classifier_3class_rectangle_ndim_gaussian.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib notebook\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import math\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "\n",
    "from tensorflow import set_random_seed\n",
    "# for reproducibility \n",
    "np.random.seed(1)\n",
    "set_random_seed(1)\n",
    "\n",
    "from keras.layers import Dense, Input, concatenate\n",
    "from keras.layers import Conv2D, MaxPooling2D, Flatten, Lambda, Dropout, Activation, Lambda\n",
    "from keras.layers import Reshape, Conv2DTranspose\n",
    "from keras.models import Model, Sequential\n",
    "from keras.optimizers import Adam\n",
    "from keras.datasets import mnist\n",
    "from keras.datasets import fashion_mnist\n",
    "from keras.losses import mse, binary_crossentropy\n",
    "from keras.utils import plot_model\n",
    "from keras import backend as K\n",
    "\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.utils import np_utils\n",
    "\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build a classifier with outlier samples from the rectangle\n",
    "\n",
    "class Classifier:\n",
    "\n",
    "    # 3 Class classifier - class 2 represents outliers and class 0,1 represent inliers.\n",
    "    # Class 0 sampled from a circle centered at (x, y) = (-2, 0) with radius 0.5\n",
    "    # Class 1 sampled from a circle centered at (x, y) = (+2, 0) with radius 0.5\n",
    "    # Class 2 sampled from rectagle with origin (x, y) = (-3, -1) with h=2, w=6 excluding regions covered by inliers.\n",
    "\n",
    "    # Generate n points uniformly within a circle of radius r, centered at (x, y)\n",
    "    # return as a numpy array n x 2 of dtype = float32\n",
    "    def __init__(self):\n",
    "        self.number_of_classes = 3\n",
    "        self.input_dim = 2\n",
    "        self.epochs = 200\n",
    "        self.batch_size = 32\n",
    "        # There are 2 inlier circles\n",
    "        # Inlier circle centers on the 1st axis; for all other dimensions, the centers are at 0.\n",
    "        self.x1 = -10\n",
    "        self.x2 = +10\n",
    "        self.r = 2.0\n",
    "        # Number of inlier samples for each class\n",
    "        self.n = 1000\n",
    "        # width of the outlier outer square side\n",
    "        self.s = 100\n",
    "        self.data, self.labels = self.get_data()\n",
    "        print(self.data.shape, self.labels.shape)\n",
    "        self.X_train, self.X_test, self.y_train, self.y_test = train_test_split(self.data, self.labels, test_size=0.5, shuffle=True, stratify=self.labels,\n",
    "                                                            random_state=1)\n",
    "        self.model = self.get_model()\n",
    "\n",
    "        \n",
    "        self.mu = np.mean(self.X_train, axis=0)\n",
    "        self.sigma = np.std(self.X_train, axis=0)\n",
    "\n",
    "        self.data = self.preprocess_data(self.data)\n",
    "        self.X_train = self.preprocess_data(self.X_train)\n",
    "        self.X_test = self.preprocess_data(self.X_test)\n",
    "\n",
    "\n",
    "\n",
    "    # return True if point p (a list) lies inside a hypersphere centered at the origin with radius r\n",
    "    def is_in_hypersphere(self, r, p):\n",
    "        p = np.array(p)\n",
    "        return np.sum(p*p) <= r**2\n",
    "\n",
    "\n",
    "    # return True if point p (a list) lies inside a hypersphere centered at the (x, 0, 0, ...) with radius r\n",
    "    def is_in_hypersphere_centered_at(self, x, r, p):\n",
    "        p = np.array(p)\n",
    "        p[0] -= x\n",
    "        return np.sum(p*p) <= r**2\n",
    "\n",
    "    \n",
    "    # uniformly sample a point from a square centered at the origin with side = s\n",
    "    def sample_point_from_square(self, s):\n",
    "        p = []\n",
    "        # first point\n",
    "        for i in range(0, self.input_dim):\n",
    "            p.append(np.random.uniform(-1*s/2, s/2, 1))\n",
    "        return p\n",
    "\n",
    "\n",
    "    def preprocess_data(self, x):\n",
    "        #return (x-self.mu)/self.sigma\n",
    "        return x\n",
    "\n",
    "\n",
    "    # sample points from 2D gaussian centered at origin\n",
    "    def sample_point_from_gaussian(self, sigma):\n",
    "        mean = np.zeros(self.input_dim)\n",
    "        cov = np.zeros((self.input_dim, self.input_dim), np.float32)\n",
    "        np.fill_diagonal(cov, sigma*sigma)\n",
    "        p = list(np.random.multivariate_normal(mean, cov, 1).T.squeeze())\n",
    "        return p\n",
    "\n",
    "        \n",
    "    # sample from a hypersphre with rejection sampling\n",
    "    def get_gaussian_data(self, x, r, n):\n",
    "        data = np.zeros((n, self.input_dim))\n",
    "        count = 0\n",
    "        while count<n:\n",
    "            data[count, :] = self.sample_point_from_gaussian(r)\n",
    "            # update the point with offset for the 1st dimension\n",
    "            data[count, 0] += x\n",
    "            count += 1\n",
    "\n",
    "        data = data.astype(np.float32)\n",
    "        return data\n",
    "    \n",
    "    \n",
    "    \n",
    "    # generate n outlier points uniformly sampled within square centered at the origin\n",
    "    # with side=s, excuding the inlier circles\n",
    "    def get_outlier_data_sub(self, s, x1, x2, r, n):\n",
    "        count = 0\n",
    "        data = np.zeros((n, self.input_dim))\n",
    "        while count < n:\n",
    "            p = self.sample_point_from_square(s)\n",
    "            if (not self.is_in_hypersphere_centered_at(x1, r, p)) and (not self.is_in_hypersphere_centered_at(x2, r, p)):\n",
    "                data[count, :] = p\n",
    "                count += 1\n",
    "        return data\n",
    "\n",
    "\n",
    "    def get_outlier_data(self):\n",
    "        data = self.get_outlier_data_sub(self.s, self.x1, self.x2, self.r*3, self.n)\n",
    "        return data.astype(np.float32)\n",
    "\n",
    "\n",
    "    def get_data(self):\n",
    "        # class 0 data\n",
    "        data = self.get_gaussian_data(self.x1, self.r, self.n)\n",
    "        labels = np.ones((self.n, 1))*0\n",
    "        print(labels.shape)\n",
    "        # class 1 data\n",
    "        data = np.concatenate((data, self.get_gaussian_data(self.x2, self.r, self.n)), axis=0)\n",
    "        labels = np.concatenate((labels, np.ones((self.n, 1))*1), axis=0)\n",
    "        print(labels.shape)\n",
    "        # outlier data\n",
    "        data = np.concatenate((data, self.get_outlier_data()), axis=0)\n",
    "        labels = np.concatenate((labels, np.ones((self.n, 1))*2), axis=0)\n",
    "        print(labels.shape)\n",
    "        return data.astype(np.float32), labels.astype(np.float32)\n",
    "    \n",
    "    def get_model(self):\n",
    "        model = Sequential()\n",
    "        model.add(Dense(500, input_shape=(self.input_dim,)))\n",
    "        model.add(Activation('relu'))\n",
    "        model.add(Dense(500))\n",
    "        model.add(Activation('relu'))\n",
    "        model.add(Dense(self.number_of_classes))\n",
    "        # model.add(Convolution2D(10,3,3, border_mode='same'))\n",
    "        # model.add(GlobalAveragePooling2D())\n",
    "        model.add(Activation('softmax'))\n",
    "        model.compile(loss='binary_crossentropy', optimizer=Adam(0.001), metrics=['accuracy'])\n",
    "        return model\n",
    "\n",
    "\n",
    "    def train(self):\n",
    "        # fit the model\n",
    "        Y_train = np_utils.to_categorical(self.y_train, self.number_of_classes)\n",
    "        Y_test = np_utils.to_categorical(self.y_test, self.number_of_classes)\n",
    "        self.model.fit(self.X_train, Y_train, epochs=self.epochs, validation_data=(self.X_test, Y_test), batch_size=self.batch_size)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1000, 1)\n",
      "(2000, 1)\n",
      "(3000, 1)\n",
      "((3000, 2), (3000, 1))\n",
      "Train on 1500 samples, validate on 1500 samples\n",
      "Epoch 1/200\n",
      "1500/1500 [==============================] - 0s 227us/step - loss: 0.3854 - acc: 0.8716 - val_loss: 0.2651 - val_acc: 0.9193\n",
      "Epoch 2/200\n",
      "1500/1500 [==============================] - 0s 93us/step - loss: 0.2186 - acc: 0.9436 - val_loss: 0.1710 - val_acc: 0.9613\n",
      "Epoch 3/200\n",
      "1500/1500 [==============================] - 0s 89us/step - loss: 0.1526 - acc: 0.9573 - val_loss: 0.1351 - val_acc: 0.9693\n",
      "Epoch 4/200\n",
      "1500/1500 [==============================] - 0s 86us/step - loss: 0.0815 - acc: 0.9798 - val_loss: 0.0531 - val_acc: 0.9858\n",
      "Epoch 5/200\n",
      "1500/1500 [==============================] - 0s 87us/step - loss: 0.0558 - acc: 0.9856 - val_loss: 0.0449 - val_acc: 0.9842\n",
      "Epoch 6/200\n",
      "1500/1500 [==============================] - 0s 87us/step - loss: 0.0285 - acc: 0.9916 - val_loss: 0.0321 - val_acc: 0.9880\n",
      "Epoch 7/200\n",
      "1500/1500 [==============================] - 0s 90us/step - loss: 0.0193 - acc: 0.9949 - val_loss: 0.0239 - val_acc: 0.9920\n",
      "Epoch 8/200\n",
      "1500/1500 [==============================] - 0s 85us/step - loss: 0.0154 - acc: 0.9947 - val_loss: 0.0212 - val_acc: 0.9953\n",
      "Epoch 9/200\n",
      "1500/1500 [==============================] - 0s 86us/step - loss: 0.0152 - acc: 0.9942 - val_loss: 0.0208 - val_acc: 0.9951\n",
      "Epoch 10/200\n",
      "1500/1500 [==============================] - 0s 88us/step - loss: 0.0115 - acc: 0.9976 - val_loss: 0.0289 - val_acc: 0.9911\n",
      "Epoch 11/200\n",
      "1500/1500 [==============================] - 0s 86us/step - loss: 0.0158 - acc: 0.9953 - val_loss: 0.0224 - val_acc: 0.9924\n",
      "Epoch 12/200\n",
      "1500/1500 [==============================] - 0s 88us/step - loss: 0.0159 - acc: 0.9956 - val_loss: 0.0174 - val_acc: 0.9924\n",
      "Epoch 13/200\n",
      "1500/1500 [==============================] - 0s 88us/step - loss: 0.0122 - acc: 0.9953 - val_loss: 0.0150 - val_acc: 0.9956\n",
      "Epoch 14/200\n",
      "1500/1500 [==============================] - 0s 88us/step - loss: 0.0115 - acc: 0.9967 - val_loss: 0.0190 - val_acc: 0.9933\n",
      "Epoch 15/200\n",
      "1500/1500 [==============================] - 0s 87us/step - loss: 0.0109 - acc: 0.9967 - val_loss: 0.0130 - val_acc: 0.9969\n",
      "Epoch 16/200\n",
      "1500/1500 [==============================] - 0s 88us/step - loss: 0.0138 - acc: 0.9940 - val_loss: 0.0148 - val_acc: 0.9951\n",
      "Epoch 17/200\n",
      "1500/1500 [==============================] - 0s 85us/step - loss: 0.0154 - acc: 0.9949 - val_loss: 0.0207 - val_acc: 0.9938\n",
      "Epoch 18/200\n",
      "1500/1500 [==============================] - 0s 84us/step - loss: 0.0091 - acc: 0.9973 - val_loss: 0.0231 - val_acc: 0.9929\n",
      "Epoch 19/200\n",
      "1500/1500 [==============================] - 0s 86us/step - loss: 0.0101 - acc: 0.9958 - val_loss: 0.0210 - val_acc: 0.9931\n",
      "Epoch 20/200\n",
      "1500/1500 [==============================] - 0s 85us/step - loss: 0.0068 - acc: 0.9980 - val_loss: 0.0212 - val_acc: 0.9924\n",
      "Epoch 21/200\n",
      "1500/1500 [==============================] - 0s 85us/step - loss: 0.0149 - acc: 0.9958 - val_loss: 0.0254 - val_acc: 0.9924\n",
      "Epoch 22/200\n",
      "1500/1500 [==============================] - 0s 87us/step - loss: 0.0107 - acc: 0.9971 - val_loss: 0.0157 - val_acc: 0.9956\n",
      "Epoch 23/200\n",
      "1500/1500 [==============================] - 0s 87us/step - loss: 0.0066 - acc: 0.9982 - val_loss: 0.0161 - val_acc: 0.9956\n",
      "Epoch 24/200\n",
      "1500/1500 [==============================] - 0s 85us/step - loss: 0.0114 - acc: 0.9969 - val_loss: 0.0176 - val_acc: 0.9938\n",
      "Epoch 25/200\n",
      "1500/1500 [==============================] - 0s 88us/step - loss: 0.0060 - acc: 0.9976 - val_loss: 0.0226 - val_acc: 0.9924\n",
      "Epoch 26/200\n",
      "1500/1500 [==============================] - 0s 95us/step - loss: 0.0112 - acc: 0.9960 - val_loss: 0.0224 - val_acc: 0.9924\n",
      "Epoch 27/200\n",
      "1500/1500 [==============================] - 0s 85us/step - loss: 0.0105 - acc: 0.9960 - val_loss: 0.0173 - val_acc: 0.9947\n",
      "Epoch 28/200\n",
      "1500/1500 [==============================] - 0s 89us/step - loss: 0.0070 - acc: 0.9973 - val_loss: 0.0224 - val_acc: 0.9938\n",
      "Epoch 29/200\n",
      "1500/1500 [==============================] - 0s 89us/step - loss: 0.0065 - acc: 0.9982 - val_loss: 0.0140 - val_acc: 0.9956\n",
      "Epoch 30/200\n",
      "1500/1500 [==============================] - 0s 89us/step - loss: 0.0072 - acc: 0.9964 - val_loss: 0.0149 - val_acc: 0.9956\n",
      "Epoch 31/200\n",
      "1500/1500 [==============================] - 0s 88us/step - loss: 0.0061 - acc: 0.9976 - val_loss: 0.0326 - val_acc: 0.9893\n",
      "Epoch 32/200\n",
      "1500/1500 [==============================] - 0s 97us/step - loss: 0.0141 - acc: 0.9958 - val_loss: 0.0261 - val_acc: 0.9924\n",
      "Epoch 33/200\n",
      "1500/1500 [==============================] - 0s 106us/step - loss: 0.0109 - acc: 0.9969 - val_loss: 0.0209 - val_acc: 0.9947\n",
      "Epoch 34/200\n",
      "1500/1500 [==============================] - 0s 98us/step - loss: 0.0117 - acc: 0.9960 - val_loss: 0.0129 - val_acc: 0.9964\n",
      "Epoch 35/200\n",
      "1500/1500 [==============================] - 0s 104us/step - loss: 0.0198 - acc: 0.9942 - val_loss: 0.0266 - val_acc: 0.9924\n",
      "Epoch 36/200\n",
      "1500/1500 [==============================] - 0s 105us/step - loss: 0.0090 - acc: 0.9960 - val_loss: 0.0179 - val_acc: 0.9942\n",
      "Epoch 37/200\n",
      "1500/1500 [==============================] - 0s 102us/step - loss: 0.0083 - acc: 0.9978 - val_loss: 0.0144 - val_acc: 0.9960\n",
      "Epoch 38/200\n",
      "1500/1500 [==============================] - 0s 96us/step - loss: 0.0064 - acc: 0.9978 - val_loss: 0.0145 - val_acc: 0.9969\n",
      "Epoch 39/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0067 - acc: 0.9976 - val_loss: 0.0177 - val_acc: 0.9947\n",
      "Epoch 40/200\n",
      "1500/1500 [==============================] - 0s 94us/step - loss: 0.0100 - acc: 0.9958 - val_loss: 0.0133 - val_acc: 0.9951\n",
      "Epoch 41/200\n",
      "1500/1500 [==============================] - 0s 100us/step - loss: 0.0090 - acc: 0.9960 - val_loss: 0.0135 - val_acc: 0.9964\n",
      "Epoch 42/200\n",
      "1500/1500 [==============================] - 0s 101us/step - loss: 0.0085 - acc: 0.9964 - val_loss: 0.0194 - val_acc: 0.9947\n",
      "Epoch 43/200\n",
      "1500/1500 [==============================] - 0s 94us/step - loss: 0.0208 - acc: 0.9940 - val_loss: 0.0274 - val_acc: 0.9916\n",
      "Epoch 44/200\n",
      "1500/1500 [==============================] - 0s 112us/step - loss: 0.0083 - acc: 0.9973 - val_loss: 0.0171 - val_acc: 0.9947\n",
      "Epoch 45/200\n",
