Tree_final.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from scipy.optimize import linprog \n",
    "from twinsvm import twinsvmclassifier\n",
    "import random\n",
    "import sys"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def optimize(X,y,delta, alpha):\n",
    "    (m,n) = X.shape\n",
    "    K = (y == 0).sum()\n",
    "    M = (y == 1).sum()\n",
    "    A = np.zeros((2*n,m))\n",
    "    B = np.zeros((m))\n",
    "    for i in range(m):\n",
    "        if y[i] == 0:     \n",
    "            V = np.zeros((2,m))\n",
    "            V[0][i] = 1\n",
    "            V[1][i] = -1\n",
    "            A = np.vstack((A,V))\n",
    "            for j in range(n):\n",
    "                A[j][i] = X[i][j]\n",
    "                A[j+n][i] = -1*X[i][j]\n",
    "                B[i] = -delta\n",
    "        else:  \n",
    "            V = np.zeros((2,m))\n",
    "            V[0][i] = -1\n",
    "            V[1][i] = 1\n",
    "            A = np.vstack((A,V)) \n",
    "            for j in range(n):\n",
    "                A[j][i] = X[i][j]\n",
    "                A[j+n][i] = -1*X[i][j]\n",
    "                B[i] = delta\n",
    "    A = np.vstack((A, np.zeros((n,m))))   # dimension of A is 2n+2m+n, m\n",
    "    s1 = np.zeros((3*n+2*m))\n",
    "    s2 = np.zeros((3*n+2*m))\n",
    "    W = np.zeros((2*n+2*m+n))\n",
    "    for i in range(m):\n",
    "        if y[i] == 0:\n",
    "            W[2*i+2*n] = 0                  #\n",
    "            W[2*i+2*n+1] = 1/K              #\n",
    "            s2[2*n+2*i] = -1\n",
    "        else:\n",
    "            W[2*i+2*n] = 0                  #\n",
    "            W[2*i+2*n+1] = 1/M              #\n",
    "            s1[2*n+2*i] = -1\n",
    "    for i in range(n):\n",
    "        W[i+2*n+2*m] = alpha   ##########\n",
    "    A = A.T\n",
    "    #Ax=B the equality contraints\n",
    "    C = np.zeros((2*n, 3*n+2*m))\n",
    "    D = np.zeros((2*n+2))\n",
    "    D[2*n+2-1] = -1\n",
    "    D[2*n+2-2] = -1\n",
    "    for i in range (n):\n",
    "        C[i,i] = 1\n",
    "        C[i+n,i] = -1\n",
    "        C[i, i+n] = -1\n",
    "        C[i+n, i+n] = 1\n",
    "        C[i, i+2*n+2*m] = -1\n",
    "        C[i+n, i+2*n+2*m] = -1\n",
    "        # Cx<=D\n",
    "    C = np.vstack((C, s1))\n",
    "    C = np.vstack((C, s2))\n",
    "    \n",
    "    res = linprog(W, A_ub=C, b_ub=D, A_eq=A, b_eq=B,  bounds=(0, None), method = 'simplex') \n",
    "    wd = res.x[:n]-res.x[n:2*n]\n",
    "    si1 = np.zeros((m))\n",
    "    si2 = np.zeros((m))\n",
    "    for i in range(m):\n",
    "        si1[i] = res.x[i*2+2*n]\n",
    "    for i in range(m):\n",
    "        si2[i] = res.x[i*2+2*n+1]\n",
    "    print(\"wd:\"+str(wd))\n",
    "    return np.matrix(wd),None\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def y_hat(X, weights, delta, y,twin_svm_en,clf):\n",
    "    if(twin_svm_en == 0):\n",
    "        n = len(X)\n",
    "        y_res = np.asarray(np.dot(weights, X.T))\n",
    "        for i in range(n):\n",
    "            if y_res[0][i] >= delta:\n",
    "                y_res[0][i] = 1\n",
    "            elif y_res[0][i] <= -delta:\n",
    "                y_res[0][i] = 0\n",
    "            else :\n",
    "                y_res[0][i] = 1-y[i]\n",
    "    else:\n",
    "        Xx = X[:,1:len(X[0])]\n",
    "        y_pred = clf.predict(Xx)\n",
    "        y_res = np.array([y_pred])\n",
    "        \n",
    "    return y_res[0]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def classes(y,y_res):\n",
    "    C1 = []\n",
    "    C2 = []\n",
    "    C3 = []\n",
    "    C4 = []\n",
    "    y = y.astype(int)\n",
    "    y_res = y_res.astype(int)\n",
    "    for i in range(len(y_res)): \n",
    "        if y[i] == 0 and y_res[i] == 0:\n",
    "            C1.append(i)\n",
    "        if y[i] == 1 and y_res[i] == 1:\n",
    "            C2.append(i)\n",
    "        if y[i] == 1 and y_res[i] == 0:\n",
    "            C3.append(i)\n",
    "        if y[i] == 0 and y_res[i] == 1:\n",
    "            C4.append(i)\n",
    "    return C1, C2, C3, C4\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "#plane1 < plane2 => class 0\n",
    "#plane1 >= plane2 => class 1\n",
    "# plane1-plane2 < 0 => class 0 \n",
    "# plane1-plane2 >= 0 => class 1\n",
    "#C3 => desired 1 we get 0\n",
    "#so plane1 < plane2 \n",
    "#   plane1-plane2 < 0\n",
    "#   Wa decided by min(plane1-plane2)\n",
    "def find_Wa(X, C3, weights):\n",
    "    res = sys.maxsize\n",
    "    y=0\n",
    "    ans =0\n",
    "    for i in C3:\n",
    "        if(weights.shape[1]==X.shape[1]):\n",
    "            y = np.asarray(np.dot(weights, X[i].T))\n",
    "            res = min(res, y[0][0])\n",
    "            ans = 1-res+0.0001\n",
    "        else:\n",
    "            print(\"twinsvm\")\n",
    "            XX = X[i,1:len(X[i])]\n",
    "            weights = np.asarray(weights)\n",
    "            X= np.asarray(X)\n",
    "            b1 = weights[0][0]\n",
    "            w1 = weights[0][1:X.shape[1]]\n",
    "            b2 = weights[0][X.shape[1]]\n",
    "            w2 = weights[0][X.shape[1]+1:2*(X.shape[1])]\n",
    "            w1mod = np.linalg.norm(w1)\n",
    "            w2mod = np.linalg.norm(w2)\n",
    "            y1 = np.dot(XX,w1)+ b1\n",
    "            y2 = np.dot(XX,w2)+ b2\n",
    "            distFromPlane1 = y1/w1mod \n",
    "            distFromPlane2 = y2/w2mod \n",
    "            dis = (distFromPlane1 - distFromPlane2)\n",
    "            res = min(res, dis)\n",
    "            ans = -res+0.000001\n",
    "    return ans\n",
    "\n",
    "#C4 => desired 0 we get 1\n",
    "#so plane1 >= plane2\n",
    "#   plane1-plane2 >= 0\n",
    "#   Wb decided by max(plane1-plane2)\n",
    "def find_Wb(X, C4, weights):\n",
    "    res = -sys.maxsize-1\n",
    "    y=0\n",
    "    ans = 0\n",
    "    for i in C4:\n",
    "        if(weights.shape[1]==X.shape[1]):\n",
    "            y = np.asarray(np.dot(weights, X[i].T))\n",
    "            res = max(res, y[0][0])\n",
    "            ans = -1-res-0.0000001\n",
    "        else:\n",
    "            print(\"twinsvm\")\n",
    "            XX = X[i,1:len(X[i])]\n",
    "            weights = np.asarray(weights)\n",
    "            X= np.asarray(X)\n",
    "            b1 = weights[0][0]\n",
    "            w1 = weights[0][1:X.shape[1]]\n",
    "            b2 = weights[0][X.shape[1]]\n",
    "            w2 = weights[0][X.shape[1]+1:2*(X.shape[1])]\n",
    "            w1mod = np.linalg.norm(w1)\n",
    "            w2mod = np.linalg.norm(w2)\n",
    "            y1 = np.dot(XX,w1)+ b1\n",
    "            y2 = np.dot(XX,w2)+ b2\n",
    "            distFromPlane1 = y1/w1mod \n",
    "            distFromPlane2 = y2/w2mod \n",
    "            dis = (distFromPlane1 - distFromPlane2)\n",
    "            res = max(res, dis)\n",
    "            ans = -res-0.000001\n",
    "    return ans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def quantize(a,b,twin_node,twin_classifier,XX):\n",
    "    if(twin_node == 0):\n",
    "        res = np.dot(a,b.T)\n",
    "        print(\"res_before_quant:\"+str(res))\n",
    "        res = res.astype(int)\n",
    "        for j in range(res.shape[0]):\n",
    "            if(res[j][0] >= delta):\n",
    "                res[j][0] = 1\n",
    "            elif(res[j][0] <= -delta):\n",
    "                res[j][0] = 0\n",
    "            else:\n",
    "                print(\"delta condition not met\")\n",
    "        print(\"res_after_quant:\"+str(res))\n",
    "    else:\n",
    "        Xd = a[:,1:XX.shape[1]]\n",
    "        if(a.shape[1] > XX.shape[1]):\n",
    "            XAB = a[:,XX.shape[1]:a.shape[1]]\n",
    "        print(\"twin_svm_res\")\n",
    "        b1 = b[0][0]\n",
    "        w1 = b[0][1:XX.shape[1]].reshape(XX.shape[1]-1,1)\n",
    "        b2 = b[0][XX.shape[1]]\n",
    "        w2 = b[0][XX.shape[1]+1:2*(XX.shape[1])].reshape(XX.shape[1]-1,1)\n",
    "        wAB = b[0][2*(XX.shape[1]):b.shape[1]].reshape(-1,1)\n",
    "        w1mod = np.linalg.norm(w1)\n",
    "        w2mod = np.linalg.norm(w2)\n",
    "        y1 = np.dot(Xd,w1)+ b1*np.ones((XX.shape[0],1))\n",
    "        y2 = np.dot(Xd,w2)+ b2*np.ones((XX.shape[0],1))\n",
    "        distFromPlane1 = y1/w1mod \n",
    "        distFromPlane2 = y2/w2mod \n",
    "        res = (distFromPlane1 - distFromPlane2) + np.dot(XAB,wAB)\n",
    "        print(\"res_twin_bef:\"+str(res))\n",
    "        for j in range(res.shape[0]):\n",
    "            if(res[j][0] >= 0):\n",
    "                res[j][0] = 1\n",
    "            elif(res[j][0] < 0):\n",
    "                res[j][0] = 0\n",
    "        print(\"res_twin_aft:\"+str(res))\n",
    "    return res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "class Neuron:\n",
    "    \n",
    "    def __init__(self,n,twin_node,twin_classifier):\n",
    "        self.X = np.zeros((1,n))\n",
    "        self.inp = n\n",
    "        self.A = None\n",
    "        self.B = None\n",
    "        self.twin_node = twin_node\n",
    "        self.twin_classifier = twin_classifier\n",
    "        self.weight = 0\n",
    "        \n",
    "    def insert(self,neuron_type, weight):     #### Type is a restricted variable, Type was always set to A, weigth to 1\n",
    "        if neuron_type == 'A':\n",
    "            self.A = Neuron(self.inp,0,None)\n",
    "            self.A.weight = weight\n",
    "            self.X = np.hstack((self.X, np.array([[weight]])))\n",
    "        else:\n",
    "            self.B = Neuron(self.inp,0,None)\n",
