Comparing different Machine learning Models.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd  \n",
    "import numpy as np  \n",
    "import matplotlib.pyplot as plt  \n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn import metrics\n",
    "from sklearn.gaussian_process import GaussianProcessRegressor\n",
    "from sklearn.gaussian_process.kernels import RBF, ConstantKernel as C\n",
    "from sklearn.svm import SVR\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.linear_model import Ridge\n",
    "from sklearn.model_selection import cross_validate\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.model_selection import validation_curve\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "import random\n",
    "%matplotlib inline\n",
    "#import elm\n",
    "#from glm import GLM\n",
    "from sklearn.preprocessing import normalize\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "#getting data from the CSV file\n",
    "dataset = pd.read_csv('C:/Users/sneha/Desktop/Application job/Frauhofer task/Task 2/20200709_Results_Exp_Sim.csv', sep=';', decimal=',')\n",
    "\n",
    "dataset_Exp = dataset[0:60]\n",
    "dataset_Sim = dataset[40:6z0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ML_Model:\n",
    "    def __init__(self, X, y, MP):\n",
    "        self.X = X\n",
    "        self.y = y\n",
    "        self.X_train, self.X_test, self.y_train, self.y_test = train_test_split(self.X, self.y, test_size=.2, shuffle= True)\n",
    "        self.Mean_Squared_Error={}\n",
    "        self.MP = MP\n",
    "    def Linear_regression(self):\n",
    "        regressor = LinearRegression()  \n",
    "        self.Error= cross_validate(regressor, X, y, cv=5, scoring=('neg_mean_squared_error'))\n",
    "        self.Mean_Squared_Error.update({'Linear_regression.................':-self.Error[\"test_score\"]})\n",
    "        \n",
    "    def Gaussian_regression(self):\n",
    "        kernel = C(1.0, (1e-3, 1e3)) * RBF(10, (1e-2, 1e2))\n",
    "        gp = GaussianProcessRegressor(kernel=kernel, alpha=9,n_restarts_optimizer=10)\n",
    "        self.Error= cross_validate(gp, X, y, cv=5, scoring=('neg_mean_squared_error')) \n",
    "        self.Mean_Squared_Error.update({'Gaussian_regression...............':-self.Error[\"test_score\"]})\n",
    "        \n",
    "    def SupportVector_regression(self):\n",
    "        svr_lin = SVR(kernel='linear', C=1, gamma='auto')\n",
    "        svr_rbf = SVR(kernel='rbf', C=100, gamma=0.1, epsilon=.1)\n",
    "        svr_poly = SVR(kernel='poly', C=100, gamma= 'scale', degree=2, epsilon=.1, coef0=1)\n",
    "        self.Error = cross_validate(svr_lin, X, y, cv=5, scoring=('neg_mean_squared_error'))\n",
    "        self.Mean_Squared_Error.update({'SupportVector_regression_lin......':-self.Error[\"test_score\"]})\n",
    "        self.Error = cross_validate(svr_rbf, X, y, cv=5, scoring=('neg_mean_squared_error'))\n",
    "        self.Mean_Squared_Error.update({'SupportVector_regression_rbf......':-self.Error[\"test_score\"]})\n",
    "        self.Error = cross_validate(svr_poly, X, y, cv=5, scoring=('neg_mean_squared_error'))\n",
    "        self.Mean_Squared_Error.update({'SupportVector_regression_poly......':-self.Error[\"test_score\"]})\n",
    "    \n",
    "    def DecisionTree_Regressor(self):\n",
    "        regr = DecisionTreeRegressor(max_depth=6)\n",
    "        self.Error = cross_validate(regr, X, y, cv=5, scoring=('neg_mean_squared_error')) \n",
    "        self.Mean_Squared_Error.update({'DecisionTree_Regressor............':-self.Error[\"test_score\"]})\n",
    "        \n",
    "    \n",
    "    def covariance(self):\n",
    "        print(\"The covarience matrix is\")\n",
    "        A= np.dot(np.transpose(X),X)\n",
    "        with np.printoptions(precision=7, suppress=True):\n",
    "            print(np.linalg.inv(A))\n",
    "    \n",
    "\n",
    "        \n",
    "    #def ELM_(self):\n",
    "        #dataset = normalize(dataset)\n",
    "        #dataset.target = [1 if i == 1 else -1 for i in  dataset.target.astype(int)]\n",
    "        #elm = ELM(hid_num=10).fit(self.X_train, self.y_train)\n",
    "        #y_pred = elm.predict(self.X_test)\n",
    "        #self.Mean_Squared_Error.update({'ELM...............................': metrics.mean_squared_error(self.y_test, y_pred)})\n",
    "\n",
    "        \n",
    "    \n",
    "    def Print(self):\n",
    "        print(\"The mean squred error for\",self.MP, \"\\n\")\n",
    "        for key, value in self.Mean_Squared_Error.items() :\n",
    "            print (key, value, \"\\n\")\n",
    "            \n",
    "    def plot(self):\n",
    "        fig = plt.figure()\n",
    "        fig.suptitle('EXP+SIM '+self.MP+'.jpeg')\n",
    "        ax = fig.add_subplot(111)\n",
    "        plt.boxplot(self.Mean_Squared_Error.values())\n",
    "        ax.set_xticklabels(['LR','GR','SVR-l','SVR-RBF','SVR-POLY','DTR'])\n",
    "        plt.show()\n",
    "    \n",
    "    def func_call(self):\n",