      "1500/1500 [==============================] - 0s 96us/step - loss: 0.0050 - acc: 0.9982 - val_loss: 0.0098 - val_acc: 0.9969\n",
      "Epoch 46/200\n",
      "1500/1500 [==============================] - 0s 98us/step - loss: 0.0076 - acc: 0.9978 - val_loss: 0.0137 - val_acc: 0.9951\n",
      "Epoch 47/200\n",
      "1500/1500 [==============================] - 0s 102us/step - loss: 0.0049 - acc: 0.9987 - val_loss: 0.0128 - val_acc: 0.9951\n",
      "Epoch 48/200\n",
      "1500/1500 [==============================] - 0s 92us/step - loss: 0.0040 - acc: 0.9978 - val_loss: 0.0201 - val_acc: 0.9942\n",
      "Epoch 49/200\n",
      "1500/1500 [==============================] - 0s 98us/step - loss: 0.0075 - acc: 0.9973 - val_loss: 0.0136 - val_acc: 0.9956\n",
      "Epoch 50/200\n",
      "1500/1500 [==============================] - 0s 103us/step - loss: 0.0058 - acc: 0.9978 - val_loss: 0.0127 - val_acc: 0.9956\n",
      "Epoch 51/200\n",
      "1500/1500 [==============================] - 0s 106us/step - loss: 0.0109 - acc: 0.9964 - val_loss: 0.0155 - val_acc: 0.9947\n",
      "Epoch 52/200\n",
      "1500/1500 [==============================] - 0s 99us/step - loss: 0.0074 - acc: 0.9982 - val_loss: 0.0175 - val_acc: 0.9933\n",
      "Epoch 53/200\n",
      "1500/1500 [==============================] - 0s 88us/step - loss: 0.0043 - acc: 0.9982 - val_loss: 0.0119 - val_acc: 0.9964\n",
      "Epoch 54/200\n",
      "1500/1500 [==============================] - 0s 93us/step - loss: 0.0036 - acc: 0.9987 - val_loss: 0.0149 - val_acc: 0.9947\n",
      "Epoch 55/200\n",
      "1500/1500 [==============================] - 0s 88us/step - loss: 0.0087 - acc: 0.9960 - val_loss: 0.0130 - val_acc: 0.9956\n",
      "Epoch 56/200\n",
      "1500/1500 [==============================] - 0s 91us/step - loss: 0.0143 - acc: 0.9938 - val_loss: 0.0130 - val_acc: 0.9953\n",
      "Epoch 57/200\n",
      "1500/1500 [==============================] - 0s 87us/step - loss: 0.0067 - acc: 0.9973 - val_loss: 0.0195 - val_acc: 0.9951\n",
      "Epoch 58/200\n",
      "1500/1500 [==============================] - 0s 100us/step - loss: 0.0055 - acc: 0.9982 - val_loss: 0.0300 - val_acc: 0.9920\n",
      "Epoch 59/200\n",
      "1500/1500 [==============================] - 0s 89us/step - loss: 0.0078 - acc: 0.9964 - val_loss: 0.0157 - val_acc: 0.9951\n",
      "Epoch 60/200\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1500/1500 [==============================] - 0s 103us/step - loss: 0.0040 - acc: 0.9978 - val_loss: 0.0144 - val_acc: 0.9956\n",
      "Epoch 61/200\n",
      "1500/1500 [==============================] - 0s 96us/step - loss: 0.0032 - acc: 0.9991 - val_loss: 0.0126 - val_acc: 0.9964\n",
      "Epoch 62/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0072 - acc: 0.9973 - val_loss: 0.0124 - val_acc: 0.9960\n",
      "Epoch 63/200\n",
      "1500/1500 [==============================] - 0s 82us/step - loss: 0.0045 - acc: 0.9980 - val_loss: 0.0123 - val_acc: 0.9960\n",
      "Epoch 64/200\n",
      "1500/1500 [==============================] - 0s 85us/step - loss: 0.0055 - acc: 0.9980 - val_loss: 0.0177 - val_acc: 0.9942\n",
      "Epoch 65/200\n",
      "1500/1500 [==============================] - 0s 84us/step - loss: 0.0059 - acc: 0.9978 - val_loss: 0.0138 - val_acc: 0.9964\n",
      "Epoch 66/200\n",
      "1500/1500 [==============================] - 0s 83us/step - loss: 0.0047 - acc: 0.9987 - val_loss: 0.0102 - val_acc: 0.9960\n",
      "Epoch 67/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0044 - acc: 0.9982 - val_loss: 0.0140 - val_acc: 0.9951\n",
      "Epoch 68/200\n",
      "1500/1500 [==============================] - 0s 84us/step - loss: 0.0073 - acc: 0.9960 - val_loss: 0.0127 - val_acc: 0.9969\n",
      "Epoch 69/200\n",
      "1500/1500 [==============================] - 0s 99us/step - loss: 0.0063 - acc: 0.9978 - val_loss: 0.0149 - val_acc: 0.9956\n",
      "Epoch 70/200\n",
      "1500/1500 [==============================] - 0s 98us/step - loss: 0.0042 - acc: 0.9984 - val_loss: 0.0180 - val_acc: 0.9938\n",
      "Epoch 71/200\n",
      "1500/1500 [==============================] - 0s 92us/step - loss: 0.0124 - acc: 0.9931 - val_loss: 0.0209 - val_acc: 0.9929\n",
      "Epoch 72/200\n",
      "1500/1500 [==============================] - 0s 91us/step - loss: 0.0127 - acc: 0.9962 - val_loss: 0.0170 - val_acc: 0.9942\n",
      "Epoch 73/200\n",
      "1500/1500 [==============================] - 0s 97us/step - loss: 0.0072 - acc: 0.9973 - val_loss: 0.0168 - val_acc: 0.9933\n",
      "Epoch 74/200\n",
      "1500/1500 [==============================] - 0s 86us/step - loss: 0.0114 - acc: 0.9969 - val_loss: 0.0116 - val_acc: 0.9960\n",
      "Epoch 75/200\n",
      "1500/1500 [==============================] - 0s 88us/step - loss: 0.0040 - acc: 0.9982 - val_loss: 0.0101 - val_acc: 0.9960\n",
      "Epoch 76/200\n",
      "1500/1500 [==============================] - 0s 100us/step - loss: 0.0048 - acc: 0.9978 - val_loss: 0.0121 - val_acc: 0.9960\n",
      "Epoch 77/200\n",
      "1500/1500 [==============================] - 0s 110us/step - loss: 0.0043 - acc: 0.9987 - val_loss: 0.0108 - val_acc: 0.9960\n",
      "Epoch 78/200\n",
      "1500/1500 [==============================] - 0s 98us/step - loss: 0.0026 - acc: 0.9987 - val_loss: 0.0152 - val_acc: 0.9951\n",
      "Epoch 79/200\n",
      "1500/1500 [==============================] - 0s 108us/step - loss: 0.0039 - acc: 0.9987 - val_loss: 0.0119 - val_acc: 0.9960\n",
      "Epoch 80/200\n",
      "1500/1500 [==============================] - 0s 108us/step - loss: 0.0034 - acc: 0.9982 - val_loss: 0.0186 - val_acc: 0.9951\n",
      "Epoch 81/200\n",
      "1500/1500 [==============================] - 0s 99us/step - loss: 0.0046 - acc: 0.9978 - val_loss: 0.0144 - val_acc: 0.9956\n",
      "Epoch 82/200\n",
      "1500/1500 [==============================] - 0s 106us/step - loss: 0.0043 - acc: 0.9973 - val_loss: 0.0107 - val_acc: 0.9967\n",
      "Epoch 83/200\n",
      "1500/1500 [==============================] - 0s 95us/step - loss: 0.0071 - acc: 0.9978 - val_loss: 0.0121 - val_acc: 0.9960\n",
      "Epoch 84/200\n",
      "1500/1500 [==============================] - 0s 101us/step - loss: 0.0035 - acc: 0.9987 - val_loss: 0.0174 - val_acc: 0.9947\n",
      "Epoch 85/200\n",
      "1500/1500 [==============================] - 0s 96us/step - loss: 0.0072 - acc: 0.9973 - val_loss: 0.0483 - val_acc: 0.9907\n",
      "Epoch 86/200\n",
      "1500/1500 [==============================] - 0s 108us/step - loss: 0.0105 - acc: 0.9951 - val_loss: 0.0497 - val_acc: 0.9889\n",
      "Epoch 87/200\n",
      "1500/1500 [==============================] - 0s 101us/step - loss: 0.0090 - acc: 0.9973 - val_loss: 0.0108 - val_acc: 0.9969\n",
      "Epoch 88/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0054 - acc: 0.9978 - val_loss: 0.0221 - val_acc: 0.9933\n",
      "Epoch 89/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0074 - acc: 0.9969 - val_loss: 0.0147 - val_acc: 0.9951\n",
      "Epoch 90/200\n",
      "1500/1500 [==============================] - 0s 100us/step - loss: 0.0070 - acc: 0.9964 - val_loss: 0.0136 - val_acc: 0.9956\n",
      "Epoch 91/200\n",
      "1500/1500 [==============================] - 0s 104us/step - loss: 0.0044 - acc: 0.9978 - val_loss: 0.0115 - val_acc: 0.9971\n",
      "Epoch 92/200\n",
      "1500/1500 [==============================] - 0s 94us/step - loss: 0.0052 - acc: 0.9969 - val_loss: 0.0113 - val_acc: 0.9960\n",
      "Epoch 93/200\n",
      "1500/1500 [==============================] - 0s 102us/step - loss: 0.0060 - acc: 0.9978 - val_loss: 0.0128 - val_acc: 0.9951\n",
      "Epoch 94/200\n",
      "1500/1500 [==============================] - 0s 101us/step - loss: 0.0108 - acc: 0.9973 - val_loss: 0.0881 - val_acc: 0.9842\n",
      "Epoch 95/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0127 - acc: 0.9956 - val_loss: 0.0175 - val_acc: 0.9947\n",
      "Epoch 96/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0197 - acc: 0.9949 - val_loss: 0.0165 - val_acc: 0.9971\n",
      "Epoch 97/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0096 - acc: 0.9967 - val_loss: 0.0147 - val_acc: 0.9951\n",
      "Epoch 98/200\n",
      "1500/1500 [==============================] - 0s 86us/step - loss: 0.0049 - acc: 0.9987 - val_loss: 0.0095 - val_acc: 0.9969\n",
      "Epoch 99/200\n",
      "1500/1500 [==============================] - 0s 82us/step - loss: 0.0049 - acc: 0.9987 - val_loss: 0.0132 - val_acc: 0.9956\n",
      "Epoch 100/200\n",
      "1500/1500 [==============================] - 0s 86us/step - loss: 0.0050 - acc: 0.9982 - val_loss: 0.0124 - val_acc: 0.9960\n",
      "Epoch 101/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0036 - acc: 0.9982 - val_loss: 0.0110 - val_acc: 0.9964\n",
      "Epoch 102/200\n",
      "1500/1500 [==============================] - 0s 88us/step - loss: 0.0035 - acc: 0.9987 - val_loss: 0.0139 - val_acc: 0.9956\n",
      "Epoch 103/200\n",
      "1500/1500 [==============================] - 0s 89us/step - loss: 0.0081 - acc: 0.9969 - val_loss: 0.0096 - val_acc: 0.9964\n",
      "Epoch 104/200\n",
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0040 - acc: 0.9982 - val_loss: 0.0265 - val_acc: 0.9898\n",
      "Epoch 105/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0128 - acc: 0.9964 - val_loss: 0.0124 - val_acc: 0.9956\n",
      "Epoch 106/200\n",
      "1500/1500 [==============================] - 0s 85us/step - loss: 0.0047 - acc: 0.9987 - val_loss: 0.0105 - val_acc: 0.9964\n",
      "Epoch 107/200\n",
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0047 - acc: 0.9978 - val_loss: 0.0140 - val_acc: 0.9958\n",
      "Epoch 108/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0030 - acc: 0.9984 - val_loss: 0.0103 - val_acc: 0.9964\n",
      "Epoch 109/200\n",
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0038 - acc: 0.9987 - val_loss: 0.0169 - val_acc: 0.9947\n",
      "Epoch 110/200\n",
      "1500/1500 [==============================] - 0s 85us/step - loss: 0.0082 - acc: 0.9960 - val_loss: 0.0136 - val_acc: 0.9956\n",
      "Epoch 111/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0051 - acc: 0.9978 - val_loss: 0.0153 - val_acc: 0.9947\n",
      "Epoch 112/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0050 - acc: 0.9973 - val_loss: 0.0117 - val_acc: 0.9960\n",
      "Epoch 113/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0047 - acc: 0.9982 - val_loss: 0.0154 - val_acc: 0.9942\n",
      "Epoch 114/200\n",
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0049 - acc: 0.9973 - val_loss: 0.0122 - val_acc: 0.9944\n",
      "Epoch 115/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0046 - acc: 0.9978 - val_loss: 0.0114 - val_acc: 0.9967\n",
      "Epoch 116/200\n",
      "1500/1500 [==============================] - 0s 97us/step - loss: 0.0033 - acc: 0.9987 - val_loss: 0.0097 - val_acc: 0.9964\n",
      "Epoch 117/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0028 - acc: 0.9987 - val_loss: 0.0098 - val_acc: 0.9969\n",
      "Epoch 118/200\n",
      "1500/1500 [==============================] - 0s 82us/step - loss: 0.0028 - acc: 0.9982 - val_loss: 0.0144 - val_acc: 0.9951\n",
      "Epoch 119/200\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0026 - acc: 0.9982 - val_loss: 0.0111 - val_acc: 0.9964\n",
      "Epoch 120/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0033 - acc: 0.9978 - val_loss: 0.0114 - val_acc: 0.9956\n",
      "Epoch 121/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0021 - acc: 0.9991 - val_loss: 0.0123 - val_acc: 0.9960\n",
      "Epoch 122/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0024 - acc: 0.9991 - val_loss: 0.0121 - val_acc: 0.9969\n",
      "Epoch 123/200\n",
      "1500/1500 [==============================] - 0s 88us/step - loss: 0.0037 - acc: 0.9982 - val_loss: 0.0119 - val_acc: 0.9964\n",
      "Epoch 124/200\n",
      "1500/1500 [==============================] - 0s 89us/step - loss: 0.0024 - acc: 0.9987 - val_loss: 0.0124 - val_acc: 0.9964\n",
      "Epoch 125/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0051 - acc: 0.9969 - val_loss: 0.0143 - val_acc: 0.9951\n",
      "Epoch 126/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0041 - acc: 0.9980 - val_loss: 0.0124 - val_acc: 0.9956\n",
      "Epoch 127/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0021 - acc: 0.9991 - val_loss: 0.0192 - val_acc: 0.9938\n",
      "Epoch 128/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0050 - acc: 0.9982 - val_loss: 0.0113 - val_acc: 0.9960\n",
      "Epoch 129/200\n",
      "1500/1500 [==============================] - 0s 85us/step - loss: 0.0030 - acc: 0.9991 - val_loss: 0.0110 - val_acc: 0.9964\n",
      "Epoch 130/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0026 - acc: 0.9987 - val_loss: 0.0103 - val_acc: 0.9964\n",
      "Epoch 131/200\n",
      "1500/1500 [==============================] - 0s 91us/step - loss: 0.0139 - acc: 0.9942 - val_loss: 0.0350 - val_acc: 0.9929\n",
      "Epoch 132/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0117 - acc: 0.9942 - val_loss: 0.0118 - val_acc: 0.9964\n",
      "Epoch 133/200\n",
      "1500/1500 [==============================] - 0s 90us/step - loss: 0.0031 - acc: 0.9978 - val_loss: 0.0137 - val_acc: 0.9956\n",
      "Epoch 134/200\n",
      "1500/1500 [==============================] - 0s 103us/step - loss: 0.0061 - acc: 0.9978 - val_loss: 0.0149 - val_acc: 0.9960\n",
      "Epoch 135/200\n",
      "1500/1500 [==============================] - 0s 111us/step - loss: 0.0051 - acc: 0.9973 - val_loss: 0.0143 - val_acc: 0.9969\n",
      "Epoch 136/200\n",
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0046 - acc: 0.9982 - val_loss: 0.0124 - val_acc: 0.9951\n",
      "Epoch 137/200\n",
      "1500/1500 [==============================] - 0s 95us/step - loss: 0.0030 - acc: 0.9982 - val_loss: 0.0122 - val_acc: 0.9960\n",
      "Epoch 138/200\n",
      "1500/1500 [==============================] - 0s 107us/step - loss: 0.0048 - acc: 0.9978 - val_loss: 0.0149 - val_acc: 0.9951\n",
      "Epoch 139/200\n",
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0071 - acc: 0.9969 - val_loss: 0.0205 - val_acc: 0.9942\n",
      "Epoch 140/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0079 - acc: 0.9964 - val_loss: 0.0082 - val_acc: 0.9969\n",
      "Epoch 141/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0029 - acc: 0.9987 - val_loss: 0.0223 - val_acc: 0.9960\n",
      "Epoch 142/200\n",
      "1500/1500 [==============================] - 0s 82us/step - loss: 0.0029 - acc: 0.9991 - val_loss: 0.0119 - val_acc: 0.9964\n",
      "Epoch 143/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0023 - acc: 0.9987 - val_loss: 0.0145 - val_acc: 0.9962\n",
      "Epoch 144/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0072 - acc: 0.9969 - val_loss: 0.0175 - val_acc: 0.9940\n",
      "Epoch 145/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0029 - acc: 0.9982 - val_loss: 0.0125 - val_acc: 0.9964\n",