    "            self.B.weight = weight\n",
    "            self.X = np.hstack((self.X, np.array([[weight]])))\n",
    "            \n",
    "    def calculate(self,XX):\n",
    "        if self.A != None and self.B != None:\n",
    "            print(\"AB\")\n",
    "            y_1 = self.A.calculate(XX)\n",
    "            y_2 = self.B.calculate(XX)\n",
    "            print(\"y_1:\"+str(y_1))\n",
    "            print(\"y_2:\"+str(y_2))\n",
    "            res = quantize(np.hstack((XX, y_1, y_2)),self.X,self.twin_node,self.twin_classifier,XX)\n",
    "            return res\n",
    "        elif self.A != None:\n",
    "            print(\"A\")\n",
    "            y_1 = self.A.calculate(XX)\n",
    "            print(\"y_1:\"+str(y_1))\n",
    "            res = quantize(np.hstack((XX, y_1)),self.X,self.twin_node,self.twin_classifier,XX)\n",
    "            return res\n",
    "        elif self.B != None:\n",
    "            print(\"B\")\n",
    "            y_2 = self.B.calculate(XX)\n",
    "            print(\"y_2:\"+str(y_2))\n",
    "            res = quantize(np.hstack((XX, y_2)),self.X,self.twin_node,self.twin_classifier,XX)\n",
    "            return res\n",
    "        else:\n",
    "            print(\"no child node\")\n",
    "            res = quantize(XX,self.X,self.twin_node,self.twin_classifier,XX)\n",
    "            return res\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def get_twin_classifier(X,y):\n",
    "    while(True):\n",
    "        C1value = random.uniform(0,5)\n",
    "        C2value = random.uniform(0,5)\n",
    "        params = {'Epsilon1': 0.1, 'Epsilon2': 0.1, 'C1': C1value, 'C2': C2value,'kernel_type':0,'kernel_param': 2}\n",
    "        clf=twinsvmclassifier(**params)\n",
    "        clf.fit(X,y)\n",
    "        y_pred = clf.predict(X)\n",
    "        C1,C2,C3,C4 = classes(y, y_pred)\n",
    "        if(not(len(C1)==0 or len(C2)==0)):\n",
    "            print(\"#######\")\n",
    "            print(C1)\n",
    "            print(C2)\n",
    "            print(C3)\n",
    "            print(C4)\n",
    "            print(\"#######\")\n",
    "            print(\"wb:\"+str(clf.getwb()))\n",
    "            return clf    \n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def twin(X,y):\n",
    "    XX = X[:,1:len(X[0])]\n",
    "    yy = y\n",
    "    print(\"XX:\"+str(XX))\n",
    "    print(\"yy:\"+str(yy))\n",
    "    clf = get_twin_classifier(XX, yy)\n",
    "    clf.fit(XX,yy)\n",
    "    w1,b1,w2,b2 = clf.getwb()\n",
    "    w1 = w1.T\n",
    "    w2 = w2.T\n",
    "    b1 = np.array([b1])\n",
    "    b2 = np.array([b2])\n",
    "    wd = np.hstack((b1,w1,b2,w2))\n",
    "    print(\"wdtwin:\"+str(wd))\n",
    "    return wd,clf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def addneuron_A(X, y, W, C1, C3, C4, delta, alpha, curr_Neuron):\n",
    "    n = X.shape[1]\n",
    "    print('A')\n",
    "    m = len(C1) + len(C3) + len(C4)\n",
    "#     print(m)\n",
    "    print(\"C1:\")\n",
    "    print((C1))\n",
    "    print(\"C3:\")\n",
    "    print((C3))\n",
    "    print(\"C4:\")\n",
    "    print((C4))\n",
    "    y_A = np.zeros((m))\n",
    "    X_A = np.zeros((m, n))\n",
    "    for i in range(len(C1)):\n",
    "        X_A[i] = X[C1[i]]\n",
    "        y_A[i] = 0\n",
    "\n",
    "    for i in range(len(C3)):\n",
    "        X_A[i + len(C1)] = X[C3[i]]\n",
    "        y_A[i + len(C1)] = 1\n",
    "\n",
    "    for i in range(len(C4)):\n",
    "        X_A[i + len(C1)+ len(C3)] = X[C4[i]]\n",
    "        y_A[i + len(C1)+ len(C3)] = 0\n",
    "    print(\"X_A:\")\n",
    "    print(X_A)\n",
    "    print(\"y_A:\")\n",
    "    print(y_A)\n",
    "    # print(\"X:\"+str(X))\n",
    "    # print(\"C3:\"+str(C3))\n",
    "    print(\"W:\"+str(W))\n",
    "    Wa = find_Wa(X, C3, W)\n",
    "    print(\"Wa:\"+str(Wa))\n",
    "#         print(Wa)     \n",
    "    W_A,clf = optimize(X_A, y_A, delta, alpha)\n",
    "    y_res = y_hat(X_A, W_A, delta, y_A,0,clf)\n",
    "    Cc1, Cc2, Cc3, Cc4 = classes(y_A, y_res)\n",
    "    twin_svm_enable = int(len(Cc1)==0 or len(Cc2)==0)\n",
    "    print(\"twin_svm_enable:\"+str(twin_svm_enable))\n",
    "    if(twin_svm_enable):\n",
    "        W_A,clf = twin(X_A,y_A)\n",
    "    curr_Neuron.insert('A', Wa)\n",
    "    curr_Neuron.A.X = W_A\n",
    "    curr_Neuron.A.twin_node = twin_svm_enable\n",
    "    curr_Neuron.A.twin_classifier = clf\n",
    "    return X_A, y_A, W_A,twin_svm_enable,clf\n",
    "\n",
    "\n",
    "def addneuron_B(X, y, W, C2, C3, C4, delta, alpha, curr_Neuron):\n",
    "    n = X.shape[1]\n",
    "    print('B')\n",
    "    m = len(C2) + len(C3) + len(C4)\n",
    "    # print(m)\n",
    "    y_B = np.zeros((m))\n",
    "    X_B = np.zeros((m, n))\n",
    "#         y_B = []\n",
    "    print(\"C2:\")\n",
    "    print((C2))\n",
    "    print(\"C3:\")\n",
    "    print((C3))\n",
    "    print(\"C4:\")\n",
    "    print((C4))\n",
    "    for i in range(len(C2)):\n",
    "        X_B[i] = X[C2[i]]\n",
    "        y_B[i] = 0\n",
    "\n",
    "    for i in range(len(C3)):\n",
    "        X_B[i + len(C2)] = X[C3[i]]\n",
    "        y_B[i + len(C2)] = 0\n",
    "\n",
    "    for i in range(len(C4)):\n",
    "        X_B[i + len(C2)+ len(C3)] = X[C4[i]]\n",
    "        y_B[i + len(C2)+ len(C3)] = 1\n",
    "\n",
    "#     print(X_B)\n",
    "    print(\"X_B:\")\n",
    "    print(X_B)\n",
    "    print(\"y_B:\")\n",
    "    print(y_B)\n",
    "    # print(\"X:\"+str(X))\n",
    "    # print(\"C4:\"+str(C4))\n",
    "    print(\"W:\"+str(W))\n",
    "    Wb = find_Wb(X, C4, W)\n",
    "    print(\"Wb:\"+str(Wb))\n",
    "    W_B,clf = optimize(X_B, y_B, delta, alpha)\n",
    "    y_res = y_hat(X_B, W_B, delta, y_B,0,clf)\n",
    "    Cc1, Cc2, Cc3, Cc4 = classes(y_B, y_res)\n",
    "    twin_svm_enable = int(len(Cc1)==0 or len(Cc2)==0)\n",
    "    print(\"twin_svm_enable:\"+str(twin_svm_enable))\n",
    "    if(twin_svm_enable):\n",
    "        W_B,clf = twin(X_B,y_B)\n",
    "    curr_Neuron.insert('B', Wb)\n",
    "    curr_Neuron.B.X = W_B\n",
    "    curr_Neuron.B.twin_node = twin_svm_enable\n",
    "    curr_Neuron.B.twin_classifier = clf\n",
    "    return X_B, y_B, W_B,twin_svm_enable,clf\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def accuracy(y_pred, y):\n",
    "    count = 0\n",
    "    if(len(y_pred) == len(y)):\n",
    "        l = len(y)\n",
    "        for i in range(l):\n",
    "            if (y_pred[i] == y[i]):\n",
    "                count+=1\n",
    "        return count/l        \n",
    "    else :\n",
    "        print(\"Unequal lenghts of arrays!!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy(np.array([0,1,0,0,0,0]).reshape(6,1), np.array([1,0,0,0,0,1]))\n",
    "# np.array([0,1,0,0,0,0]).shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def classifier(X, y, W, delta, alpha, curr_Neuron,twin_en,clff):\n",
    "    print(\"y:\")\n",
    "    print(y)\n",
    "    y_res = y_hat(X, W, delta, y,twin_en,clff)\n",
    "    print(\"y_res:\")\n",
    "    print(y_res)\n",
    "    C1, C2, C3, C4 = classes(y, y_res)\n",
    "    print(\"C1len:\"+str(len(C1)))\n",
    "    print(\"C2len:\"+str(len(C2)))\n",
    "    print(\"twin_en:\"+str(twin_en))\n",
    "    if len(C3)!= 0:\n",
    "        X_A, y_A, W_A,twin_enable,clf = addneuron_A(X, y, W, C1, C3, C4, delta, alpha, curr_Neuron)\n",
    "        classifier(X_A, y_A, W_A, delta, alpha, curr_Neuron.A,twin_enable,clf)\n",
    "    if len(C4)!= 0:\n",
    "        X_B, y_B, W_B,twin_enable,clf = addneuron_B(X, y, W, C2, C3, C4, delta, alpha, curr_Neuron)\n",
    "        classifier(X_B, y_B, W_B, delta, alpha, curr_Neuron.B,twin_enable,clf)\n",
    "    else:\n",
    "        print(\"terminated\")\n",
    "        return 'Terminated'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "def classifier2(X, y, W, delta, alpha, curr_Neuron, twin_en, clff):\n",
    "    y_res = y_hat(X, W, delta, y,twin_en,clff)\n",
    "    C1, C2, C3, C4 = classes(y, y_res)\n",
    "    training_error = []\n",
    "    \n",
    "    if len(C3)!= 0:\n",
    "        X_A, y_A, W_A,twin_enable,clf = addneuron_A(X, y, W, C1, C3, C4, delta, alpha, curr_Neuron)\n",
    "        a = curr_Neuron.calculate(X)\n",
    "        training_error = np.concatenate((training_error, [accuracy(a, y)]))\n",
    "        classifier2(X_A, y_A, W_A, delta, alpha, curr_Neuron.A,twin_enable,clf)\n",
    "    \n",
    "    \n",
    "    if len(C4)!= 0:\n",
    "        X_B, y_B, W_B,twin_enable,clf = addneuron_B(X, y, W, C2, C3, C4, delta, alpha, curr_Neuron)\n",
    "        a = curr_Neuron.calculate(X)\n",
    "        training_error = np.concatenate((training_error, [accuracy(a, y)]))\n",
    "        classifier2(X_B, y_B, W_B, delta, alpha, curr_Neuron.B,twin_enable,clf)\n",
    "    \n",
    "    \n",
    "    else:\n",
    "        print(\"terminated\")\n",
    "        return training_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "wd:[-1.60714286  0.375      -0.14285714]\n",
      "A\n",
      "C1:\n",
      "[8, 9]\n",
      "C3:\n",
      "[0, 1, 2, 3, 4, 5]\n",
      "C4:\n",
      "[10, 11]\n",
      "X_A:\n",
      "[[1. 2. 6.]\n",
      " [1. 2. 3.]\n",
      " [1. 1. 8.]\n",
      " [1. 1. 1.]\n",
      " [1. 4. 5.]\n",
      " [1. 4. 4.]\n",
      " [1. 6. 5.]\n",
      " [1. 6. 4.]\n",
      " [1. 8. 6.]\n",
      " [1. 8. 3.]]\n",
      "y_A:\n",
      "[0. 0. 1. 1. 1. 1. 1. 1. 0. 0.]\n",
      "W:[[-1.60714286  0.375      -0.14285714]]\n",
      "Wa:3.3750999999999984\n",
      "wd:[ 1.28571429 -0.42857143  0.14285714]\n",