    "        self.Linear_regression()\n",
    "        self.Gaussian_regression()\n",
    "        self.SupportVector_regression()\n",
    "        self.DecisionTree_Regressor()\n",
    "        self.covariance()\n",
    "        self.plot()\n",
    "        self.Print()\n",
    "       "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60, 3)\n",
      "(60,)\n",
      "The covarience matrix is\n",
      "[[ 0.0000001 -0.0000108 -0.0000216]\n",
      " [-0.0000108  0.0049142 -0.0000179]\n",
      " [-0.0000216 -0.0000179  0.0051452]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The mean squred error for MP1 \n",
      "\n",
      "Linear_regression................. [ 832.25171214  907.93380709 1352.80921063 1200.64742493  532.72525903] \n",
      "\n",
      "Gaussian_regression............... [498.91375523 407.52239568 936.27616955 326.92680316 530.49998721] \n",
      "\n",
      "SupportVector_regression_lin...... [1035.87901674  901.038562   1241.6351823  1333.25238256  572.74003169] \n",
      "\n",
      "SupportVector_regression_rbf...... [457.83418779 318.58859204 723.81322526 411.07319555 558.01527349] \n",
      "\n",
      "SupportVector_regression_poly...... [3268.37051471 1243.96253944 2029.61672449 3912.91335812  791.7236342 ] \n",
      "\n",
      "DecisionTree_Regressor............ [505.26041667 412.88541667 993.53125    326.73958333 525.58333333] \n",
      "\n",
      "The covarience matrix is\n",
      "[[ 0.0000001 -0.0000108 -0.0000216]\n",
      " [-0.0000108  0.0049142 -0.0000179]\n",
      " [-0.0000216 -0.0000179  0.0051452]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The mean squred error for MP2 \n",
      "\n",
      "Linear_regression................. [187.66772177 548.29559514 156.22101049 108.91506193 340.30537881] \n",
      "\n",
      "Gaussian_regression............... [189.01421648 501.05409235 224.45217078 146.08937755 435.25775109] \n",
      "\n",
      "SupportVector_regression_lin...... [197.40606163 615.22421455 165.14086322 123.7674511  365.0175    ] \n",
      "\n",
      "SupportVector_regression_rbf...... [162.14172345 636.48406778 236.33888756 172.31020207 369.65638222] \n",
      "\n",
      "SupportVector_regression_poly...... [ 584.48617692 1389.61795142  758.49363276  797.3883621   336.10478696] \n",
      "\n",
      "DecisionTree_Regressor............ [219.84375    534.79166667 261.88541667 153.47916667 461.91666667] \n",
      "\n",
      "The covarience matrix is\n",
      "[[ 0.0000001 -0.0000108 -0.0000216]\n",
      " [-0.0000108  0.0049142 -0.0000179]\n",
      " [-0.0000216 -0.0000179  0.0051452]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The mean squred error for MP3 \n",
      "\n",
      "Linear_regression................. [189.63410457 152.3163916  459.10377255 341.94142982 118.43762935] \n",
      "\n",
      "Gaussian_regression............... [171.86527123 163.82217052 376.44271865 303.56451198 124.3318101 ] \n",
      "\n",
      "SupportVector_regression_lin...... [334.69188842  47.61914195 474.31934228 563.92872079 102.53089166] \n",
      "\n",
      "SupportVector_regression_rbf...... [236.13451576  48.29788431 406.25551205 498.87222576 127.99130703] \n",
      "\n",
      "SupportVector_regression_poly...... [467.47107267  32.66233176 569.12130675 745.14192704 144.54454958] \n",
      "\n",
      "DecisionTree_Regressor............ [153.42708333 165.08333333 429.71875    271.20833333 124.79166667] \n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\gaussian_process\\gpr.py:469: ConvergenceWarning: fmin_l_bfgs_b terminated abnormally with the  state: {'grad': array([-4.12404381e+02,  1.80931232e-05]), 'task': b'ABNORMAL_TERMINATION_IN_LNSRCH', 'funcalls': 49, 'nit': 6, 'warnflag': 2}\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The covarience matrix is\n",
      "[[ 0.0000001 -0.0000108 -0.0000216]\n",
      " [-0.0000108  0.0049142 -0.0000179]\n",
      " [-0.0000216 -0.0000179  0.0051452]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The mean squred error for MP4 \n",
      "\n",
      "Linear_regression................. [1580.10243329 1212.12024002  909.73052797 1411.7336681   897.59153988] \n",
      "\n",
      "Gaussian_regression............... [1754.35539502 1396.01203071 1405.47509838  714.09201768  783.65934116] \n",
      "\n",
      "SupportVector_regression_lin...... [1391.82666667 1949.54333333  924.66585509 1528.69318528  899.37299741] \n",
      "\n",
      "SupportVector_regression_rbf...... [1341.68492028 1199.60421945 1036.23659524 1130.92488664  725.17536398] \n",
      "\n",
      "SupportVector_regression_poly...... [2399.66165488 3805.88833296 1910.25629605 3791.3917941  2911.14965323] \n",
      "\n",
      "DecisionTree_Regressor............ [1853.95833333 1411.15625    1439.10416667  736.71875     743.58333333] \n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\gaussian_process\\gpr.py:469: ConvergenceWarning: fmin_l_bfgs_b terminated abnormally with the  state: {'grad': array([-2.78549754e+02, -3.50784367e-05]), 'task': b'ABNORMAL_TERMINATION_IN_LNSRCH', 'funcalls': 48, 'nit': 5, 'warnflag': 2}\n",