      "Epoch 146/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0027 - acc: 0.9987 - val_loss: 0.0110 - val_acc: 0.9969\n",
      "Epoch 147/200\n",
      "1500/1500 [==============================] - 0s 85us/step - loss: 0.0022 - acc: 0.9991 - val_loss: 0.0110 - val_acc: 0.9973\n",
      "Epoch 148/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0023 - acc: 0.9996 - val_loss: 0.0143 - val_acc: 0.9960\n",
      "Epoch 149/200\n",
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0037 - acc: 0.9991 - val_loss: 0.0115 - val_acc: 0.9964\n",
      "Epoch 150/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0023 - acc: 0.9989 - val_loss: 0.0103 - val_acc: 0.9964\n",
      "Epoch 151/200\n",
      "1500/1500 [==============================] - 0s 86us/step - loss: 0.0030 - acc: 0.9982 - val_loss: 0.0122 - val_acc: 0.9960\n",
      "Epoch 152/200\n",
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0044 - acc: 0.9978 - val_loss: 0.0138 - val_acc: 0.9962\n",
      "Epoch 153/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0031 - acc: 0.9984 - val_loss: 0.0087 - val_acc: 0.9969\n",
      "Epoch 154/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0047 - acc: 0.9973 - val_loss: 0.0113 - val_acc: 0.9964\n",
      "Epoch 155/200\n",
      "1500/1500 [==============================] - 0s 92us/step - loss: 0.0040 - acc: 0.9978 - val_loss: 0.0110 - val_acc: 0.9973\n",
      "Epoch 156/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0050 - acc: 0.9982 - val_loss: 0.0104 - val_acc: 0.9960\n",
      "Epoch 157/200\n",
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0042 - acc: 0.9978 - val_loss: 0.0107 - val_acc: 0.9964\n",
      "Epoch 158/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0027 - acc: 0.9987 - val_loss: 0.0102 - val_acc: 0.9960\n",
      "Epoch 159/200\n",
      "1500/1500 [==============================] - 0s 86us/step - loss: 0.0020 - acc: 0.9991 - val_loss: 0.0101 - val_acc: 0.9969\n",
      "Epoch 160/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0019 - acc: 0.9991 - val_loss: 0.0121 - val_acc: 0.9956\n",
      "Epoch 161/200\n",
      "1500/1500 [==============================] - 0s 94us/step - loss: 0.0022 - acc: 0.9987 - val_loss: 0.0113 - val_acc: 0.9960\n",
      "Epoch 162/200\n",
      "1500/1500 [==============================] - 0s 103us/step - loss: 0.0022 - acc: 0.9984 - val_loss: 0.0111 - val_acc: 0.9956\n",
      "Epoch 163/200\n",
      "1500/1500 [==============================] - 0s 106us/step - loss: 0.0027 - acc: 0.9987 - val_loss: 0.0095 - val_acc: 0.9964\n",
      "Epoch 164/200\n",
      "1500/1500 [==============================] - 0s 83us/step - loss: 0.0024 - acc: 0.9991 - val_loss: 0.0106 - val_acc: 0.9964\n",
      "Epoch 165/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0025 - acc: 0.9982 - val_loss: 0.0101 - val_acc: 0.9969\n",
      "Epoch 166/200\n",
      "1500/1500 [==============================] - 0s 95us/step - loss: 0.0022 - acc: 0.9991 - val_loss: 0.0172 - val_acc: 0.9942\n",
      "Epoch 167/200\n",
      "1500/1500 [==============================] - 0s 104us/step - loss: 0.0061 - acc: 0.9969 - val_loss: 0.0167 - val_acc: 0.9953\n",
      "Epoch 168/200\n",
      "1500/1500 [==============================] - 0s 94us/step - loss: 0.0038 - acc: 0.9987 - val_loss: 0.0087 - val_acc: 0.9960\n",
      "Epoch 169/200\n",
      "1500/1500 [==============================] - 0s 96us/step - loss: 0.0052 - acc: 0.9991 - val_loss: 0.0352 - val_acc: 0.9907\n",
      "Epoch 170/200\n",
      "1500/1500 [==============================] - 0s 102us/step - loss: 0.0226 - acc: 0.9924 - val_loss: 0.0147 - val_acc: 0.9973\n",
      "Epoch 171/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0116 - acc: 0.9956 - val_loss: 0.0213 - val_acc: 0.9947\n",
      "Epoch 172/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0086 - acc: 0.9973 - val_loss: 0.0147 - val_acc: 0.9969\n",
      "Epoch 173/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0076 - acc: 0.9973 - val_loss: 0.0132 - val_acc: 0.9956\n",
      "Epoch 174/200\n",
      "1500/1500 [==============================] - 0s 82us/step - loss: 0.0152 - acc: 0.9956 - val_loss: 0.0170 - val_acc: 0.9951\n",
      "Epoch 175/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0050 - acc: 0.9987 - val_loss: 0.0183 - val_acc: 0.9960\n",
      "Epoch 176/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0033 - acc: 0.9982 - val_loss: 0.0141 - val_acc: 0.9960\n",
      "Epoch 177/200\n",
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0059 - acc: 0.9973 - val_loss: 0.0150 - val_acc: 0.9969\n",
      "Epoch 178/200\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0034 - acc: 0.9987 - val_loss: 0.0162 - val_acc: 0.9964\n",
      "Epoch 179/200\n",
      "1500/1500 [==============================] - 0s 79us/step - loss: 0.0025 - acc: 0.9987 - val_loss: 0.0139 - val_acc: 0.9967\n",
      "Epoch 180/200\n",
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0016 - acc: 0.9991 - val_loss: 0.0147 - val_acc: 0.9971\n",
      "Epoch 181/200\n",
      "1500/1500 [==============================] - 0s 80us/step - loss: 0.0025 - acc: 0.9991 - val_loss: 0.0130 - val_acc: 0.9969\n",
      "Epoch 182/200\n",
      "1500/1500 [==============================] - 0s 104us/step - loss: 0.0028 - acc: 0.9982 - val_loss: 0.0141 - val_acc: 0.9969\n",
      "Epoch 183/200\n",
      "1500/1500 [==============================] - 0s 102us/step - loss: 0.0019 - acc: 0.9991 - val_loss: 0.0135 - val_acc: 0.9973\n",
      "Epoch 184/200\n",
      "1500/1500 [==============================] - 0s 91us/step - loss: 0.0044 - acc: 0.9987 - val_loss: 0.0123 - val_acc: 0.9969\n",
      "Epoch 185/200\n",
      "1500/1500 [==============================] - 0s 101us/step - loss: 0.0039 - acc: 0.9982 - val_loss: 0.0156 - val_acc: 0.9964\n",
      "Epoch 186/200\n",
      "1500/1500 [==============================] - 0s 101us/step - loss: 0.0037 - acc: 0.9982 - val_loss: 0.0108 - val_acc: 0.9978\n",
      "Epoch 187/200\n",
      "1500/1500 [==============================] - 0s 99us/step - loss: 0.0043 - acc: 0.9980 - val_loss: 0.0157 - val_acc: 0.9949\n",
      "Epoch 188/200\n",
      "1500/1500 [==============================] - 0s 100us/step - loss: 0.0039 - acc: 0.9980 - val_loss: 0.0134 - val_acc: 0.9960\n",
      "Epoch 189/200\n",
      "1500/1500 [==============================] - 0s 99us/step - loss: 0.0020 - acc: 0.9996 - val_loss: 0.0120 - val_acc: 0.9973\n",
      "Epoch 190/200\n",
      "1500/1500 [==============================] - 0s 98us/step - loss: 0.0037 - acc: 0.9982 - val_loss: 0.0167 - val_acc: 0.9951\n",
      "Epoch 191/200\n",
      "1500/1500 [==============================] - 0s 91us/step - loss: 0.0035 - acc: 0.9987 - val_loss: 0.0255 - val_acc: 0.9956\n",
      "Epoch 192/200\n",
      "1500/1500 [==============================] - 0s 114us/step - loss: 0.0026 - acc: 0.9987 - val_loss: 0.0115 - val_acc: 0.9973\n",
      "Epoch 193/200\n",
      "1500/1500 [==============================] - 0s 94us/step - loss: 0.0018 - acc: 0.9989 - val_loss: 0.0115 - val_acc: 0.9973\n",
      "Epoch 194/200\n",
      "1500/1500 [==============================] - 0s 82us/step - loss: 0.0051 - acc: 0.9978 - val_loss: 0.0101 - val_acc: 0.9960\n",
      "Epoch 195/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0084 - acc: 0.9969 - val_loss: 0.0202 - val_acc: 0.9933\n",
      "Epoch 196/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0052 - acc: 0.9980 - val_loss: 0.0153 - val_acc: 0.9960\n",
      "Epoch 197/200\n",
      "1500/1500 [==============================] - 0s 83us/step - loss: 0.0023 - acc: 0.9987 - val_loss: 0.0127 - val_acc: 0.9969\n",
      "Epoch 198/200\n",
      "1500/1500 [==============================] - 0s 78us/step - loss: 0.0020 - acc: 0.9991 - val_loss: 0.0121 - val_acc: 0.9973\n",
      "Epoch 199/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0020 - acc: 0.9987 - val_loss: 0.0145 - val_acc: 0.9969\n",
      "Epoch 200/200\n",
      "1500/1500 [==============================] - 0s 81us/step - loss: 0.0030 - acc: 0.9982 - val_loss: 0.0123 - val_acc: 0.9973\n"
     ]
    }
   ],
   "source": [
    "classifier = Classifier()\n",
    "\n",
    "classifier.train()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "window.mpl = {};\n",
       "\n",
       "\n",
       "mpl.get_websocket_type = function() {\n",
       "    if (typeof(WebSocket) !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert('Your browser does not have WebSocket support.' +\n",
       "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "              'Firefox 4 and 5 are also supported but you ' +\n",
       "              'have to enable WebSockets in about:config.');\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = (this.ws.binaryType != undefined);\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById(\"mpl-warnings\");\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent = (\n",
       "                \"This browser does not support binary websocket messages. \" +\n",
       "                    \"Performance may be slow.\");\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = $('<div/>');\n",
       "    this._root_extra_style(this.root)\n",
       "    this.root.attr('style', 'display: inline-block');\n",
       "\n",
       "    $(parent_element).append(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen =  function () {\n",
       "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
       "            fig.send_message(\"send_image_mode\", {});\n",
       "            if (mpl.ratio != 1) {\n",
       "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
       "            }\n",
       "            fig.send_message(\"refresh\", {});\n",
       "        }\n",
       "\n",
       "    this.imageObj.onload = function() {\n",
       "            if (fig.image_mode == 'full') {\n",
       "                // Full images could contain transparency (where diff images\n",
       "                // almost always do), so we need to clear the canvas so that\n",
       "                // there is no ghosting.\n",
       "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "            }\n",
       "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "        };\n",
       "\n",
       "    this.imageObj.onunload = function() {\n",
       "        fig.ws.close();\n",
       "    }\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_header = function() {\n",
       "    var titlebar = $(\n",
       "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
       "        'ui-helper-clearfix\"/>');\n",
       "    var titletext = $(\n",
       "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
       "        'text-align: center; padding: 3px;\"/>');\n",
       "    titlebar.append(titletext)\n",
       "    this.root.append(titlebar);\n",
       "    this.header = titletext[0];\n",
       "}\n",
       "\n",
       "\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = $('<div/>');\n",
       "\n",
       "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
       "\n",
       "    function canvas_keyboard_event(event) {\n",
       "        return fig.key_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
       "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
       "    this.canvas_div = canvas_div\n",
       "    this._canvas_extra_style(canvas_div)\n",
       "    this.root.append(canvas_div);\n",
       "\n",
       "    var canvas = $('<canvas/>');\n",
       "    canvas.addClass('mpl-canvas');\n",
       "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
       "\n",
       "    this.canvas = canvas[0];\n",
       "    this.context = canvas[0].getContext(\"2d\");\n",
       "\n",
       "    var backingStore = this.context.backingStorePixelRatio ||\n",
       "\tthis.context.webkitBackingStorePixelRatio ||\n",
       "\tthis.context.mozBackingStorePixelRatio ||\n",
       "\tthis.context.msBackingStorePixelRatio ||\n",
       "\tthis.context.oBackingStorePixelRatio ||\n",
       "\tthis.context.backingStorePixelRatio || 1;\n",
       "\n",
       "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband = $('<canvas/>');\n",
       "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
       "\n",
       "    var pass_mouse_events = true;\n",
       "\n",
       "    canvas_div.resizable({\n",
       "        start: function(event, ui) {\n",
       "            pass_mouse_events = false;\n",
       "        },\n",
       "        resize: function(event, ui) {\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "        stop: function(event, ui) {\n",
       "            pass_mouse_events = true;\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "    });\n",
       "\n",
       "    function mouse_event_fn(event) {\n",
       "        if (pass_mouse_events)\n",
       "            return fig.mouse_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    rubberband.mousedown('button_press', mouse_event_fn);\n",
       "    rubberband.mouseup('button_release', mouse_event_fn);\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
       "\n",
       "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
       "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
       "\n",
       "    canvas_div.on(\"wheel\", function (event) {\n",
       "        event = event.originalEvent;\n",
       "        event['data'] = 'scroll'\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        mouse_event_fn(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.append(canvas);\n",
       "    canvas_div.append(rubberband);\n",
       "\n",
       "    this.rubberband = rubberband;\n",
       "    this.rubberband_canvas = rubberband[0];\n",
       "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
       "    this.rubberband_context.strokeStyle = \"#000000\";\n",
       "\n",
       "    this._resize_canvas = function(width, height) {\n",
       "        // Keep the size of the canvas, canvas container, and rubber band\n",
       "        // canvas in synch.\n",
       "        canvas_div.css('width', width)\n",
       "        canvas_div.css('height', height)\n",
       "\n",
       "        canvas.attr('width', width * mpl.ratio);\n",
       "        canvas.attr('height', height * mpl.ratio);\n",
       "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
       "\n",
       "        rubberband.attr('width', width);\n",
       "        rubberband.attr('height', height);\n",
       "    }\n",
       "\n",
       "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
       "    // upon first draw.\n",
       "    this._resize_canvas(600, 600);\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus () {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            // put a spacer in here.\n",
       "            continue;\n",
       "        }\n",
       "        var button = $('<button/>');\n",
       "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
       "                        'ui-button-icon-only');\n",
       "        button.attr('role', 'button');\n",
       "        button.attr('aria-disabled', 'false');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "\n",
