{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Chapter 4 - Classification"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- [Load dataset](#Load-dataset)\n",
    "- [The Default data set](#Figure-4.1---Default-data-set)\n",
    "- [4.3 Logistic Regression](#4.3-Logistic-Regression)\n",
    "- [4.4 Linear Discriminant Analysis](#4.4-Linear-Discriminant-Analysis)\n",
    "- [Lab: 4.6.3 Linear Discriminant Analysis](#4.6.3-Linear-Discriminant-Analysis)\n",
    "- [Lab: 4.6.4 Quadratic Discriminant Analysis](#4.6.4-Quadratic-Discriminant-Analysis)\n",
    "- [Lab: 4.6.5 K-Nearest Neighbors](#4.6.5-K-Nearest-Neighbors)\n",
    "- [Lab: 4.6.6 An Application to Caravan Insurance Data](#4.6.6-An-Application-to-Caravan-Insurance-Data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# %load ../standard_import.txt\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import sklearn.linear_model as skl_lm\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
    "from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis\n",
    "from sklearn.metrics import confusion_matrix, classification_report, precision_score\n",
    "from sklearn import preprocessing\n",
    "from sklearn import neighbors\n",
    "\n",
    "import statsmodels.api as sm\n",
    "import statsmodels.formula.api as smf\n",
    "\n",
    "%matplotlib inline\n",
    "plt.style.use('seaborn-white')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>default</th>\n",
       "      <th>student</th>\n",
       "      <th>balance</th>\n",
       "      <th>income</th>\n",
       "      <th>default2</th>\n",
       "      <th>student2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>729.526495</td>\n",
       "      <td>44361.625074</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>817.180407</td>\n",
       "      <td>12106.134700</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>1073.549164</td>\n",
       "      <td>31767.138947</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  default student      balance        income  default2  student2\n",
       "1      No      No   729.526495  44361.625074         0         0\n",
       "2      No     Yes   817.180407  12106.134700         0         1\n",
       "3      No      No  1073.549164  31767.138947         0         0"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# In R, I exported the dataset from package 'ISLR' to an Excel file\n",
    "df = pd.read_excel('Data/Default.xlsx')\n",
    "\n",
    "# Note: factorize() returns two objects: a label array and an array with the unique values.\n",
    "# We are only interested in the first object. \n",
    "df['default2'] = df.default.factorize()[0]\n",
    "df['student2'] = df.student.factorize()[0]\n",
    "df.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  Figure 4.1 - Default data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f407198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(12,5))\n",
    "gs = mpl.gridspec.GridSpec(1, 4)\n",
    "ax1 = plt.subplot(gs[0,:-2])\n",
    "ax2 = plt.subplot(gs[0,-2])\n",
    "ax3 = plt.subplot(gs[0,-1])\n",
    "\n",
    "# Take a fraction of the samples where target value (default) is 'no'\n",
    "df_no = df[df.default2 == 0].sample(frac=0.15)\n",
    "# Take all samples  where target value is 'yes'\n",
    "df_yes = df[df.default2 == 1]\n",
    "df_ = df_no.append(df_yes)\n",
    "\n",
    "ax1.scatter(df_[df_.default == 'Yes'].balance, df_[df_.default == 'Yes'].income, s=40, c='orange', marker='+',\n",
    "            linewidths=1)\n",
    "ax1.scatter(df_[df_.default == 'No'].balance, df_[df_.default == 'No'].income, s=40, marker='o', linewidths='1',\n",
    "            edgecolors='lightblue', facecolors='white', alpha=.6)\n",
    "\n",
    "ax1.set_ylim(ymin=0)\n",
    "ax1.set_ylabel('Income')\n",
    "ax1.set_xlim(xmin=-100)\n",
    "ax1.set_xlabel('Balance')\n",
    "\n",
    "c_palette = {'No':'lightblue', 'Yes':'orange'}\n",
    "sns.boxplot('default', 'balance', data=df, orient='v', ax=ax2, palette=c_palette)\n",
    "sns.boxplot('default', 'income', data=df, orient='v', ax=ax3, palette=c_palette)\n",
    "gs.tight_layout(plt.gcf())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.3 Logistic Regression\n",
    "### Figure 4.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f6c2080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_train = df.balance.values.reshape(-1,1) \n",
    "y = df.default2\n",
    "\n",
    "# Create array of test data. Calculate the classification probability\n",
    "# and predicted classification.\n",
    "X_test = np.arange(df.balance.min(), df.balance.max()).reshape(-1,1)\n",
