smiling.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "start_time = time.time()\n",
    "import progressbar\n",
    "import os\n",
    "import numpy as np\n",
    "from keras.preprocessing import image\n",
    "import cv2\n",
    "import dlib\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import sys\n",
    "from scipy import stats\n",
    "from sklearn import svm, datasets\n",
    "from sklearn.dummy import DummyClassifier\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.utils import shuffle\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV, learning_curve, ShuffleSplit\n",
    "from sklearn.metrics import classification_report, accuracy_score, confusion_matrix, recall_score, precision_score, f1_score\n",
    "from sklearn.metrics import make_scorer, roc_auc_score\n",
    "from sklearn.model_selection import GridSearchCV, RandomizedSearchCV\n",
    "from sklearn import preprocessing\n",
    "from sklearn.calibration import CalibratedClassifierCV\n",
    "from sklearn.ensemble import RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import landmarks_v2 as l2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_data():\n",
    "    return l2.extract_features_labels()\n",
    "    \n",
    "def split_data(X, Y):\n",
    "\n",
    "    X, Y = shuffle(X,Y)\n",
    "    tr_X, te_X, tr_Y, te_Y = train_test_split(X, Y, train_size=0.7)\n",
    "    \n",
    "    return tr_X, tr_Y, te_X, te_Y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4800, 68, 2)\n",
      "(4800,)\n"
     ]
    }
   ],
   "source": [
    "X,y = get_data()\n",
    "\n",
    "print(X.shape)\n",
    "print(y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "Y = np.array([y, -(y - 1)]).T \n",
    "\n",
    "tr_X, tr_Y, te_X, te_Y = split_data(X, Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3360, 68, 2)\n",
      "(3360, 2)\n",
      "(1440, 68, 2)\n",
      "(1440, 2)\n"
     ]
    }
   ],
   "source": [
    "print(tr_X.shape)\n",
    "print(tr_Y.shape)\n",
    "print(te_X.shape)\n",
    "print(te_Y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def reshapeX(X):\n",
    "    return X.reshape((X.shape[0], X.shape[1] * X.shape[2]))\n",
    "\n",
    "def reshapeY(y):\n",
    "    return list(zip(*y))[0]\n",
    "\n",
    "tr_X = reshapeX(tr_X)\n",
    "te_X = reshapeX(te_X)\n",
    "tr_Y = reshapeY(tr_Y)\n",
    "te_Y = reshapeY(te_Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3360, 136)\n"
     ]
    }
   ],
   "source": [
    "print(tr_X.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.1 Linear: 0.9013888888888889\n",
      "0.1 scale Rbf: 0.8555555555555555\n",
      "0.1 1 Poly 0.8604166666666667\n",
      "0.1 Linear: 0.9013888888888889\n",
      "0.1 scale Rbf: 0.8555555555555555\n",
      "0.1 2 Poly 0.8763888888888889\n",
      "0.1 Linear: 0.9013888888888889\n",
      "0.1 scale Rbf: 0.8555555555555555\n",
      "0.1 3 Poly 0.8958333333333334\n",
      "0.1 Linear: 0.9013888888888889\n",
      "0.1 scale Rbf: 0.8555555555555555\n",
      "0.1 4 Poly 0.8951388888888889\n",
      "1 Linear: 0.9027777777777778\n",
      "1 scale Rbf: 0.875\n",
      "1 1 Poly 0.8673611111111111\n",
      "1 Linear: 0.9027777777777778\n",
      "1 scale Rbf: 0.875\n",
      "1 2 Poly 0.8895833333333333\n",
      "1 Linear: 0.9027777777777778\n",
      "1 scale Rbf: 0.875\n",
      "1 3 Poly 0.8944444444444445\n",
      "1 Linear: 0.9027777777777778\n",
      "1 scale Rbf: 0.875\n",
      "1 4 Poly 0.9027777777777778\n"
     ]
    }
   ],
   "source": [
    "gamma='scale'\n",
    "\n",
    "for C in [0.1,1]:\n",
    "    for degree in [1, 2, 3, 4]:\n",