       "        var icon_img = $('<span/>');\n",
       "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
       "        icon_img.addClass(image);\n",
       "        icon_img.addClass('ui-corner-all');\n",
       "\n",
       "        var tooltip_span = $('<span/>');\n",
       "        tooltip_span.addClass('ui-button-text');\n",
       "        tooltip_span.html(tooltip);\n",
       "\n",
       "        button.append(icon_img);\n",
       "        button.append(tooltip_span);\n",
       "\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    var fmt_picker_span = $('<span/>');\n",
       "\n",
       "    var fmt_picker = $('<select/>');\n",
       "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
       "    fmt_picker_span.append(fmt_picker);\n",
       "    nav_element.append(fmt_picker_span);\n",
       "    this.format_dropdown = fmt_picker[0];\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = $(\n",
       "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
       "        fmt_picker.append(option)\n",
       "    }\n",
       "\n",
       "    // Add hover states to the ui-buttons\n",
       "    $( \".ui-button\" ).hover(\n",
       "        function() { $(this).addClass(\"ui-state-hover\");},\n",
       "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
       "    );\n",
       "\n",
       "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_message = function(type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function() {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
       "    }\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1]);\n",
       "        fig.send_message(\"refresh\", {});\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
       "    var x0 = msg['x0'] / mpl.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
       "    var x1 = msg['x1'] / mpl.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0, 0, fig.canvas.width, fig.canvas.height);\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch(cursor)\n",
       "    {\n",
       "    case 0:\n",
       "        cursor = 'pointer';\n",
       "        break;\n",
       "    case 1:\n",
       "        cursor = 'default';\n",
       "        break;\n",
       "    case 2:\n",
       "        cursor = 'crosshair';\n",
       "        break;\n",
       "    case 3:\n",
       "        cursor = 'move';\n",
       "        break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message(\"ack\", {});\n",
       "}\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = \"image/png\";\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src);\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data);\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig[\"handle_\" + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "}\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function(e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e)\n",
       "        e = window.event;\n",
       "    if (e.target)\n",
       "        targ = e.target;\n",
       "    else if (e.srcElement)\n",
       "        targ = e.srcElement;\n",
       "    if (targ.nodeType == 3) // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "\n",
       "    // jQuery normalizes the pageX and pageY\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    // offset() returns the position of the element relative to the document\n",
       "    var x = e.pageX - $(targ).offset().left;\n",
       "    var y = e.pageY - $(targ).offset().top;\n",
       "\n",
       "    return {\"x\": x, \"y\": y};\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys (original) {\n",
       "  return Object.keys(original).reduce(function (obj, key) {\n",
       "    if (typeof original[key] !== 'object')\n",
       "        obj[key] = original[key]\n",
       "    return obj;\n",
       "  }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function(event, name) {\n",
       "    var canvas_pos = mpl.findpos(event)\n",
       "\n",
       "    if (name === 'button_press')\n",
       "    {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * mpl.ratio;\n",
       "    var y = canvas_pos.y * mpl.ratio;\n",
       "\n",
       "    this.send_message(name, {x: x, y: y, button: event.button,\n",
       "                             step: event.step,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.key_event = function(event, name) {\n",
       "\n",
       "    // Prevent repeat events\n",
       "    if (name == 'key_press')\n",
       "    {\n",
       "        if (event.which === this._key)\n",
       "            return;\n",
       "        else\n",
       "            this._key = event.which;\n",
       "    }\n",
       "    if (name == 'key_release')\n",
       "        this._key = null;\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which != 17)\n",
       "        value += \"ctrl+\";\n",
       "    if (event.altKey && event.which != 18)\n",
       "        value += \"alt+\";\n",
       "    if (event.shiftKey && event.which != 16)\n",
       "        value += \"shift+\";\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, {key: value,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
       "    if (name == 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message(\"toolbar_button\", {name: name});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function() {\n",
       "        comm.close()\n",
       "    };\n",
       "    ws.send = function(m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function(msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data'])\n",
       "    });\n",
       "    return ws;\n",
       "}\n",
       "\n",
       "mpl.mpl_figure_comm = function(comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = $(\"#\" + id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm)\n",
       "\n",
       "    function ondownload(figure, format) {\n",
       "        window.open(figure.imageObj.src);\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy,\n",
       "                           ondownload,\n",
       "                           element.get(0));\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element.get(0);\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error(\"Failed to find cell for figure\", id, fig);\n",
       "        return;\n",
       "    }\n",
       "\n",
       "    var output_index = fig.cell_info[2]\n",
       "    var cell = fig.cell_info[0];\n",
       "\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
       "    var width = fig.canvas.width/mpl.ratio\n",
       "    fig.root.unbind('remove')\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable()\n",
       "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
       "    fig.close_ws(fig, msg);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.close_ws = function(fig, msg){\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width/mpl.ratio\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message(\"ack\", {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items){\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) { continue; };\n",
       "\n",
       "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
       "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
       "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
       "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
       "    buttongrp.append(button);\n",
       "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
       "    titlebar.prepend(buttongrp);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(el){\n",
       "    var fig = this\n",
       "    el.on(\"remove\", function(){\n",
       "\tfig.close_ws(fig, {});\n",
       "    });\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(el){\n",
       "    // this is important to make the div 'focusable\n",
       "    el.attr('tabindex', 0)\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    }\n",
       "    else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager)\n",
       "        manager = IPython.keyboard_manager;\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which == 13) {\n",
       "        this.canvas_div.blur();\n",
       "        event.shiftKey = false;\n",
       "        // Send a \"J\" for go to next cell\n",
       "        event.which = 74;\n",
       "        event.keyCode = 74;\n",
       "        manager.command_mode();\n",
       "        manager.handle_keydown(event);\n",
       "    }\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.find_output_cell = function(html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i=0; i<ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code'){\n",
       "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] == html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "}\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel != null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"638.888905813665\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Plot of input data\n"
     ]
    }
   ],
   "source": [
    "# Plot input data\n",
    "plt.figure()\n",
    "plt.scatter(classifier.data[:, 0], classifier.data[:, 1], s=4, c=classifier.labels.squeeze())\n",
    "plt.show()\n",
    "print(\"Plot of input data\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(200, 1)\n",
      "(400, 1)\n",
      "(600, 1)\n",
      "(300, 2)\n"
     ]
    }
   ],
   "source": [
    "# Create image for paper\n",
    "classifier.n = 200\n",
    "data, labels = classifier.get_data()\n",
    "labels = labels.squeeze()\n",
    "ind_0 = np.where(labels == 0)[0]\n",
    "\n",
    "ind = np.random.randint(0, ind_0.shape[0], 50)\n",
    "ind_0 = ind_0[ind]\n",
    "\n",
    "ind_1 = np.where(labels == 1)[0]\n",
    "\n",
    "ind = np.random.randint(0, ind_1.shape[0], 50)\n",
    "ind_1 = ind_1[ind]\n",
    "\n",
    "ind_2 = np.where(labels == 2)[0]\n",
    "\n",
    "ind = np.concatenate((ind_0, ind_1, ind_2), axis=0)\n",
    "data = data[ind, :]\n",
    "labels = labels[ind]\n",
    "\n",
    "print(data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "window.mpl = {};\n",
       "\n",
       "\n",
       "mpl.get_websocket_type = function() {\n",
       "    if (typeof(WebSocket) !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert('Your browser does not have WebSocket support.' +\n",
       "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "              'Firefox 4 and 5 are also supported but you ' +\n",
       "              'have to enable WebSockets in about:config.');\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = (this.ws.binaryType != undefined);\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById(\"mpl-warnings\");\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent = (\n",
       "                \"This browser does not support binary websocket messages. \" +\n",
       "                    \"Performance may be slow.\");\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = $('<div/>');\n",
       "    this._root_extra_style(this.root)\n",
       "    this.root.attr('style', 'display: inline-block');\n",
       "\n",
       "    $(parent_element).append(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen =  function () {\n",
       "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
       "            fig.send_message(\"send_image_mode\", {});\n",
       "            if (mpl.ratio != 1) {\n",
       "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
       "            }\n",
       "            fig.send_message(\"refresh\", {});\n",
       "        }\n",
       "\n",
       "    this.imageObj.onload = function() {\n",
       "            if (fig.image_mode == 'full') {\n",
       "                // Full images could contain transparency (where diff images\n",
       "                // almost always do), so we need to clear the canvas so that\n",
       "                // there is no ghosting.\n",
       "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "            }\n",
       "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "        };\n",
       "\n",
       "    this.imageObj.onunload = function() {\n",
       "        fig.ws.close();\n",
       "    }\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_header = function() {\n",
       "    var titlebar = $(\n",
       "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
       "        'ui-helper-clearfix\"/>');\n",
       "    var titletext = $(\n",
       "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
       "        'text-align: center; padding: 3px;\"/>');\n",
       "    titlebar.append(titletext)\n",
       "    this.root.append(titlebar);\n",
       "    this.header = titletext[0];\n",
       "}\n",
       "\n",
       "\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = $('<div/>');\n",
       "\n",
       "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
       "\n",
       "    function canvas_keyboard_event(event) {\n",
       "        return fig.key_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
       "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
       "    this.canvas_div = canvas_div\n",
       "    this._canvas_extra_style(canvas_div)\n",
       "    this.root.append(canvas_div);\n",
       "\n",
       "    var canvas = $('<canvas/>');\n",
       "    canvas.addClass('mpl-canvas');\n",
       "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
       "\n",
       "    this.canvas = canvas[0];\n",
       "    this.context = canvas[0].getContext(\"2d\");\n",
       "\n",
       "    var backingStore = this.context.backingStorePixelRatio ||\n",
       "\tthis.context.webkitBackingStorePixelRatio ||\n",
       "\tthis.context.mozBackingStorePixelRatio ||\n",
       "\tthis.context.msBackingStorePixelRatio ||\n",
       "\tthis.context.oBackingStorePixelRatio ||\n",
       "\tthis.context.backingStorePixelRatio || 1;\n",
       "\n",
       "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband = $('<canvas/>');\n",
       "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
       "\n",
       "    var pass_mouse_events = true;\n",
       "\n",
       "    canvas_div.resizable({\n",
       "        start: function(event, ui) {\n",
       "            pass_mouse_events = false;\n",
       "        },\n",
       "        resize: function(event, ui) {\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "        stop: function(event, ui) {\n",
       "            pass_mouse_events = true;\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "    });\n",
       "\n",
       "    function mouse_event_fn(event) {\n",
       "        if (pass_mouse_events)\n",
       "            return fig.mouse_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    rubberband.mousedown('button_press', mouse_event_fn);\n",
       "    rubberband.mouseup('button_release', mouse_event_fn);\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
       "\n",
       "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
       "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
