71738267e-01,\n",
       "        -1.70806751e-01, -1.70213968e-01, -1.69451818e-01, -1.68774351e-01,\n",
       "        -1.68012202e-01, -1.67362958e-01, -1.66657254e-01, -1.65979788e-01,\n",
       "        -1.64991811e-01, -1.64201438e-01, -1.63580418e-01, -1.62790045e-01,\n",
       "        -1.62027895e-01, -1.61152825e-01, -1.60277769e-01, -1.59459159e-01,\n",
       "        -1.58555865e-01, -1.57737255e-01, -1.56805739e-01, -1.55958906e-01,\n",
       "        -1.54999152e-01, -1.53926492e-01, -1.52853832e-01, -1.51724726e-01,\n",
       "        -1.50539145e-01, -1.49438262e-01, -1.48647889e-01, -1.47151813e-01,\n",
       "        -1.46022692e-01, -1.44752443e-01, -1.43284589e-01, -1.42099023e-01,\n",
       "        -1.40715852e-01, -1.39248013e-01, -1.37921304e-01, -1.36538133e-01,\n",
       "        -1.34646863e-01, -1.33094341e-01, -1.31400660e-01, -1.29706994e-01,\n",
       "        -1.27815723e-01, -1.25839785e-01, -1.23892054e-01, -1.21718511e-01,\n",
       "        -1.19686104e-01, -1.17569014e-01, -1.15226105e-01, -1.12629145e-01,\n",
       "        -1.10229775e-01, -1.07773952e-01, -1.05176985e-01, -1.02523573e-01,\n",
       "        -9.98136997e-02, -9.71320495e-02, -9.41963494e-02, -9.14017931e-02,\n",
       "        -8.82685035e-02, -8.52763504e-02, -8.25382471e-02, -7.94896409e-02,\n",
       "        -7.64410347e-02, -7.33641982e-02, -7.03438148e-02, -6.71540648e-02,\n",
       "        -6.43312857e-02, -6.12544455e-02, -5.81776090e-02, -5.51290028e-02,\n",
       "        -5.21650761e-02, -4.93705198e-02, -4.63219099e-02, -4.32168469e-02,\n",
       "        -4.02246937e-02, -3.72889973e-02, -3.41839343e-02, -3.14458311e-02,\n",
       "        -2.84819063e-02, -2.55179815e-02, -2.25822851e-02, -1.96183603e-02,\n",
       "        -1.69367138e-02, -1.41421556e-02, -1.11500025e-02, -8.44012853e-03,\n",
       "        -5.70202665e-03, -2.79455772e-03, -1.12911424e-04,  2.65341834e-03,\n",
       "         5.36329253e-03,  8.01671110e-03,  1.06701292e-02,  1.34364590e-02,\n",
       "         1.60616487e-02,  1.85739286e-02,  2.11991202e-02,  2.37960815e-02,\n",
       "         2.62801331e-02,  2.87924130e-02,  3.15869711e-02,  3.39863375e-02,\n",
       "         3.64986174e-02,  3.90673503e-02,  4.15796302e-02,  4.40636836e-02,\n",
       "         4.64630499e-02,  4.88906465e-02,  5.13464697e-02,  5.38305193e-02,\n",
       "         5.60887493e-02,  5.85445724e-02,  6.10003956e-02,  6.32303953e-02,\n",
       "         6.55168518e-02,  6.78315386e-02,  7.01744482e-02,  7.24326745e-02,\n",
       "         7.46909082e-02,  7.69209042e-02,  7.91791305e-02,  8.12962204e-02,\n",
       "         8.35544541e-02,  8.57562274e-02,  8.79015401e-02,  8.99903998e-02,\n",
       "         9.21639502e-02,  9.42810401e-02,  9.63698998e-02,  9.84869897e-02,\n",
       "         1.00462936e-01,  1.02551796e-01,  1.04499519e-01,  1.06503695e-01,\n",
       "         1.08564332e-01,  1.10512055e-01,  1.12459779e-01,  1.14322811e-01,\n",
       "         1.16101168e-01,  1.17992438e-01,  1.19827248e-01,  1.21492691e-01,\n",
       "         1.23299271e-01,  1.25077620e-01,  1.26630157e-01,  1.28210917e-01,\n",