    "                \n",
    "        lin_svc = svm.SVC(kernel='linear', C=C).fit(tr_X, tr_Y)\n",
    "        print(C, \"Linear:\", lin_svc.score(te_X, te_Y))\n",
    "\n",
    "        rbf_svc = svm.SVC(kernel='rbf', gamma=gamma, C=C).fit(tr_X, tr_Y)\n",
    "        print( C, gamma, \"Rbf:\", rbf_svc.score(te_X, te_Y))\n",
    "\n",
    "        poly_svc = svm.SVC(kernel='poly', degree=degree, C=C).fit(tr_X, tr_Y)\n",
    "        print( C, degree, \"Poly\", poly_svc.score(te_X, te_Y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.svm import SVC as svc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  16 tasks      | elapsed:  4.8min\n",
      "[Parallel(n_jobs=-1)]: Done  50 out of  50 | elapsed: 26.3min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'mean_fit_time': array([ 42.86319351, 134.54837728, 234.11385045,  98.40222597,\n",
      "       415.13487439,  85.59347501,  52.37077231, 439.03298478,\n",
      "       402.08766651, 410.49716034]), 'std_fit_time': array([  8.71799884,  14.19251686,  43.27063944,   6.85742035,\n",
      "       116.21376088,  15.90081105,   7.24157948,  92.9295045 ,\n",
      "        61.37841869,  29.78693504]), 'mean_score_time': array([0.11260791, 0.10731282, 0.11289811, 0.10990577, 0.11948047,\n",
      "       0.11675262, 0.12366853, 0.12110705, 0.11164503, 0.08616881]), 'std_score_time': array([0.00897267, 0.00312904, 0.00465255, 0.00317916, 0.00074606,\n",
      "       0.00806031, 0.01133617, 0.00506764, 0.0232839 , 0.01638196]), 'param_C': masked_array(data=[1.251229016933845, 3.6904316102385772,\n",
      "                   5.763890114534913, 2.747810778444424, 9.57287812914673,\n",
      "                   2.0171501071852695, 1.3024006556404557,\n",
      "                   10.48800752126697, 9.758728053614494,\n",
      "                   10.839995989809484],\n",
      "             mask=[False, False, False, False, False, False, False, False,\n",
      "                   False, False],\n",
      "       fill_value='?',\n",
      "            dtype=object), 'param_kernel': masked_array(data=['linear', 'linear', 'linear', 'linear', 'linear',\n",
      "                   'linear', 'linear', 'linear', 'linear', 'linear'],\n",
      "             mask=[False, False, False, False, False, False, False, False,\n",
      "                   False, False],\n",
      "       fill_value='?',\n",
      "            dtype=object), 'params': [{'C': 1.251229016933845, 'kernel': 'linear'}, {'C': 3.6904316102385772, 'kernel': 'linear'}, {'C': 5.763890114534913, 'kernel': 'linear'}, {'C': 2.747810778444424, 'kernel': 'linear'}, {'C': 9.57287812914673, 'kernel': 'linear'}, {'C': 2.0171501071852695, 'kernel': 'linear'}, {'C': 1.3024006556404557, 'kernel': 'linear'}, {'C': 10.48800752126697, 'kernel': 'linear'}, {'C': 9.758728053614494, 'kernel': 'linear'}, {'C': 10.839995989809484, 'kernel': 'linear'}], 'split0_test_score': array([0.88988095, 0.88988095, 0.88988095, 0.88839286, 0.89136905,\n",
      "       0.88988095, 0.88839286, 0.88988095, 0.88988095, 0.88988095]), 'split1_test_score': array([0.8764881, 0.8764881, 0.8764881, 0.875    , 0.8764881, 0.875    ,\n",
      "       0.8764881, 0.8764881, 0.8764881, 0.8764881]), 'split2_test_score': array([0.88690476, 0.88690476, 0.88690476, 0.88690476, 0.88690476,\n",
      "       0.88690476, 0.88690476, 0.88690476, 0.88690476, 0.88690476]), 'split3_test_score': array([0.875    , 0.875    , 0.8764881, 0.875    , 0.875    , 0.875    ,\n",
      "       0.875    , 0.875    , 0.875    , 0.8735119]), 'split4_test_score': array([0.89285714, 0.89136905, 0.89136905, 0.89136905, 0.89136905,\n",