       "\n",
       "    canvas_div.on(\"wheel\", function (event) {\n",
       "        event = event.originalEvent;\n",
       "        event['data'] = 'scroll'\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        mouse_event_fn(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.append(canvas);\n",
       "    canvas_div.append(rubberband);\n",
       "\n",
       "    this.rubberband = rubberband;\n",
       "    this.rubberband_canvas = rubberband[0];\n",
       "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
       "    this.rubberband_context.strokeStyle = \"#000000\";\n",
       "\n",
       "    this._resize_canvas = function(width, height) {\n",
       "        // Keep the size of the canvas, canvas container, and rubber band\n",
       "        // canvas in synch.\n",
       "        canvas_div.css('width', width)\n",
       "        canvas_div.css('height', height)\n",
       "\n",
       "        canvas.attr('width', width * mpl.ratio);\n",
       "        canvas.attr('height', height * mpl.ratio);\n",
       "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
       "\n",
       "        rubberband.attr('width', width);\n",
       "        rubberband.attr('height', height);\n",
       "    }\n",
       "\n",
       "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
       "    // upon first draw.\n",
       "    this._resize_canvas(600, 600);\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus () {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            // put a spacer in here.\n",
       "            continue;\n",
       "        }\n",
       "        var button = $('<button/>');\n",
       "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
       "                        'ui-button-icon-only');\n",
       "        button.attr('role', 'button');\n",
       "        button.attr('aria-disabled', 'false');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "\n",
       "        var icon_img = $('<span/>');\n",
       "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
       "        icon_img.addClass(image);\n",
       "        icon_img.addClass('ui-corner-all');\n",
       "\n",
       "        var tooltip_span = $('<span/>');\n",
       "        tooltip_span.addClass('ui-button-text');\n",
       "        tooltip_span.html(tooltip);\n",
       "\n",
       "        button.append(icon_img);\n",
       "        button.append(tooltip_span);\n",
       "\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    var fmt_picker_span = $('<span/>');\n",
       "\n",
       "    var fmt_picker = $('<select/>');\n",
       "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
       "    fmt_picker_span.append(fmt_picker);\n",
       "    nav_element.append(fmt_picker_span);\n",
       "    this.format_dropdown = fmt_picker[0];\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = $(\n",
       "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
       "        fmt_picker.append(option)\n",
       "    }\n",
       "\n",
       "    // Add hover states to the ui-buttons\n",
       "    $( \".ui-button\" ).hover(\n",
       "        function() { $(this).addClass(\"ui-state-hover\");},\n",
       "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
       "    );\n",
       "\n",
       "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_message = function(type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function() {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
       "    }\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1]);\n",
       "        fig.send_message(\"refresh\", {});\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
       "    var x0 = msg['x0'] / mpl.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
       "    var x1 = msg['x1'] / mpl.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0, 0, fig.canvas.width, fig.canvas.height);\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch(cursor)\n",
       "    {\n",
       "    case 0:\n",
       "        cursor = 'pointer';\n",
       "        break;\n",
       "    case 1:\n",
       "        cursor = 'default';\n",
       "        break;\n",
       "    case 2:\n",
       "        cursor = 'crosshair';\n",
       "        break;\n",
       "    case 3:\n",
       "        cursor = 'move';\n",
       "        break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message(\"ack\", {});\n",
       "}\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = \"image/png\";\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src);\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data);\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig[\"handle_\" + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "}\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function(e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e)\n",
       "        e = window.event;\n",
       "    if (e.target)\n",
       "        targ = e.target;\n",
       "    else if (e.srcElement)\n",
       "        targ = e.srcElement;\n",
       "    if (targ.nodeType == 3) // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "\n",
       "    // jQuery normalizes the pageX and pageY\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    // offset() returns the position of the element relative to the document\n",
       "    var x = e.pageX - $(targ).offset().left;\n",
       "    var y = e.pageY - $(targ).offset().top;\n",
       "\n",
       "    return {\"x\": x, \"y\": y};\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys (original) {\n",
       "  return Object.keys(original).reduce(function (obj, key) {\n",
       "    if (typeof original[key] !== 'object')\n",
       "        obj[key] = original[key]\n",
       "    return obj;\n",
       "  }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function(event, name) {\n",
       "    var canvas_pos = mpl.findpos(event)\n",
       "\n",
       "    if (name === 'button_press')\n",
       "    {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * mpl.ratio;\n",
       "    var y = canvas_pos.y * mpl.ratio;\n",
       "\n",
       "    this.send_message(name, {x: x, y: y, button: event.button,\n",
       "                             step: event.step,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.key_event = function(event, name) {\n",
       "\n",
       "    // Prevent repeat events\n",
       "    if (name == 'key_press')\n",
       "    {\n",
       "        if (event.which === this._key)\n",
       "            return;\n",
       "        else\n",
       "            this._key = event.which;\n",
       "    }\n",
       "    if (name == 'key_release')\n",
       "        this._key = null;\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which != 17)\n",
       "        value += \"ctrl+\";\n",
       "    if (event.altKey && event.which != 18)\n",
       "        value += \"alt+\";\n",
       "    if (event.shiftKey && event.which != 16)\n",
       "        value += \"shift+\";\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, {key: value,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
       "    if (name == 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message(\"toolbar_button\", {name: name});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function() {\n",
       "        comm.close()\n",
       "    };\n",
       "    ws.send = function(m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function(msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data'])\n",
       "    });\n",
       "    return ws;\n",
       "}\n",
       "\n",
       "mpl.mpl_figure_comm = function(comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = $(\"#\" + id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm)\n",
       "\n",
       "    function ondownload(figure, format) {\n",
       "        window.open(figure.imageObj.src);\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy,\n",
       "                           ondownload,\n",
       "                           element.get(0));\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element.get(0);\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error(\"Failed to find cell for figure\", id, fig);\n",
       "        return;\n",
       "    }\n",
       "\n",
       "    var output_index = fig.cell_info[2]\n",
       "    var cell = fig.cell_info[0];\n",
       "\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
       "    var width = fig.canvas.width/mpl.ratio\n",
       "    fig.root.unbind('remove')\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable()\n",
       "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
       "    fig.close_ws(fig, msg);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.close_ws = function(fig, msg){\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width/mpl.ratio\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message(\"ack\", {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items){\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) { continue; };\n",
       "\n",
       "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
       "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
       "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
       "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
       "    buttongrp.append(button);\n",
       "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
       "    titlebar.prepend(buttongrp);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(el){\n",
       "    var fig = this\n",
       "    el.on(\"remove\", function(){\n",
       "\tfig.close_ws(fig, {});\n",
       "    });\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(el){\n",
       "    // this is important to make the div 'focusable\n",
       "    el.attr('tabindex', 0)\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    }\n",
       "    else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager)\n",
       "        manager = IPython.keyboard_manager;\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which == 13) {\n",
       "        this.canvas_div.blur();\n",
       "        event.shiftKey = false;\n",
       "        // Send a \"J\" for go to next cell\n",
       "        event.which = 74;\n",
       "        event.keyCode = 74;\n",
       "        manager.command_mode();\n",
       "        manager.handle_keydown(event);\n",
       "    }\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.find_output_cell = function(html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i=0; i<ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code'){\n",
       "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] == html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "}\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel != null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"498.88890210493145\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(5,5))\n",
    "ind = np.where(labels == 0)[0]\n",
    "plt.scatter(data[ind, 0], data[ind, 1], s=10, c='#FFA500', marker='D')\n",
    "ind = np.where(labels == 1)[0]\n",
    "plt.scatter(data[ind, 0], data[ind, 1], s=10, c='b', marker='o')\n",
    "ind = np.where(labels == 2)[0]\n",
    "plt.scatter(data[ind, 0], data[ind, 1], s=10, c='#FFFF33', marker='*')\n",
    "#plt.xlim(-18, +18)\n",
    "#plt.ylim(-18, +18)\n",
    "#plt.gca().set_aspect('equal', adjustable='box')\n",
    "plt.legend(('$class\\ 0$','$class\\ 1$','$class\\ 2\\ (OOD\\ sample)$'), fontsize='x-small');\n",
    "plt.show()\n",
    "plt.savefig(\"exp2_a\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "window.mpl = {};\n",
       "\n",
       "\n",
       "mpl.get_websocket_type = function() {\n",
       "    if (typeof(WebSocket) !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert('Your browser does not have WebSocket support.' +\n",
       "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "              'Firefox 4 and 5 are also supported but you ' +\n",
       "              'have to enable WebSockets in about:config.');\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = (this.ws.binaryType != undefined);\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById(\"mpl-warnings\");\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent = (\n",
       "                \"This browser does not support binary websocket messages. \" +\n",
       "                    \"Performance may be slow.\");\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = $('<div/>');\n",
       "    this._root_extra_style(this.root)\n",
       "    this.root.attr('style', 'display: inline-block');\n",
       "\n",
       "    $(parent_element).append(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen =  function () {\n",
       "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
       "            fig.send_message(\"send_image_mode\", {});\n",
       "            if (mpl.ratio != 1) {\n",
       "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
       "            }\n",
       "            fig.send_message(\"refresh\", {});\n",
       "        }\n",
       "\n",
       "    this.imageObj.onload = function() {\n",
       "            if (fig.image_mode == 'full') {\n",
       "                // Full images could contain transparency (where diff images\n",
       "                // almost always do), so we need to clear the canvas so that\n",
       "                // there is no ghosting.\n",
       "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "            }\n",
       "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "        };\n",
       "\n",
       "    this.imageObj.onunload = function() {\n",
       "        fig.ws.close();\n",
       "    }\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_header = function() {\n",
       "    var titlebar = $(\n",
       "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
       "        'ui-helper-clearfix\"/>');\n",
       "    var titletext = $(\n",
       "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
       "        'text-align: center; padding: 3px;\"/>');\n",
       "    titlebar.append(titletext)\n",
       "    this.root.append(titlebar);\n",
       "    this.header = titletext[0];\n",
       "}\n",
       "\n",
       "\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = $('<div/>');\n",
       "\n",
       "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
       "\n",
       "    function canvas_keyboard_event(event) {\n",
       "        return fig.key_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
       "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
       "    this.canvas_div = canvas_div\n",
       "    this._canvas_extra_style(canvas_div)\n",
       "    this.root.append(canvas_div);\n",
       "\n",
       "    var canvas = $('<canvas/>');\n",
       "    canvas.addClass('mpl-canvas');\n",
       "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
       "\n",
       "    this.canvas = canvas[0];\n",
       "    this.context = canvas[0].getContext(\"2d\");\n",
       "\n",
       "    var backingStore = this.context.backingStorePixelRatio ||\n",