       "         1.29735216e-01,  1.31259531e-01,  1.32642686e-01,  1.34167001e-01,\n",
       "         1.35493711e-01,  1.36594594e-01,  1.37949526e-01,  1.39078647e-01,\n",
       "         1.40264213e-01,  1.41534463e-01,  1.42437756e-01,  1.43312827e-01,\n",
       "         1.44244343e-01,  1.45091176e-01,  1.45938009e-01,  1.46841303e-01,\n",
       "         1.47575229e-01,  1.48224473e-01,  1.48873702e-01,  1.49522945e-01,\n",
       "         1.50228649e-01,  1.50764972e-01,  1.51329532e-01,  1.51922315e-01,\n",
       "         1.52430415e-01,  1.53023198e-01,  1.53390169e-01,  1.53870046e-01,\n",
       "         1.54265225e-01,  1.54575735e-01,  1.55027375e-01,  1.55422568e-01,\n",
       "         1.55874223e-01,  1.56241179e-01,  1.56523451e-01,  1.56833962e-01,\n",
       "         1.57088012e-01,  1.57398522e-01,  1.57765478e-01,  1.58019528e-01,\n",
       "         1.58273578e-01,  1.58499405e-01,  1.58781692e-01,  1.59035742e-01,\n",
       "         1.59233332e-01,  1.59515604e-01,  1.59713209e-01,  1.59826115e-01,\n",
       "         1.60108402e-01,  1.60277769e-01,  1.60390675e-01,  1.60672948e-01,\n",
       "         1.60926998e-01,  1.61039919e-01,  1.61209285e-01,  1.61350414e-01,\n",
       "         1.61576241e-01,  1.61886752e-01,  1.61689162e-01,  1.61322191e-01,\n",
       "         1.60955235e-01,  1.60616502e-01,  1.60362452e-01,  1.59939021e-01,\n",
       "         1.59600288e-01,  1.59205109e-01,  1.58922821e-01,  1.58584088e-01,\n",
       "         1.58104211e-01,  1.57793716e-01,  1.57454982e-01,  1.57003328e-01,\n",
       "         1.56608135e-01,  1.56212956e-01,  1.55704856e-01,  1.55309662e-01,\n",
       "         1.54942691e-01,  1.54378146e-01,  1.53954729e-01,  1.53559536e-01,\n",
       "         1.53079659e-01,  1.52543336e-01,  1.52176365e-01,  1.51668265e-01,\n",
       "         1.51019022e-01,  1.50595605e-01,  1.50115728e-01,  1.49466500e-01,\n",
       "         1.48789018e-01,  1.48252696e-01,  1.47659913e-01,  1.46982446e-01,\n",
       "         1.46276742e-01,  1.45655736e-01,  1.44921809e-01,  1.44187883e-01,\n",
       "         1.43566877e-01,  1.42720029e-01,  1.41873196e-01,  1.40969902e-01,\n",
       "         1.40066624e-01,  1.39248013e-01,  1.38316497e-01,  1.37413204e-01,\n",
       "         1.36284083e-01,  1.35154977e-01,  1.34082317e-01,  1.32981420e-01,\n",
       "         1.31824091e-01,  1.30469143e-01,  1.29227117e-01,  1.27900407e-01,\n",
       "         1.26376107e-01,  1.24908261e-01,  1.23440415e-01,  1.21746741e-01,\n",
       "         1.20278895e-01,  1.18585221e-01,  1.16891548e-01,  1.15169652e-01,\n",
       "         1.13391295e-01,  1.11725852e-01,  1.09919265e-01,  1.08028002e-01,\n",
       "         1.06164962e-01,  1.04330152e-01,  1.02382429e-01,  1.00547619e-01,\n",
       "         9.85716730e-02,  9.66239497e-02,  9.47609097e-02,  9.27285030e-02,\n",
       "         9.06960964e-02,  8.86636898e-02,  8.66877437e-02,  8.45142007e-02,\n",
       "         8.25100243e-02,  8.05623010e-02,  7.83887506e-02,  7.63281211e-02,\n",