    "\n",
    "clf = skl_lm.LogisticRegression(solver='newton-cg')\n",
    "clf.fit(X_train,y)\n",
    "prob = clf.predict_proba(X_test)\n",
    "\n",
    "fig, (ax1, ax2) = plt.subplots(1,2, figsize=(12,5))\n",
    "# Left plot\n",
    "sns.regplot(df.balance, df.default2, order=1, ci=None,\n",
    "            scatter_kws={'color':'orange'},\n",
    "            line_kws={'color':'lightblue', 'lw':2}, ax=ax1)\n",
    "# Right plot\n",
    "ax2.scatter(X_train, y, color='orange')\n",
    "ax2.plot(X_test, prob[:,1], color='lightblue')\n",
    "\n",
    "for ax in fig.axes:\n",
    "    ax.hlines(1, xmin=ax.xaxis.get_data_interval()[0],\n",
    "              xmax=ax.xaxis.get_data_interval()[1], linestyles='dashed', lw=1)\n",
    "    ax.hlines(0, xmin=ax.xaxis.get_data_interval()[0],\n",
    "              xmax=ax.xaxis.get_data_interval()[1], linestyles='dashed', lw=1)\n",
    "    ax.set_ylabel('Probability of default')\n",
    "    ax.set_xlabel('Balance')\n",
    "    ax.set_yticks([0, 0.25, 0.5, 0.75, 1.])\n",
    "    ax.set_xlim(xmin=-100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Table 4.1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = df.default2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#####  scikit-learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
      "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
      "          penalty='l2', random_state=None, solver='newton-cg', tol=0.0001,\n",
      "          verbose=0, warm_start=False)\n",
      "classes:  [0 1]\n",
      "coefficients:  [[ 0.00549891]]\n",
      "intercept : [-10.65131761]\n"
     ]
    }
   ],
   "source": [
    "# Using newton-cg solver, the coefficients are equal/closest to the ones in the book. \n",
    "# I do not know the details on the differences between the solvers.\n",
    "clf = skl_lm.LogisticRegression(solver='newton-cg')\n",
    "X_train = df.balance.values.reshape(-1,1)\n",
    "clf.fit(X_train,y)\n",
    "print(clf)\n",
    "print('classes: ',clf.classes_)\n",
    "print('coefficients: ',clf.coef_)\n",
    "print('intercept :', clf.intercept_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### statsmodels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.\n",
      "         Current function value: 0.079823\n",
      "         Iterations 10\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Coef.</th>\n",
       "      <th>Std.Err.</th>\n",
       "      <th>z</th>\n",
       "      <th>P&gt;|z|</th>\n",
       "      <th>[0.025</th>\n",
       "      <th>0.975]</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>const</th>\n",
       "      <td>-10.651331</td>\n",
       "      <td>0.361169</td>\n",
       "      <td>-29.491287</td>\n",
       "      <td>3.723665e-191</td>\n",
       "      <td>-11.359208</td>\n",
       "      <td>-9.943453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>balance</th>\n",
       "      <td>0.005499</td>\n",
       "      <td>0.000220</td>\n",
       "      <td>24.952404</td>\n",
       "      <td>2.010855e-137</td>\n",
       "      <td>0.005067</td>\n",
       "      <td>0.005931</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Coef.  Std.Err.          z          P>|z|     [0.025    0.975]\n",
       "const   -10.651331  0.361169 -29.491287  3.723665e-191 -11.359208 -9.943453\n",
       "balance   0.005499  0.000220  24.952404  2.010855e-137   0.005067  0.005931"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train = sm.add_constant(df.balance)\n",
    "est = smf.Logit(y.ravel(), X_train).fit()\n",
    "est.summary2().tables[1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  Table 4.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.\n",
      "         Current function value: 0.145434\n",
      "         Iterations 7\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Coef.</th>\n",
       "      <th>Std.Err.</th>\n",
       "      <th>z</th>\n",
       "      <th>P&gt;|z|</th>\n",
       "      <th>[0.025</th>\n",
       "      <th>0.975]</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>const</th>\n",
       "      <td>-3.504128</td>\n",
       "      <td>0.070713</td>\n",
       "      <td>-49.554094</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-3.642723</td>\n",
       "      <td>-3.365532</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>student2</th>\n",
       "      <td>0.404887</td>\n",
       "      <td>0.115019</td>\n",
       "      <td>3.520177</td>\n",
       "      <td>0.000431</td>\n",
       "      <td>0.179454</td>\n",
       "      <td>0.630320</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Coef.  Std.Err.          z     P>|z|    [0.025    0.975]\n",