       "\tthis.context.webkitBackingStorePixelRatio ||\n",
       "\tthis.context.mozBackingStorePixelRatio ||\n",
       "\tthis.context.msBackingStorePixelRatio ||\n",
       "\tthis.context.oBackingStorePixelRatio ||\n",
       "\tthis.context.backingStorePixelRatio || 1;\n",
       "\n",
       "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband = $('<canvas/>');\n",
       "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
       "\n",
       "    var pass_mouse_events = true;\n",
       "\n",
       "    canvas_div.resizable({\n",
       "        start: function(event, ui) {\n",
       "            pass_mouse_events = false;\n",
       "        },\n",
       "        resize: function(event, ui) {\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "        stop: function(event, ui) {\n",
       "            pass_mouse_events = true;\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "    });\n",
       "\n",
       "    function mouse_event_fn(event) {\n",
       "        if (pass_mouse_events)\n",
       "            return fig.mouse_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    rubberband.mousedown('button_press', mouse_event_fn);\n",
       "    rubberband.mouseup('button_release', mouse_event_fn);\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
       "\n",
       "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
       "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
       "\n",
       "    canvas_div.on(\"wheel\", function (event) {\n",
       "        event = event.originalEvent;\n",
       "        event['data'] = 'scroll'\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        mouse_event_fn(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.append(canvas);\n",
       "    canvas_div.append(rubberband);\n",
       "\n",
       "    this.rubberband = rubberband;\n",
       "    this.rubberband_canvas = rubberband[0];\n",
       "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
       "    this.rubberband_context.strokeStyle = \"#000000\";\n",
       "\n",
       "    this._resize_canvas = function(width, height) {\n",
       "        // Keep the size of the canvas, canvas container, and rubber band\n",
       "        // canvas in synch.\n",
       "        canvas_div.css('width', width)\n",
       "        canvas_div.css('height', height)\n",
       "\n",
       "        canvas.attr('width', width * mpl.ratio);\n",
       "        canvas.attr('height', height * mpl.ratio);\n",
       "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
       "\n",
       "        rubberband.attr('width', width);\n",
       "        rubberband.attr('height', height);\n",
       "    }\n",
       "\n",
       "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
       "    // upon first draw.\n",
       "    this._resize_canvas(600, 600);\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus () {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            // put a spacer in here.\n",
       "            continue;\n",
       "        }\n",
       "        var button = $('<button/>');\n",
       "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
       "                        'ui-button-icon-only');\n",
       "        button.attr('role', 'button');\n",
       "        button.attr('aria-disabled', 'false');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "\n",
       "        var icon_img = $('<span/>');\n",
       "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
       "        icon_img.addClass(image);\n",
       "        icon_img.addClass('ui-corner-all');\n",
       "\n",
       "        var tooltip_span = $('<span/>');\n",
       "        tooltip_span.addClass('ui-button-text');\n",
       "        tooltip_span.html(tooltip);\n",
       "\n",
       "        button.append(icon_img);\n",
       "        button.append(tooltip_span);\n",
       "\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    var fmt_picker_span = $('<span/>');\n",
       "\n",
       "    var fmt_picker = $('<select/>');\n",
       "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
       "    fmt_picker_span.append(fmt_picker);\n",
       "    nav_element.append(fmt_picker_span);\n",
       "    this.format_dropdown = fmt_picker[0];\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = $(\n",
       "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
       "        fmt_picker.append(option)\n",
       "    }\n",
       "\n",
       "    // Add hover states to the ui-buttons\n",
       "    $( \".ui-button\" ).hover(\n",
       "        function() { $(this).addClass(\"ui-state-hover\");},\n",
       "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
       "    );\n",
       "\n",
       "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_message = function(type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function() {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
       "    }\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1]);\n",
       "        fig.send_message(\"refresh\", {});\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
       "    var x0 = msg['x0'] / mpl.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
       "    var x1 = msg['x1'] / mpl.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0, 0, fig.canvas.width, fig.canvas.height);\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch(cursor)\n",
       "    {\n",
       "    case 0:\n",
       "        cursor = 'pointer';\n",
       "        break;\n",
       "    case 1:\n",
       "        cursor = 'default';\n",
       "        break;\n",
       "    case 2:\n",
       "        cursor = 'crosshair';\n",
       "        break;\n",
       "    case 3:\n",
       "        cursor = 'move';\n",
       "        break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message(\"ack\", {});\n",
       "}\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = \"image/png\";\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src);\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data);\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig[\"handle_\" + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "}\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function(e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e)\n",
       "        e = window.event;\n",
       "    if (e.target)\n",
       "        targ = e.target;\n",
       "    else if (e.srcElement)\n",
       "        targ = e.srcElement;\n",
       "    if (targ.nodeType == 3) // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "\n",
       "    // jQuery normalizes the pageX and pageY\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    // offset() returns the position of the element relative to the document\n",
       "    var x = e.pageX - $(targ).offset().left;\n",
       "    var y = e.pageY - $(targ).offset().top;\n",
       "\n",
       "    return {\"x\": x, \"y\": y};\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys (original) {\n",
       "  return Object.keys(original).reduce(function (obj, key) {\n",
       "    if (typeof original[key] !== 'object')\n",
       "        obj[key] = original[key]\n",
       "    return obj;\n",
       "  }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function(event, name) {\n",
       "    var canvas_pos = mpl.findpos(event)\n",
       "\n",
       "    if (name === 'button_press')\n",
       "    {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * mpl.ratio;\n",
       "    var y = canvas_pos.y * mpl.ratio;\n",
       "\n",
       "    this.send_message(name, {x: x, y: y, button: event.button,\n",
       "                             step: event.step,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.key_event = function(event, name) {\n",
       "\n",
       "    // Prevent repeat events\n",
       "    if (name == 'key_press')\n",
       "    {\n",
       "        if (event.which === this._key)\n",
       "            return;\n",
       "        else\n",
       "            this._key = event.which;\n",
       "    }\n",
       "    if (name == 'key_release')\n",
       "        this._key = null;\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which != 17)\n",
       "        value += \"ctrl+\";\n",
       "    if (event.altKey && event.which != 18)\n",
       "        value += \"alt+\";\n",
       "    if (event.shiftKey && event.which != 16)\n",
       "        value += \"shift+\";\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, {key: value,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
       "    if (name == 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message(\"toolbar_button\", {name: name});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function() {\n",
       "        comm.close()\n",
       "    };\n",
       "    ws.send = function(m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function(msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data'])\n",
       "    });\n",
       "    return ws;\n",
       "}\n",
       "\n",
       "mpl.mpl_figure_comm = function(comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = $(\"#\" + id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm)\n",
       "\n",
       "    function ondownload(figure, format) {\n",
       "        window.open(figure.imageObj.src);\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy,\n",
       "                           ondownload,\n",
       "                           element.get(0));\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element.get(0);\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error(\"Failed to find cell for figure\", id, fig);\n",
       "        return;\n",
       "    }\n",
       "\n",
       "    var output_index = fig.cell_info[2]\n",
       "    var cell = fig.cell_info[0];\n",
       "\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
       "    var width = fig.canvas.width/mpl.ratio\n",
       "    fig.root.unbind('remove')\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable()\n",
       "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
       "    fig.close_ws(fig, msg);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.close_ws = function(fig, msg){\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width/mpl.ratio\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message(\"ack\", {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items){\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) { continue; };\n",
       "\n",
       "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
       "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
       "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
       "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
       "    buttongrp.append(button);\n",
       "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
       "    titlebar.prepend(buttongrp);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(el){\n",
       "    var fig = this\n",
       "    el.on(\"remove\", function(){\n",
       "\tfig.close_ws(fig, {});\n",
       "    });\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(el){\n",
       "    // this is important to make the div 'focusable\n",
       "    el.attr('tabindex', 0)\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    }\n",
       "    else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager)\n",
       "        manager = IPython.keyboard_manager;\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which == 13) {\n",
       "        this.canvas_div.blur();\n",
       "        event.shiftKey = false;\n",
       "        // Send a \"J\" for go to next cell\n",
       "        event.which = 74;\n",
       "        event.keyCode = 74;\n",
       "        manager.command_mode();\n",
       "        manager.handle_keydown(event);\n",
       "    }\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.find_output_cell = function(html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i=0; i<ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code'){\n",
       "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] == html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "}\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel != null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"638.888905813665\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7fecdc287810>"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Plot classified data - for test data\n",
    "\n",
    "y_pred = classifier.model.predict(classifier.X_test)\n",
    "y_max = y_pred.max(axis = -1)\n",
    "y_labels = y_pred.argmax(axis = -1)\n",
    "\n",
    "fig = plt.figure()\n",
    "#ax = fig.add_subplot(111, projection='3d')\n",
    "#ax = Axes3D(fig)\n",
    "\n",
    "#ax.scatter(classifier.X_test[:, 0], classifier.X_test[:, 1], y_max, c=classifier.y_test.squeeze())\n",
    "plt.scatter(classifier.X_test[:, 0], classifier.X_test[:, 1], s=4, c=y_labels.squeeze())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(200, 1)\n",
      "(400, 1)\n",
      "(600, 1)\n",
      "(300, 2)\n"
     ]
    }
   ],
   "source": [
    "# Create image for paper\n",
    "classifier.n = 200\n",
    "data, labels = classifier.get_data()\n",
    "labels = labels.squeeze()\n",
    "ind_0 = np.where(labels == 0)[0]\n",
    "\n",
    "ind = np.random.randint(0, ind_0.shape[0], 50)\n",
    "ind_0 = ind_0[ind]\n",
    "\n",
    "ind_1 = np.where(labels == 1)[0]\n",
    "\n",
    "ind = np.random.randint(0, ind_1.shape[0], 50)\n",
    "ind_1 = ind_1[ind]\n",
    "\n",
    "ind_2 = np.where(labels == 2)[0]\n",
    "\n",
    "ind = np.concatenate((ind_0, ind_1, ind_2), axis=0)\n",
    "data = data[ind, :]\n",
    "labels = labels[ind]\n",
    "\n",
    "print(data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "/* Put everything inside the global mpl namespace */\n",
       "window.mpl = {};\n",
       "\n",
       "\n",
       "mpl.get_websocket_type = function() {\n",
       "    if (typeof(WebSocket) !== 'undefined') {\n",
       "        return WebSocket;\n",
       "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
       "        return MozWebSocket;\n",
       "    } else {\n",
       "        alert('Your browser does not have WebSocket support.' +\n",
       "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
       "              'Firefox 4 and 5 are also supported but you ' +\n",
       "              'have to enable WebSockets in about:config.');\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
       "    this.id = figure_id;\n",
       "\n",
       "    this.ws = websocket;\n",
       "\n",
       "    this.supports_binary = (this.ws.binaryType != undefined);\n",
       "\n",
       "    if (!this.supports_binary) {\n",
       "        var warnings = document.getElementById(\"mpl-warnings\");\n",
       "        if (warnings) {\n",
       "            warnings.style.display = 'block';\n",
       "            warnings.textContent = (\n",
       "                \"This browser does not support binary websocket messages. \" +\n",