      "twin_svm_enable:0\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[ 2.        ]\n",
      " [ 1.        ]\n",
      " [ 0.28571429]\n",
      " [ 0.14285714]\n",
      " [-0.57142857]\n",
      " [-0.71428571]\n",
      " [-1.85714286]\n",
      " [-2.85714286]\n",
      " [ 1.28571429]\n",
      " [ 0.85714286]\n",
      " [-1.28571429]\n",
      " [-1.71428571]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ 1.0001    ]\n",
      " [-1.375     ]\n",
      " [-0.82142857]\n",
      " [-0.67857143]\n",
      " [-0.07142857]\n",
      " [ 0.07142857]\n",
      " [ 1.        ]\n",
      " [ 2.        ]\n",
      " [ 1.66081429]\n",
      " [-1.28571429]\n",
      " [ 0.53571429]\n",
      " [ 0.96428571]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "A\n",
      "C1:\n",
      "[8, 9]\n",
      "C3:\n",
      "[3, 4, 5, 6, 7]\n",
      "C4:\n",
      "[0, 1]\n",
      "X_A:\n",
      "[[1. 8. 6.]\n",
      " [1. 8. 3.]\n",
      " [1. 1. 1.]\n",
      " [1. 4. 5.]\n",
      " [1. 4. 4.]\n",
      " [1. 6. 5.]\n",
      " [1. 6. 4.]\n",
      " [1. 2. 6.]\n",
      " [1. 2. 3.]]\n",
      "y_A:\n",
      "[0. 0. 1. 1. 1. 1. 1. 0. 0.]\n",
      "W:[[ 1.28571429 -0.42857143  0.14285714]]\n",
      "Wa:1.7143857142857144\n",
      "wd:[ 2.73333333 -0.16666667 -0.56666667]\n",
      "twin_svm_enable:0\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[-1.        ]\n",
      " [ 0.7       ]\n",
      " [-1.96666667]\n",
      " [ 2.        ]\n",
      " [-0.76666667]\n",
      " [-0.2       ]\n",
      " [-1.1       ]\n",
      " [-0.53333333]\n",
      " [-2.        ]\n",
      " [-0.3       ]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ 1.28571429]\n",
      " [ 0.85714286]\n",
      " [ 2.        ]\n",
      " [ 2.71438571]\n",
      " [ 0.28571429]\n",
      " [ 0.14285714]\n",
      " [-0.57142857]\n",
      " [-0.71428571]\n",
      " [-1.28571429]\n",
      " [-1.71428571]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "A\n",
      "C1:\n",
      "[0, 7]\n",
      "C3:\n",
      "[3, 4, 5, 6]\n",
      "C4:\n",
      "[1, 8]\n",
      "X_A:\n",
      "[[1. 8. 6.]\n",
      " [1. 2. 6.]\n",
      " [1. 4. 5.]\n",
      " [1. 4. 4.]\n",
      " [1. 6. 5.]\n",
      " [1. 6. 4.]\n",
      " [1. 8. 3.]\n",
      " [1. 2. 3.]]\n",
      "y_A:\n",
      "[0. 0. 1. 1. 1. 1. 0. 0.]\n",
      "W:[[ 2.73333333 -0.16666667 -0.56666667]]\n",
      "Wa:2.1001000000000016\n",
      "wd:[ 0.          0.5        -0.33333333]\n",
      "twin_svm_enable:0\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[ 2.00000000e+00]\n",
      " [ 3.00000000e+00]\n",
      " [ 1.66666667e-01]\n",
      " [ 3.33333333e-01]\n",
      " [ 6.66666667e-01]\n",
      " [ 1.33333333e+00]\n",
      " [ 1.66666667e+00]\n",
      " [-1.00000000e+00]\n",
      " [-4.44089210e-16]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ 0.1001    ]\n",
      " [ 1.8001    ]\n",
      " [ 2.        ]\n",
      " [-0.76666667]\n",
      " [-0.2       ]\n",
      " [ 1.0001    ]\n",
      " [ 1.56676667]\n",
      " [-1.        ]\n",
      " [ 0.7       ]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "A\n",
      "C1:\n",
      "[1]\n",
      "C3:\n",
      "[2, 3]\n",
      "C4:\n",
      "[0, 6, 7]\n",
      "X_A:\n",
      "[[1. 2. 6.]\n",
      " [1. 4. 5.]\n",
      " [1. 4. 4.]\n",
      " [1. 8. 6.]\n",
      " [1. 8. 3.]\n",
      " [1. 2. 3.]]\n",
      "y_A:\n",
      "[0. 1. 1. 0. 0. 0.]\n",
      "W:[[ 0.          0.5        -0.33333333]]\n",
      "Wa:0.6667666666666674\n",
      "wd:[ 3.         -0.75        0.33333333]\n",
      "twin_svm_enable:0\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[-1.        ]\n",
      " [ 3.5       ]\n",
      " [ 1.66666667]\n",
      " [ 1.33333333]\n",
      " [ 0.16666667]\n",
      " [-0.16666667]\n",
      " [-2.        ]\n",
      " [ 2.5       ]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]]\n",
      "y_1:[[0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]]\n",
      "res_before_quant:[[ 2.        ]\n",
      " [-0.33323333]\n",
      " [ 1.0001    ]\n",
      " [ 1.33343333]\n",
      " [ 1.33333333]\n",
      " [ 1.66666667]\n",
      " [ 3.        ]\n",
      " [ 0.66676667]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]]\n",
      "B\n",
      "C2:\n",
      "[1, 2]\n",
      "C3:\n",
      "[]\n",
      "C4:\n",
      "[0, 3, 5]\n",
      "X_B:\n",
      "[[1. 4. 5.]\n",
      " [1. 4. 4.]\n",
      " [1. 2. 6.]\n",
      " [1. 8. 6.]\n",
      " [1. 2. 3.]]\n",
      "y_B:\n",
      "[0. 0. 1. 1. 1.]\n",
      "W:[[ 3.         -0.75        0.33333333]]\n",
      "Wb:-4.5000001\n",
      "wd:[-8.66666667 -0.16666667  1.83333333]\n",
      "twin_svm_enable:0\n",
      "B\n",
      "no child node\n",
      "res_before_quant:[[ 2.        ]\n",
      " [-0.16666667]\n",
      " [-2.        ]\n",
      " [ 1.        ]\n",
      " [-4.5       ]\n",
      " [-3.5       ]]\n",
      "delta condition not met\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[-1.0000001 ]\n",
      " [ 1.66666667]\n",
      " [ 1.33333333]\n",
      " [-5.5000001 ]\n",
      " [-2.        ]\n",
      " [ 2.5       ]]\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]]\n",
      "A\n",
      "C1:\n",
      "[1]\n",
      "C3:\n",
      "[4]\n",
      "C4:\n",
      "[0]\n",
      "X_A:\n",
      "[[1. 4. 4.]\n",
      " [1. 2. 3.]\n",
      " [1. 4. 5.]]\n",
      "y_A:\n",
      "[0. 1. 0.]\n",
      "W:[[-8.66666667 -0.16666667  1.83333333]]\n",
      "Wa:4.500099999999999\n",
      "wd:[ 4.  -2.5  1. ]\n",
      "twin_svm_enable:0\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[ -1.]\n",
      " [ -2.]\n",
      " [  5.]\n",
      " [-10.]\n",
      " [  2.]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]]\n",
      "y_1:[[0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]]\n",
      "res_before_quant:[[-0.16666667]\n",
      " [-2.        ]\n",
      " [ 6.5001    ]\n",
      " [ 1.        ]\n",
      " [ 1.0001    ]]\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]]\n",
      "terminated\n",
      "training_error\n",
      "B\n",
      "C2:\n",
      "[2, 3]\n",
      "C3:\n",
      "[4]\n",
      "C4:\n",
      "[0]\n",
      "X_B:\n",
      "[[1. 2. 6.]\n",
      " [1. 8. 6.]\n",
      " [1. 2. 3.]\n",
      " [1. 4. 5.]]\n",
      "y_B:\n",
      "[0. 0. 0. 1.]\n",
      "W:[[-8.66666667 -0.16666667  1.83333333]]\n",
      "Wb:-0.8333334333333339\n",
      "wd:[ 0.          0.         -0.16666667]\n",
      "twin_svm_enable:1\n",
      "XX:[[2. 6.]\n",
      " [8. 6.]\n",
      " [2. 3.]\n",
      " [4. 5.]]\n",
      "yy:[0. 0. 0. 1.]\n",
      "Total_Data:[[4. 5.]\n",
      " [2. 6.]\n",
      " [8. 6.]\n",
      " [2. 3.]]\n",
      "A:[[4. 5.]]\n",
      "B:[[2. 6.]\n",
      " [8. 6.]\n",
      " [2. 3.]]\n",
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -6.1110e-01 -4.4394e+00  4e+00  3e-17  3e-16\n",
      " 1: -6.8747e-01 -1.0065e+00  3e-01  1e-16  7e-16\n",
      " 2: -8.1796e-01 -8.4847e-01  3e-02  8e-17  3e-16\n",
      " 3: -8.2200e-01 -8.2365e-01  2e-03  1e-16  6e-17\n",
      " 4: -8.2210e-01 -8.2213e-01  3e-05  2e-16  1e-16\n",
      " 5: -8.2210e-01 -8.2210e-01  3e-07  5e-17  8e-17\n",
      "Optimal solution found.\n",
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -5.6814e-01 -8.8289e-01  4e+00  2e+00  2e-16\n",
      " 1: -3.4388e-01 -7.1240e-01  4e-01  2e-16  0e+00\n",
      " 2: -4.0340e-01 -4.2595e-01  2e-02  6e-17  0e+00\n",
      " 3: -4.2279e-01 -4.2302e-01  2e-04  8e-17  0e+00\n",
      " 4: -4.2298e-01 -4.2299e-01  2e-06  6e-17  2e-16\n",
      " 5: -4.2298e-01 -4.2298e-01  2e-08  8e-17  1e-16\n",
      "Optimal solution found.\n",
      "w1:[[ 8.32382312e-07]\n",
      " [-9.67654418e-02]]\n",
      "b1:[-0.41941415]\n",
      "w2:[[ 0.00044472]\n",
      " [-0.02245831]]\n",
      "b2:[-0.03090797]\n",
      "self:<twinsvm.twinsvmclassifier object at 0x00000221536C8448>\n",
      "Total_Data:[[4. 5.]\n",
      " [2. 6.]\n",
      " [8. 6.]\n",
      " [2. 3.]]\n",
      "A:[[4. 5.]]\n",
      "B:[[2. 6.]\n",
      " [8. 6.]\n",
      " [2. 3.]]\n",
      "     pcost       dcost       gap    pres   dres\n",
      " 0:  2.8576e-01 -1.4788e+01  2e+01  4e-17  1e-15\n",
      " 1: -6.1172e-01 -1.5723e+00  1e+00  2e-16  7e-16\n",
      " 2: -8.5796e-01 -9.9055e-01  1e-01  1e-16  4e-16\n",
      " 3: -9.1944e-01 -9.2558e-01  6e-03  1e-16  2e-16\n",
      " 4: -9.1999e-01 -9.2006e-01  7e-05  2e-16  1e-16\n",
      " 5: -9.2000e-01 -9.2000e-01  7e-07  2e-16  2e-16\n",
      "Optimal solution found.\n",
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -1.2513e+00 -4.7606e+00  4e+00  0e+00  1e-16\n",
      " 1: -1.4032e+00 -1.7412e+00  3e-01  0e+00  4e-16\n",
      " 2: -1.5625e+00 -1.6153e+00  5e-02  2e-16  1e-16\n",
      " 3: -1.5917e+00 -1.5969e+00  5e-03  0e+00  3e-17\n",
      " 4: -1.5962e+00 -1.5963e+00  1e-04  2e-16  4e-17\n",
      " 5: -1.5963e+00 -1.5963e+00  1e-06  2e-16  1e-16\n",
      "Optimal solution found.\n",
      "w1:[[ 0.07999922]\n",
      " [-0.15999956]]\n",
      "b1:[-0.67999975]\n",
      "w2:[[ 0.00286238]\n",
      " [-0.14455022]]\n",
      "b2:[-0.19893546]\n",
      "self:<twinsvm.twinsvmclassifier object at 0x00000221537245C8>\n",
      "Total_Data:[[4. 5.]\n",
      " [2. 6.]\n",
      " [8. 6.]\n",
      " [2. 3.]]\n",
      "A:[[4. 5.]]\n",
      "B:[[2. 6.]\n",
      " [8. 6.]\n",
      " [2. 3.]]\n",
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -4.9816e-01 -5.0514e-01  1e+01  3e+00  1e-16\n",
      " 1: -2.6755e-01 -3.9012e-01  6e-01  1e-01  2e-16\n",
      " 2: -1.7639e-01 -2.6128e-01  8e-02  6e-17  7e-16\n",