      "       0.89136905, 0.89285714, 0.89136905, 0.89136905, 0.88988095]), 'mean_test_score': array([0.88422619, 0.88392857, 0.88422619, 0.88333333, 0.88422619,\n",
      "       0.88363095, 0.88392857, 0.88392857, 0.88392857, 0.88333333]), 'std_test_score': array([0.00719229, 0.00685171, 0.00647963, 0.00695436, 0.00713045,\n",
      "       0.00719229, 0.00697979, 0.00685171, 0.00685171, 0.00695436]), 'rank_test_score': array([ 1,  5,  1,  9,  1,  8,  4,  5,  5, 10])}\n",
      "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  16 tasks      | elapsed:    6.5s\n",
      "[Parallel(n_jobs=-1)]: Done  50 out of  50 | elapsed:   19.5s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'mean_fit_time': array([2.47757311, 1.46727414, 3.00875077, 1.56153088, 3.38936224,\n",
      "       3.27145863, 3.20283093, 2.68343296, 3.07948828, 1.08700142]), 'std_fit_time': array([0.0644307 , 0.01821018, 0.05338659, 0.04568227, 0.08029478,\n",
      "       0.07065161, 0.04802832, 0.20827623, 0.32589268, 0.1739924 ]), 'mean_score_time': array([0.28802872, 0.18031864, 0.54924426, 0.2233058 , 0.50076294,\n",
      "       0.57187014, 0.54753551, 0.29481249, 0.47713289, 0.15379496]), 'std_score_time': array([0.0222432 , 0.0090401 , 0.01890543, 0.03528421, 0.02528669,\n",
      "       0.02416321, 0.02749514, 0.02499024, 0.09296489, 0.0279862 ]), 'param_C': masked_array(data=[3.210775005312925, 10.186203809374742,\n",
      "                   8.029937216685447, 3.002185814141008,\n",
      "                   9.758123901586497, 8.424677569295659,\n",
      "                   3.096245900909357, 7.474366347208107,\n",
      "                   9.143364980836731, 6.032974563111113],\n",
      "             mask=[False, False, False, False, False, False, False, False,\n",
      "                   False, False],\n",
      "       fill_value='?',\n",
      "            dtype=object), 'param_gamma': masked_array(data=[0.0028671778742097075, 0.0015232477810465882,\n",
      "                   0.009043799085678416, 0.0018993158840944523,\n",
      "                   0.006159016025216311, 0.009014752118313159,\n",
      "                   0.009402796207786885, 0.0024273693646190905,\n",
      "                   0.008814834110087362, 0.0009873139147704966],\n",
      "             mask=[False, False, False, False, False, False, False, False,\n",
      "                   False, False],\n",
      "       fill_value='?',\n",
      "            dtype=object), 'param_kernel': masked_array(data=['rbf', 'rbf', 'rbf', 'rbf', 'rbf', 'rbf', 'rbf', 'rbf',\n",
      "                   'rbf', 'rbf'],\n",
      "             mask=[False, False, False, False, False, False, False, False,\n",
      "                   False, False],\n",
      "       fill_value='?',\n",
      "            dtype=object), 'params': [{'C': 3.210775005312925, 'gamma': 0.0028671778742097075, 'kernel': 'rbf'}, {'C': 10.186203809374742, 'gamma': 0.0015232477810465882, 'kernel': 'rbf'}, {'C': 8.029937216685447, 'gamma': 0.009043799085678416, 'kernel': 'rbf'}, {'C': 3.002185814141008, 'gamma': 0.0018993158840944523, 'kernel': 'rbf'}, {'C': 9.758123901586497, 'gamma': 0.006159016025216311, 'kernel': 'rbf'}, {'C': 8.424677569295659, 'gamma': 0.009014752118313159, 'kernel': 'rbf'}, {'C': 3.096245900909357, 'gamma': 0.009402796207786885, 'kernel': 'rbf'}, {'C': 7.474366347208107, 'gamma': 0.0024273693646190905, 'kernel': 'rbf'}, {'C': 9.143364980836731, 'gamma': 0.008814834110087362, 'kernel': 'rbf'}, {'C': 6.032974563111113, 'gamma': 0.0009873139147704966, 'kernel': 'rbf'}], 'split0_test_score': array([0.85714286, 0.8735119 , 0.82291667, 0.8735119 , 0.82886905,\n",