       "const    -3.504128  0.070713 -49.554094  0.000000 -3.642723 -3.365532\n",
       "student2  0.404887  0.115019   3.520177  0.000431  0.179454  0.630320"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train = sm.add_constant(df.student2)\n",
    "y = df.default2\n",
    "\n",
    "est = smf.Logit(y, X_train).fit()\n",
    "est.summary2().tables[1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  Table 4.3 - Multiple Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.\n",
      "         Current function value: 0.078577\n",
      "         Iterations 10\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Coef.</th>\n",
       "      <th>Std.Err.</th>\n",
       "      <th>z</th>\n",
       "      <th>P&gt;|z|</th>\n",
       "      <th>[0.025</th>\n",
       "      <th>0.975]</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>const</th>\n",
       "      <td>-10.869045</td>\n",
       "      <td>0.492273</td>\n",
       "      <td>-22.079320</td>\n",
       "      <td>4.995499e-108</td>\n",
       "      <td>-11.833882</td>\n",
       "      <td>-9.904209</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>balance</th>\n",
       "      <td>0.005737</td>\n",
       "      <td>0.000232</td>\n",
       "      <td>24.736506</td>\n",
       "      <td>4.331521e-135</td>\n",
       "      <td>0.005282</td>\n",
       "      <td>0.006191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>income</th>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000008</td>\n",
       "      <td>0.369808</td>\n",
       "      <td>7.115254e-01</td>\n",
       "      <td>-0.000013</td>\n",
       "      <td>0.000019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>student2</th>\n",
       "      <td>-0.646776</td>\n",
       "      <td>0.236257</td>\n",
       "      <td>-2.737595</td>\n",
       "      <td>6.189022e-03</td>\n",
       "      <td>-1.109831</td>\n",
       "      <td>-0.183721</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Coef.  Std.Err.          z          P>|z|     [0.025    0.975]\n",
       "const    -10.869045  0.492273 -22.079320  4.995499e-108 -11.833882 -9.904209\n",
       "balance    0.005737  0.000232  24.736506  4.331521e-135   0.005282  0.006191\n",
       "income     0.000003  0.000008   0.369808   7.115254e-01  -0.000013  0.000019\n",
       "student2  -0.646776  0.236257  -2.737595   6.189022e-03  -1.109831 -0.183721"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train = sm.add_constant(df[['balance', 'income', 'student2']])\n",
    "est = smf.Logit(y, X_train).fit()\n",
    "est.summary2().tables[1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Figure 4.3 - Confounding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# balance and default vectors for students\n",
    "X_train = df[df.student == 'Yes'].balance.values.reshape(df[df.student == 'Yes'].balance.size,1) \n",
    "y = df[df.student == 'Yes'].default2\n",
    "\n",
    "# balance and default vectors for non-students\n",
    "X_train2 = df[df.student == 'No'].balance.values.reshape(df[df.student == 'No'].balance.size,1) \n",
    "y2 = df[df.student == 'No'].default2\n",
    "\n",
    "# Vector with balance values for plotting\n",
    "X_test = np.arange(df.balance.min(), df.balance.max()).reshape(-1,1)\n",
    "\n",
    "clf = skl_lm.LogisticRegression(solver='newton-cg')\n",
    "clf2 = skl_lm.LogisticRegression(solver='newton-cg')\n",
    "\n",
    "clf.fit(X_train,y)\n",
    "clf2.fit(X_train2,y2)\n",
    "\n",
    "prob = clf.predict_proba(X_test)\n",
    "prob2 = clf2.predict_proba(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>default</th>\n",
       "      <th>No</th>\n",
       "      <th>Yes</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>student</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>No</th>\n",
       "      <td>6850</td>\n",
       "      <td>206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>2817</td>\n",
       "      <td>127</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "default    No  Yes\n",
       "student           \n",
       "No       6850  206\n",
       "Yes      2817  127"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(['student','default']).size().unstack('default')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f4372b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# creating plot\n",
    "fig, (ax1, ax2) = plt.subplots(1,2, figsize=(12,5))\n",
    "\n",
    "# Left plot\n",
    "ax1.plot(X_test, pd.DataFrame(prob)[1], color='orange', label='Student')\n",
    "ax1.plot(X_test, pd.DataFrame(prob2)[1], color='lightblue', label='Non-student')\n",