      " 3: -1.8564e-01 -1.8786e-01  2e-03  2e-17  2e-16\n",
      " 4: -1.8686e-01 -1.8688e-01  2e-05  2e-17  2e-16\n",
      " 5: -1.8687e-01 -1.8687e-01  2e-07  3e-17  1e-16\n",
      " 6: -1.8687e-01 -1.8687e-01  2e-09  4e-17  2e-16\n",
      "Optimal solution found.\n",
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -8.5818e-01 -2.4444e+00  2e+00  0e+00  2e-16\n",
      " 1: -9.7153e-01 -1.0882e+00  1e-01  5e-16  3e-16\n",
      " 2: -1.0692e+00 -1.0708e+00  2e-03  0e+00  0e+00\n",
      " 3: -1.0705e+00 -1.0705e+00  2e-05  2e-16  2e-16\n",
      " 4: -1.0705e+00 -1.0705e+00  2e-07  2e-16  3e-16\n",
      "Optimal solution found.\n",
      "w1:[[-0.01934894]\n",
      " [-0.02418618]]\n",
      "b1:[-0.00483724]\n",
      "w2:[[ 0.00132512]\n",
      " [-0.0669186 ]]\n",
      "b2:[-0.09209589]\n",
      "self:<twinsvm.twinsvmclassifier object at 0x0000022153E937C8>\n",
      "Total_Data:[[4. 5.]\n",
      " [2. 6.]\n",
      " [8. 6.]\n",
      " [2. 3.]]\n",
      "A:[[4. 5.]]\n",
      "B:[[2. 6.]\n",
      " [8. 6.]\n",
      " [2. 3.]]\n",
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -5.4791e-01 -1.3988e+00  9e+00  3e+00  3e-16\n",
      " 1: -4.2771e-01 -1.1730e+00  7e-01  5e-16  1e-16\n",
      " 2: -4.6177e-01 -5.1131e-01  5e-02  7e-17  3e-16\n",
      " 3: -4.8770e-01 -4.9153e-01  4e-03  1e-16  1e-16\n",
      " 4: -4.8971e-01 -4.8994e-01  2e-04  1e-16  2e-16\n",
      " 5: -4.8974e-01 -4.8974e-01  3e-06  1e-16  6e-16\n",
      " 6: -4.8974e-01 -4.8974e-01  3e-08  6e-17  2e-16\n",
      "Optimal solution found.\n",
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -4.1195e-01 -7.1619e-02  4e+00  2e+00  0e+00\n",
      " 1: -2.5898e-02 -4.3185e-02  6e-02  2e-02  0e+00\n",
      " 2: -2.0077e-02 -2.7419e-02  7e-03  7e-18  6e-16\n",
      " 3: -2.1951e-02 -2.2109e-02  2e-04  3e-18  3e-16\n",
      " 4: -2.2054e-02 -2.2056e-02  2e-06  3e-18  2e-16\n",
      " 5: -2.2055e-02 -2.2055e-02  2e-08  3e-18  1e-16\n",
      "Optimal solution found.\n",
      "w1:[[-0.03144101]\n",
      " [-0.1200697 ]]\n",
      "b1:[-0.02805236]\n",
      "w2:[[ 2.16233906e-05]\n",
      " [-1.09198122e-03]]\n",
      "b2:[-0.00150283]\n",
      "self:<twinsvm.twinsvmclassifier object at 0x0000022153E93EC8>\n",
      "#######\n",
      "[1]\n",
      "[3]\n",
      "[]\n",
      "[0, 2]\n",
      "#######\n",
      "wb:(array([[-0.03144101],\n",
      "       [-0.1200697 ]]), array([-0.02805236]), array([[ 2.16233906e-05],\n",
      "       [-1.09198122e-03]]), array([-0.00150283]))\n",
      "Total_Data:[[4. 5.]\n",
      " [2. 6.]\n",
      " [8. 6.]\n",
      " [2. 3.]]\n",
      "A:[[4. 5.]]\n",
      "B:[[2. 6.]\n",
      " [8. 6.]\n",
      " [2. 3.]]\n",
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -5.4791e-01 -1.3988e+00  9e+00  3e+00  3e-16\n",
      " 1: -4.2771e-01 -1.1730e+00  7e-01  5e-16  1e-16\n",
      " 2: -4.6177e-01 -5.1131e-01  5e-02  7e-17  3e-16\n",
      " 3: -4.8770e-01 -4.9153e-01  4e-03  1e-16  1e-16\n",
      " 4: -4.8971e-01 -4.8994e-01  2e-04  1e-16  2e-16\n",
      " 5: -4.8974e-01 -4.8974e-01  3e-06  1e-16  6e-16\n",
      " 6: -4.8974e-01 -4.8974e-01  3e-08  6e-17  2e-16\n",
      "Optimal solution found.\n",
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -4.1195e-01 -7.1619e-02  4e+00  2e+00  0e+00\n",
      " 1: -2.5898e-02 -4.3185e-02  6e-02  2e-02  0e+00\n",
      " 2: -2.0077e-02 -2.7419e-02  7e-03  7e-18  6e-16\n",
      " 3: -2.1951e-02 -2.2109e-02  2e-04  3e-18  3e-16\n",
      " 4: -2.2054e-02 -2.2056e-02  2e-06  3e-18  2e-16\n",
      " 5: -2.2055e-02 -2.2055e-02  2e-08  3e-18  1e-16\n",
      "Optimal solution found.\n",
      "w1:[[-0.03144101]\n",
      " [-0.1200697 ]]\n",
      "b1:[-0.02805236]\n",
      "w2:[[ 2.16233906e-05]\n",
      " [-1.09198122e-03]]\n",
      "b2:[-0.00150283]\n",
      "self:<twinsvm.twinsvmclassifier object at 0x0000022153E93EC8>\n",
      "wdtwin:[[-2.80523630e-02 -3.14410150e-02 -1.20069698e-01 -1.50282564e-03\n",
      "   2.16233906e-05 -1.09198122e-03]]\n",
      "AB\n",
      "no child node\n",
      "res_before_quant:[[ -1.]\n",
      " [ -2.]\n",
      " [  5.]\n",
      " [-10.]\n",
      " [  2.]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]]\n",
      "no child node\n",
      "twin_svm_res\n"
     ]
    },
    {
     "ename": "UnboundLocalError",
     "evalue": "local variable 'XAB' referenced before assignment",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mUnboundLocalError\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-22-94da7c930936>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m     20\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mX\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mW\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     21\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 22\u001b[1;33m \u001b[0mclassifier2\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mW\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdelta\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mclf1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m<ipython-input-21-cc500957a361>\u001b[0m in \u001b[0;36mclassifier2\u001b[1;34m(X, y, W, delta, alpha, curr_Neuron, twin_en, clff)\u001b[0m\n\u001b[0;32m      8\u001b[0m         \u001b[0ma\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcalculate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      9\u001b[0m         \u001b[0mtraining_error\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconcatenate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtraining_error\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0maccuracy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0ma\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 10\u001b[1;33m         \u001b[0mclassifier2\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX_A\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my_A\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mW_A\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdelta\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mA\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mtwin_enable\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mclf\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     11\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     12\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-21-cc500957a361>\u001b[0m in \u001b[0;36mclassifier2\u001b[1;34m(X, y, W, delta, alpha, curr_Neuron, twin_en, clff)\u001b[0m\n\u001b[0;32m      8\u001b[0m         \u001b[0ma\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcalculate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      9\u001b[0m         \u001b[0mtraining_error\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconcatenate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtraining_error\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0maccuracy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0ma\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 10\u001b[1;33m         \u001b[0mclassifier2\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX_A\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my_A\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mW_A\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdelta\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mA\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mtwin_enable\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mclf\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     11\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     12\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-21-cc500957a361>\u001b[0m in \u001b[0;36mclassifier2\u001b[1;34m(X, y, W, delta, alpha, curr_Neuron, twin_en, clff)\u001b[0m\n\u001b[0;32m      8\u001b[0m         \u001b[0ma\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcalculate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      9\u001b[0m         \u001b[0mtraining_error\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconcatenate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtraining_error\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0maccuracy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0ma\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 10\u001b[1;33m         \u001b[0mclassifier2\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX_A\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my_A\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mW_A\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdelta\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mA\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mtwin_enable\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mclf\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     11\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     12\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-21-cc500957a361>\u001b[0m in \u001b[0;36mclassifier2\u001b[1;34m(X, y, W, delta, alpha, curr_Neuron, twin_en, clff)\u001b[0m\n\u001b[0;32m      8\u001b[0m         \u001b[0ma\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcalculate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      9\u001b[0m         \u001b[0mtraining_error\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconcatenate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtraining_error\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0maccuracy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0ma\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 10\u001b[1;33m         \u001b[0mclassifier2\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX_A\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my_A\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mW_A\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdelta\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mA\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mtwin_enable\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mclf\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     11\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     12\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-21-cc500957a361>\u001b[0m in \u001b[0;36mclassifier2\u001b[1;34m(X, y, W, delta, alpha, curr_Neuron, twin_en, clff)\u001b[0m\n\u001b[0;32m     15\u001b[0m         \u001b[0ma\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcalculate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     16\u001b[0m         \u001b[0mtraining_error\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconcatenate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtraining_error\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0maccuracy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0ma\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 17\u001b[1;33m         \u001b[0mclassifier2\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX_B\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my_B\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mW_B\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdelta\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mB\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mtwin_enable\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mclf\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     18\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     19\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-21-cc500957a361>\u001b[0m in \u001b[0;36mclassifier2\u001b[1;34m(X, y, W, delta, alpha, curr_Neuron, twin_en, clff)\u001b[0m\n\u001b[0;32m     13\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mC4\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m!=\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     14\u001b[0m         \u001b[0mX_B\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my_B\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mW_B\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mtwin_enable\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mclf\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0maddneuron_B\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mW\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mC2\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mC3\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mC4\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdelta\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 15\u001b[1;33m         \u001b[0ma\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcalculate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     16\u001b[0m         \u001b[0mtraining_error\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconcatenate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtraining_error\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0maccuracy\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0ma\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     17\u001b[0m         \u001b[0mclassifier2\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX_B\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my_B\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mW_B\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdelta\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcurr_Neuron\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mB\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mtwin_enable\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mclf\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-7-859861a6fe15>\u001b[0m in \u001b[0;36mcalculate\u001b[1;34m(self, XX)\u001b[0m\n\u001b[0;32m     24\u001b[0m             \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"AB\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     25\u001b[0m             \u001b[0my_1\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mA\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcalculate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mXX\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 26\u001b[1;33m             \u001b[0my_2\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mB\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcalculate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mXX\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     27\u001b[0m             \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"y_1:\"\u001b[0m\u001b[1;33m+\u001b[0m\u001b[0mstr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     28\u001b[0m             \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"y_2:\"\u001b[0m\u001b[1;33m+\u001b[0m\u001b[0mstr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_2\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-7-859861a6fe15>\u001b[0m in \u001b[0;36mcalculate\u001b[1;34m(self, XX)\u001b[0m\n\u001b[0;32m     43\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     44\u001b[0m             \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"no child node\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 45\u001b[1;33m             \u001b[0mres\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mquantize\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mXX\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mX\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtwin_node\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtwin_classifier\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mXX\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     46\u001b[0m             \u001b[1;32mreturn\u001b[0m \u001b[0mres\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m<ipython-input-6-24d83039218b>\u001b[0m in \u001b[0;36mquantize\u001b[1;34m(a, b, twin_node, twin_classifier, XX)\u001b[0m\n\u001b[0;32m     28\u001b[0m         \u001b[0mdistFromPlane1\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0my1\u001b[0m\u001b[1;33m/\u001b[0m\u001b[0mw1mod\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     29\u001b[0m         \u001b[0mdistFromPlane2\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0my2\u001b[0m\u001b[1;33m/\u001b[0m\u001b[0mw2mod\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 30\u001b[1;33m         \u001b[0mres\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mdistFromPlane1\u001b[0m \u001b[1;33m-\u001b[0m \u001b[0mdistFromPlane2\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m+\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdot\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mXAB\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mwAB\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     31\u001b[0m         \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m\"res_twin_bef:\"\u001b[0m\u001b[1;33m+\u001b[0m\u001b[0mstr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mres\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     32\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0mj\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mres\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mUnboundLocalError\u001b[0m: local variable 'XAB' referenced before assignment"
     ]
    }
   ],
   "source": [
    "X = np.array([[1,1,8],[1,1,1],[1,4,5],[1,4,4],[1,6,5],[1,6,4],[1,10,8],[1,10,1],[1,2,6],[1,2,3],[1,8,6],[1,8,3]])\n",
    "y=np.array([1,1,1,1,1,1,1,1,0,0,0,0])\n",
    "\n",
    "# annots = loadmat('Archive/1.mat')\n",
    "# X = annots['x']\n",
    "# y = annots['y']\n",
    "\n",
    "# X = np.concatenate((np.ones((X.shape[0],1)), X), axis = 1)   # Adding bias to the input\n",
    "# y = y.reshape(y.shape[0])   # Reshaping the output array\n",
    "# for i in range(len(y)):   # Converting the output into a (0,1) output rather than (-1,1)\n",
    "#     if(y[i]==-1):\n",
    "#         y[i]=0\n",
    "\n",
    "\n",
    "        \n",
    "alpha = 0\n",
    "delta = 1\n",
    "curr_Neuron = Neuron(3,0,None)\n",
    "W, clf1 = optimize(X, y, delta, alpha)\n",
    "curr_Neuron.X = W\n",
    "\n",
    "a = classifier2(X, y, W, delta, alpha, curr_Neuron,0,clf1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "###Verification###\n",
      "AB\n",
      "AB\n",
      "AB\n",
      "AB\n",
      "B\n",
      "AB\n",
      "no child node\n",
      "res_before_quant:[[  9.5]\n",
      " [  2.5]\n",
      " [ -1. ]\n",
      " [ -2. ]\n",
      " [ -6. ]\n",
      " [ -7. ]\n",
      " [-13. ]\n",
      " [-20. ]\n",
      " [  5. ]\n",
      " [  2. ]\n",
      " [-10. ]\n",
      " [-13. ]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "B\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[ 40.]\n",
      " [ -2.]\n",
      " [ -2.]\n",
      " [ -8.]\n",
      " [-18.]\n",
      " [-24.]\n",
      " [-32.]\n",
      " [-74.]\n",
      " [ 20.]\n",
      " [  2.]\n",
      " [-28.]\n",
      " [-46.]]\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[  4.5001    ]\n",
      " [  2.16666667]\n",
      " [ -2.        ]\n",
      " [ -2.33333333]\n",
      " [ -5.66666667]\n",
      " [ -6.        ]\n",
      " [-12.        ]\n",
      " [-14.33333333]\n",
      " [  2.0001    ]\n",
      " [  1.0001    ]\n",
      " [ -9.        ]\n",
      " [-10.        ]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "twin_svm_res\n",
      "res_twin_bef:[[ 1.88168454e-02]\n",
      " [ 2.01888564e-02]\n",