      "  ConvergenceWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The covarience matrix is\n",
      "[[ 0.0000001 -0.0000108 -0.0000216]\n",
      " [-0.0000108  0.0049142 -0.0000179]\n",
      " [-0.0000216 -0.0000179  0.0051452]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The mean squred error for MP5 \n",
      "\n",
      "Linear_regression................. [923.63324628 605.17380425 794.77133333 947.05514137 477.57014619] \n",
      "\n",
      "Gaussian_regression............... [937.92258325 572.77667107 537.71735216 449.49904062 317.17396921] \n",
      "\n",
      "SupportVector_regression_lin...... [2014.87028429  799.68255963 1444.96839607 1448.52496077  427.37855031] \n",
      "\n",
      "SupportVector_regression_rbf...... [1281.85010027  580.75973244  692.84144451  402.02469821  252.36937299] \n",
      "\n",
      "SupportVector_regression_poly...... [1347.08978029  836.63272422 3060.27183605 2865.91931522   98.67858727] \n",
      "\n",
      "DecisionTree_Regressor............ [1222.23958333  539.45833333  623.88541667  367.54166667  350.5       ] \n",
      "\n",
      "The covarience matrix is\n",
      "[[ 0.0000001 -0.0000108 -0.0000216]\n",
      " [-0.0000108  0.0049142 -0.0000179]\n",
      " [-0.0000216 -0.0000179  0.0051452]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The mean squred error for MP6 \n",
      "\n",
      "Linear_regression................. [ 843.28964073  585.03098085  599.31039933 1079.60916169  541.2562505 ] \n",
      "\n",
      "Gaussian_regression............... [206.37933466 209.52404872 129.23667136 268.29211764  53.38291712] \n",
      "\n",
      "SupportVector_regression_lin...... [1680.91085521  720.31083859 1226.59483878 2269.12341614  206.20059338] \n",
      "\n",
      "SupportVector_regression_rbf...... [ 93.25616162 149.84207949 184.58618538 255.37784413  34.76201796] \n",
      "\n",
      "SupportVector_regression_poly...... [2603.38677438  100.40954603 2205.35448003 3452.30589208   37.05534242] \n",
      "\n",
      "DecisionTree_Regressor............ [ 91.90625    123.59375    179.28125    138.26041667  43.33333333] \n",
      "\n",
      "PCA \n",
      "\n",
      "Explained variance ratio: \n",
      " [0.97997641 0.01037999 0.0096436 ]\n",
      "Singular values: \n",
      " [138.5749971   14.26183719  13.7466425 ]\n"
     ]
    }
   ],
   "source": [
    "X_exp = dataset_Exp[['T_O', 'delta_t_h','t_a']].values\n",
    "X_sim = dataset_Sim[['T_O', 'delta_t_h', 't_a']].values\n",
    "X = np.concatenate([X_exp, X_sim])\n",
    "#X = np.transpose(np.array([X[:,0], X[:,1], X[:,2], X[:,1]*X[:,1], X[:,2]*X[:,2] ]))\n",
    "y = np.concatenate([dataset_Exp['HV_MP1_exp'].values, dataset_Sim['HV_MP1_sim'].values])\n",
    "print(X.shape)\n",
    "print(y.shape)\n",
    "MP1=ML_Model(X,y,'MP1')\n",
    "MP1.func_call()\n",
    "            \n",
    "y = np.concatenate([dataset_Exp['HV_MP2_exp'].values, dataset_Sim['HV_MP2_sim'].values])\n",
    "MP2=ML_Model(X,y,'MP2')\n",
    "MP2.func_call()\n",
    "\n",
    "y = np.concatenate([dataset_Exp['HV_MP3_exp'].values, dataset_Sim['HV_MP3_sim'].values])\n",
    "MP3=ML_Model(X,y,'MP3')\n",
    "MP3.func_call()\n",
    "\n",
    "y = np.concatenate([dataset_Exp['HV_MP4_exp'].values, dataset_Sim['HV_MP4_sim'].values])\n",
    "MP4=ML_Model(X,y,'MP4')\n",
    "MP4.func_call()\n",
    "\n",
    "y = np.concatenate([dataset_Exp['HV_MP5_exp'].values, dataset_Sim['HV_MP5_sim'].values])\n",
    "MP5=ML_Model(X,y,'MP5')\n",
    "MP5.func_call()\n",
    "\n",
    "y = np.concatenate([dataset_Exp['HV_MP6_exp'].values, dataset_Sim['HV_MP6_sim'].values])\n",
    "MP6=ML_Model(X,y,'MP6')\n",
    "MP6.func_call()\n",
    "\n",
    "#PCA\n",
    "pca = PCA(n_components=3, svd_solver='randomized')\n",
    "X= pca.fit_transform(X)\n",
    "print(\"PCA \\n\")\n",
    "print(\"Explained variance ratio: \\n\", pca.explained_variance_ratio_)\n",
    "print(\"Singular values: \\n\", pca.singular_values_)\n",
    "    \n",
    "#ELM\n",
    "#https://towardsdatascience.com/build-an-extreme-learning-machine-in-python-91d1e8958599\n",
    "#https://github.com/masaponto/Python-ELM\n",
    "\n",
    "\n",
    "#GLM\n",
    "#https://towardsdatascience.com/generalized-linear-models-9cbf848bb8ab            \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_exp = dataset_Exp[['T_O', 'delta_t_h','t_a']].values\n",
    "X_sim = dataset_Sim[['T_O', 'delta_t_h', 't_a']].values\n",
    "X = np.concatenate([X_exp, X_sim])\n",
    "#X = np.transpose(np.array([X[:,0], X[:,1], X[:,2], X[:,1]*X[:,1], X[:,2]*X[:,2] ]))\n",
    "y = np.concatenate([dataset_Exp['HV_MP2_exp'].values, dataset_Sim['HV_MP2_sim'].values])\n",
    "#svr_poly = SVR(kernel='poly', C=100, gamma= 'scale', degree=2, epsilon=.1, coef0=1)\n",
    "\n",
    "train_scores, test_scores = validation_curve(DecisionTreeRegressor(), X, y,'max_depth',[x for x in range(1,10)],cv=5, scoring='neg_mean_squared_error')\n",