      "       0.82291667, 0.82142857, 0.86011905, 0.82142857, 0.8735119 ]), 'split1_test_score': array([0.85863095, 0.85863095, 0.84672619, 0.86755952, 0.84375   ,\n",
      "       0.84672619, 0.83630952, 0.85416667, 0.8452381 , 0.86458333]), 'split2_test_score': array([0.87946429, 0.8764881 , 0.84970238, 0.89136905, 0.86309524,\n",
      "       0.84970238, 0.84821429, 0.88244048, 0.85267857, 0.89732143]), 'split3_test_score': array([0.87202381, 0.85565476, 0.84672619, 0.86904762, 0.84970238,\n",
      "       0.84672619, 0.84821429, 0.86309524, 0.84672619, 0.87202381]), 'split4_test_score': array([0.84970238, 0.85119048, 0.82142857, 0.86011905, 0.8422619 ,\n",
      "       0.82142857, 0.82142857, 0.84970238, 0.82142857, 0.86607143]), 'mean_test_score': array([0.86339286, 0.86309524, 0.8375    , 0.87232143, 0.84553571,\n",
      "       0.8375    , 0.83511905, 0.86190476, 0.8375    , 0.87470238]), 'std_test_score': array([0.01078843, 0.01004878, 0.01257066, 0.01045486, 0.011112  ,\n",
      "       0.01257066, 0.01199372, 0.0112703 , 0.01335643, 0.01180764]), 'rank_test_score': array([ 3,  4,  7,  2,  6,  7, 10,  5,  7,  1])}\n",
      "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  16 tasks      | elapsed:    3.8s\n",
      "[Parallel(n_jobs=-1)]: Done  50 out of  50 | elapsed:    9.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'mean_fit_time': array([0.81721425, 1.46052704, 1.47876759, 1.45815568, 1.08589764,\n",
      "       1.57661033, 0.89062819, 0.98098707, 1.00940814, 0.86430364]), 'std_fit_time': array([0.09046234, 0.10942306, 0.06010334, 0.05098638, 0.02979232,\n",
      "       0.03327095, 0.06855781, 0.07664366, 0.02717374, 0.07676315]), 'mean_score_time': array([0.12207408, 0.11967969, 0.12906775, 0.2277905 , 0.12147822,\n",
      "       0.12366872, 0.12999544, 0.12716303, 0.14341683, 0.0935606 ]), 'std_score_time': array([0.02282628, 0.01259958, 0.01296756, 0.01965157, 0.0077054 ,\n",
      "       0.01414488, 0.00704894, 0.00713357, 0.01238311, 0.00862368]), 'param_C': masked_array(data=[4.385023066183646, 8.697770499061054,\n",
      "                   6.9624105619906445, 3.942389544865043,\n",
      "                   9.673848787494284, 8.458458576970397,\n",
      "                   5.131725800641961, 3.560314375842936, 2.74720688528629,\n",
      "                   7.185559640131561],\n",
      "             mask=[False, False, False, False, False, False, False, False,\n",
      "                   False, False],\n",
      "       fill_value='?',\n",
      "            dtype=object), 'param_degree': masked_array(data=[2.9229448394082636, 3.7738751227404004,\n",
      "                   3.6447423182499357, 1.1093276222220125,\n",
      "                   2.3175615269288605, 3.648227218292908,\n",
      "                   2.536640657064369, 2.8969161975869664,\n",
      "                   2.770103733905792, 2.2054879059228685],\n",
      "             mask=[False, False, False, False, False, False, False, False,\n",
      "                   False, False],\n",
      "       fill_value='?',\n",
      "            dtype=object), 'param_kernel': masked_array(data=['poly', 'poly', 'poly', 'poly', 'poly', 'poly', 'poly',\n",
      "                   'poly', 'poly', 'poly'],\n",
      "             mask=[False, False, False, False, False, False, False, False,\n",
      "                   False, False],\n",
      "       fill_value='?',\n",