      " [ 2.45141588e-02]\n",
      " [ 2.47101604e-02]\n",
      " [-1.50815476e-02]\n",
      " [-1.48855460e-02]\n",
      " [-9.48609652e-02]\n",
      " [-9.34889542e-02]\n",
      " [-5.89004710e-04]\n",
      " [-1.00000000e-06]\n",
      " [-5.48732556e-02]\n",
      " [-5.42852509e-02]]\n",
      "res_twin_aft:[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]]\n",
      "res_before_quant:[[ 9.5000999 ]\n",
      " [-3.33323343]\n",
      " [-1.0000001 ]\n",
      " [-2.83333343]\n",
      " [-0.5       ]\n",
      " [-2.33333333]\n",
      " [ 4.33333333]\n",
      " [-8.5       ]\n",
      " [ 6.5001    ]\n",
      " [ 1.0001    ]\n",
      " [ 1.        ]\n",
      " [-4.5       ]]\n",
      "delta condition not met\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]]\n",
      "res_before_quant:[[ 0.41666657]\n",
      " [ 2.58333333]\n",
      " [ 1.66666667]\n",
      " [ 1.33333333]\n",
      " [ 0.16666667]\n",
      " [-0.16666667]\n",
      " [-6.33333343]\n",
      " [-4.16666667]\n",
      " [-1.0000001 ]\n",
      " [-2.0000001 ]\n",
      " [-5.5000001 ]\n",
      " [-2.        ]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "AB\n",
      "AB\n",
      "B\n",
      "no child node\n",
      "res_before_quant:[[-20.5]\n",
      " [  0.5]\n",
      " [ -1. ]\n",
      " [  2. ]\n",
      " [  6. ]\n",
      " [  9. ]\n",
      " [ 11. ]\n",
      " [ 32. ]\n",
      " [-11. ]\n",
      " [ -2. ]\n",
      " [ 10. ]\n",
      " [ 19. ]]\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[  3.5      ]\n",
      " [  3.5      ]\n",
      " [ -1.       ]\n",
      " [ -1.0000001]\n",
      " [ -4.0000001]\n",
      " [ -4.0000001]\n",
      " [-10.0000001]\n",
      " [-10.0000001]\n",
      " [  2.       ]\n",
      " [  2.       ]\n",
      " [ -7.0000001]\n",
      " [ -7.0000001]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "A\n",
      "B\n",
      "no child node\n",
      "res_before_quant:[[-50.5]\n",
      " [ -1.5]\n",
      " [-13. ]\n",
      " [ -6. ]\n",
      " [ -2. ]\n",
      " [  5. ]\n",
      " [ -1. ]\n",
      " [ 48. ]\n",
      " [-31. ]\n",
      " [-10. ]\n",
      " [  2. ]\n",
      " [ 23. ]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[-3.       ]\n",
      " [-3.       ]\n",
      " [ 0.       ]\n",
      " [ 0.       ]\n",
      " [ 2.       ]\n",
      " [-3.0000001]\n",
      " [ 6.       ]\n",
      " [ 0.9999999]\n",
      " [-2.       ]\n",
      " [-2.       ]\n",
      " [-1.0000001]\n",
      " [-1.0000001]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[-20.5       ]\n",
      " [  2.83333333]\n",
      " [ -5.        ]\n",
      " [ -1.66666667]\n",
      " [  1.0001    ]\n",
      " [  2.        ]\n",
      " [ -1.66656667]\n",
      " [ 19.33333333]\n",
      " [-12.        ]\n",
      " [ -2.        ]\n",
      " [ -1.        ]\n",
      " [  9.        ]]\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]]\n",
      "twin_svm_res\n",
      "res_twin_bef:[[-8.37078842e-01]\n",
      " [-8.66663194e-01]\n",
      " [-2.65562561e-01]\n",
      " [-1.79374200e-01]\n",
      " [-8.61893608e-02]\n",
      " [-1.00000000e-06]\n",
      " [ 1.91270059e+00]\n",
      " [ 1.88311624e+00]\n",
      " [-2.58564082e-01]\n",
      " [ 1.00000000e-06]\n",
      " [ 1.27280123e+00]\n",
      " [ 8.98463437e-01]]\n",
      "res_twin_aft:[[0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "B\n",
      "no child node\n",
      "res_before_quant:[[ 20.5       ]\n",
      " [ -2.83333333]\n",
      " [  5.        ]\n",
      " [  1.66666667]\n",
      " [  1.33333333]\n",
      " [ -2.        ]\n",
      " [  4.        ]\n",
      " [-19.33333333]\n",
      " [ 12.        ]\n",
      " [  2.        ]\n",
      " [  1.        ]\n",
      " [ -9.        ]]\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]]\n",
      "twin_svm_res\n",
      "res_twin_bef:[[ 4.13721839e+00]\n",
      " [ 1.74458673e+00]\n",
      " [-3.86034544e-03]\n",
      " [-6.93075298e-01]\n",
      " [-1.38614960e+00]\n",
      " [ 3.56508463e-01]\n",
      " [-2.08308324e+00]\n",
      " [-4.47571489e+00]\n",
      " [ 2.06764386e+00]\n",
      " [-1.00000000e-06]\n",
      " [-2.07922389e+00]\n",
      " [-1.71499574e+00]]\n",
      "res_twin_aft:[[1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]]\n",
      "y_1:[[0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "y_2:[[1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]]\n",
      "res_before_quant:[[-7.5000001e+00]\n",
      " [-5.0000010e-01]\n",
      " [-2.0000000e+00]\n",
      " [-1.0000000e+00]\n",
      " [-1.0000000e+00]\n",
      " [-1.0000001e+00]\n",
      " [ 1.0000000e-04]\n",
      " [ 7.0001000e+00]\n",
      " [-5.0000001e+00]\n",
      " [ 1.0001000e+00]\n",
      " [ 1.0001000e+00]\n",
      " [ 4.0001000e+00]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]]\n",
      "y_1:[[0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[-2.16666667]\n",
      " [ 0.83343333]\n",
      " [ 1.0001    ]\n",
      " [ 1.33343333]\n",
      " [ 1.33333333]\n",
      " [ 1.66666667]\n",
      " [ 2.33333333]\n",
      " [ 0.66666657]\n",
      " [-1.        ]\n",
      " [-4.0000001 ]\n",
      " [-2.0000001 ]\n",
      " [-1.0000001 ]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "AB\n",
      "no child node\n",
      "res_before_quant:[[ 33.5]\n",
      " [ -1.5]\n",
      " [ -1. ]\n",
      " [ -6. ]\n",
      " [-14. ]\n",
      " [-19. ]\n",
      " [-25. ]\n",
      " [-60. ]\n",
      " [ 17. ]\n",
      " [  2. ]\n",
      " [-22. ]\n",
      " [-37. ]]\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "AB\n",
      "no child node\n",
      "res_before_quant:[[17.5       ]\n",
      " [-8.16666667]\n",
      " [ 6.        ]\n",
      " [ 2.33333333]\n",
      " [ 5.66666667]\n",
      " [ 2.        ]\n",
      " [16.        ]\n",
      " [-9.66666667]\n",
      " [10.        ]\n",
      " [-1.        ]\n",
      " [ 9.        ]\n",
      " [-2.        ]]\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]]\n",
      "B\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[-12.]\n",
      " [  2.]\n",
      " [ -6.]\n",
      " [ -4.]\n",
      " [ -6.]\n",
      " [ -4.]\n",
      " [-12.]\n",
      " [  2.]\n",
      " [ -8.]\n",
      " [ -2.]\n",
      " [ -8.]\n",
      " [ -2.]]\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[-22.33333333]\n",
      " [  1.0001    ]\n",
      " [ -1.33333333]\n",
      " [  2.        ]\n",
      " [  6.        ]\n",
      " [  9.33333333]\n",
      " [ 10.66666667]\n",
      " [ 34.0001    ]\n",
      " [-12.        ]\n",
      " [ -2.        ]\n",
      " [ 10.        ]\n",
      " [ 20.        ]]\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "y_2:[[0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[ 20.       ]\n",
      " [ -1.0000001]\n",
      " [  2.       ]\n",
      " [ -1.0000001]\n",
      " [ -4.0000001]\n",
      " [ -7.0000001]\n",
      " [ -7.0000001]\n",
      " [-28.0000001]\n",
      " [ 11.       ]\n",
      " [  2.       ]\n",
      " [ -7.0000001]\n",
      " [-16.0000001]]\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ 5.53343323]\n",
      " [ 1.        ]\n",
      " [ 0.13343323]\n",
      " [ 2.46676667]\n",
      " [ 2.06676667]\n",
      " [ 1.0001    ]\n",
      " [ 2.33343333]\n",
      " [-5.6       ]\n",
      " [ 2.66676657]\n",
      " [-1.0000001 ]\n",
      " [ 1.66676667]\n",
      " [-2.        ]]\n",
      "delta condition not met\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]]\n",
      "res_before_quant:[[-11.4999001 ]\n",
      " [ -1.0000001 ]\n",
      " [ -1.66666667]\n",
      " [ -5.0000001 ]\n",
      " [ -6.16666677]\n",
      " [ -4.66666677]\n",
      " [-10.0000001 ]\n",
      " [  5.33333333]\n",
      " [ -8.33323343]\n",
      " [  1.0001    ]\n",
      " [ -7.33333343]\n",
      " [  2.        ]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]]\n",
      "y_1:[[0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]]\n",
      "res_before_quant:[[-1.96666667]\n",
      " [ 2.        ]\n",
      " [ 1.33343333]\n",
      " [ 1.9001    ]\n",
      " [ 1.0001    ]\n",
      " [ 1.56676667]\n",
      " [-1.36656667]\n",
      " [-1.2000001 ]\n",
      " [-1.        ]\n",
      " [-1.0000001 ]\n",
      " [-2.        ]\n",
      " [-2.0000001 ]]\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "B\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[  2.        ]\n",
      " [  1.        ]\n",
      " [ -7.42857143]\n",
      " [ -7.57142857]\n",
      " [-13.42857143]\n",
      " [-13.57142857]\n",
      " [-25.        ]\n",
      " [-26.        ]\n",
      " [ -1.28571429]\n",
      " [ -1.71428571]\n",
      " [-19.28571429]\n",
      " [-19.71428571]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ 1.0001    ]\n",
      " [ 3.33343333]\n",
      " [ 1.33333333]\n",
      " [ 1.66666667]\n",
      " [ 4.33333333]\n",
      " [ 4.66666667]\n",
      " [ 9.33333333]\n",
      " [11.66666667]\n",
      " [-2.        ]\n",
      " [-1.        ]\n",
      " [ 7.        ]\n",
      " [ 8.        ]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "y_2:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[ -1.0000001]\n",
      " [ -1.0000001]\n",
      " [ -4.7500001]\n",
      " [ -4.7500001]\n",
      " [ -7.2500001]\n",
      " [ -7.2500001]\n",
      " [-12.2500001]\n",
      " [-12.2500001]\n",
      " [  1.5      ]\n",
      " [  1.5      ]\n",
      " [ -9.7500001]\n",
      " [ -9.7500001]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ 2.        ]\n",
      " [ 2.71438571]\n",
      " [ 2.0001    ]\n",