    "#train_scores, test_scores = validation_curve(SVR(), X, y,'degree',[x for x in range(1,5)],cv=5, scoring='neg_mean_squared_error')\n",
    "\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",
    "\n",
    "plt.title(\"Validation Curve with Decision tree\")\n",
    "plt.xlabel(r\"$\\gamma$\")\n",
    "plt.ylabel(\"Score\")\n",
    "plt.ylim(0.0, 2000)\n",
    "plt.plot([x for x in range(1,10)], train_scores_mean, label=\"Training score\",color=\"darkorange\", lw=2)\n",
    "plt.fill_between([x for x in range(1,10)], train_scores_mean - train_scores_std,train_scores_mean + train_scores_std, alpha=0.2,color=\"darkorange\", lw=2)\n",
    "plt.plot([x for x in range(1,10)], test_scores_mean, label=\"Cross-validation score\",color=\"navy\", lw=2)\n",
    "plt.fill_between([x for x in range(1,10)], test_scores_mean - test_scores_std, test_scores_mean + test_scores_std, alpha=0.2, color=\"navy\", lw=2)\n",
    "plt.legend(loc=\"best\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:1978: FutureWarning: The default value of cv will change from 3 to 5 in version 0.22. Specify it explicitly to silence this warning.\n",
      "  warnings.warn(CV_WARNING, FutureWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:1360: RuntimeWarning: Removed duplicate entries from 'train_sizes'. Number of ticks will be less than the size of 'train_sizes' 18 instead of 20).\n",
      "  % (train_sizes_abs.shape[0], n_ticks), RuntimeWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:1978: FutureWarning: The default value of cv will change from 3 to 5 in version 0.22. Specify it explicitly to silence this warning.\n",
      "  warnings.warn(CV_WARNING, FutureWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:1360: RuntimeWarning: Removed duplicate entries from 'train_sizes'. Number of ticks will be less than the size of 'train_sizes' 18 instead of 20).\n",
      "  % (train_sizes_abs.shape[0], n_ticks), RuntimeWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:1978: FutureWarning: The default value of cv will change from 3 to 5 in version 0.22. Specify it explicitly to silence this warning.\n",
      "  warnings.warn(CV_WARNING, FutureWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:1360: RuntimeWarning: Removed duplicate entries from 'train_sizes'. Number of ticks will be less than the size of 'train_sizes' 18 instead of 20).\n",
      "  % (train_sizes_abs.shape[0], n_ticks), RuntimeWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:1978: FutureWarning: The default value of cv will change from 3 to 5 in version 0.22. Specify it explicitly to silence this warning.\n",
      "  warnings.warn(CV_WARNING, FutureWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:1360: RuntimeWarning: Removed duplicate entries from 'train_sizes'. Number of ticks will be less than the size of 'train_sizes' 18 instead of 20).\n",
      "  % (train_sizes_abs.shape[0], n_ticks), RuntimeWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_split.py:1978: FutureWarning: The default value of cv will change from 3 to 5 in version 0.22. Specify it explicitly to silence this warning.\n",
      "  warnings.warn(CV_WARNING, FutureWarning)\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:1360: RuntimeWarning: Removed duplicate entries from 'train_sizes'. Number of ticks will be less than the size of 'train_sizes' 18 instead of 20).\n",
      "  % (train_sizes_abs.shape[0], n_ticks), RuntimeWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAagAAAEYCAYAAAAJeGK1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nO3deXhU5dn48e+dyUpICISdIJsIorIoIooL7riiFhR3qy2tVV8tXcTaur31/Vmrdalb0SparYo7WtxQKS64BIkKYd8DyBIgC2Sbmfv3xzmTTMIkmYRMZia5P9d1rjn73HMCc8/znOc8j6gqxhhjTKxJiHYAxhhjTCiWoIwxxsQkS1DGGGNikiUoY4wxMckSlDHGmJhkCcoYY0xMsgRl2jQReVdErox2HK1JRH4hIg82sH2diJzSmjG1FhHpISJLRSQl2rGY/WcJykRErHwJquoZqvpstONoLSKSDPwR+Gu0Y4kUEblQRL4Qkb0iMi94m6puBT4BpkYlONOiLEGZuCUiidGOYX9F4DNMBJap6qYWPm+jWvHvsRN4ELinnu0vAL9opVhMBFmCMq1ORM4WkTwR2e3+Eh4etG26iKwWkRIRyReR84O2XSUin4vIAyKyE7jDXfeZiNwnIrtEZK2InBF0zDwR+VnQ8Q3tO0BE5rvvPVdEHhWR5xv4HBPdz1HsxjzBXV+r9CgidwTOIyL9RURF5BoR2QB8LCLvicj1dc79nYhc4M4PFZEPRWSniCwXkQsbuLxnAP+tc67LRWS9iBSKyK11tiUEXfNCEZklIl2Ctl8RdOyfgj+b+7leFZHnRaQYuCqM8411/+a73c84voHPEpKqzlXVWcDmenb5ChgoIv2aem4TWyxBmVYlIocDT+P8ws0G/gHMDrpnsBo4DugE3Ak8LyK9gk5xFLAG6A7cHbRuOdAVuBf4p4hIPSE0tO+/ga/duO4ALm/gc4wBngN+B2QBxwPrGvv8QU4ADgZOd9/34qBzDwP6Af8RkXTgQ3ef7u5+j4nIIfWc9zD38wWf63H3s/R2P1tO0P7/A5znxtMb2AU8GnTsY8ClQC+cv0mfOu83EXgV5xq80Mj5+gD/Af4MdAF+C7wmIt3c7Y+5iSvU9H1DFzOYqnqBVcCIcI8xMUpVbbKpxSecL+tTQqx/HPjfOuuWAyfUc548YKI7fxWwoc72q4BVQcsdAAV6usvzgJ81ti9wAOAFOgRtfx54vp64/gE8EM5nx0l2z7vz/d33HBi0PQPYA/Rzl+8GnnbnLwI+DfHet9fz3iuBCUHLtwEvBS2nA5WB+IClwMlB23sBVUCie+yLda5X8LF3APPrvH9D57sZ+Fed/d8Hrmzmv7GfAfPq2fY5cEW0/x/YtH+TlaBMa+sH/Cb41zHQF+fXdqBKKS9o26E4pZ2AjSHO+WNgRlX3urMd63n/+vbtDewMWlffewX0xSntNVf1uVW1BKdkMcVdNQWnNALO9TqqzvW6FCephrILJ+EF9K7zXnuAwqDt/YA3gs69FPABPUIcu7fOsbU+Rxjn6wdMrvNZjsVJYi0tA9gdgfOaVhT3N5lN