       "         7.41828009e-02,  7.20657110e-02,  6.99768513e-02,  6.77186251e-02,\n",
       "         6.55733049e-02,  6.35691285e-02,  6.13673590e-02,  5.91938123e-02,\n",
       "         5.70484959e-02,  5.49031794e-02,  5.26449494e-02,  5.03584929e-02,\n",
       "         4.82696332e-02,  4.60114032e-02,  4.37814035e-02,  4.16078605e-02,\n",
       "         3.93214040e-02,  3.69502641e-02,  3.47484909e-02,  3.24902609e-02,\n",
       "         3.01755778e-02,  2.78608929e-02,  2.55179815e-02,  2.32032966e-02,\n",
       "         2.08321568e-02,  1.85457002e-02,  1.61745604e-02,  1.38316490e-02,\n",
       "         1.14322817e-02,  9.00468603e-03,  6.54886244e-03,  4.20595054e-03],\n",
       "       dtype=float32),\n",
       "        array([ 0.00197595, -0.00056456, -0.00302038, -0.00539152, -0.00793203,\n",
       "        -0.01041608, -0.01281545, -0.01521481, -0.01789646, -0.02040874,\n",
       "        -0.02280811, -0.02534861, -0.02805849, -0.03059899, -0.03299836,\n",
       "        -0.03559532, -0.0381076 , -0.04061988, -0.04313216, -0.04575735,\n",
       "        -0.04829786, -0.05083837, -0.05343533, -0.05603229, -0.05865748,\n",
       "        -0.0613109 , -0.06379496, -0.06633546, -0.06896065, -0.07150116,\n",
       "        -0.07392875, -0.07638457, -0.07889686, -0.08118331, -0.08355445,\n",
       "        -0.08589736, -0.08807091, -0.09035736, -0.09255914, -0.09445041,\n",
       "        -0.09648281, -0.09848699, -0.10037825, -0.10218484, -0.10385028,\n",
       "        -0.10557218, -0.10698357, -0.10850788, -0.11008864, -0.11116129,\n",
       "        -0.11243155, -0.11381471, -0.11502851, -0.11621408, -0.11739965,\n",
       "        -0.11852876, -0.11943205, -0.12047648, -0.12143623, -0.12239598,\n",
       "        -0.1233275 , -0.12417433, -0.12490826, -0.12572686, -0.12643257,\n",
       "        -0.12694067, -0.12775928, -0.1284085 , -0.12911421, -0.12967877,\n",
       "        -0.13032801, -0.13080788, -0.13140066, -0.13190877, -0.13252978,\n",
       "        -0.13303788, -0.13343307, -0.1339694 , -0.13439281, -0.13481623,\n",
       "        -0.13535257, -0.13588889, -0.13605826, -0.13662282, -0.13687687,\n",
       "        -0.13713092, -0.13758257, -0.13780838, -0.13817535, -0.13840118,\n",
       "        -0.13859877, -0.13885282, -0.1393327 , -0.1394456 , -0.13969965,\n",
       "        -0.14001016, -0.14029244, -0.14060295, -0.14077231, -0.14111105,\n",
       "        -0.1415627 , -0.1416756 , -0.1418732 , -0.14184497, -0.14128041,\n",
       "        -0.14082877, -0.1404618 , -0.13998194, -0.1393327 , -0.13876814,\n",
       "        -0.13834472, -0.13800599, -0.13732852, -0.13687687, -0.13628408,\n",
       "        -0.13571952, -0.13512674, -0.13450573, -0.13413876, -0.1334613 ,\n",
       "        -0.13292497, -0.13227573, -0.13173941, -0.13109016, -0.13029978,\n",
       "        -0.12967877, -0.1290013 , -0.12832382, -0.12758991, -0.12702535,\n",
       "        -0.12640433, -0.12547281, -0.12488003, -0.12408965, -0.12321459,\n",
       "        -0.12253712, -0.12171851, -0.12078699, -0.11991193, -0.119178  ,\n",
       "        -0.11810534, -0.11717383, -0.11607294, -0.11511319, -0.11406876,\n",