       "                    \"Performance may be slow.\");\n",
       "        }\n",
       "    }\n",
       "\n",
       "    this.imageObj = new Image();\n",
       "\n",
       "    this.context = undefined;\n",
       "    this.message = undefined;\n",
       "    this.canvas = undefined;\n",
       "    this.rubberband_canvas = undefined;\n",
       "    this.rubberband_context = undefined;\n",
       "    this.format_dropdown = undefined;\n",
       "\n",
       "    this.image_mode = 'full';\n",
       "\n",
       "    this.root = $('<div/>');\n",
       "    this._root_extra_style(this.root)\n",
       "    this.root.attr('style', 'display: inline-block');\n",
       "\n",
       "    $(parent_element).append(this.root);\n",
       "\n",
       "    this._init_header(this);\n",
       "    this._init_canvas(this);\n",
       "    this._init_toolbar(this);\n",
       "\n",
       "    var fig = this;\n",
       "\n",
       "    this.waiting = false;\n",
       "\n",
       "    this.ws.onopen =  function () {\n",
       "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
       "            fig.send_message(\"send_image_mode\", {});\n",
       "            if (mpl.ratio != 1) {\n",
       "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
       "            }\n",
       "            fig.send_message(\"refresh\", {});\n",
       "        }\n",
       "\n",
       "    this.imageObj.onload = function() {\n",
       "            if (fig.image_mode == 'full') {\n",
       "                // Full images could contain transparency (where diff images\n",
       "                // almost always do), so we need to clear the canvas so that\n",
       "                // there is no ghosting.\n",
       "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
       "            }\n",
       "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
       "        };\n",
       "\n",
       "    this.imageObj.onunload = function() {\n",
       "        fig.ws.close();\n",
       "    }\n",
       "\n",
       "    this.ws.onmessage = this._make_on_message_function(this);\n",
       "\n",
       "    this.ondownload = ondownload;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_header = function() {\n",
       "    var titlebar = $(\n",
       "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
       "        'ui-helper-clearfix\"/>');\n",
       "    var titletext = $(\n",
       "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
       "        'text-align: center; padding: 3px;\"/>');\n",
       "    titlebar.append(titletext)\n",
       "    this.root.append(titlebar);\n",
       "    this.header = titletext[0];\n",
       "}\n",
       "\n",
       "\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_canvas = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var canvas_div = $('<div/>');\n",
       "\n",
       "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
       "\n",
       "    function canvas_keyboard_event(event) {\n",
       "        return fig.key_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
       "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
       "    this.canvas_div = canvas_div\n",
       "    this._canvas_extra_style(canvas_div)\n",
       "    this.root.append(canvas_div);\n",
       "\n",
       "    var canvas = $('<canvas/>');\n",
       "    canvas.addClass('mpl-canvas');\n",
       "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
       "\n",
       "    this.canvas = canvas[0];\n",
       "    this.context = canvas[0].getContext(\"2d\");\n",
       "\n",
       "    var backingStore = this.context.backingStorePixelRatio ||\n",
       "\tthis.context.webkitBackingStorePixelRatio ||\n",
       "\tthis.context.mozBackingStorePixelRatio ||\n",
       "\tthis.context.msBackingStorePixelRatio ||\n",
       "\tthis.context.oBackingStorePixelRatio ||\n",
       "\tthis.context.backingStorePixelRatio || 1;\n",
       "\n",
       "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
       "\n",
       "    var rubberband = $('<canvas/>');\n",
       "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
       "\n",
       "    var pass_mouse_events = true;\n",
       "\n",
       "    canvas_div.resizable({\n",
       "        start: function(event, ui) {\n",
       "            pass_mouse_events = false;\n",
       "        },\n",
       "        resize: function(event, ui) {\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "        stop: function(event, ui) {\n",
       "            pass_mouse_events = true;\n",
       "            fig.request_resize(ui.size.width, ui.size.height);\n",
       "        },\n",
       "    });\n",
       "\n",
       "    function mouse_event_fn(event) {\n",
       "        if (pass_mouse_events)\n",
       "            return fig.mouse_event(event, event['data']);\n",
       "    }\n",
       "\n",
       "    rubberband.mousedown('button_press', mouse_event_fn);\n",
       "    rubberband.mouseup('button_release', mouse_event_fn);\n",
       "    // Throttle sequential mouse events to 1 every 20ms.\n",
       "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
       "\n",
       "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
       "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
       "\n",
       "    canvas_div.on(\"wheel\", function (event) {\n",
       "        event = event.originalEvent;\n",
       "        event['data'] = 'scroll'\n",
       "        if (event.deltaY < 0) {\n",
       "            event.step = 1;\n",
       "        } else {\n",
       "            event.step = -1;\n",
       "        }\n",
       "        mouse_event_fn(event);\n",
       "    });\n",
       "\n",
       "    canvas_div.append(canvas);\n",
       "    canvas_div.append(rubberband);\n",
       "\n",
       "    this.rubberband = rubberband;\n",
       "    this.rubberband_canvas = rubberband[0];\n",
       "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
       "    this.rubberband_context.strokeStyle = \"#000000\";\n",
       "\n",
       "    this._resize_canvas = function(width, height) {\n",
       "        // Keep the size of the canvas, canvas container, and rubber band\n",
       "        // canvas in synch.\n",
       "        canvas_div.css('width', width)\n",
       "        canvas_div.css('height', height)\n",
       "\n",
       "        canvas.attr('width', width * mpl.ratio);\n",
       "        canvas.attr('height', height * mpl.ratio);\n",
       "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
       "\n",
       "        rubberband.attr('width', width);\n",
       "        rubberband.attr('height', height);\n",
       "    }\n",
       "\n",
       "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
       "    // upon first draw.\n",
       "    this._resize_canvas(600, 600);\n",
       "\n",
       "    // Disable right mouse context menu.\n",
       "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
       "        return false;\n",
       "    });\n",
       "\n",
       "    function set_focus () {\n",
       "        canvas.focus();\n",
       "        canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    window.setTimeout(set_focus, 100);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items) {\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) {\n",
       "            // put a spacer in here.\n",
       "            continue;\n",
       "        }\n",
       "        var button = $('<button/>');\n",
       "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
       "                        'ui-button-icon-only');\n",
       "        button.attr('role', 'button');\n",
       "        button.attr('aria-disabled', 'false');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "\n",
       "        var icon_img = $('<span/>');\n",
       "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
       "        icon_img.addClass(image);\n",
       "        icon_img.addClass('ui-corner-all');\n",
       "\n",
       "        var tooltip_span = $('<span/>');\n",
       "        tooltip_span.addClass('ui-button-text');\n",
       "        tooltip_span.html(tooltip);\n",
       "\n",
       "        button.append(icon_img);\n",
       "        button.append(tooltip_span);\n",
       "\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    var fmt_picker_span = $('<span/>');\n",
       "\n",
       "    var fmt_picker = $('<select/>');\n",
       "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
       "    fmt_picker_span.append(fmt_picker);\n",
       "    nav_element.append(fmt_picker_span);\n",
       "    this.format_dropdown = fmt_picker[0];\n",
       "\n",
       "    for (var ind in mpl.extensions) {\n",
       "        var fmt = mpl.extensions[ind];\n",
       "        var option = $(\n",
       "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
       "        fmt_picker.append(option)\n",
       "    }\n",
       "\n",
       "    // Add hover states to the ui-buttons\n",
       "    $( \".ui-button\" ).hover(\n",
       "        function() { $(this).addClass(\"ui-state-hover\");},\n",
       "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
       "    );\n",
       "\n",
       "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
       "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
       "    // which will in turn request a refresh of the image.\n",
       "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_message = function(type, properties) {\n",
       "    properties['type'] = type;\n",
       "    properties['figure_id'] = this.id;\n",
       "    this.ws.send(JSON.stringify(properties));\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.send_draw_message = function() {\n",
       "    if (!this.waiting) {\n",
       "        this.waiting = true;\n",
       "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
       "    }\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    var format_dropdown = fig.format_dropdown;\n",
       "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
       "    fig.ondownload(fig, format);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
       "    var size = msg['size'];\n",
       "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
       "        fig._resize_canvas(size[0], size[1]);\n",
       "        fig.send_message(\"refresh\", {});\n",
       "    };\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
       "    var x0 = msg['x0'] / mpl.ratio;\n",
       "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
       "    var x1 = msg['x1'] / mpl.ratio;\n",
       "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
       "    x0 = Math.floor(x0) + 0.5;\n",
       "    y0 = Math.floor(y0) + 0.5;\n",
       "    x1 = Math.floor(x1) + 0.5;\n",
       "    y1 = Math.floor(y1) + 0.5;\n",
       "    var min_x = Math.min(x0, x1);\n",
       "    var min_y = Math.min(y0, y1);\n",
       "    var width = Math.abs(x1 - x0);\n",
       "    var height = Math.abs(y1 - y0);\n",
       "\n",
       "    fig.rubberband_context.clearRect(\n",
       "        0, 0, fig.canvas.width, fig.canvas.height);\n",
       "\n",
       "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
       "    // Updates the figure title.\n",
       "    fig.header.textContent = msg['label'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
       "    var cursor = msg['cursor'];\n",
       "    switch(cursor)\n",
       "    {\n",
       "    case 0:\n",
       "        cursor = 'pointer';\n",
       "        break;\n",
       "    case 1:\n",
       "        cursor = 'default';\n",
       "        break;\n",
       "    case 2:\n",
       "        cursor = 'crosshair';\n",
       "        break;\n",
       "    case 3:\n",
       "        cursor = 'move';\n",
       "        break;\n",
       "    }\n",
       "    fig.rubberband_canvas.style.cursor = cursor;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
       "    fig.message.textContent = msg['message'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
       "    // Request the server to send over a new figure.\n",
       "    fig.send_draw_message();\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
       "    fig.image_mode = msg['mode'];\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Called whenever the canvas gets updated.\n",
       "    this.send_message(\"ack\", {});\n",
       "}\n",
       "\n",
       "// A function to construct a web socket function for onmessage handling.\n",
       "// Called in the figure constructor.\n",
       "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
       "    return function socket_on_message(evt) {\n",
       "        if (evt.data instanceof Blob) {\n",
       "            /* FIXME: We get \"Resource interpreted as Image but\n",
       "             * transferred with MIME type text/plain:\" errors on\n",
       "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
       "             * to be part of the websocket stream */\n",
       "            evt.data.type = \"image/png\";\n",
       "\n",
       "            /* Free the memory for the previous frames */\n",
       "            if (fig.imageObj.src) {\n",
       "                (window.URL || window.webkitURL).revokeObjectURL(\n",
       "                    fig.imageObj.src);\n",
       "            }\n",
       "\n",
       "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
       "                evt.data);\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
       "            fig.imageObj.src = evt.data;\n",
       "            fig.updated_canvas_event();\n",
       "            fig.waiting = false;\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        var msg = JSON.parse(evt.data);\n",
       "        var msg_type = msg['type'];\n",
       "\n",
       "        // Call the  \"handle_{type}\" callback, which takes\n",
       "        // the figure and JSON message as its only arguments.\n",
       "        try {\n",
       "            var callback = fig[\"handle_\" + msg_type];\n",
       "        } catch (e) {\n",
       "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
       "            return;\n",
       "        }\n",
       "\n",
       "        if (callback) {\n",
       "            try {\n",
       "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
       "                callback(fig, msg);\n",
       "            } catch (e) {\n",
       "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
       "            }\n",
       "        }\n",
       "    };\n",
       "}\n",
       "\n",
       "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
       "mpl.findpos = function(e) {\n",
       "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
       "    var targ;\n",
       "    if (!e)\n",
       "        e = window.event;\n",
       "    if (e.target)\n",
       "        targ = e.target;\n",
       "    else if (e.srcElement)\n",
       "        targ = e.srcElement;\n",
       "    if (targ.nodeType == 3) // defeat Safari bug\n",
       "        targ = targ.parentNode;\n",