3NkI3K2qd9fd4N4zeBI4GVigqj4RyQOCq+si1f3+FqCLiHQISlJ9G9h/IzConm17cEobAaGSSd3P8SJwu4jMB9JwWqIF3ue/qnpqQ8EH+R44KGh5C05VIgAi0gGnmi9gI3C1qn5e90QisgUYErScVufYUJ+jofNtxClB/TxU4CLyBHBZqG3AelWtr1qz7nkSgQOB78LZ38QuK0GZSEoSkdSgKREnAf1SRI4SR7qInCUiGTjVTwpsBxCRn+KUoCJOVdcDuTgNL5JF5GjgnAYO+SfwUxE52W0Y0EdEhrrb8oApIpIkIqOBSWGEMAenhHEX8LKq+t317wAHuQ0dktzpSBE5uIHznBC0/CpwtogcK04T9Luo/f/+CeDuQIMCEekmIhODjj1HRI5xj72T2j8WQmnofM+75ztdRDzuv4nxIpIDoKq/VNWO9UzVySlwLM4P7AT3PElBMYwB1rl/UxPHLEGZSJoDlAVNd6hqLvBz4BGc6qhVOPeGUNV84H5gAbAV54b/Pr/EI+hS4Gicaqw/Ay8DFaF2VNWvgZ8CDwBFOC3nAq3G/oRTutqF86X+78beWFUrgNeBU4L3d6v/TsOp9tuMU0X5F6C+B1HfBoaKSG/3+CXAde45t7gxFQTt/xAwG/hAREqAL3EakgSOvQF4yT22BNhW3zUJ43wbcRpV/AHnR8hGnEYmTf0euhzn39PjOA1qynB++ARcipMoTZwTVRuw0JhQRORlnGeKbo92LE0hIlOBYap6UwuftyPOfZ3Bqrq2Jc/dUkSkO86PhVGqWh7teMz+sQRljEtEjsR5CHQtTqnlTeBoVV0U1cCiSETOAT7Cqdq7H6c0dLjaF4dpBRGr4hORp0Vkm4gsrme7iMjDIrJKRL53n48xJpp64jRLLwUeBq5tz8nJNRGnanEzMBiYYsnJtJaIlaBE5Hic/+jPqeo+N7pF5Eyc+u0zcX6VPaSqR0UkGGOMMXEnYiUoVZ2PU11Sn4k4yUtV9Usgq06PAcYYY9qxaD4H1YfaD/kVuOu21N3Rvek7FSA9Pf2IoUOH1t3FGGNMnFq4cOEOVe1Wd300E1So5ylC1jeq6gxgBsDo0aM1Nzc3knEZY4xpRSIS8pm1aD4HVUDtJ/VzqL93YmOMMe1MNBPUbOAKtzXfWKBIVfep3jPGGNM+RayKT0ReBMYDXUWkALgdSAJQ1Sdwehk4E6cngb04T+UbY4wxQAQTlKpe3Mh2xemCxRhjjNmH9cVnjDEmJlmCMsYYE5MsQRljjIlJlqCMMcbEJEtQxhhjYpIlKGOMMTHJEpQxxpiYZAnKGGNMTLIEZYwxJiZZgjLGGBOTLEEZY4yJSe0vQfm8sG1ZtKMwxhjTiPaVoEq2wkMjYOaZUFUe7WiMMcY0oH0lqI7doUMX2FsIS16PdjTGGGMa0L4SlAgc9Qtn/qt/gIYcYd4YY0wMaF8JCuDQn0BaF9iSBwXfRDsaY4wx9YjYgIUxKykNjrgSPnvAKUX1HRPtiIwx8cZXBRUlULknaApeLnVeK0pr5qvXl9Y+zlsO6d0hsxdk9ILM3s6U0dtd19u5NSES7U/d6tpfggIYfQ18/hDkvwkld0NGz2hHZEx88/ude7slm6F4s/PFKwnOl6okANLAsjSyvaX2F/BWNJI0ghPLnn3XB159lS17/fZsh21L6t/uSalJVpm99k1gmb2d7zFPUsvGFWXtM0Fl9YWhZ8HStyH3GTjxlmhHZEzs8nmhdKuTeIo3QckW57V4MxRvqVnX0l/asUw8kNIRkgNTuju58ymh1gctp2TUzCckwZ5tta9l8ZaaZF+8BSqKYNc6Z6o/KEjvVjtp1UpqfZwSWmpmK12k/dc+ExTAmF84CWrhM3DcbyAxOdoRGVVY/bHzSzy1E6RmQVpWzXxSauvH5PM6VTdVZU71cEomJHhaP45IqSoP+iIMngJflJud5KT+xs+VmlVTPZWS4TZCUudY1TrLgXX+Ovu479Pg9vrO2dj+OP+GghPJPq+hkktG7UST0hE8yS1b5ZbZC3qNqH975Z46CWxzzWtgvnSrk+j2bIMt39V/ruSOdaoSQ8ynd4uJf+ftN0H1Pxa6D4Nt+ZD/FgyfHO2I9p/PC9/OhK+fchqDHPcbSIiTdjBlu2H2DbB0dv37JKbWSVxu8qpvPrWTc6+gsqSmWqeixH0tranaqaj7WlJTleMN8bxckvsLODXTeU3JcBJXSmZ465M6OOfR4C/r+qa6+/ga2R5i8vucKqRA4ineUjNftjOMP45Axx7ul1efml/mgfnAL/Tk9Gb96U0YktOh64HOVJ9ASbe+BBZ4rSyFwpXOVB/xuH/v+u6LueuS0lr+swaHoXHW1Hr06NGam5vbMifLfQbeuQlyjoSfzW2Zc0bLyrnwwa2wPaiXjIPPhfOfiP0vjo3fwKtXQ9EG58v8wFOgothJWuW7obzImfdXtX5skuD8gk5Kc0pRFcU4P8XbiITEEF9AgamP82XUsafVMLQVqs7/qeoqxEDi2ly7WnFvYXjnS+sMh10IZ967X2GJyEJVHV13ffstQQEMvxDm3u40N9/0LfQ5PNoRNd3WfPjgj7D6I2c5qx8cfoXTCGTpbNi5Fi7+N2QdEN04Q/H7YcHf4aO7wO+F3qNg0jPQZcC++6pC1d6aZBWcuMp3u69FdeaLnJvGKRk1VTPVrxm1q3P22cddTkqrXZXj90PVHqeUVV7svFYUu1NJeOsr99TczK93amR7QkPHhji+Q9c6pR53Pr1b/JSyzf4TcZJKWmfoMaz+/bwVNffC6rsvVrIFynY5/3cjFW67LkEBvH8rLHgERlzslDbiRel2mPd/sHCmU42TkgnH/855EDkxBbavgBenwM7VzpfTRc9Dv6OjHXWNPTvgjV/Cqg8B+HHYNbzU6Rq+WFdMx5RERuRkMaJvJ0bkZNE53X69GxNzAi03UaeXnv1QXwnKEtTOtfDwKOeX9q/zoWO3ljt3JFSVw1ePw/z7nXsr4oHRP4Xxt0B619r7lu2CV34Kaz5xWgqddb/zDFi0rf0U76s/I3HPj5QmZHCz71f8p6L+G8T9szswom+Wm7SyOKR3JqlJ0b+Ba4xpGZagGvLvKbDiXTjpj04pJFatnAv/+TXs3uAsDz4NTv1f6D60/mN8XvjwT/DlY87yUb+E0+4GT+vW7haXV7Fg1TYSP7uP8T/OxIOfr/xDubHyOn4km4Hd0jl+cDeOPbArZVU+8jbu5ruNu/lhUxEV3totyBIThIN7ZVaXsEb2zWJQt44kJLS/BxmNaQssQTVk9cfwr/OdG8Q3fR+bD7t9/SS8+3unOq/bwXD63XDgyeEf/+2/4J1fOw0NBo537vV06BKpaPH5le8LdvPpyh18unI7BRvW8DfPoxztycevwpNyPj8ceC3HHtSTYwd3Jadzh5DnqfL5WbG1hO82FvHdxt3kbdzNim0l+3Sj2DElkeE5napLWqMOyKJHZhSapRtjmswSVENU4dExsGMFTJ4Jh5zfsuffH34/fHQnfP6gs3zCzXD875tXAtrwJbx8mdPkuMtAmPJiw6WvJirYtbc6IX2+qpCiMqfV3fiEPO5PepxsKWFPUjabT36YgWPOwtPMEk9phZfFm5yE9V3BbvI27GZz0b7NwXtmpjK6f2eOHpTN0QOzGdA1HWmH3cUYE+ssQTXm6ydhzm/hgGPg6ndb/vzN4a2At66DH15xmgOf8xCMumz/zrl7I7x0Mfz4g9O6K2cMDJkAQ86Ergc16+HDx+et5pWFG1mzfU+t9QO7JHNnx9c5btu/3RUnwgUz9vuGaijbisv5riAoaW3cTUl57dZFPTNTq5PV0YOy6dsldKnNGNO6LEE1pqIE7j/YaXjwy8+g52EN7793J+T+03mAceRlLd9Ut2y3U9pZ96nT3PnCZ53ng1pC5R6Y8zv4flbtZ4s6D4AhZzjTAUeHVdW5tbico/7PaeLeMSWRYwZlc9xB3TixRxk5c6+DTblOQ46T/gjjbmq1Js1+v7JmRylfrtnJgtWFfLmmkMI9tbvi6ZOVVith9c6K7EOHxpjQLEGF492b4asnYNTlMPGR0Pvs3ek0S//qH84T2QADjoeJjzl9/LWEogJ4fhJsX+o8JHnprIa7QWmuihLn/tvyd2HF+7V7FUjtBAee6iSrA09xemgI4aOlW7nm2VxG9+vMi1PHkuRJcHrmeOsGp/+wzByY9E84YGzLx98EqsqKraUsWL2DL1YX8tXandVVkAH9sjtwzKBsxroJq3uG3cMypjVEJUGJyATgIcADPKWq99TZfgDwLJDl7jNdVec0dM6IJqgdq+CRI5wudaYtrd2IYE+hk5i+nhGUmE6ArUtg7w7nOaQz/uI8T7U/9zl+/AFemOw8BNd1CFz2aus8ZOv3OQ8sL58Dy9+DHctrtiUkOiWqQOmqy8DqTQ9/tJK/fbiCa44dwJ9OH+j0ZvHNU87GIWc5iT6CjTGay+dXlm4p5ss1hSxwE1ZpRe0qwUHd0t0SVlfGDuxCdseUKEVrTNvW6glKRDzACuBUoAD4BrhYVfOD9pkBLFLVx0VkGDBHVfs3dN6IJiiA538Cq+bCKXfCsTeFTkwHngInTIe+RzoPzL5zEyx7x9k25Cw458Hm3WdZ/TG8fIVTzdhvHEx5wXniOxoKV8OK95zS1fovnD7gAroOqb5v9ct5wnv5O5hxZianLbkFtv7gdKR52p9hzNS4GcPG6/OzeHMxC1YXsmBNId+s3UlZla/WPjmd0zikdybDenViWO9MDumdSa9Oqdbwwpj9FI0EdTRwh6qe7i7fAqCq/y9on38Aa1T1L+7+96vqMQ2dN+IJasUH8O/J0KkvHDa5/sQUTBW+e8lpBl5RDB2y4ewHYNjE8N6zci/8MAv+8xun25BDLnB6tUiMkV/sZbucZ7BWvOu8VhRVb9pNJp/5hnFmyvckePc6patJz0DvkVEMeP9Vev18X7C7OmEtXL9rn+exALI6JDGsl5OshrnJa1C3dBI91n2QiU+qSoXXT1mljz2VXsoqfeytnrzsrfS567zsqfQxtGcGJx/cY7/eMxoJahIwQVV/5i5fDhylqtcH7dML+ADoDKQDp6jqwhDnmgpMBTjggAOOWL9+fURiBpxm3Y8cATvX1KyrLzHVVVTgtLpbM89ZHn4RnHFv7fs3qs5+G79yqtQ2fuVU6wX6szrmf5zSW6z2j+arckpUy9/Ft/xdPLvX1Ww7dJKTmONovJlweX1+Vm/fQ/6WIvI3F5O/pZglm4vZvXffDmyTExMY2jOjVuIa2jOT9JT23fWlaVlVPj97K3zsrapJGnsqvOyt8gUlFW91cilzE0pZnfWBbcFJyN+EtDDpiBzum7x/98ijkaAmA6fXSVBjVPWGoH2muTHc75ag/gkcqlr/4DMRL0EBfPcyvDE1/MQUzO93Wvd98CfwljkP/55yh/PsUSAplWypfYwkQPdDYMzPY6MrojAtWLWDP/3zNS7usoJrzjnJaarejqq7VJUtReVBCauI/C3FbNxZts++IjAgO52De2cyuHtHUpM8eERISBA8Ap6EwHzNa/A6TwIk7LNOqtfV2l69zplPTBBSkzx0SPHQIcljpbtGqCp+de5T+vyK1++vnneWQ88H9vWr4vUFbVfF53Pm/Ro4xo/PDz6/nyqf1iSUKq+TdCp9lLmJJ1QSKqvyUeWLXPuBZE8CackeOiR7SEv2kJ6cWL3sTInV20bkZHHmYb326/2i0Zt5ARDcrC0H2Fxnn2uACQCqukBEUoGuwLYIxtW4ERfBoRc0r0eJhAQn0Qw6yekMteBrJ9kFS82CvmOcZ5D6jnF6UU/JaJnYW9GSLcWs0hxWDToGhjbSLL8NEhF6Z6XROyuNU4bVVHEUlVWxzC1h5W8pJn9zMSu3lbBmxx7W7NjTwBlbR3JiAh2CvnTS63wJVb+m1HwR1f1SSg+sT0mkg5v8kj0JDd6P8/qcL+NKr58Kn696vtLrp8rnpyJovtLrp9JXz3p3W2XQcvAxlV51X/d9jyq/H7+fWknHG/Tqd1/jgSdBav1t0pKcv1lasvs3SXZ/lLjbQv0d6/59A0koKUZ+xEQyQX0DDBaRAcAmYApwSZ19NgAnAzNF5GAgFdgewZjCt7/dHWUPgqvfcxpYLJsD3Q5yE9JRkH1g7FbhNUH+5mIADund9qr09kentCSOGpjNUQOzq9dVev2s3FZC/uZi1u7YU+uXt19rv/r81Myr86VZa7tSva56e9Crz0+tdV6/Ul5VU30T+MIOVT25P4K/MD0iVLrJJZA84uR7H4AEt1TrSRASExJIEEj0JLjLbsnUIyGWE0gMKuEG9gleDpRqq/f1iJNckt3kUudHQFqdxBJY19gPgrYgYglKVb0icj3wPk4T8qdVdYmI3AXkqups4DfAkyLya5xR4K7SeHswqyEJHhh3ozO1QUssQYUtOTGBQ3p34pDenaIaR+AG+N6gqqI9FV73hnj99ywC9zb2Vnj3uWEemK/yKSXl3n168AhIEOc6JHkSSElMINmTQJL7GlifnOhsS/LUt15ITkwg2eMhKVFI9gTtn1hn/xDnT/IEJZ0ESEyoSTqBhBSoZjXRF9G7tu4zTXPqrLstaD4fGBfJGExklFf5WLW9lASBoT0tQcULEed+VGqShy4tPM5WZVDLLwWSPEKKx+MmB7F7X6bJrFmRaZblP5bg8yuDu3ckLdnGZjJUl2A6dYjB0QBMXLKfNKZZrHrPGBNplqBMsyzZ7DysG+17KsaYtssSlGkWK0EZYyLNEpRpMp9fWfajk6CGWYIyxkSIJSjTZGu2l1Je5adPVhpZHVq2JZgxxgRYgjJNZtV7xpjWYAnKNJk1kDDGtAZLUKbJrARljGkNlqBMk6hqTYLqYwnKGBM5lqBMk2zaXUZRWRVd0pPpmZka7XCMMW2YJSjTJMHVe229J2VjTHRZgjJNEkhQ9vyTMSbSLEGZJsl3W/AN62UJyhgTWZagTJPUVPFZE3NjTGRZgjJh27mnki1F5aQleRjQNT3a4Rhj2jhLUCZsgQd0D+6VgcdGHDXGRJglKBM2q94zxrQmS1AmbNaDhDGmNTWaoESkg4j8SUSedJcHi8jZkQ/NxBrrg88Y05rCKUE9A1QAR7vLBcCfIxaRiUl7Krys3bGHxAThoJ4dox2OMaYdCCdBDVLVe4EqAFUtA+wOeTuz7MdiVOHA7h1JSfREOxxjTDsQToKqFJE0QAFEZBBOicq0I9ZAwhjT2hLD2Od24D2gr4i8AIwDropkUCb25FsDCWNMK2s0QanqhyLyLTAWp2rvRlXdEfHITEyxFnzGmNYWTiu+8wGvqv5HVd8BvCJyXuRDM7Giyudn+Y8lgHUSa4xpPeHcg7pdVYsCC6q6G6faz7QTq7aVUunz0y+7AxmpSdEOxxjTToSToELtE869K9NGWPWeMSYawklQuSLyNxEZJCIDReQBYGGkAzOxwx7QNcZEQzgJ6gagEngZeAUoB66LZFAmttgghcaYaAinFd8eYHorxGJikN+vLLUqPmNMFITTiu8gEZkhIh+IyMeBKZyTi8gEEVkuIqtEJGSSE5ELRSRfRJaIyL+b+gFMZG3ctZeSCi/dMlLonpEa7XCMMe1IOI0dXgGeAJ4CfOGeWEQ8wKPAqTj9930jIrNVNT9on8HALcA4Vd0lIt2bEryJPGsgYYyJlnASlFdVH2/GuccAq1R1DYCIvARMBPKD9vk58Kiq7gJQ1W3NeB8TQTUNJCxBGWNaVziNJN4WkV+JSC8R6RKYwjiuD7AxaLnAXRfsIOAgEflcRL4UkQmhTiQiU0UkV0Ryt2/fHsZbm5ZiffAZY6IlnBLUle7r74LWKTCwkeNC9XiuId5/MDAeyAE+FZFD3YeBaw5SnQHMABg9enTdc5gIsio+Y0y0hNOKb0Azz10A9A1azgE2h9jnS1WtAtaKyHKchPVNM9/TtKBtJeVsL6kgIyWRvp07RDscY0w7E1aPECJyKDAMqG7GparPNXLYN8BgERkAbAKmAJfU2edN4GJgpoh0xanyWxNe6CbSAqWng3tnkpBgQ4AZY1pXowlKRG7HqYIbBswBzgA+AxpMUKrqFZHrgfcBD/C0qi4RkbuAXFWd7W47TUTycVoI/k5VC/fj85gWZENsGGOiKZwS1CRgBLBIVX8qIj1wmpw3SlXn4CS14HW3Bc0rMM2dTIyxLo6MMdEUTiu+MlX14wyzkQlso/EGEqYNsAYSxphoCqcElSsiWcCTOJ3ElgJfRzQqE3XF5VWsL9xLcmICB3bvGO1wjDHtUDit+H7lzj4hIu8Bmar6fWTDMtEW6H9vSI8MkjzhFLSNMaZl1ZugRGSoqi4TkcNDbDtcVb+NbGgmmqx6zxgTbQ2VoKYBU4H7Q2xT4KSIRGRigg2xYYyJtnoTlKpOFZEE4I+q+nkrxmRigPXBZ4yJtgZvLrit9+5rpVhMjKjw+li1rRQRGNrTEpQxJjrCufv9gYj8RESsK4F2YsWPpXj9yoCu6aSnhNXZiDHGtLhwvn2mAek4z0GV43QCq6pqP63bKHtA1xgTC8JpZp7RGoGY2GEt+IwxsSDczmI74/QyHtxZ7PxIBWWiyxpIGGNiQTidxf4MuBFnuIw8YCywAGtm3ib5/MrSLSWAVfEZY6IrnEYSNwJHAutV9URgFGDD2rZRa3fsoazKR69OqXRJT452OMaYdiycBFWuquUAIpKiqsuAIZENy0RL/ha7/2SMiQ3h3IMqcDuLfRP4UER2se/IuKaNCNx/GmbVe8aYKAunFd/57uwdIvIJ0Al4L6JRmaixQQqNMbEinEYSDwEvq+oXqvrfVojJRImqWhNzY0zMCOce1LfAH0VklYj8VURGRzooEx0/Fpezc08lndKS6JOVFu1wjDHtXKMJSlWfVdUzgTHACuAvIrIy4pGZVrdkU03pyXq2MsZEW1NGojsQGAr0B5ZFJBoTVVa9Z4yJJY0mKBEJlJjuAhYDR6jqORGPzLQ664PPGBNLwmlmvhY4WlV3RDoYE11WgjLGxJJwmpk/0RqBmOjavbeSTbvLSE1KYGC3jtEOxxhjmnQPyrRhgeefhvbMxJNgDSSMMdFnCcoAVr1njIk99VbxiUiXhg5U1Z0tH46JFmsgYYyJNQ3dg1oIKM4IugcAu9z5LGADMCDi0ZlWYyUoY0ysqbeKT1UHqOpA4H3gHFXtqqrZwNnA660VoIm8skofq7eX4kkQhvS0AZSNMbEhnHtQR6rqnMCCqr4LnBC5kExrW/ZjMX6FA7t1JDXJE+1wjDEGCO85qB0i8kfgeZwqv8uAwohGZVqVVe8ZY2JROCWoi4FuwBvu1M1dZ9qIQIIaZgnKGBNDwuksdqeq3ggcp6qHq+pN4bbgE5EJIrLc7Ql9egP7TRIRtZ7SoyPfWvAZY2JQOH3xHSMi+UC+uzxCRB4L4zgP8ChwBjAMuFhEhoXYLwP4H+CrJsZuWoDX52fZjyWAlaCMMbElnCq+B4DTce87qep3wPFhHDcGWKWqa1S1EngJmBhiv/8F7gXKw4rYtKjV2/dQ4fXTt/KdcGkAABvYSURBVEsandKSoh2OMcZUC6snCVXdWGeVL4zD+gDBxxW466qJyCigr6q+09CJRGSqiOSKSO727dvDCdmEqfoB3V5WvWeMiS3hJKiNInIMoCKSLCK/BZaGcVyoDt20eqNIAk7p7DeNnUhVZ6jqaFUd3a1btzDe2oTLWvAZY2JVOAnql8B1OKWfAmCku9yYAqBv0HIOsDloOQM4FJgnIuuAscBsayjRugIlKLv/ZIyJNeEMt7EDuLQZ5/4GGCwiA4BNwBTgkqDzFgFdA8siMg/4rarmNuO9TDOoanUv5taCzxgTaxpNUCLSDfg5zlDv1fur6tUNHaeqXhG5HqerJA/wtKouEZG7gFxVnb0/gZv9V7CrjOJyL9npyfTITIl2OMYYU0s4PUm8BXwKzCW8xhHV3C6S5tRZd1s9+45vyrnN/guu3hOxMaCMMbElnATVQVVvjngkptUtseo9Y0wMC6eRxDsicmbEIzGtzlrwGWNiWTgJ6kacJFUmIsUiUiIixZEOzERezSCFlqCMMbEnnFZ8NkBQG7SjtIKtxRWkJ3von50e7XCMMWYfDQ35PlRVl4nI4aG2q+q3kQvLRFqgefnBvTJJSLAGEsaY2NNQCWoaMBW4P8Q2BU6KSESmVdj9J2NMrKs3QanqVPf1xNYLx7SWJTbEhjEmxoXTzBwRORRnyIzUwDpVfS5SQZnIqvT6WbRhN2BdHBljYlc4PUncDozHSVBzcMZ3+gywBBWn7v9gOZt2l5HTOY0hPa0NjDEmNoXTzHwScDLwo6r+FBgBWL84cWr+iu38Y/4aPAnCQ1NGkuQJa8QVY4xpdeF8O5Wpqh/wikgmsA0YGNmwTCRsL6lg2qzvAPj1KYM5ol+XKEdkjDH1C+ceVK6IZAFPAguBUuDriEZlWpzfr0yblceO0grGDuzCteMPjHZIxhjToHAe1P2VO/uEiLwHZKrq95ENy7S0pz5bw6crd9C5QxIPXjQKjz37ZIyJcQ09qBvyAd3ANntQN358t3E39763HID7Jo+gZ6fURo4wxpjoa6gEFeoB3QB7UDdOlJRXccOLi/D6lauO6c/JB/eIdkjGGBOWhh7UtQd045yq8sc3F7Nh516G9crkljOHRjskY4wJWzjPQaUCvwKOxSk5fQo8oarlEY7N7KfXvt3EW3mbSUvy8PdLRpGS6Il2SMYYE7ZwWvE9B5QAf3eXLwb+BUyOVFBm/63ZXsptby0G4M6JhzCoW8coR2SMMU0TToIaoqojgpY/EZHvIhWQ2X8VXh83vLiIvZU+zh3Rm8lH5EQ7JGOMabJwHtRdJCJjAwsichTweeRCMvvrL+8uZ8nmYvp2SePu8w9FxJqUG2PiTzglqKOAK0Rkg7t8ALBURH4AVFWHRyw602S563by9OdrSUwQ/n7x4WSkJkU7JGOMaZZwEtSEiEdhWswruQUAXHPcAEb2zYpyNMYY03zhJKjBqjo3eIWIXKmqz0YoJtNMlV4/7y35EYCfHG73nYwx8S2ce1C3icjjIpIuIj1E5G3gnEgHZpru89U7KCqr4qAeHTmohw2jYYyJb+EkqBOA1UAezjhQ/1bVSRGNyjTLO99tAeCsw3pHORJjjNl/4SSozjgNJVYDFUA/sWZhMafC6+ODfKd676zhvaIcjTHG7L9wEtSXwLuqOgE4EuiNNTOPOZ+u2EFJuZehPTM4sLs9lGuMiX/hNJI4RVU3AKhqGfA/InJ8ZMMyTfWfH5zqvXNGWPWeMaZtCKcEtVFELhOR2wBE5ADA+uGLIeVVPj7M3wrAWYdZ9Z4xpm0IJ0E9BhyN0wcfOP3yPRqxiEyTzVu+ndIKL4f2yaR/1/Roh2OMMS0irJ4kVPVwEVkEoKq7RCQ5wnGZJghU75093Kr3jDFtRzglqCoR8eAMtYGIdAP84ZxcRCaIyHIRWSUi00NsnyYi+SLyvYh8JCL9mhS9oazSx0dLrXrPGNP2hJOgHgbeALqLyN04z0L9X2MHuUntUeAMYBhwsYgMq7PbImC025/fq8C9TYjdAJ8s38beSh8j+mbRt0uHaIdjjDEtptEqPlV9QUQWAicDApynqkvDOPcYYJWqrgEQkZeAiUB+0Lk/Cdr/S+CyJsRugHe+3wzAOfbskzGmjQnnHhSqugxY1sRz9wE2Bi0X4DzwW59rgHdDbRCRqcBUgAMOOKCJYbRdeyq8fLxsGwBnWvWeMaaNCaeKr7lC9TahIXcUuQwYDfw11HZVnaGqo1V1dLdu3VowxPj20bJtlFf5OaJfZ3pnpUU7HGOMaVFhlaCaqQDoG7ScA2yuu5OInALcCpygqhURjKfNeec753Ja4whjTFsUyRLUN8BgERngNkufAswO3kFERgH/AM5V1W0RjKXNKSmvYt6K7YhY33vGmLYpYglKVb3A9cD7wFJglqouEZG7RORcd7e/Ah2BV0QkT0Rm13M6U8fcpVup9Po5sn8XemSmRjscY4xpcZGs4kNV5wBz6qy7LWj+lEi+f1sWGFrjbCs9GWPaqEhW8ZkIKSqrYv7K7SQITDi0Z7TDMcaYiLAEFYc+WPIjVT7lqAHZdM+w6j1jTNtkCSoOVfe9N8Kq94wxbZclqDiza08ln63cgSdBmHCIVe8ZY9ouS1Bx5oP8H/H6lWMGZZPdMSXa4RhjTMRYgooz73xvrfeMMe2DJag4UlhawRerC0lMEE636j1jTBtnCSqOvLfkR3x+5djBXcnqYGNGGmPaNktQceT9JTYwoTGm/bAEFUe27C4DYHhOVpQjMcaYyLMEFUeKyqoAyEyLaA9VxhgTEyxBxZHicidBdUpLinIkxhgTeZag4kSF10d5lZ/EBCEtyRPtcIwxJuIsQcWJ4jIvAJlpSYiEGqzYGGPaFktQcSJw/8mq94wx7YUlqDgRuP+UmWoNJIwx7YN928WJmhZ8VoIy7UtVVRUFBQWUl5dHOxSzn1JTU8nJySEpKbzvMUtQcaLYEpRppwoKCsjIyKB///52/zWOqSqFhYUUFBQwYMCAsI6xKr44UWz3oEw7VV5eTnZ2tiWnOCciZGdnN6kkbAkqThSXu634Ui1BmfbHklPb0NS/oyWoOGElKGNMe2MJKk5YN0fGmIbMmzePs88+G4DZs2dzzz33hNyvY8eODZ5n9+7dPPbYY9XLmzdvZtKkSS0XaBNYgooT1s2RMbHL6/VGO4Razj33XKZPn96sY+smqN69e/Pqq6+2VGhNYj/H40R1CcruQZl2rP/0/0TkvOvuOavRfZ577jnuu+8+RIThw4fj8Xjo0qULixYt4vDDD+fWW2/l6quvZs2aNXTo0IEZM2YwfPhw/vvf/3LjjTcCzj2Y+fPnU1paykUXXURxcTFer5fHH3+c4447rtb7HXXUUTz99NMccsghAIwfP577778fn8/HTTfdRFlZGWlpaTzzzDMMGTKk1rEzZ84kNzeXRx55hLVr13LJJZfg9XqZMGFC9T6lpaVMnDiRXbt2UVVVxZ///GcmTpzI9OnTWb16NSNHjuTUU0/luuuu4+yzz2bx4sWUl5dz7bXXkpubS2JiIn/729848cQTmTlzJrNnz2bv3r2sXr2a888/n3vvvXd//yyWoOJFoKsjK0EZ0/qWLFnC3Xffzeeff07Xrl3ZuXMn06ZNY8WKFcydOxePx8MNN9zAqFGjePPNN/n444+54ooryMvL47777uPRRx9l3LhxlJaWkpqayowZMzj99NO59dZb8fl87N27d5/3nDJlCrNmzeLOO+9ky5YtbN68mSOOOILi4mLmz59PYmIic+fO5Q9/+AOvvfZavbHfeOONXHvttVxxxRU8+uij1etTU1N54403yMzMZMeOHYwdO5Zzzz2Xe+65h8WLF5OXlwfAunXrqo8JHP/DDz+wbNkyTjvtNFasWAFAXl4eixYtIiUlhSFDhnDDDTfQt2/f/brulqDihD2oa0x4JZ1I+Pjjj5k0aRJdu3YFoEuXLgBMnjwZj8fpvPmzzz6rThQnnXQShYWFFBUVMW7cOKZNm8all17KBRdcQE5ODkceeSRXX301VVVVnHfeeYwcOXKf97zwwgs59dRTufPOO5k1axaTJ08GoKioiCuvvJKVK1ciIlRVVTUY++eff14d1+WXX87NN98MOM8l/eEPf2D+/PkkJCSwadMmtm7d2uC5PvvsM2644QYAhg4dSr9+/aoT1Mknn0ynTp0AGDZsGOvXr9/vBGX3oOKE3YMyJnpUNWQT6fT09Fr71CUiTJ8+naeeeoqysjLGjh3LsmXLOP7445k/fz59+vTh8ssv57nnnuONN95g5MiRjBw5ktzcXPr06UN2djbff/89L7/8MlOmTAHgT3/6EyeeeCKLFy/m7bffDuu5olCxv/DCC2zfvp2FCxeSl5dHjx49Gj1XqM8YkJKSUj3v8Xha5L6cJag44PdrdTPzDOuLz5hWd/LJJzNr1iwKCwsB2Llz5z77HH/88bzwwguA06Kua9euZGZmsnr1ag477DBuvvlmRo8ezbJly1i/fj3du3fn5z//Oddccw3ffvst559/Pnl5eeTl5TF69GjAqea79957KSoq4rDDDgOcElSfPn0A515TY8aNG8dLL70EUB1f4Dzdu3cnKSmJTz75hPXr1wOQkZFBSUlJyHMFf8YVK1awYcOGfe5/tSRLUHFgT6UXv0KHZA9JHvuTGdPaDjnkEG699VZOOOEERowYwbRp0/bZ54477iA3N5fhw4czffp0nn32WQAefPBBDj30UEaMGEFaWhpnnHEG8+bNY+TIkYwaNYrXXnutuhFFXZMmTeKll17iwgsvrF73+9//nltuuYVx48bh8/kajf2hhx7i0Ucf5cgjj6SoqKh6/aWXXkpubi6jR4/mhRdeYOjQoQBkZ2czbtw4Dj30UH73u9/VOtevfvUrfD4fhx12GBdddBEzZ86sVXJqadJQkS0WjR49WnNzc6MdRqsq2LWXY//yCb06pbLglpOjHY4xrWrp0qUcfPDB0Q7DtJBQf08RWaiqo+vuaz/H40D1YIXWxNwY045ENEGJyAQRWS4iq0Rkn6fGRCRFRF52t38lIv0jGU+8sgYSxpj2KGIJSkQ8wKPAGcAw4GIRGVZnt2uAXap6IPAA8JdIxRPPrJsjY0x7FMkS1BhglaquUdVK4CVgYp19JgLPuvOvAieLdVu8DxsLyhjTHkUyQfUBNgYtF7jrQu6jql6gCMiueyIRmSoiuSKSu3379giFG7usmyNjTHsUyQQVqiRUt8lgOPugqjNUdbSqju7WrVuLBBdPAmNB2T0oY0x7EskEVQAE93ORA2yubx8RSQQ6Afs+AdfOWRWfMdFVt4fvcJ155pns3r27wX1uu+025s6d29zQ2rRIJqhvgMEiMkBEkoEpwOw6+8wGrnTnJwEfa7w9mNUKbLBCY6KrvgTV2IOyc+bMISsrq8F97rrrLk455ZT9iq+tilizMFX1isj1wPuAB3haVZeIyF1ArqrOBv4J/EtEVuGUnKZEKp54VnMPylrxmXbujk4ROm9Rg5uDh6BISkqiY8eO9OrVi7y8PPLz8znvvPPYuHEj5eXl3HjjjUydOhWA/v37k5ubS2lpKWeccQbHHnssX3zxBX369OGtt94iLS2Nq666irPPPptJkybRv39/rrzySt5++22qqqp45ZVXGDp0KNu3b+eSSy6hsLCQI488kvfee4+FCxdWd17bVkX0OShVnaOqB6nqIFW92113m5ucUNVyVZ2sqgeq6hhVXRPJeOKVPQdlTHTdc889DBo0iLy8PP7617/y9ddfc/fdd5Ofnw/A008/zcKFC8nNzeXhhx+u7rMv2MqVK7nuuutYsmQJWVlZ9Q6R0bVrV7799luuvfZa7rvvPgDuvPNOTjrppOo++zZs2BC5DxtD7Cd5HLChNoxxNVLSaS1jxoxhwIAB1csPP/wwb7zxBgAbN25k5cqVZGfXbpA8YMCA6mE1jjjiiFrjLAW74IILqvd5/fXXAWeYi8D5J0yYQOfOnVv088QqS1BxoLqrI0tQxsSE4GE25s2bx9y5c1mwYAEdOnRg/PjxIYetqDscRVlZWchzB/YLHrKivd6at7744oBV8RkTXQ0NQVFUVETnzp3p0KEDy5Yt48svv2zx9z/22GOZNWsWAB988AG7du1q8feIRVaCinFVPj97K314EoT0ZE+0wzGmXQoegiItLY0ePXpUb5swYQJPPPEEw4cPZ8iQIYwdO7bF3//222/n4osv5uWXX+aEE06gV69eZGRktPj7xBobbiPGFZZWcMSf59K5QxKLbjst2uEY0+psuA2oqKjA4/GQmJjIggULuPbaa8nLy4t2WM3SlOE2rAQV46yBhDFmw4YNXHjhhfj9fpKTk3nyySejHVKrsAQV46ybI2PM4MGDWbRoUbTDaHXWSCLGWUexxrTfVmxtTVP/jpagYpx1c2Tau9TUVAoLCy1JxTlVpbCwkNTU1LCPsSq+GGeDFZr2Licnh4KCAtrjUDttTWpqKjk5OWHvb996MS7wDJQ1kjDtVVJSUq1eG0z7YVV8Mc7uQRlj2itLUDEu0M2R3YMyxrQ3lqBinFXxGWPaq7jrSUJEtgPrW+GtugI7WuF9IiWe47fYo8Nij454jh1aJv5+qtqt7sq4S1CtRURyQ3W9ES/iOX6LPTos9uiI59ghsvFbFZ8xxpiYZAnKGGNMTLIEVb8Z0Q5gP8Vz/BZ7dFjs0RHPsUME47d7UMYYY2KSlaCMMcbEJEtQxhhjYpIlqBBEZJ2I/CAieSIS08P3isjTIrJNRBYHresiIh+KyEr3tXM0Y2xIPfHfISKb3OufJyJnRjPGUESkr4h8IiJLRWSJiNzoro+La99A/PFw7VNF5GsR+c6N/U53/QAR+cq99i+LSHK0Y62rgdhnisjaoOs+Mtqx1kdEPCKySETecZcjdt0tQdXvRFUdGQfPJ8wEJtRZNx34SFUHAx+5y7FqJvvGD/CAe/1HquqcVo4pHF7gN6p6MDAWuE5EhhE/176++CH2r30FcJKqjgBGAhNEZCzwF5zYBwO7gGuiGGN96osd4HdB1z2Wx3O/EVgatByx624JKs6p6nxgZ53VE4Fn3flngfNaNagmqCf+mKeqW1T1W3e+BOc/bB/i5No3EH/MU0epu5jkTgqcBLzqro/Ja99A7HFBRHKAs4Cn3GUhgtfdElRoCnwgIgtFZGq0g2mGHqq6BZwvIqB7lONpjutF5Hu3CjAmq8kCRKQ/MAr4iji89nXihzi49m41Ux6wDfgQWA3sVlWvu0sBMZpw68auqoHrfrd73R8QkZQohtiQB4HfA353OZsIXndLUKGNU9XDgTNwqj6Oj3ZA7czjwCCcKpAtwP3RDad+ItIReA24SVWLox1PU4WIPy6uvar6VHUkkAOMAQ4OtVvrRhWeurGLyKHALcBQ4EigC3BzFEMMSUTOBrap6sLg1SF2bbHrbgkqBFXd7L5uA97A+Q8QT7aKSC8A93VblONpElXd6v4n9gNPEqPXX0SScL7cX1DV193VcXPtQ8UfL9c+QFV3A/Nw7qNliUhgENYcYHO04gpHUOwT3CpXVdUK4Bli87qPA84VkXXASzhVew8SwetuCaoOEUkXkYzAPHAasLjho2LObOBKd/5K4K0oxtJkgS941/nE4PV3697/CSxV1b8FbYqLa19f/HFy7buJSJY7nwacgnMP7RNgkrtbTF77emJfFvSjRnDu4cTcdVfVW1Q1R1X7A1OAj1X1UiJ43a0niTpEZCBOqQkgEfi3qt4dxZAaJCIvAuNxurzfCtwOvAnMAg4ANgCTVTUmGyLUE/94nComBdYBvwjc14kVInIs8CnwAzX18X/AuY8T89e+gfgvJvav/XCcm/EenB/Zs1T1Lvf/7ks4VWSLgMvcEknMaCD2j4FuOFVmecAvgxpTxBwRGQ/8VlXPjuR1twRljDEmJlkVnzHGmJhkCcoYY0xMsgRljDEmJlmCMsYYE5MsQRljjIlJlqBMuyYi80Qk4h0Ci8j/uD2Hv1Bn/cjm9BguIr1F5NXG94wOt1f030Y7DhPfLEEZ00xBT8+H41fAme6DjcFGAiETVEPnV9XNqjqpvu3GtAWWoEzME5H+bunjSXcMnQ/cp/BrlYBEpKvbDQsicpWIvCkib7vj7FwvItPccWy+FJEuQW9xmYh8ISKLRWSMe3y621nqN+4xE4PO+4qIvA18ECLWae55FovITe66J4CBwGwR+XXQvsnAXcBF7hhAF7kljxki8gHwnPvZPxWRb93pmKBrsjgoptdF5D1xxuS5t57reI+I5Lsdkt7nrjtHnLF8FonIXBHp4a6/Q0Seda/1OhG5QETuFWectPfcbpICY6f9RZwxjr4WkQNDvO8g95iF7mcZ6q6f7F6n70Rkfnj/Gky7oqo22RTTE9AfZ/yike7yLJyn1cHpy2y0O98VWOfOXwWsAjJwntAvwnk6H+ABnM5RA8c/6c4fDyx25/8v6D2ygBVAunveAqBLiDiPwOmZIR3oCCwBRrnb1gFdQxxzFfBI0PIdwEIgzV3uAKS684OB3KBrsjjoHGuATkAqsB7oW+d9ugDLqXk4P8t97Ry07mfA/UFxfIYzHMQIYC9whrvtDeC8oM91qzt/BfBO0PG/dec/Aga780fhdJGDe636BMdjk03BU1OqKIyJprVaM4jbQpwv6MZ8os5YRyUiUgS87a7/ARgetN+L4IxNJSKZbl9pp+F0jBm4j5KK030ROEMkhOq+6FjgDVXdAyAirwPH4XT/0hSzVbXMnU8CHhFnhFUfcFA9x3ykqkXu++YD/YCNQduLgXLgKRH5D/COuz4HeNntCy4ZWBt0zLuqWiUiP+B0zfOeu/4Hal//F4NeHwgOSpze0o8BXnG6mQMgMJTE58BMEZkFvI4xdViCMvEiuG8vH5DmznupqapObeAYf9Cyn9r/9uv296U4faL9RFWXB28QkaOAPfXEGGrogeYIPv+vcfooHIHzOcvrOabu9an1f1tVvW715ck4HX1ej9Mb9d+Bv6nqbLd/tTvqnlNV/SJSpaqB69TQ9at7LRNwxgvaZwhzVf2lez3PAvJEZKSqFtbz+Uw7ZPegTLxbh1O1BjU9KjfVRVDdgWqRWxJ5H7hB3J/9IjIqjPPMB84TkQ7i9IR/Pk6HrA0pwamGrE8nYIs6w19cjlOSaTK3JNNJnSHcb8JpnBE4/yZ3/spQx4bhoqDXBcEb1Bljaq2ITHbjEBEZ4c4PUtWvVPU2YAfQt5nvb9ooK0GZeHcfMEtELgc+buY5donIF0AmcLW77n9xxrr53k1S64CzGzqJqn4rIjOBr91VT6lqY9V7nwDTxRlh9f+F2P4Y8Jr7Bf8J9ZfeGpMBvCUiqTglvUBjjTtwqt82AV8CA5px7hQR+QrnB+/FIbZfCjwuIn/EqbJ8CfgO+KuIDHbj+chdZ0w1683cGNNsbqvJ0aq6I9qxmLbHqviMMcbEJCtBGWOMiUlWgjLGGBOTLEEZY4yJSZagjDHGxCRLUMYYY2KSJShjjDEx6f8DhAtkyvr55MEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# A learning curve for d=1, 5, 15\n",
    "for d in [1,2,3,4,5]:\n",
    "    model = make_pipeline(PolynomialFeatures(degree=d), LinearRegression())\n",
    "\n",
    "    from sklearn.model_selection import learning_curve\n",
    "    train_sizes, train_scores, validation_scores = learning_curve(model, X, y,train_sizes=(np.logspace(-1, 0, 20)))\n",
    "    # Plot the mean train error and validation error across folds\n",
    "    plt.figure(figsize=(6, 4))\n",
    "    plt.plot(train_sizes, validation_scores.mean(axis=1),lw=2, label='cross-validation')\n",
    "    plt.plot(train_sizes, train_scores.mean(axis=1),lw=2, label='training')\n",
    "    plt.ylim(ymin=-.1, ymax=1)\n",
    "\n",
    "    plt.legend(loc='best')\n",
    "    plt.xlabel('number of train samples')\n",
    "    plt.ylabel('explained variance')\n",
    "    plt.title('Learning curve (degree=%i)' % d)\n",
    "    plt.tight_layout()\n",
    "    \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}