KNNandSVMEigenfaces.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import seaborn as sns; sns.set()\n",
    "import cv2\n",
    "import os\n",
    "from matplotlib import pyplot as plt\n",
    "from matplotlib import animation\n",
    "from JSAnimation import IPython_display"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_Images(path):\n",
    "    i = 0\n",
    "    img_struct = []\n",
    "    for (path, dirs, files) in os.walk(path):\n",
    "        for fname in files:\n",
    "            if fname.endswith('jpg'):\n",
    "                src = os.path.join(path, fname)\n",
    "                #print(src)\n",
    "                img = cv2.imread(src, 0)\n",
    "                d1, d2 = img.shape\n",
    "                img = img.flatten()\n",
    "                img_struct.append(img)\n",
    "    return d1,d2,img_struct"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "d1, d2, male_Test = get_Images('C:/Users/samya/Downloads/face/face/male/test')\n",
    "d1, d2, female_Test = get_Images('C:/Users/samya/Downloads/face/face/female/test')\n",
    "d1, d2, male_Train = get_Images('C:/Users/samya/Downloads/face/face/male/train')\n",
    "d1, d2, female_Train = get_Images('C:/Users/samya/Downloads/face/face/female/train')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "male_Test = np.transpose(np.array(male_Test, dtype = np.float32))\n",
    "female_Test = np.transpose(np.array(female_Test, dtype = np.float32))\n",
    "male_Train = np.transpose(np.array(male_Train, dtype = np.float32))\n",
    "female_Train = np.transpose(np.array(female_Train, dtype = np.float32))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[   0.,    0.,    0., ...,    0.,  130.,    0.],\n",
       "       [   0.,    0.,    0., ...,    0.,  130.,    0.],\n",
       "       [   0.,    2.,    0., ...,    0.,  130.,    0.],\n",
       "       ..., \n",
       "       [   0.,   36.,    0., ...,    0.,   28.,    0.],\n",
       "       [   0.,   36.,    0., ...,    0.,   27.,    0.],\n",
       "       [   0.,   36.,    0., ...,    0.,   27.,    0.]], dtype=float32)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "male_Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(62500, 1000)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "female_Test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(([], <a list of 0 Text xticklabel objects>),\n",
       " ([], <a list of 0 Text yticklabel objects>))"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAE5CAYAAADr4VfxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkhvKUIbyXMtQyQ1lKEN5ruV/AFdZmuuHR8aKAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x208166ecc50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "face_1 = male_Test[:,0].reshape(d1,d2)\n",
    "plt.imshow(face_1, cmap = plt.get_cmap('gray'))\n",
    "plt.xticks([]), plt.yticks([])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "d_testSet = np.hstack((male_Test, female_Test))\n",
    "d_trainSet = np.hstack((male_Train, female_Train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(62500, 2000)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d_testSet.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "mean_Face = np.mean(d_trainSet, axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(([], <a list of 0 Text xticklabel objects>),\n",
       " ([], <a list of 0 Text yticklabel objects>))"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAE5CAYAAADr4VfxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x208176f6ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(mean_Face.reshape(d1,d2), cmap = 'gray')\n",
    "plt.xticks([]), plt.yticks([])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "d_trainSetC = d_trainSet - mean_Face[:, np.newaxis]\n",
    "d_testSetC = d_testSet - mean_Face[:,np.newaxis]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "corr_TrainSet = np.matmul(np.transpose(d_trainSetC), d_trainSetC)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "eigVal, eigVec = np.linalg.eig(corr_TrainSet)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Container object of 100 artists>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeEAAAFTCAYAAAAOSLRwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\newAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x20884fd59b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "xi = [i for i in range(0, len(eigVal[0:100]))]\n",