       "        -0.11302433, -0.11209282, -0.1109637 , -0.10969345, -0.10839497,\n",
       "        -0.10737876, -0.10591091, -0.10466889, -0.10328572, -0.10178965,\n",
       "        -0.10040648, -0.09865636, -0.09704737, -0.09532547, -0.09354711,\n",
       "        -0.0917123 , -0.08979281, -0.08784509, -0.08586913, -0.08394964,\n",
       "        -0.0817761 , -0.07946141, -0.0773161 , -0.07508609, -0.07274318,\n",
       "        -0.07025913, -0.06788799, -0.0655733 , -0.06300458, -0.06049229,\n",
       "        -0.05803647, -0.0556371 , -0.05315305, -0.05075368, -0.04821318,\n",
       "        -0.0457009 , -0.0432733 , -0.04067634, -0.03813583, -0.03573646,\n",
       "        -0.03345001, -0.03085305, -0.02839722, -0.02599785, -0.02365494,\n",
       "        -0.02125558, -0.0187433 , -0.01637216, -0.0138881 , -0.01168633,\n",
       "        -0.00945633, -0.00711342, -0.0046576 , -0.00225823, -0.00016937,\n",
       "         0.00208886,  0.00440355,  0.00666177,  0.00889177,  0.01112178,\n",
       "         0.01335178,  0.01546886,  0.01769887,  0.01990064,  0.02196127,\n",
       "         0.02410659,  0.0262519 ,  0.02831254,  0.03037317,  0.03246203,\n",
       "         0.03452267,  0.03661153,  0.03864393,  0.04064811,  0.04256761,\n",
       "         0.04454356,  0.04660419,  0.04858014,  0.05044318,  0.05241913,\n",
       "         0.05445153,  0.0563428 ,  0.05823407,  0.06021002,  0.06204483,\n",
       "         0.06387964,  0.06568622,  0.06757749,  0.06944052,  0.07119065,\n",
       "         0.07299723,  0.07480381,  0.07663862,  0.07827584,  0.08011065,\n",
       "         0.08186078,  0.08361091,  0.08530458,  0.08697002,  0.08852255,\n",
       "         0.09024445,  0.09182521,  0.0934342 ,  0.09507142,  0.09656749,\n",
       "         0.09820471,  0.09975724,  0.10122509,  0.10272116,  0.10416079,\n",
       "         0.10560041,  0.1070118 ,  0.10831028,  0.10969345,  0.1109637 ,\n",
       "         0.1121775 ,  0.11344775,  0.11452041,  0.11559307,  0.11672218,\n",
       "         0.1176537 ,  0.11864167,  0.1195732 ,  0.1203918 ,  0.12126686,\n",
       "         0.12214193,  0.12290408,  0.12355332,  0.12437193,  0.12502117,\n",
       "         0.12550105,  0.12609383,  0.12682775,  0.1272794 ,  0.1277875 ,\n",
       "         0.1282956 ,  0.12877548,  0.1291989 ,  0.12956585,  0.12996104,\n",
       "         0.13038446,  0.1307232 ,  0.13106193,  0.13137244,  0.13171117,\n",
       "         0.13202168,  0.13233219,  0.13272737,  0.13298142,  0.13315079,\n",
       "         0.13351776,  0.13380003,  0.13394117,  0.13422345,  0.13439281,\n",
       "         0.13450573,  0.13475978,  0.13501383,  0.13523966,  0.13532434,\n",
       "         0.13549371,  0.13574776,  0.13583244,  0.13600181,  0.13611472,\n",
       "         0.13634054,  0.13650991,  0.13662282,  0.13679218,  0.13698979,\n",
       "         0.13713092,  0.13713092,  0.13707447,  0.13687687,  0.13650991,\n",
       "         0.13625586,  0.13588889,  0.13566308,  0.13538079,  0.1350703 ,\n",
       "         0.134788  ,  0.13453396,  0.13425168,  0.13385649,  0.13360244,\n",