    "ax1.hlines(127/2817, colors='orange', label='Overall Student',\n",
    "           xmin=ax1.xaxis.get_data_interval()[0],\n",
    "           xmax=ax1.xaxis.get_data_interval()[1], linestyles='dashed')\n",
    "ax1.hlines(206/6850, colors='lightblue', label='Overall Non-Student',\n",
    "           xmin=ax1.xaxis.get_data_interval()[0],\n",
    "           xmax=ax1.xaxis.get_data_interval()[1], linestyles='dashed')\n",
    "ax1.set_ylabel('Default Rate')\n",
    "ax1.set_xlabel('Credit Card Balance')\n",
    "ax1.set_yticks([0, 0.2, 0.4, 0.6, 0.8, 1.])\n",
    "ax1.set_xlim(450,2500)\n",
    "ax1.legend(loc=2)\n",
    "\n",
    "# Right plot\n",
    "sns.boxplot('student', 'balance', data=df, orient='v', ax=ax2,  palette=c_palette);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.4 Linear Discriminant Analysis\n",
    "### Table 4.4 \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>True default status</th>\n",
       "      <th>No</th>\n",
       "      <th>Yes</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Predicted default status</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>No</th>\n",
       "      <td>9645</td>\n",
       "      <td>254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>22</td>\n",
       "      <td>79</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "True default status         No  Yes\n",
       "Predicted default status           \n",
       "No                        9645  254\n",
       "Yes                         22   79"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = df[['balance', 'income', 'student2']].as_matrix()\n",
    "y = df.default2.as_matrix()\n",
    "\n",
    "lda = LinearDiscriminantAnalysis(solver='svd')\n",
    "y_pred = lda.fit(X, y).predict(X)\n",
    "\n",
    "df_ = pd.DataFrame({'True default status': y,\n",
    "                    'Predicted default status': y_pred})\n",
    "df_.replace(to_replace={0:'No', 1:'Yes'}, inplace=True)\n",
    "\n",
    "df_.groupby(['Predicted default status','True default status']).size().unstack('True default status')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "         No       0.97      1.00      0.99      9667\n",
      "        Yes       0.78      0.24      0.36       333\n",
      "\n",
      "avg / total       0.97      0.97      0.97     10000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y, y_pred, target_names=['No', 'Yes']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Table 4.5\n",
    "Instead of using the probability of 50% as decision boundary, we say that a probability of default of 20% is to be classified as 'Yes'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>True default status</th>\n",
       "      <th>No</th>\n",
       "      <th>Yes</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Predicted default status</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>No</th>\n",
       "      <td>9435</td>\n",
       "      <td>140</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Yes</th>\n",
       "      <td>232</td>\n",
       "      <td>193</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "True default status         No  Yes\n",
       "Predicted default status           \n",
       "No                        9435  140\n",
       "Yes                        232  193"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "decision_prob = 0.2\n",
    "y_prob = lda.fit(X, y).predict_proba(X)\n",
    "\n",
    "df_ = pd.DataFrame({'True default status': y,\n",
    "                    'Predicted default status': y_prob[:,1] > decision_prob})\n",
    "df_.replace(to_replace={0:'No', 1:'Yes', 'True':'Yes', 'False':'No'}, inplace=True)\n",
    "\n",
    "df_.groupby(['Predicted default status','True default status']).size().unstack('True default status')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#  Lab"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.6.3 Linear Discriminant Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('Data/Smarket.csv', usecols=range(1,10), index_col=0, parse_dates=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = df[:'2004'][['Lag1','Lag2']]\n",
    "y_train = df[:'2004']['Direction']\n",
    "\n",
    "X_test = df['2005':][['Lag1','Lag2']]\n",
    "y_test = df['2005':]['Direction']\n",
    "\n",
    "lda = LinearDiscriminantAnalysis()\n",
    "pred = lda.fit(X_train, y_train).predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.49198397,  0.50801603])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda.priors_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.04279022,  0.03389409],\n",