      " [ 1.85724286]\n",
      " [ 1.14295714]\n",
      " [ 1.0001    ]\n",
      " [-1.85714286]\n",
      " [-2.85714286]\n",
      " [-1.0000001 ]\n",
      " [-1.42857153]\n",
      " [-1.28571429]\n",
      " [-1.71428571]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "AB\n",
      "B\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[ 50.5]\n",
      " [  1.5]\n",
      " [ 13. ]\n",
      " [  6. ]\n",
      " [  2. ]\n",
      " [ -5. ]\n",
      " [  1. ]\n",
      " [-48. ]\n",
      " [ 31. ]\n",
      " [ 10. ]\n",
      " [ -2. ]\n",
      " [-23. ]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ 8.5484871 ]\n",
      " [15.09687419]\n",
      " [ 8.83880968]\n",
      " [ 9.77429355]\n",
      " [ 7.16139032]\n",
      " [ 1.5483871 ]\n",
      " [ 1.0001    ]\n",
      " [ 1.        ]\n",
      " [ 9.58074516]\n",
      " [12.38719677]\n",
      " [-2.        ]\n",
      " [ 0.80645161]]\n",
      "delta condition not met\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ -7.09677429]\n",
      " [-13.19354849]\n",
      " [ -7.67741945]\n",
      " [ -8.5483872 ]\n",
      " [ -6.32258075]\n",
      " [ -7.19354849]\n",
      " [ -1.0000001 ]\n",
      " [ -7.09677429]\n",
      " [ -8.16129042]\n",
      " [-10.77419365]\n",
      " [  2.        ]\n",
      " [ -0.61290323]]\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]]\n",
      "A\n",
      "B\n",
      "B\n",
      "no child node\n",
      "res_before_quant:[[-51.]\n",
      " [ -2.]\n",
      " [-15.]\n",
      " [ -8.]\n",
      " [ -5.]\n",
      " [  2.]\n",
      " [ -6.]\n",
      " [ 43.]\n",
      " [-32.]\n",
      " [-11.]\n",
      " [ -2.]\n",
      " [ 19.]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]]\n",
      "res_before_quant:[[-50.5      ]\n",
      " [ -1.5      ]\n",
      " [-13.       ]\n",
      " [ -6.       ]\n",
      " [ -2.       ]\n",
      " [ -1.0000001]\n",
      " [ -1.       ]\n",
      " [ 41.9999999]\n",
      " [-31.       ]\n",
      " [-10.       ]\n",
      " [  2.       ]\n",
      " [ 16.9999999]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[10.       ]\n",
      " [ 3.       ]\n",
      " [ 4.       ]\n",
      " [ 3.       ]\n",
      " [ 2.       ]\n",
      " [ 1.       ]\n",
      " [ 1.       ]\n",
      " [-8.0000001]\n",
      " [ 7.       ]\n",
      " [ 4.       ]\n",
      " [-1.0000001]\n",
      " [-4.0000001]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "twin_svm_res\n",
      "res_twin_bef:[[-3.03532543e-01]\n",
      " [ 1.49907842e+00]\n",
      " [ 5.70192859e-01]\n",
      " [ 8.27708711e-01]\n",
      " [ 6.37644758e-01]\n",
      " [ 8.95160609e-01]\n",
      " [ 1.00000000e-06]\n",
      " [ 3.93470585e-01]\n",
      " [ 2.45225109e-01]\n",
      " [ 1.01777266e+00]\n",
      " [-9.61560570e-01]\n",
      " [-1.89013016e-01]]\n",
      "res_twin_aft:[[0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]]\n",
      "y_1:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]]\n",
      "y_2:[[0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]]\n",
      "res_before_quant:[[-4.33333333]\n",
      " [-7.0000001 ]\n",
      " [-6.33333343]\n",
      " [-6.0000001 ]\n",
      " [-5.0000001 ]\n",
      " [-4.66666677]\n",
      " [-3.33333343]\n",
      " [-1.0000001 ]\n",
      " [-8.0000001 ]\n",
      " [-7.0000001 ]\n",
      " [ 1.0001    ]\n",
      " [ 2.        ]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[ 1.0001    ]\n",
      " [ 2.0001    ]\n",
      " [ 2.55367143]\n",
      " [ 2.69652857]\n",
      " [ 3.30367143]\n",
      " [ 3.44652857]\n",
      " [ 1.        ]\n",
      " [ 2.        ]\n",
      " [-1.71428571]\n",
      " [-1.28571429]\n",
      " [-1.42857153]\n",
      " [-1.0000001 ]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "AB\n",
      "AB\n",
      "AB\n",
      "AB\n",
      "B\n",
      "AB\n",
      "no child node\n",
      "res_before_quant:[[  9.5]\n",
      " [  2.5]\n",
      " [ -1. ]\n",
      " [ -2. ]\n",
      " [ -6. ]\n",
      " [ -7. ]\n",
      " [-13. ]\n",
      " [-20. ]\n",
      " [  5. ]\n",
      " [  2. ]\n",
      " [-10. ]\n",
      " [-13. ]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "B\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[ 40.]\n",
      " [ -2.]\n",
      " [ -2.]\n",
      " [ -8.]\n",
      " [-18.]\n",
      " [-24.]\n",
      " [-32.]\n",
      " [-74.]\n",
      " [ 20.]\n",
      " [  2.]\n",
      " [-28.]\n",
      " [-46.]]\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[  4.5001    ]\n",
      " [  2.16666667]\n",
      " [ -2.        ]\n",
      " [ -2.33333333]\n",
      " [ -5.66666667]\n",
      " [ -6.        ]\n",
      " [-12.        ]\n",
      " [-14.33333333]\n",
      " [  2.0001    ]\n",
      " [  1.0001    ]\n",
      " [ -9.        ]\n",
      " [-10.        ]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "twin_svm_res\n",
      "res_twin_bef:[[ 1.88168454e-02]\n",
      " [ 2.01888564e-02]\n",
      " [ 2.45141588e-02]\n",
      " [ 2.47101604e-02]\n",
      " [-1.50815476e-02]\n",
      " [-1.48855460e-02]\n",
      " [-9.48609652e-02]\n",
      " [-9.34889542e-02]\n",
      " [-5.89004710e-04]\n",
      " [-1.00000000e-06]\n",
      " [-5.48732556e-02]\n",
      " [-5.42852509e-02]]\n",
      "res_twin_aft:[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]]\n",
      "res_before_quant:[[ 9.5000999 ]\n",
      " [-3.33323343]\n",
      " [-1.0000001 ]\n",
      " [-2.83333343]\n",
      " [-0.5       ]\n",
      " [-2.33333333]\n",
      " [ 4.33333333]\n",
      " [-8.5       ]\n",
      " [ 6.5001    ]\n",
      " [ 1.0001    ]\n",
      " [ 1.        ]\n",
      " [-4.5       ]]\n",
      "delta condition not met\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]]\n",
      "res_before_quant:[[ 0.41666657]\n",
      " [ 2.58333333]\n",
      " [ 1.66666667]\n",
      " [ 1.33333333]\n",
      " [ 0.16666667]\n",
      " [-0.16666667]\n",
      " [-6.33333343]\n",
      " [-4.16666667]\n",
      " [-1.0000001 ]\n",
      " [-2.0000001 ]\n",
      " [-5.5000001 ]\n",
      " [-2.        ]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "AB\n",
      "AB\n",
      "B\n",
      "no child node\n",
      "res_before_quant:[[-20.5]\n",
      " [  0.5]\n",
      " [ -1. ]\n",
      " [  2. ]\n",
      " [  6. ]\n",
      " [  9. ]\n",
      " [ 11. ]\n",
      " [ 32. ]\n",
      " [-11. ]\n",
      " [ -2. ]\n",
      " [ 10. ]\n",
      " [ 19. ]]\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[  3.5      ]\n",
      " [  3.5      ]\n",
      " [ -1.       ]\n",
      " [ -1.0000001]\n",
      " [ -4.0000001]\n",
      " [ -4.0000001]\n",
      " [-10.0000001]\n",
      " [-10.0000001]\n",
      " [  2.       ]\n",
      " [  2.       ]\n",
      " [ -7.0000001]\n",
      " [ -7.0000001]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "A\n",
      "B\n",
      "no child node\n",
      "res_before_quant:[[-50.5]\n",
      " [ -1.5]\n",
      " [-13. ]\n",
      " [ -6. ]\n",
      " [ -2. ]\n",
      " [  5. ]\n",
      " [ -1. ]\n",
      " [ 48. ]\n",
      " [-31. ]\n",
      " [-10. ]\n",
      " [  2. ]\n",
      " [ 23. ]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[-3.       ]\n",
      " [-3.       ]\n",
      " [ 0.       ]\n",
      " [ 0.       ]\n",
      " [ 2.       ]\n",
      " [-3.0000001]\n",
      " [ 6.       ]\n",
      " [ 0.9999999]\n",
      " [-2.       ]\n",
      " [-2.       ]\n",
      " [-1.0000001]\n",
      " [-1.0000001]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[-20.5       ]\n",
      " [  2.83333333]\n",
      " [ -5.        ]\n",
      " [ -1.66666667]\n",
      " [  1.0001    ]\n",
      " [  2.        ]\n",
      " [ -1.66656667]\n",
      " [ 19.33333333]\n",
      " [-12.        ]\n",
      " [ -2.        ]\n",
      " [ -1.        ]\n",
      " [  9.        ]]\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]]\n",
      "twin_svm_res\n",
      "res_twin_bef:[[-8.37078842e-01]\n",
      " [-8.66663194e-01]\n",
      " [-2.65562561e-01]\n",
      " [-1.79374200e-01]\n",
      " [-8.61893608e-02]\n",
      " [-1.00000000e-06]\n",
      " [ 1.91270059e+00]\n",
      " [ 1.88311624e+00]\n",
      " [-2.58564082e-01]\n",
      " [ 1.00000000e-06]\n",
      " [ 1.27280123e+00]\n",
      " [ 8.98463437e-01]]\n",
      "res_twin_aft:[[0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "B\n",
      "no child node\n",
      "res_before_quant:[[ 20.5       ]\n",
      " [ -2.83333333]\n",
      " [  5.        ]\n",
      " [  1.66666667]\n",
      " [  1.33333333]\n",
      " [ -2.        ]\n",
      " [  4.        ]\n",
      " [-19.33333333]\n",
      " [ 12.        ]\n",
      " [  2.        ]\n",
      " [  1.        ]\n",
      " [ -9.        ]]\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]]\n",
      "twin_svm_res\n",
      "res_twin_bef:[[ 4.13721839e+00]\n",
      " [ 1.74458673e+00]\n",
      " [-3.86034544e-03]\n",
      " [-6.93075298e-01]\n",
      " [-1.38614960e+00]\n",
      " [ 3.56508463e-01]\n",
      " [-2.08308324e+00]\n",
      " [-4.47571489e+00]\n",
      " [ 2.06764386e+00]\n",
      " [-1.00000000e-06]\n",
      " [-2.07922389e+00]\n",
      " [-1.71499574e+00]]\n",
      "res_twin_aft:[[1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]]\n",
      "y_1:[[0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]]\n",
      "y_2:[[1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]\n",
      " [0.]]\n",
      "res_before_quant:[[-7.5000001e+00]\n",
      " [-5.0000010e-01]\n",
      " [-2.0000000e+00]\n",
      " [-1.0000000e+00]\n",