    "plt.bar(xi, eigVal[0:100])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[249]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[idx for idx,val in enumerate(eigVal) if val < 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "eig_Faces = np.matmul(d_trainSetC, eigVec)\n",
    "eig_Faces = eig_Faces[:, 0:200]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(62500, 200)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "eig_Faces.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "eig_FacesNorm = eig_Faces / np.sqrt(eigVal[np.newaxis ,0:200])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Finding Weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "weights_Train = np.matmul(np.transpose(d_trainSetC), eig_FacesNorm)\n",
    "weights_Test = np.matmul(np.transpose(d_testSetC), eig_FacesNorm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3868, 200)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weights_Train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# K Nearest Neighbours"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3868"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sel_TestWeights = weights_Test[1,:]\n",
    "err = np.sqrt(np.sum(np.square(weights_Train - sel_TestWeights[np.newaxis, :]), axis = 1))\n",
    "len(err)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "err = np.argsort(err)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "err[err <= (3868/2) - 1] = -1\n",
    "err[err >= (3868/2)] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.any(err > 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-2"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fin_Class = np.sum(err[0:10])\n",
    "fin_Class"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that __.append()__ will return an empty object. So something like __A = A.append(i)__ will just return an empty object. We just need to do __A.append()__. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def k_NN(weights_Test, weights_Train, k):\n",
    "    miss_ClassM = []\n",
    "    miss_ClassF = []\n",
    "    for NN in k:\n",
    "        k_NNMale = []\n",
    "        k_NNFemale = []\n",
    "        for i in range(len(weights_Test[:,1])):\n",
    "            sel_TestWeights = weights_Test[i,:]\n",
    "            err = np.sqrt(np.sum(np.square(weights_Train - sel_TestWeights[np.newaxis, :]), axis = 1))\n",
    "\n",
    "            err = np.argsort(err)\n",
    "\n",
    "            err[err <= (3868/2) - 1] = -1\n",
    "            err[err >= (3868/2)] = 1\n",
    "\n",
    "            fin_Class = np.sum(err[0:NN])\n",
    "\n",
    "            if i <= (2000/2) - 1:\n",
    "                k_NNMale.append(fin_Class)\n",
    "            else:\n",
    "                k_NNFemale.append(fin_Class)\n",
    "\n",
    "        k_NNMale = [0 if p <= 0 else 1 for p in k_NNMale]\n",
    "        k_NNFemale = [1 if p < 0 else 0 for p in k_NNFemale]\n",
    "\n",
    "        err_Male = sum(k_NNMale)\n",
    "        err_Female = sum(k_NNFemale)\n",
    "        \n",
    "        miss_ClassM.append(err_Male)\n",
    "        miss_ClassF.append(err_Female)\n",
    "    \n",
    "    return miss_ClassM, miss_ClassF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "k = [1, 5, 10, 50, 100]\n",
    "miss_ClassM, miss_ClassF = k_NN(weights_Test, weights_Train, k)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x2093eda7400>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAFXCAYAAACP5RboAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAIAhCG0AAAzx/9S82/ibc7oJAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2093b7e5c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ind = np.array([1,2,3,4,5])  # the x locations for the groups\n",
    "width = 0.27 # the width of the bars\n",
    "\n",
    "fig = plt.figure(facecolor = 'white')\n",
    "ax = fig.add_subplot(111)\n",
    "\n",
    "yvals = [(i/1000)*100 for i in miss_ClassM]\n",
    "rects1 = ax.bar(ind, yvals, width)\n",
    "zvals = [(i/1000)*100 for i in miss_ClassF]\n",
    "rects2 = ax.bar(ind+width, zvals, width)\n",
    "\n",
    "ax.set_ylabel('Misclassified Values')\n",
    "ax.set_xlabel('K - Nearest Neighbours')\n",
    "ax.set_xticks(ind+ width/2)\n",
    "ax.set_xticklabels( ('1', '5', '10', '50', '100') )\n",
    "ax.legend( (rects1[0], rects2[0]), ('Male', 'Female') )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "miss_ClassM, miss_ClassF = k_NN(weights_Test, weights_Train, [1, 5, 10, 50, 100])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import sklearn\n",
    "from sklearn import svm, datasets\n",
    "\n",
    "iris = datasets.load_iris()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "iris"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X = iris['data'][:,0:2]\n",
    "#X = iris.data[:,:2]\n",
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Y = iris['target']\n",
    "Y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "d1_iris, d2_iris = X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "C = 1.0\n",
    "models = (svm.SVC(kernel='linear', C=C),\n",
    "          svm.LinearSVC(C=C),\n",
    "          svm.SVC(kernel='rbf', gamma=0.7, C=C),\n",
    "          svm.SVC(kernel='poly', degree=3, C=C))\n",
    "\n",
    "models = (clf.fit(X, Y) for clf in models)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "C = 1.0  # SVM regularization parameter\n",
    "numModels = 4\n",
    "models = (svm.SVC(kernel='linear', C=C),\n",
    "          svm.LinearSVC(C=C),\n",
    "          svm.SVC(kernel='rbf', gamma=0.7, C=C),\n",
    "          svm.SVC(kernel='poly', degree=3, C=C))\n",
    "\n",
    "models = (clf.fit(X, Y) for clf in models)\n",
    "\n",
    "fin_ClassIris = np.array([])\n",
    "for model in models:\n",
    "    temp = np.array([])\n",
    "    for i in range(d1_iris):\n",
    "        pred = model.predict(iris.data[i,:2].reshape(1,-1))\n",
    "        temp = np.append(temp, pred)\n",
    "    \n",
    "    fin_ClassIris = np.concatenate((fin_ClassIris, temp))\n",
    "    \n",
    "fin_ClassIris = fin_ClassIris.reshape((numModels, d1_iris))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fin_ClassIris"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def SVM_auto(test_Weights, train_Weights, labels, C):\n",
    "    #C = 1.0\n",
    "    numModels = 4\n",
    "    models = (svm.SVC(kernel='linear', C=C),\n",
    "              svm.LinearSVC(C=C),\n",
    "              svm.SVC(kernel='rbf', gamma=0.7, C=C),\n",
    "              svm.SVC(kernel='poly', degree=3, C=C))\n",
    "\n",
    "    models = (clf.fit(train_Weights, labels) for clf in models)\n",
    "\n",
    "    fin_Class = np.array([])\n",
    "    for model in models:\n",
    "        temp_Arr = np.array([])\n",
    "        for i in range(2000):\n",
    "            pred = model.predict(test_Weights[i,:].reshape(1,-1))\n",
    "            temp_Arr = np.append(temp_Arr, pred)\n",
    "\n",
    "        fin_Class = np.concatenate((fin_Class, temp_Arr))\n",
    "\n",
    "    fin_Class = fin_ClassIris.reshape((numModels, 2000))\n",
    "    return fin_Class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "C = 1.0\n",
    "labels = np.concatenate((np.ones(1934), np.zeros(1934)))\n",
    "labels.shape    \n",
    "numModels = 2\n",
    "models = (svm.SVC(kernel='linear', C=C))\n",
    "\n",
    "models = (clf.fit(weights_Train, labels) for clf in models)\n",
    "\n",
    "fin_Class = np.array([])\n",
    "for model in models:\n",
    "    temp_Arr = np.array([])\n",
    "    for i in range(2000):\n",
    "        pred = model.predict(weights_Test[i,:].reshape(1,-1))\n",
    "        temp_Arr = np.append(temp_Arr, pred)\n",
    "\n",
    "        fin_Class = np.concatenate((fin_Class, temp_Arr))\n",
    "\n",
    "    fin_Class = fin_ClassIris.reshape((numModels, 2000))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3868,)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels = np.concatenate((np.ones(1934), np.zeros(1934)))\n",
    "labels.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "numModels = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2000, 200)"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weights_Test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3868, 200)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weights_Train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#fin_Class = SVM_auto(weights_Test, weights_Train, labels, 1.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "A = np.array([])\n",
    "B = np.array([1, 2, 3, 4, 5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X = np.concatenate((A,B))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}