      "            dtype=object), 'params': [{'C': 4.385023066183646, 'degree': 2.9229448394082636, 'kernel': 'poly'}, {'C': 8.697770499061054, 'degree': 3.7738751227404004, 'kernel': 'poly'}, {'C': 6.9624105619906445, 'degree': 3.6447423182499357, 'kernel': 'poly'}, {'C': 3.942389544865043, 'degree': 1.1093276222220125, 'kernel': 'poly'}, {'C': 9.673848787494284, 'degree': 2.3175615269288605, 'kernel': 'poly'}, {'C': 8.458458576970397, 'degree': 3.648227218292908, 'kernel': 'poly'}, {'C': 5.131725800641961, 'degree': 2.536640657064369, 'kernel': 'poly'}, {'C': 3.560314375842936, 'degree': 2.8969161975869664, 'kernel': 'poly'}, {'C': 2.74720688528629, 'degree': 2.770103733905792, 'kernel': 'poly'}, {'C': 7.185559640131561, 'degree': 2.2054879059228685, 'kernel': 'poly'}], 'split0_test_score': array([0.88988095, 0.88541667, 0.88541667, 0.8764881 , 0.89136905,\n",
      "       0.88541667, 0.88988095, 0.89136905, 0.89285714, 0.89285714]), 'split1_test_score': array([0.88988095, 0.88541667, 0.88690476, 0.87053571, 0.88988095,\n",
      "       0.88392857, 0.88988095, 0.89136905, 0.89136905, 0.89136905]), 'split2_test_score': array([0.89285714, 0.90029762, 0.9047619 , 0.90029762, 0.89434524,\n",
      "       0.90029762, 0.89285714, 0.89434524, 0.89732143, 0.89285714]), 'split3_test_score': array([0.88244048, 0.87946429, 0.88244048, 0.87797619, 0.88392857,\n",
      "       0.87946429, 0.88541667, 0.88095238, 0.88095238, 0.88541667]), 'split4_test_score': array([0.90625   , 0.9077381 , 0.9077381 , 0.88392857, 0.9047619 ,\n",
      "       0.9077381 , 0.90625   , 0.89880952, 0.89732143, 0.90029762]), 'mean_test_score': array([0.8922619 , 0.89166667, 0.89345238, 0.88184524, 0.89285714,\n",
      "       0.89136905, 0.89285714, 0.89136905, 0.89196429, 0.89255952]), 'std_test_score': array([0.00779512, 0.01058118, 0.01058955, 0.01016272, 0.00685171,\n",
      "       0.010772  , 0.00710556, 0.00587751, 0.00599686, 0.00474327]), 'rank_test_score': array([ 5,  7,  1, 10,  2,  8,  3,  8,  6,  4])}\n",
      "Optimum hyperparameters for linear kernel: \n",
      "{'C': 1.251229016933845, 'kernel': 'linear'}\n",
      "Optimum hyperparameters for rbf kernel: \n",
      "{'C': 6.032974563111113, 'gamma': 0.0009873139147704966, 'kernel': 'rbf'}\n",
      "Optimum hyperparameters for polynomial kernel: \n",
      "{'C': 6.9624105619906445, 'degree': 3.6447423182499357, 'kernel': 'poly'}\n"
     ]
    }
   ],
   "source": [
    "lin = {'C': stats.uniform(1, 10), 'kernel': ['linear']}\n",
    "\n",
    "rbf = {'C': stats.uniform(1, 10), 'gamma': stats.uniform(0.0001, 0.01), 'kernel': ['rbf']}\n",
    "\n",
    "poly = {'C': stats.uniform(1, 10), 'degree': stats.uniform(1, 4), 'kernel': ['poly']}\n",
    "\n",
    "def randomSearch(X, y, param_kernel):\n",
    "    param_distributions = param_kernel \n",
    "    searchrand = RandomizedSearchCV(SVC(), param_distributions, n_iter=10, n_jobs=-1, refit=True, verbose=3)\n",
    "    searchrand.fit(X, y)\n",
    "#   printing statistical scores \n",
    "    searchrand.cv_results_\n",
    "    \n",
    "#    print(searchrand.cv_results_)\n",
    "    return searchrand.best_params_, searchrand.best_estimator_  \n",
    "\n",
    "# Obtaining optimum hyperparameters and classifier for different kernel\n",
    "linSVC_param, lin_SVC = randomSearch(tr_X, tr_Y, lin)\n",
    "rbfSVC_param, rbf_SVC = randomSearch(tr_X, tr_Y, rbf)\n",
    "polySVC_param, poly_SVC = randomSearch(tr_X, tr_Y, poly)\n",
    "#print(datetime.now() - start)\n",
    "\n",
    "# Display optimum hyperparameters for SVC kernel\n",
    "print('Optimum hyperparameters for linear kernel: ')\n",
    "print(linSVC_param)\n",
    "print('Optimum hyperparameters for rbf kernel: ')\n",