       "         0.1332637 ,  0.1329532 ,  0.13264269,  0.13227573,  0.1320499 ,\n",
       "         0.13165471,  0.13120307,  0.13086434,  0.13046914,  0.13018687,\n",
       "         0.12973522,  0.1291989 ,  0.12888838,  0.12846497,  0.12804155,\n",
       "         0.12753345,  0.12711003,  0.12665838,  0.1260656 ,  0.12552927,\n",
       "         0.12513408,  0.12462598,  0.12397674,  0.12341218,  0.12279117,\n",
       "         0.12217016,  0.12152091,  0.12081522,  0.1200813 ,  0.11934737,\n",
       "         0.11861344,  0.11782306,  0.11703268,  0.1162423 ,  0.11531079,\n",
       "         0.11435104,  0.11347598,  0.11237509,  0.11130244,  0.11039914,\n",
       "         0.10918535,  0.10808446,  0.10681421,  0.10545927,  0.10421724,\n",
       "         0.10294699,  0.10164851,  0.10006775,  0.09868458,  0.09735787,\n",
       "         0.09597471,  0.09445041,  0.0929261 ,  0.09140179,  0.08984926,\n",
       "         0.08835319,  0.08671597,  0.08516344,  0.08363914,  0.08203015,\n",
       "         0.08033647,  0.07867103,  0.07706205,  0.07536837,  0.07370293,\n",
       "         0.07198103,  0.07020268,  0.06856546,  0.06690001,  0.06512167,\n",
       "         0.06331508,  0.06167786,  0.05992774,  0.05812115,  0.05637103,\n",
       "         0.0546209 ,  0.05284255,  0.05103596,  0.04922938,  0.04747925,\n",
       "         0.04564444,  0.04389431,  0.04208773,  0.04008355,  0.0383052 ,\n",
       "         0.03652684,  0.03466381,  0.03274431,  0.0309095 ,  0.02899001,\n",
       "         0.02718342,  0.02526393,  0.02331621,  0.02142494,  0.01953368,\n",
       "         0.01761418,  0.01552532,  0.01369051,  0.01168633,  0.00965393,\n",
       "         0.0077062 ,  0.00575848,  0.00381076], dtype=float32)],\n",
       "       dtype=object),\n",
       " 'washoutLength': 50}"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['winner']['net']['p']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "figsize(20,10)\n",
    "import librosa.display\n",
    "subplot(2,1,1)\n",
    "plt.subplots_adjust(left=None, bottom=None, right=None, top=None, wspace=None, hspace=0.3)\n",
    "CQT = librosa.amplitude_to_db(np.abs(librosa.stft(data['winner']['waveform'], n_fft=1024, hop_length=64)), ref=np.max)\n",
    "ax = librosa.display.specshow(CQT, x_axis='time', y_axis='log', sr=22050, cmap=plt.get_cmap('viridis'))\n",
    "plt.colorbar(format='%+2.0f dB')\n",
    "ax.set_ylabel(\"frequency (Hz)\", fontsize=16)\n",
    "ax.set_xlabel(\"time (s)\", fontsize=16)\n",
    "ax.set_title('Reconstruction', fontsize=22)\n",
    "\n",
    "subplot(2,1,2)\n",
    "CQT = librosa.amplitude_to_db(np.abs(librosa.stft(y, n_fft=1024, hop_length=64)), ref=np.max)\n",
    "ax = librosa.display.specshow(CQT, x_axis='time', y_axis='log', sr=22050, cmap=plt.get_cmap('viridis'))\n",
    "plt.colorbar(format='%+2.0f dB')\n",
    "ax.set_title('Original', fontsize=22)\n",
    "ax.set_ylabel(\"frequency (Hz)\", fontsize=16)\n",
    "ax.set_xlabel(\"time (s)\", fontsize=16)\n",
    "plt.savefig(\"tmp/kickSpectra.pdf\", bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}