       "       [-0.03954635, -0.03132544]])"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda.means_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.05544078, -0.0443452 ]])"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# These do not seem to correspond to the values from the R output in the book?\n",
    "lda.coef_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 35,  35],\n",
       "       [ 76, 106]])"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "confusion_matrix(y_test, pred).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "       Down      0.500     0.315     0.387       111\n",
      "         Up      0.582     0.752     0.656       141\n",
      "\n",
      "avg / total      0.546     0.560     0.538       252\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, pred, digits=3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_p = lda.predict_proba(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([False,  True], dtype=bool), array([ 70, 182]))"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(pred_p[:,1]>0.5, return_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([False], dtype=bool), array([252]))"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(pred_p[:,1]>0.9, return_counts=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.6.4 Quadratic Discriminant Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "qda = QuadraticDiscriminantAnalysis()\n",
    "pred = qda.fit(X_train, y_train).predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.49198397,  0.50801603])"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "qda.priors_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.04279022,  0.03389409],\n",
       "       [-0.03954635, -0.03132544]])"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "qda.means_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 30,  20],\n",
       "       [ 81, 121]])"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "confusion_matrix(y_test, pred).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "       Down      0.600     0.270     0.373       111\n",
      "         Up      0.599     0.858     0.706       141\n",
      "\n",
      "avg / total      0.599     0.599     0.559       252\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, pred, digits=3))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.6.5 K-Nearest Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[43 58]\n",
      " [68 83]]\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "       Down      0.426     0.387     0.406       111\n",
      "         Up      0.550     0.589     0.568       141\n",
      "\n",
      "avg / total      0.495     0.500     0.497       252\n",
      "\n"
     ]
    }
   ],
   "source": [
    "knn = neighbors.KNeighborsClassifier(n_neighbors=1)\n",
    "pred = knn.fit(X_train, y_train).predict(X_test)\n",
    "print(confusion_matrix(y_test, pred).T)\n",
    "print(classification_report(y_test, pred, digits=3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[48 55]\n",
      " [63 86]]\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "       Down      0.466     0.432     0.449       111\n",
      "         Up      0.577     0.610     0.593       141\n",
      "\n",
      "avg / total      0.528     0.532     0.529       252\n",
      "\n"
     ]
    }
   ],
   "source": [
    "knn = neighbors.KNeighborsClassifier(n_neighbors=3)\n",
    "pred = knn.fit(X_train, y_train).predict(X_test)\n",
    "print(confusion_matrix(y_test, pred).T)\n",
    "print(classification_report(y_test, pred, digits=3))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.6.6 An Application to Caravan Insurance Data\n",
    "\n",
    "#### K-Nearest Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# In R, I exported the dataset from package 'ISLR' to a csv file\n",
    "df = pd.read_csv('Data/Caravan.csv')\n",
    "y = df.Purchase\n",
    "X = df.drop('Purchase', axis=1).astype('float64')\n",
    "X_scaled = preprocessing.scale(X)\n",
    "\n",
    "X_train = X_scaled[1000:,:]\n",
    "y_train = y[1000:]\n",
    "X_test = X_scaled[:1000,:]\n",
    "y_test = y[:1000]\n",
    "\n",
    "def KNN(n_neighbors=1, weights='uniform'):\n",
    "    clf = neighbors.KNeighborsClassifier(n_neighbors, weights)\n",
    "    clf.fit(X_train, y_train)\n",
    "    pred = clf.predict(X_test)\n",
    "    score = clf.score(X_test, y_test)\n",
    "    return(pred, score, clf.classes_)\n",
    "\n",
    "def plot_confusion_matrix(cm, classes, n_neighbors, title='Confusion matrix (Normalized)',\n",
    "                          cmap=plt.cm.Blues):\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=plt.cm.Blues)\n",