      " [-1.0000000e+00]\n",
      " [-1.0000001e+00]\n",
      " [ 1.0000000e-04]\n",
      " [ 7.0001000e+00]\n",
      " [-5.0000001e+00]\n",
      " [ 1.0001000e+00]\n",
      " [ 1.0001000e+00]\n",
      " [ 4.0001000e+00]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]]\n",
      "y_1:[[0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[-2.16666667]\n",
      " [ 0.83343333]\n",
      " [ 1.0001    ]\n",
      " [ 1.33343333]\n",
      " [ 1.33333333]\n",
      " [ 1.66666667]\n",
      " [ 2.33333333]\n",
      " [ 0.66666657]\n",
      " [-1.        ]\n",
      " [-4.0000001 ]\n",
      " [-2.0000001 ]\n",
      " [-1.0000001 ]]\n",
      "delta condition not met\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "AB\n",
      "no child node\n",
      "res_before_quant:[[ 33.5]\n",
      " [ -1.5]\n",
      " [ -1. ]\n",
      " [ -6. ]\n",
      " [-14. ]\n",
      " [-19. ]\n",
      " [-25. ]\n",
      " [-60. ]\n",
      " [ 17. ]\n",
      " [  2. ]\n",
      " [-22. ]\n",
      " [-37. ]]\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "AB\n",
      "no child node\n",
      "res_before_quant:[[17.5       ]\n",
      " [-8.16666667]\n",
      " [ 6.        ]\n",
      " [ 2.33333333]\n",
      " [ 5.66666667]\n",
      " [ 2.        ]\n",
      " [16.        ]\n",
      " [-9.66666667]\n",
      " [10.        ]\n",
      " [-1.        ]\n",
      " [ 9.        ]\n",
      " [-2.        ]]\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]]\n",
      "B\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[-12.]\n",
      " [  2.]\n",
      " [ -6.]\n",
      " [ -4.]\n",
      " [ -6.]\n",
      " [ -4.]\n",
      " [-12.]\n",
      " [  2.]\n",
      " [ -8.]\n",
      " [ -2.]\n",
      " [ -8.]\n",
      " [ -2.]]\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[-22.33333333]\n",
      " [  1.0001    ]\n",
      " [ -1.33333333]\n",
      " [  2.        ]\n",
      " [  6.        ]\n",
      " [  9.33333333]\n",
      " [ 10.66666667]\n",
      " [ 34.0001    ]\n",
      " [-12.        ]\n",
      " [ -2.        ]\n",
      " [ 10.        ]\n",
      " [ 20.        ]]\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "y_2:[[0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[ 20.       ]\n",
      " [ -1.0000001]\n",
      " [  2.       ]\n",
      " [ -1.0000001]\n",
      " [ -4.0000001]\n",
      " [ -7.0000001]\n",
      " [ -7.0000001]\n",
      " [-28.0000001]\n",
      " [ 11.       ]\n",
      " [  2.       ]\n",
      " [ -7.0000001]\n",
      " [-16.0000001]]\n",
      "res_after_quant:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ 5.53343323]\n",
      " [ 1.        ]\n",
      " [ 0.13343323]\n",
      " [ 2.46676667]\n",
      " [ 2.06676667]\n",
      " [ 1.0001    ]\n",
      " [ 2.33343333]\n",
      " [-5.6       ]\n",
      " [ 2.66676657]\n",
      " [-1.0000001 ]\n",
      " [ 1.66676667]\n",
      " [-2.        ]]\n",
      "delta condition not met\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]]\n",
      "res_before_quant:[[-11.4999001 ]\n",
      " [ -1.0000001 ]\n",
      " [ -1.66666667]\n",
      " [ -5.0000001 ]\n",
      " [ -6.16666677]\n",
      " [ -4.66666677]\n",
      " [-10.0000001 ]\n",
      " [  5.33333333]\n",
      " [ -8.33323343]\n",
      " [  1.0001    ]\n",
      " [ -7.33333343]\n",
      " [  2.        ]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]]\n",
      "y_1:[[0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]]\n",
      "res_before_quant:[[-1.96666667]\n",
      " [ 2.        ]\n",
      " [ 1.33343333]\n",
      " [ 1.9001    ]\n",
      " [ 1.0001    ]\n",
      " [ 1.56676667]\n",
      " [-1.36656667]\n",
      " [-1.2000001 ]\n",
      " [-1.        ]\n",
      " [-1.0000001 ]\n",
      " [-2.        ]\n",
      " [-2.0000001 ]]\n",
      "res_after_quant:[[0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "B\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[  2.        ]\n",
      " [  1.        ]\n",
      " [ -7.42857143]\n",
      " [ -7.57142857]\n",
      " [-13.42857143]\n",
      " [-13.57142857]\n",
      " [-25.        ]\n",
      " [-26.        ]\n",
      " [ -1.28571429]\n",
      " [ -1.71428571]\n",
      " [-19.28571429]\n",
      " [-19.71428571]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ 1.0001    ]\n",
      " [ 3.33343333]\n",
      " [ 1.33333333]\n",
      " [ 1.66666667]\n",
      " [ 4.33333333]\n",
      " [ 4.66666667]\n",
      " [ 9.33333333]\n",
      " [11.66666667]\n",
      " [-2.        ]\n",
      " [-1.        ]\n",
      " [ 7.        ]\n",
      " [ 8.        ]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "y_2:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[ -1.0000001]\n",
      " [ -1.0000001]\n",
      " [ -4.7500001]\n",
      " [ -4.7500001]\n",
      " [ -7.2500001]\n",
      " [ -7.2500001]\n",
      " [-12.2500001]\n",
      " [-12.2500001]\n",
      " [  1.5      ]\n",
      " [  1.5      ]\n",
      " [ -9.7500001]\n",
      " [ -9.7500001]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[0]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ 2.        ]\n",
      " [ 2.71438571]\n",
      " [ 2.0001    ]\n",
      " [ 1.85724286]\n",
      " [ 1.14295714]\n",
      " [ 1.0001    ]\n",
      " [-1.85714286]\n",
      " [-2.85714286]\n",
      " [-1.0000001 ]\n",
      " [-1.42857153]\n",
      " [-1.28571429]\n",
      " [-1.71428571]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "AB\n",
      "B\n",
      "A\n",
      "no child node\n",
      "res_before_quant:[[ 50.5]\n",
      " [  1.5]\n",
      " [ 13. ]\n",
      " [  6. ]\n",
      " [  2. ]\n",
      " [ -5. ]\n",
      " [  1. ]\n",
      " [-48. ]\n",
      " [ 31. ]\n",
      " [ 10. ]\n",
      " [ -2. ]\n",
      " [-23. ]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ 8.5484871 ]\n",
      " [15.09687419]\n",
      " [ 8.83880968]\n",
      " [ 9.77429355]\n",
      " [ 7.16139032]\n",
      " [ 1.5483871 ]\n",
      " [ 1.0001    ]\n",
      " [ 1.        ]\n",
      " [ 9.58074516]\n",
      " [12.38719677]\n",
      " [-2.        ]\n",
      " [ 0.80645161]]\n",
      "delta condition not met\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "res_before_quant:[[ -7.09677429]\n",
      " [-13.19354849]\n",
      " [ -7.67741945]\n",
      " [ -8.5483872 ]\n",
      " [ -6.32258075]\n",
      " [ -7.19354849]\n",
      " [ -1.0000001 ]\n",
      " [ -7.09677429]\n",
      " [ -8.16129042]\n",
      " [-10.77419365]\n",
      " [  2.        ]\n",
      " [ -0.61290323]]\n",
      "delta condition not met\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]]\n",
      "A\n",
      "B\n",
      "B\n",
      "no child node\n",
      "res_before_quant:[[-51.]\n",
      " [ -2.]\n",
      " [-15.]\n",
      " [ -8.]\n",
      " [ -5.]\n",
      " [  2.]\n",
      " [ -6.]\n",
      " [ 43.]\n",
      " [-32.]\n",
      " [-11.]\n",
      " [ -2.]\n",
      " [ 19.]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]]\n",
      "res_before_quant:[[-50.5      ]\n",
      " [ -1.5      ]\n",
      " [-13.       ]\n",
      " [ -6.       ]\n",
      " [ -2.       ]\n",
      " [ -1.0000001]\n",
      " [ -1.       ]\n",
      " [ 41.9999999]\n",
      " [-31.       ]\n",
      " [-10.       ]\n",
      " [  2.       ]\n",
      " [ 16.9999999]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[10.       ]\n",
      " [ 3.       ]\n",
      " [ 4.       ]\n",
      " [ 3.       ]\n",
      " [ 2.       ]\n",
      " [ 1.       ]\n",
      " [ 1.       ]\n",
      " [-8.0000001]\n",
      " [ 7.       ]\n",
      " [ 4.       ]\n",
      " [-1.0000001]\n",
      " [-4.0000001]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]]\n",
      "twin_svm_res\n",
      "res_twin_bef:[[-3.03532543e-01]\n",
      " [ 1.49907842e+00]\n",
      " [ 5.70192859e-01]\n",
      " [ 8.27708711e-01]\n",
      " [ 6.37644758e-01]\n",
      " [ 8.95160609e-01]\n",
      " [ 1.00000000e-06]\n",
      " [ 3.93470585e-01]\n",
      " [ 2.45225109e-01]\n",
      " [ 1.01777266e+00]\n",
      " [-9.61560570e-01]\n",
      " [-1.89013016e-01]]\n",
      "res_twin_aft:[[0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]]\n",
      "y_1:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [0]]\n",
      "y_2:[[0.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [1.]\n",
      " [0.]\n",
      " [0.]]\n",
      "res_before_quant:[[-4.33333333]\n",
      " [-7.0000001 ]\n",
      " [-6.33333343]\n",
      " [-6.0000001 ]\n",
      " [-5.0000001 ]\n",
      " [-4.66666677]\n",
      " [-3.33333343]\n",
      " [-1.0000001 ]\n",
      " [-8.0000001 ]\n",
      " [-7.0000001 ]\n",
      " [ 1.0001    ]\n",
      " [ 2.        ]]\n",
      "res_after_quant:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "y_1:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n",
      "y_2:[[0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [1]\n",
      " [1]]\n",
      "res_before_quant:[[ 1.0001    ]\n",
      " [ 2.0001    ]\n",
      " [ 2.55367143]\n",
      " [ 2.69652857]\n",
      " [ 3.30367143]\n",
      " [ 3.44652857]\n",
      " [ 1.        ]\n",
      " [ 2.        ]\n",
      " [-1.71428571]\n",
      " [-1.28571429]\n",
      " [-1.42857153]\n",
      " [-1.0000001 ]]\n",
      "res_after_quant:[[1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [1]\n",
      " [0]\n",
      " [0]\n",
      " [0]\n",
      " [0]]\n"
     ]
    }
   ],
   "source": [
    "print(\"###Verification###\")\n",
    "print(curr_Neuron.calculate(X))\n",
    "a = curr_Neuron.calculate(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}