    "print(rbfSVC_param)\n",
    "print('Optimum hyperparameters for polynomial kernel: ')\n",
    "print(polySVC_param)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9034722222222222\n",
      "0.8930555555555556\n",
      "0.9034722222222222\n"
     ]
    }
   ],
   "source": [
    "# printing validation accuracy scores for each kernel\n",
    "print(lin_SVC.score(te_X, te_Y))\n",
    "print(rbf_SVC.score(te_X, te_Y))\n",
    "print(poly_SVC.score(te_X, te_Y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Automatically created module for IPython interactive environment\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x1296 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(__doc__)\n",
    "\n",
    "\n",
    "def plot_learning_curve(estimator, title, X, y, axes=None, ylim=None, cv=None,\n",
    "                        n_jobs=None, train_sizes=np.linspace(.1, 1.0, 5)):\n",
    "\n",
    "    axes.set_title(title)\n",
    "    if ylim is not None:\n",
    "        axes.set_ylim(*ylim)\n",
    "        axes.set_xlabel(\"Training examples\")\n",
    "        axes.set_ylabel(\"Score\")\n",
    "\n",
    "        train_sizes, train_scores, test_scores, fit_times, _ = \\\n",
    "            learning_curve(estimator, X, y, cv=cv, n_jobs=n_jobs,\n",
    "                           train_sizes=train_sizes,\n",
    "                           return_times=True)\n",
    "        train_scores_mean = np.mean(train_scores, axis=1)\n",
    "        train_scores_std = np.std(train_scores, axis=1)\n",
    "        test_scores_mean = np.mean(test_scores, axis=1)\n",
    "        test_scores_std = np.std(test_scores, axis=1)\n",
    "        fit_times_mean = np.mean(fit_times, axis=1)\n",
    "        fit_times_std = np.std(fit_times, axis=1)\n",
    "\n",
    "    # Plot learning curve\n",
    "        axes.grid()\n",
    "        axes.fill_between(train_sizes, train_scores_mean - train_scores_std,\n",
    "                             train_scores_mean + train_scores_std, alpha=0.1,\n",
    "                             color=\"r\")\n",
    "        axes.fill_between(train_sizes, test_scores_mean - test_scores_std,\n",
    "                             test_scores_mean + test_scores_std, alpha=0.1,\n",
    "                             color=\"g\")\n",
    "        axes.plot(train_sizes, train_scores_mean, 'o-', color=\"r\",\n",
    "                     label=\"Training score\")\n",
    "        axes.plot(train_sizes, test_scores_mean, 'o-', color=\"g\",\n",
    "                     label=\"Cross-validation score\")\n",
    "        axes.legend(loc=\"best\")\n",
    "\n",
    "    return plt\n",
    "\n",
    "plt.figure(figsize=(9,18))\n",
    "\n",
    "cv = ShuffleSplit(n_splits=30, test_size=0.2, random_state=0)\n",
    "\n",
    "\n",
    "# estimator = poly_SVC \n",
    "# plot_learning_curve(estimator, title, te_X, te_Y, axes=axes, ylim=(0.8, 1.01), cv=cv, n_jobs=-1)\n",
    "axes = plt.subplot(311)\n",
    "title = r\"Learning Curves (linear)\"\n",
    "plot_learning_curve(lin_SVC, title, te_X, te_Y, axes=axes, ylim=(0.8, 1.01), cv=cv, n_jobs=-1)\n",
    "\n",
    "axes = plt.subplot(312)\n",
    "title = r\"Learning Curves (rbf)\"\n",
    "plot_learning_curve(rbf_SVC, title, te_X, te_Y, axes=axes, ylim=(0.8, 1.01), cv=cv, n_jobs=-1)\n",
    "\n",
    "axes = plt.subplot(313)\n",
    "title = r\"Learning Curves (poly)\"\n",
    "plot_learning_curve(poly_SVC, title, te_X, te_Y, axes=axes, ylim=(0.8, 1.01), cv=cv, n_jobs=-1)\n",
    "\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.6.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}