    "    plt.title('Normalized confusion matrix: KNN-{}'.format(n_neighbors))\n",
    "    plt.colorbar()\n",
    "    plt.xticks(np.arange(2), classes)\n",
    "    plt.yticks(np.arange(2), classes)\n",
    "    plt.tight_layout()\n",
    "    plt.xlabel('True label',rotation='horizontal', ha='right')\n",
    "    plt.ylabel('Predicted label')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f41c668>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True        No  Yes\n",
      "Predicted          \n",
      "No         882   48\n",
      "Yes         59   11\n",
      "     Precision\n",
      "No    0.948387\n",
      "Yes   0.157143\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1103dc5f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True        No  Yes\n",
      "Predicted          \n",
      "No         921   53\n",
      "Yes         20    6\n",
      "     Precision\n",
      "No    0.945585\n",
      "Yes   0.230769\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVEAAAElCAYAAABdxdGVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvAOZPmwAAIABJREFUeJzt3XtcFPX+P/DXLDfFVZAUT2oi4AHOyQzRpFQ0FRQvpWQCilia91uJJihioIIgeEmt1OM1NDRNy0cePV4y9XhBfyYp5xGaZl5R+Ya3ReW2n98fxBaxsKPDwK6+nj3m8WB3Zmfeq/ni85n5zGckIYQAERE9EU1NF0BEZMkYokRECjBEiYgUYIgSESnAECUiUoAhSkSkwDMfolevXoWnpyc2b95c5v1Vq1YhKiqq2uvZunUrRo0aBQCIjo7GkSNHqmS/1fV9PvvsM7z++uuYNm3aE32+b9++uHfvXhVX9eRmzJiBzMxMo+uU/P2kp6ejT58+Zd5bs2YNOnXqhKysLKSnp8PLywuHDx8us82sWbOwZMkSAEBUVBQCAwPx4MGDMtu0bt0aV69eNXrc1NRUdOjQAX379kXfvn0xaNCgJ6qf/mBd0wWYA41Gg6SkJLRp0wZubm41XY5BfHx8TZfw2LZs2YKUlBS0bdv2iT7/zTffVHFFyhw5cgQhISFG11Xl38/ChQuxe/dupKWloUmTJkhPT4eNjQ0iIyOxfft2ODk5Gf3ctWvXEB8fL7uWU6dOISoqCm+88UaV1f6sY4gCqFWrFoYOHYopU6Zg48aNsLW1LbP+/v37iIuLQ1ZWFiRJgp+fHyIiImBtbY2WLVuiW7duyMrKQkpKCgYNGoShQ4fiyJEjePDgAcaPH49du3bh3LlzcHZ2xrJly2Bvb48tW7Zg06ZNKCwsxN27dzFixIhyrYLw8HCEhYXBysoKS5cuNbx/+fJl+Pv7Izk5GT/88ANSUlLw8OFDaDQajB8/Hl26dEFhYSHmzJmDI0eO4LnnnsNzzz2HunXrGv3+y5cvx7Zt22BtbQ0XFxckJiaibt26+OSTT7Bjxw5YWVnB1dUVMTExaNiwIcLDw+Ht7Y0ffvgB2dnZeO211zB79mxERETg5s2biI6Oxvvvv4+0tDSEhYUhMDCwzPcJDAzE4sWLsWfPHtjY2KB+/fqYO3cunJ2d4enpiaNHj8LJyemxj6/RlO1YhYeH48UXX0RGRgZyc3MRHByM//u//8Px48fx8OFDLFq0CJ6ensjIyEBycjIKCgqQk5OD9u3bIyEhAQsXLsStW7cwZcoUzJs3DykpKXBwcMAvv/yCgQMHYvfu3QgLC4OjoyPef/99bN++HQ0bNsQ777yDV199FePGjUN0dDRatmyJgQMHGv2z1+v1mDVrFrKysvDFF1+gfv36hnUuLi5o1aoVpk+fjmXLlhn9/JAhQ/DNN9/gP//5D3r06FHB/+F/OHXqFHQ6HVasWAFnZ2dMnToVnp6eJj9HlRDPuCtXrghvb29RXFwswsLCRGJiohBCiJUrV4rIyEghhBBTp04Vs2fPFnq9XuTn54thw4aJ5cuXCyGE8PDwENu2bTPsz8PDQ6xbt04IIcTy5ctF69atxY0bN0RxcbEICgoS27dvFzqdTgQHB4vc3FwhhBCnTp0S3t7eQgghvvrqKzFy5EghhBCDBw8WO3fuLFPvvn37hL+/v8jJyRF37twR3bt3F1euXBFCCHHjxg3RqVMnce3aNbF27VoxZMgQkZ+fL/Ly8kRQUJDh+/zZ3r17Rffu3cWdO3eEEEIkJCSITz/9VGzZskWEhISIvLw8IYQQixcvFsOGDTPUNXHiRFFcXCzu378vOnbsKI4ePSqEEKJLly7i9OnTRusvfX39+nXh4+Mj8vPzhRBCrFq1SuzZs8fw5/fbb7898fH/bPDgwWL8+PFCCCEyMjKEh4eH2LdvnxBCiPj4eDFjxgwhhBCTJk0Sx44dE0IIodPphK+vrzhz5ozR7zNt2rRy30cIIRYsWCCGDx8ulixZIoYNGyaKi4vL1fNnx44dEz169BARERHCw8NDfP/99+XW9+7dW+Tl5Ynu3buL1NRUIYQQcXFxYvHixUIIISIjI8XKlSvFoUOHRLt27cT169eFEEJ4e3sb/p/4s7y8PDFs2DBx/PhxIYQQO3bsEH5+fkKn01VaK1XumT8nWkqj0SA5ORlbt24tdx7q4MGDGDx4MCRJgq2tLUJDQ3Hw4EHD+r92XUtbBM2aNYOHhwcaNWoEjUaDpk2b4u7du6hTpw6WLVuGAwcOYNGiRVi2bFm581rGZGRkIDY2FsuWLUODBg2QkZGBnJwcjBs3Dn379sXIkSMhSRLOnj2Lo0ePok+fPrC1tYW9vX2F3bejR48iMDAQDg4OAIBp06ZhzJgxOHjwIN566y3Y29sDKGnxHDt2DAUFBQCALl26QKPRQKvVwsXFBXfv3pX5Jw00atQIXl5eCAoKQlJSEv7xj3/A39+/zDZVdfyAgAAAwAsvvAAA8PPzA1Dyd1P6mcTERNy/fx/Lli1DXFwc8vPzK/z7qOg0xYQJE3D37l2kpaUhOTm5XKvYmIsXL8LW1hZJSUmIiopCdnZ2uW3s7e2xYMECLFy4EOfOnTO6n44dOyIoKAgffvgh9Hp9hcezt7fHqlWr8MorrwAAevXqBQcHB5w5c8ZkrVQxhuifPP/884iLi0NkZCRu375teF+v10OSpDKvi4qKDK9L/6GXsrGxMfpzqRs3bqBfv364du0a2rRpgw8++MBkbRcvXsSECROQnJwMd3d3AEBxcTHc3d3xzTffGJZNmzahY8eO5T5vZWVldL9WVlZlvtu9e/dw9epVk9+5Vq1ahp8lSYKoYAqGP79fWFgIoOQX1vr16zF37lw4OjoiISEB8+bNK/O5qjr+X0/NGPv7GDx4MA4cOAA3NzeMGzcOzs7OFe7vr3/Xpe7fv4+cnBxIkoRLly4Z3eavmjdvjrlz56Jfv34ICAjAhAkTDL8k/uzFF1/EmDFjMHnyZOTn5xvdV0REBPLy8sp0+9PS0gwXkKKjo3Ht2jWkpqaW+ZwQAtbWPKunBEP0LwIDA9GpUyesW7fO8F7Hjh2xfv16CCFQUFCAL7/8Eu3bt3/iY2RmZsLJyQljx45Fx44dsX//fgAloWhMTk4ORowYgalTp8LX19fwvre3Ny5duoQTJ04AAH766Sf06NEDN2/ehJ+fH77++mvk5+cjPz8f//73v43uu3379tizZw90Oh0AYMmSJVi7di38/Pzw1VdfGVpkqampeOWVV8qFUmWcnJwMV7bPnz+Ps2fPAgCysrLQp08fuLu7Y9SoUXj33XfLtYaq4vhy3Lt3D2fOnMGUKVPQvXt33LhxA5cvXza06KysrMqEd0Wio6Px5ptvYu7cuZgyZQru379v8jN/DvTo6GgUFxcjLi7O6LbvvfceGjRogO3btxtdb2tri/nz52P16tV49OgRAGDgwIGGX67x8fGoXbs2Fi1ahNOnTwMADhw4gIcPH6JVq1Yma6WK8VeQETNmzMDJkyfLvJ4zZw7eeOMNFBYWws/PD6NHj37i/Xfo0AFbtmxBYGAgJElCu3bt4OTkVGELZsmSJfjtt9+wbt06rFy5EgDg7OyMf/3rX1i8eDHmzZuH/Px8CCEwb948NG3aFKGhobh8+TL69OkDR0dHuLi4GN13586dcf78ecOFjxYtWmD27Nmwt7dHdnY2BgwYAL1eDxcXF6SkpDzW9xwzZgyioqIMrbzSrrCXlxd69uyJ/v37w97eHrVq1cKMGTPKfPbtt99WfHw56tWrh5EjRyIoKAj29vZo1KgRfHx8cOnSJbz22msICAjAhx9+iNjY2Ar3sWHDBmRnZ+Pjjz+GjY0NOnbsiJiYGCxatMjkhaVSdnZ2+PjjjxEUFIRWrVqhefPmZdZLkoSkpCS8+eabFe7Dzc0NkZGR5f4sSzk5OWHRokWYOXMmCgsLodVq8cknn1T5L6ZnjSQq6rcQEZFJ7M4TESnAECUiUoAhSkSkgNlfWHr06BEyMzPRsGHDCofpEJF5Ki4uRk5ODlq2bFlmWJoSd+7cMYwmkUOr1cLR0bFKjm2M2YdoZmYmwsLCaroMIlJgw4YNTzyfwp/duXMHbX07wAqmh52VcnBwwO7du1ULUrMP0YYNGwIAbti2RbGman6TkZmp26CmKyCVWBXn4W+5ewz/jpXS6XSwQhFu2rVFkWQ6D6zFI+Du/4NOp3t2Q7S0C1+sqYViTe0aroZUYa2t6QpIZVV9Kq5IYy8vD/TqX/Yx+xAlIipHAvCn24Ir3U5lDFEisjySpmSRs53KGKJEZHkkSWZLVP2mKEOUiCwPW6JERApIGkAj42JVNUyZzBAlIsvD7jwRkQLszhMRKcCWKBGRAmyJEhEpwJYoEZECkiSzJcoQJSIyQmZ3vhqmTGaIEpHl0Ugli5ztVMYQJSLLwwtLREQK8MISEZECvLBERKQAW6JERArwnCgRkRIyW6LVMLU9Q5SILI9G5lR4GrZEiYjKY3eeiEgBXlgiIlKALVEiIgUYokRECvC580RESnAWJyKiJ8cLS0RECvCcKBGRAmyJEhE9OUmSIMkISDnbKMUQJSKLU9IQlROi6tfCECUiyyNB3vAlhigRUXnszhMRKSBBZohyKjwiovI0Gg2EjGnuNJwKj4jICJ4TJSJSQOY5UY4TJSIygheWiIgUYIgSESnAwfZEREpVQ0DKwRAlIovD7jwRkQJKQlSv1yM2NhZnz56Fra0t5syZAxcXF8P6VatWYceOHZAkCaNHj0ZAQEClx2CIEpHFURKie/fuRUFBATZt2oSMjAwkJibis88+AwDcu3cPqamp2L17Nx4+fIh+/foxRInoKaRgsP3Jkyfh5+cHAPD29kZmZqZhXe3atdG4cWM8fPgQDx8+lBXUDFEisjhKWqI6nQ5ardbw2srKCkVFRbC2LonD559/Hr1790ZxcTFGjRpl8hjq31hKRFTVfg9RU4uxMU5arRZ5eXmG13q93hCgBw8exK1bt7Bv3z58//332Lt3L06fPl1pKQxRIrI4cgK0otaqj48PDh48CADIyMiAh4eHYZ2DgwNq1aoFW1tb2NnZoW7durh3716ltbA7T0QWR4K8ZywZmwovICAAhw8fRmhoKIQQSEhIwJo1a9CsWTN069YNR44cQXBwMDQaDXx8fNChQ4dKj8EQJSKLI2kAaGSEqJG+tkajwaxZs8q85+7ubvh54sSJmDhxouxaGKJEZHEqOt9pdDuVMUSJyOIwRImIlOCkzERET44tUSIiBcwpRKtlnGh6ejratm2L7Oxsw3spKSnYunVrdRyeiJ46cseIPiUhCgA2NjaYNm0ahBDVdUgiekopGWxf1aotRF999VU4ODhgw4YNZd5fvXo1+vfvj5CQECQnJ1dXOURkyaTHWFRWrbd9xsbGYu3atfj1118BAHl5edi5cyc2btyIjRs34tKlS9i/f391lkREFuiZbIkCQP369TF9+nRERUVBr9cjPz8fL7/8MmxsbCBJEtq2bYuff/65OksiIgtU+owl04v6tVT7BCRdu3aFq6srtm3bBjs7O5w+fRpFRUUQQuDEiRNwdXWt7pKIyMKUXpyXs6itRoY4RUdH49ixY6hTpw569uyJgQMHQq/Xo02bNvD396+JkojIgsgd4lQdKVotIerr6wtfX1/Da61WW+bc59ChQ6ujDCJ6SsgevfS0tkSJiJSQJAmSjFmcxNPSEiUiqkoajcwQ1UjQq1wLQ5SILI7si0bszhMRlSd7DCi780RE5bElSkSkAFuiRESKyAtRUQ1NUYYoEVkcMxprzxAlIssjtzvPme2JiIxgS5SISIHSWZzkbKc2higRWRy2RImIFOA5USIiBdgSJSJSRO6jP9gSJSIqR6ORoJE5i5PaGKJEZHHYnSciUoAXloiIFGBLlIhIAbZEiYgUYIgSESlUHV11ORiiRGRxLKIlOn/+/AoLiIiIUK0gIiJTLOLCkpubm/pHJyJ6AhYxi1NQUBAAoKioCNu2bUN2djZ8fX3x97//Xf2qiIgqYU4tUY2pDT766CNcv34dhw8fRl5eHiIjI9WvioioEhpJkr2oXoupDS5fvoz3338ftra26Nq1K+7fv696UURElSlticpZ/kqv12PmzJkICQlBeHg4Ll26VGb9gQMHEBwcjODgYMTGxkIIUWktJkO0uLgYubm5kCQJOp0OGo3JjxARqev3q/OmFmMpunfvXhQUFGDTpk2YPHkyEhMTDet0Oh2Sk5OxbNkyfPnll2jSpAlu375daSkmhzh98MEHGDhwIHJychASEoLp06c/wTcmIqo6GgByJmgy1uQ7efIk/Pz8AADe3t7IzMw0rDt16hQ8PDyQlJSEK1euYMCAAXBycqr0GCZDtF27dti5cydu374NJyenahl3RURUGblT4RnbRqfTQavVGl5bWVmhqKgI1tbWuH37NtLT0/H111/D3t4eYWFh8Pb2hqura8XHMFXEgQMHEBAQgOHDhyMwMBDp6ekmCyciUpP0GP/9lVarRV5enuG1Xq+HtXVJe9LR0REvvfQSGjZsiDp16qBt27b46aefKq3FZIguXboUmzdvxrZt27B+/XrMnz//cb8vEVGV0kjyl7/y8fHBwYMHAQAZGRnw8PAwrGvZsiXOnTuH3NxcFBUV4ccff0SLFi0qrcVkd75OnTqGcwINGzZE7dq1H+e7EhFVPZm3fRq7sBQQEIDDhw8jNDQUQggkJCRgzZo1aNasGbp164bJkydj+PDhAIDAwMAyIWtMhSG6YMECACVX50eNGoU2bdrg9OnTsLW1NV04EZGKlAy212g0mDVrVpn33N3dDT/37t0bvXv3ll1LhSFaeiL1zydUu3XrJnvHRERqkTuQvjoG28u67fPMmTMoKiqCEAK3bt1SvSgiosqY022fJs+Jjh8/HoWFhbh16xaKi4vh7OyMPn36qF8ZEVEFJJmPTDZ2db6qmbw6r9PpsGrVKrRq1Qpbt25Ffn6+6kUREVVGyW2fVc1kS7R0/NTDhw9Rq1YtFBYWql4UEVFlJEne+U6zCNGAgAAsXboUXl5eCA4ORp06ddSvioioEtLvi5zt1GYyRMPCwgw/d+7cGc2bN1ezHiIikyzi8SAREREVFsC7loioJlV0N5Kx7dRWYYiGhoaqf3QioidgES3Rdu3aqX5wIqInYVHjRImIzI1GkmAlZyq8mmyJEhGZK4vozk+bNq3CD82dO1eVYoiI5DCnIU4V3rHUq1cv9OrVC3fv3oWbmxvefvtteHp6oqCgoBrKIiKqmEU87dPPzw9+fn549OgRRowYgTZt2uDdd99Fbm6u6kUREVXGnG77NHnv/IMHD3D06FHodDocOnSIt30SUY2T86RPuedNlTIZovHx8UhLS0NISAi+/PJLLFy4UPWiiIgqJbcVag5DnNzd3TFp0iRcvnwZnp6eaNCggfpVERFVwiImZS61fv167NmzB3fv3kVQUBAuXbqEmTNnql4YEVFFLGqw/Y4dO/DFF19gyJAheOedd9C/f3/1qzLix+1xaNKkaY0cm9Sle1RU0yWQSrKvX8Pbb3xd5fuVIG8MqFnM4iSEAPBHwXxQHRHVNA1kXNCRuY1SJkO0d+/eCAsLw/Xr1zFixAj4+/tXQ1lERBWziDuWSg0cOBDt27fHuXPn4OrqisaNG6teFBFRZSSZU+HV6DjRnJwcXLx4EYMGDYKVlRW8vLxgY2ODYcOGqV8VEVElSucTlbOorcKW6I8//oh169bh4sWLmDlzJoQQ0Gg06Nixo/pVERFVwiK68/7+/vD398eBAwfQrl071K5dGzdv3kSjRo1UL4qIqDJWmpJFznZqM3mIM2fO4OOPPwZQcvfSihUrVC+KiKgykszJR8zits/vvvsOUVFRAIDFixfju+++U70oIqLKaB5jqY5aKiVJkmH6u8LCQsO4USKimlIy2F7GUg21mBziFBoaijfeeAMeHh745ZdfMHz48Gooi4ioYhZ17/yAAQPQrVs3XLlyBS+88AKcnJxUL4qIqDIWce/8p59+irFjxxp9/jyfO09ENcmcBttXGKJdu3YFwOfPE5H5sYjufFZWFrKyslQvgIjocVlEd/7ChQsASu5cqlWrFlq3bo0zZ86gqKgI/fr1U78yIqIKyL2ls0Zv+5w8eTIA4L333iszwJ73zhNRTZN+/0/OdmozOU40NzcX9+7dAwDcvn0bd+7cUb0oIqLKlF5YMrWYxcz2o0ePRv/+/aHVaqHT6ZCQkKB+VUREldBAZnde9UpkhGiPHj3Qo0cP/Pbbb6hXrx5sbGyqoSwioopZxCxOpU6cOIG4uDgUFxcjMDAQjRs3xoABA1QvjIioIuZ0Yclka3fRokVYv349