       "\n",
       "    // jQuery normalizes the pageX and pageY\n",
       "    // pageX,Y are the mouse positions relative to the document\n",
       "    // offset() returns the position of the element relative to the document\n",
       "    var x = e.pageX - $(targ).offset().left;\n",
       "    var y = e.pageY - $(targ).offset().top;\n",
       "\n",
       "    return {\"x\": x, \"y\": y};\n",
       "};\n",
       "\n",
       "/*\n",
       " * return a copy of an object with only non-object keys\n",
       " * we need this to avoid circular references\n",
       " * http://stackoverflow.com/a/24161582/3208463\n",
       " */\n",
       "function simpleKeys (original) {\n",
       "  return Object.keys(original).reduce(function (obj, key) {\n",
       "    if (typeof original[key] !== 'object')\n",
       "        obj[key] = original[key]\n",
       "    return obj;\n",
       "  }, {});\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.mouse_event = function(event, name) {\n",
       "    var canvas_pos = mpl.findpos(event)\n",
       "\n",
       "    if (name === 'button_press')\n",
       "    {\n",
       "        this.canvas.focus();\n",
       "        this.canvas_div.focus();\n",
       "    }\n",
       "\n",
       "    var x = canvas_pos.x * mpl.ratio;\n",
       "    var y = canvas_pos.y * mpl.ratio;\n",
       "\n",
       "    this.send_message(name, {x: x, y: y, button: event.button,\n",
       "                             step: event.step,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "\n",
       "    /* This prevents the web browser from automatically changing to\n",
       "     * the text insertion cursor when the button is pressed.  We want\n",
       "     * to control all of the cursor setting manually through the\n",
       "     * 'cursor' event from matplotlib */\n",
       "    event.preventDefault();\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    // Handle any extra behaviour associated with a key event\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.key_event = function(event, name) {\n",
       "\n",
       "    // Prevent repeat events\n",
       "    if (name == 'key_press')\n",
       "    {\n",
       "        if (event.which === this._key)\n",
       "            return;\n",
       "        else\n",
       "            this._key = event.which;\n",
       "    }\n",
       "    if (name == 'key_release')\n",
       "        this._key = null;\n",
       "\n",
       "    var value = '';\n",
       "    if (event.ctrlKey && event.which != 17)\n",
       "        value += \"ctrl+\";\n",
       "    if (event.altKey && event.which != 18)\n",
       "        value += \"alt+\";\n",
       "    if (event.shiftKey && event.which != 16)\n",
       "        value += \"shift+\";\n",
       "\n",
       "    value += 'k';\n",
       "    value += event.which.toString();\n",
       "\n",
       "    this._key_event_extra(event, name);\n",
       "\n",
       "    this.send_message(name, {key: value,\n",
       "                             guiEvent: simpleKeys(event)});\n",
       "    return false;\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
       "    if (name == 'download') {\n",
       "        this.handle_save(this, null);\n",
       "    } else {\n",
       "        this.send_message(\"toolbar_button\", {name: name});\n",
       "    }\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
       "    this.message.textContent = tooltip;\n",
       "};\n",
       "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
       "\n",
       "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
       "\n",
       "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
       "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
       "    // object with the appropriate methods. Currently this is a non binary\n",
       "    // socket, so there is still some room for performance tuning.\n",
       "    var ws = {};\n",
       "\n",
       "    ws.close = function() {\n",
       "        comm.close()\n",
       "    };\n",
       "    ws.send = function(m) {\n",
       "        //console.log('sending', m);\n",
       "        comm.send(m);\n",
       "    };\n",
       "    // Register the callback with on_msg.\n",
       "    comm.on_msg(function(msg) {\n",
       "        //console.log('receiving', msg['content']['data'], msg);\n",
       "        // Pass the mpl event to the overridden (by mpl) onmessage function.\n",
       "        ws.onmessage(msg['content']['data'])\n",
       "    });\n",
       "    return ws;\n",
       "}\n",
       "\n",
       "mpl.mpl_figure_comm = function(comm, msg) {\n",
       "    // This is the function which gets called when the mpl process\n",
       "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
       "\n",
       "    var id = msg.content.data.id;\n",
       "    // Get hold of the div created by the display call when the Comm\n",
       "    // socket was opened in Python.\n",
       "    var element = $(\"#\" + id);\n",
       "    var ws_proxy = comm_websocket_adapter(comm)\n",
       "\n",
       "    function ondownload(figure, format) {\n",
       "        window.open(figure.imageObj.src);\n",
       "    }\n",
       "\n",
       "    var fig = new mpl.figure(id, ws_proxy,\n",
       "                           ondownload,\n",
       "                           element.get(0));\n",
       "\n",
       "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
       "    // web socket which is closed, not our websocket->open comm proxy.\n",
       "    ws_proxy.onopen();\n",
       "\n",
       "    fig.parent_element = element.get(0);\n",
       "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
       "    if (!fig.cell_info) {\n",
       "        console.error(\"Failed to find cell for figure\", id, fig);\n",
       "        return;\n",
       "    }\n",
       "\n",
       "    var output_index = fig.cell_info[2]\n",
       "    var cell = fig.cell_info[0];\n",
       "\n",
       "};\n",
       "\n",
       "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
       "    var width = fig.canvas.width/mpl.ratio\n",
       "    fig.root.unbind('remove')\n",
       "\n",
       "    // Update the output cell to use the data from the current canvas.\n",
       "    fig.push_to_output();\n",
       "    var dataURL = fig.canvas.toDataURL();\n",
       "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
       "    // the notebook keyboard shortcuts fail.\n",
       "    IPython.keyboard_manager.enable()\n",
       "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
       "    fig.close_ws(fig, msg);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.close_ws = function(fig, msg){\n",
       "    fig.send_message('closing', msg);\n",
       "    // fig.ws.close()\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
       "    // Turn the data on the canvas into data in the output cell.\n",
       "    var width = this.canvas.width/mpl.ratio\n",
       "    var dataURL = this.canvas.toDataURL();\n",
       "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.updated_canvas_event = function() {\n",
       "    // Tell IPython that the notebook contents must change.\n",
       "    IPython.notebook.set_dirty(true);\n",
       "    this.send_message(\"ack\", {});\n",
       "    var fig = this;\n",
       "    // Wait a second, then push the new image to the DOM so\n",
       "    // that it is saved nicely (might be nice to debounce this).\n",
       "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._init_toolbar = function() {\n",
       "    var fig = this;\n",
       "\n",
       "    var nav_element = $('<div/>')\n",
       "    nav_element.attr('style', 'width: 100%');\n",
       "    this.root.append(nav_element);\n",
       "\n",
       "    // Define a callback function for later on.\n",
       "    function toolbar_event(event) {\n",
       "        return fig.toolbar_button_onclick(event['data']);\n",
       "    }\n",
       "    function toolbar_mouse_event(event) {\n",
       "        return fig.toolbar_button_onmouseover(event['data']);\n",
       "    }\n",
       "\n",
       "    for(var toolbar_ind in mpl.toolbar_items){\n",
       "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
       "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
       "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
       "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
       "\n",
       "        if (!name) { continue; };\n",
       "\n",
       "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
       "        button.click(method_name, toolbar_event);\n",
       "        button.mouseover(tooltip, toolbar_mouse_event);\n",
       "        nav_element.append(button);\n",
       "    }\n",
       "\n",
       "    // Add the status bar.\n",
       "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
       "    nav_element.append(status_bar);\n",
       "    this.message = status_bar[0];\n",
       "\n",
       "    // Add the close button to the window.\n",
       "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
       "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
       "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
       "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
       "    buttongrp.append(button);\n",
       "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
       "    titlebar.prepend(buttongrp);\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._root_extra_style = function(el){\n",
       "    var fig = this\n",
       "    el.on(\"remove\", function(){\n",
       "\tfig.close_ws(fig, {});\n",
       "    });\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._canvas_extra_style = function(el){\n",
       "    // this is important to make the div 'focusable\n",
       "    el.attr('tabindex', 0)\n",
       "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
       "    // off when our div gets focus\n",
       "\n",
       "    // location in version 3\n",
       "    if (IPython.notebook.keyboard_manager) {\n",
       "        IPython.notebook.keyboard_manager.register_events(el);\n",
       "    }\n",
       "    else {\n",
       "        // location in version 2\n",
       "        IPython.keyboard_manager.register_events(el);\n",
       "    }\n",
       "\n",
       "}\n",
       "\n",
       "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
       "    var manager = IPython.notebook.keyboard_manager;\n",
       "    if (!manager)\n",
       "        manager = IPython.keyboard_manager;\n",
       "\n",
       "    // Check for shift+enter\n",
       "    if (event.shiftKey && event.which == 13) {\n",
       "        this.canvas_div.blur();\n",
       "        event.shiftKey = false;\n",
       "        // Send a \"J\" for go to next cell\n",
       "        event.which = 74;\n",
       "        event.keyCode = 74;\n",
       "        manager.command_mode();\n",
       "        manager.handle_keydown(event);\n",
       "    }\n",
       "}\n",
       "\n",
       "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
       "    fig.ondownload(fig, null);\n",
       "}\n",
       "\n",
       "\n",
       "mpl.find_output_cell = function(html_output) {\n",
       "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
       "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
       "    // IPython event is triggered only after the cells have been serialised, which for\n",
       "    // our purposes (turning an active figure into a static one), is too late.\n",
       "    var cells = IPython.notebook.get_cells();\n",
       "    var ncells = cells.length;\n",
       "    for (var i=0; i<ncells; i++) {\n",
       "        var cell = cells[i];\n",
       "        if (cell.cell_type === 'code'){\n",
       "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
       "                var data = cell.output_area.outputs[j];\n",
       "                if (data.data) {\n",
       "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
       "                    data = data.data;\n",
       "                }\n",
       "                if (data['text/html'] == html_output) {\n",
       "                    return [cell, data, j];\n",
       "                }\n",
       "            }\n",
       "        }\n",
       "    }\n",
       "}\n",
       "\n",
       "// Register the function which deals with the matplotlib target/channel.\n",
       "// The kernel may be null if the page has been refreshed.\n",
       "if (IPython.notebook.kernel != null) {\n",
       "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
       "}\n"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"data:image/png;base64,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\" width=\"498.88890210493145\">"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred = classifier.model.predict(data)\n",
    "y_labels = y_pred.argmax(axis = -1)\n",
    "c_list = ['#FFA500', 'b', '#FFFF33']\n",
    "marker_list = ['D', 'o', '*']\n",
    "plt.figure(figsize=(5,5))\n",
    "for i in range(0, data.shape[0]):\n",
    "    marker = marker_list[int(labels[i])]\n",
    "    c = c_list[y_labels[i]]\n",
    "    plt.scatter(data[i, 0], data[i, 1], s=10, c=c, marker=marker)\n",
    "\n",
    "plt.legend(('$class\\ 0$','$class\\ 1$','$OOD\\ sample$'), fontsize='x-small');\n",
    "import matplotlib.lines as lines\n",
    "dimond_line = lines.Line2D([0], [0], color='w', marker='D', markerfacecolor='#000000',\n",
    "                          markersize=5, label='TC 0')\n",
    "\n",
    "circle_line = lines.Line2D([0], [0], marker='o', color='w', label='TC 1',\n",
    "                          markerfacecolor='#000000', markersize=5)\n",
    "\n",
    "star_line = lines.Line2D([0], [0], color='w', marker='*', markerfacecolor='#000000',\n",
    "                          markersize=10, label='TC 2')\n",
    "\n",
    "\n",
    "orange_line = lines.Line2D([0], [0], marker='s', color='w', label='PC 0',\n",
    "                          markerfacecolor='#FFA500', markersize=5)\n",
    "\n",
    "\n",
    "blue_line = lines.Line2D([0], [0], marker='s', color='w', label='PC 1',\n",
    "                          markerfacecolor='b', markersize=5)\n",
    "\n",
    "\n",
    "yellow_line = lines.Line2D([0], [0], marker='s', color='w', label='PC 2',\n",
    "                          markerfacecolor='#FFFF33', markersize=5)\n",
    "\n",
    "# Create a legend for the first line.\n",
    "first_legend = plt.legend(handles=[dimond_line, circle_line, star_line], loc=1)\n",
    "\n",
    "\n",
    "# Add the legend manually to the current Axes.\n",
    "ax = plt.gca().add_artist(first_legend)\n",
    "\n",
    "# Create another legend for the second line.\n",
    "plt.legend(handles=[orange_line, blue_line, yellow_line], loc=4)\n",
    "\n",
    "plt.show()\n",
    "plt.savefig(\"exp2_c\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1500, 2)\n"
     ]
    }
   ],
   "source": [
    "print(classifier.X_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}