GjRogNGjRyMtLU39qoiIKqHRAFYayeSiMYep8DQaDRwdHSFJEuzs7FCnTh31qyIiqoSSme31ej1mzpyJkJAQhIeH49KlS0a3GT58uKxGo8kQbdasGebPn487d+5gxYoVfMYSEdU4WTM4VXB1fu/evSgoKMCmTZswefJkJCYmlttm0aJFuHv3rqxaTIboRx99hMaNG6NNmzaoXbs2Zs+eLWvHRERq0UCSvfzVyZMn4efnBwDw9vZGZmZmmfW7du2CJEno1KmTrFpkDXFavXq1rJ0REVUHJbd96nQ6aLVaw2srKysUFRXB2toa586dw7fffovFixfjk08+kVWLyRCtW7cu9u3bh+bNm0Pz+1laV1dXWTsnIlKDBJmzOBl5T6vVIi8vz/Bar9fD2rokCr/++mvcvHkT77zzDq5duwYbGxs0adKk0lapyRDNzc3F2rVr/yhKkvD555+brp6ISCVKZnHy8fHB/v370atXL2RkZMDDw8OwburUqYaflyxZggYNGpjs1lcaojqdDitWrEDt2rVNFktEVF2UdOcDAgJw+PBhhIaGQgiBhIQErFmzBs2aNUO3bt0eu5YKQ3T9+vVYvXo1rK2tERMTYzgRS0RU05S0RDUaDWbNmlXmPXd393LbTZgwQV4tFa349ttvsWvXLmzcuBHr1q2TtTMiouqgZIhTVauwJWprawtbW1s4OTmhsLBQ/UqIiGSSIG9yEbN42icAPiaZiMyKRUxAcv78eUyePBlCCMPPpfigOiKqSRLktTJrtCW6aNEiw898WB0RmROLeFBdu3btVD84EdGTsIiWKBGRuZIkCRoZtyyZxaTMRETmRgN5V+fN4vEgRETmxiKuzhMRmSueEyUiUqDkbiQ5LVH1a2GIEpHF4TlRIiIlZJ4TrY6mKEOUiCwOz4kSESmgZD7RqsYQJSKLU9FD6IxtpzaGKBFZHLZEiYgUkH7/T852amOIEpHFYUuUiEgBSeY5UbZEiYiMYEuUiEgBDWROysyWKBFReRqpZJGzndoYokRkcXh1nohICbnPlGdLlIioPLZEiYgU4DlRIiIFSmZxktMSVZ8qc5ZOnDgRK1asMLzOy8tDjx49kJWVpcbhiOgZUzpOVM6iNlVCNDY2FmlpaTh//jwAICkpCSEhIfDy8lLjcET0jJEeY1GbKt15JycnxMTEYMaMGYiIiMCVK1cQFxeHs2fPYs6cOQAAR0dHJCQkoLCwEB988AGEECgsLERcXBw8PT3VKIuInhKSJG+wvUU/7bNr167Ys2cPoqKikJaWBkmSEBMTg4SEBLRo0QKbN2/GypUr0bp1a9StWxfz58/H+fPnodPp1CqJiJ4Sz8zM9v369cOjR4/QqFEjAMCFCxcQFxcHACgsLISrqys6deqEX3/9FWPHjoW1tTXGjBmjZklE9DQwoxSt1qvzrq6uSEpKQuPGjXHy5Enk5OQgPT0dzs7OWL16NU6dOoUFCxYgNTW1OssiIgvzzI4TjY2NRWRkJIqLiwEA8fHxcHR0xKRJk7Bu3TpoNBqMGzeuOksiIgv0zMzi5OvrC19fX8Prli1bGm1lrl27Vs0yiOgpY0a9eQ62JyLLI0Hec+efuu48EVFVeGa680REamB3nohICTNKUYYoEVkgeUOcqiNFGaJEZHGUnBPV6/WIjY3F2bNnYWtrizlz5sDFxcWwfu3atdixYwcAoHPnzhg/fnylx1BlAhIiIjUpmYBk7969KCgowKZNmzB58mQkJiYa1l25cgXbt2/Hxo0bsWnTJvz3v/81OfscW6JEZHkUnBM9efIk/Pz8AADe3t7IzMw0rPvb3/6GlStXwsrKCgBQVFQEOzu7Sg/BECUii6Pktk+dTgetVmt4bWVlhaKiIlhbW8PGxgZOTk4QQmDevHn45z//CVdX10qPwe48EVkcJZMya7Va5OXlGV7r9XpYW//RnszPz8eUKVOQl5eHjz76yGQtDFEisjhKzon6+Pjg4MGDAICMjAx4eHgY1gkhMHbsWHh6emLWrFmGbn1l2J0nIsuj4JxoQEAADh8+jNDQUAghkJCQgDVr1qBZs2bQ6/U4fvw4CgoKcOjQIQBAREQEWrduXeEhGKJEZHGUnBPVaDSYNWtWmffc3d0NP585c+axamGIEpHF4b3zREQKVcd98XIwRInI4nAqPCIiBdidJyJSwIwmcWKIEpEFMqMUZYgSkcUpyVA550TVxxAlIovDc6JERAqYUW+eIUpEFsiMUpQhSkQWR8ltn1WNIUpElkfmOVG2RImIjDCj3jxDlIgskBmlKEOUiCwOz4kSESnAcaJERApoAGhkBGR1PP+IIUpEFsh8TooyRInI4rA7T0SkgPm0QxmiRGSJONieiOjJcYgTEZESZtSfZ4gSkcUxowxliBKR5eHVeSIiBXhOlIhICTPqzzNEicjimFGGMkSJyPLwnCgRkQI8J0pEpATvWCIienISZHbnVa+EIUpEFojdeSIiBXhhiYhIAQ5xIiJSwoxSlCFKRBanJEPlnBNVH0OUiCwOz4kSESlgRr15higRWSAzSlGzD9Hi4mIAwM0bN2q4ElLLg/zimi6BVHLrZsm/29J/x1W335uQk5Al26nL7EM0JycHADB0SFgNV0JETyonJwcuLi6K96PVauHg4PBYeeDg4ACtVqv42BWRhBBCtb1XgUePHiEzMxMNGzaElZVVTZdDRI+huLgYOTk5aNmyJWrVqlUPPSlCAAAE2klEQVQl+7xz5w50Op3s7bVaLRwdHavk2MaYfYgSEZkzTU0XQERkyRiiREQKMERrUHp6Otq2bYvs7GzDeykpKdi6dWsNVkVKTZw4EStWrDC8zsvLQ48ePZCVlVWDVZFaGKI1zMbGBtOmTQNPTT89YmNjkZaWhvPnzwMAkpKSEBISAi8vrxqujNTAEK1hr776KhwcHLBhw4Yy769evRr9+/dHSEgIkpOTa6g6ehJOTk6IiYnBjBkzcPz4cVy5cgVDhw7F2bNnER4ejvDwcEyYMAH3799Hbm4uhgwZgvDwcISGhuLs2bM1XT49JrMfJ/osiI2NxYABA9CxY0cAJd2/nTt3YuPGjbC2tsaECROwf/9+dOnSpYYrJbm6du2KPXv2ICoqCmlpaZAkCTExMUhISECLFi2wefNmrFy5Eq1bt0bdunUxf/58nD9//rGG7pB5YIiagfr162P69OmIioqCj48P8vPz8fLLL8PGxgYA0LZtW/z8888MUQvTr18/PHr0CI0aNQIAXLhwAXFxcQCAwsJCuLq6olOnTvj1118xduxYWFtbY8yYMTVZMj0BdufNRNeuXeHq6opt27bBzs4Op0+fRlFREYQQOHHiBFxdXWu6RFLI1dUVSUlJSE1NxYcffojOnTsjPT0dzs7OWL16NcaMGYMFCxbUdJn0mNgSNSPR0dE4duwY6tSpg549e2LgwIHQ6/Vo06YN/P39a7o8Uig2NhaRkZGG+8jj4+Ph6OiISZMmYd26ddBoNBg3blwNV0mPi3csEREpwO48EZECDFEiIgUYokRECjBEiYgUYIgSESnAEH3KJCYmIjw8HIGBgXj99dcRHh6OiRMnVtn+O3ToUOG6q1evIjg4WNZ+HmdbInPGcaJPmaioKADA1q1b8csvv2DKlCk1XBHR040h+oxIT09HSkoKbGxsEBwcjMWLF2Pnzp2ws7NDSkoK3Nzc8NZbb2H+/Pk4ceIEhBB499130bNnT6P7O378OJYuXQqg5BEuSUlJsLGxQW5uLkaPHo3c3Fx07twZ48aNQ3Z2NmJiYpCfnw87OzvMnj27Or86karYnX+G5Ofn44svvkC/fv2Mrj9w4ACuXr2KjRs34vPPP8eyZctw7949o9v+/PPPSE5Oxueff46uXbti165dAIAHDx4gOTkZaWlpOHToELKyspCUlITw8HCkpqbivffeQ0pKimrfkai6sSX6DKno/vvSm9bOnTuH//3vfwgPDwcAFBUV4fr166hXr165zzRq1Ajx8fGwt7fHzZs34ePjAwDw8vJC3bp1AQAvvfQSLl68iHPnzmH58uVYuXIlhBCGiVWIngYM0WeIRvNHx8PW1ha3bt1C06ZNkZWVBXd3d7i5ucHX1xezZ8+GXq/Hp59+iqZNmxrd14wZM7B3715otVpERkYagvjChQvIy8szTKISEhICNzc3DBs2DD4+Prhw4QJOnDhRLd+XqDowRJ9Rw4cPx8iRI9GkSRNDS7Nr1644fvw4Bg0ahAcPHsDf37/C53X37dsXwcHBqFevHho0aIBbt24BKHnG96RJk5Cbm4tevXqhRYsWiIyMRGxsLPLz8/Ho0SNER0dX2/ckUhsnICEiUoAXloiIFGCIEhEpwBAlIlKAIUpEpABDlIhIAYYoEZECDFEiIgX+Pwm8UE3SFFNfAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c1b130978>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True        No  Yes\n",
      "Predicted          \n",
      "No         934   55\n",
      "Yes          7    4\n",
      "     Precision\n",
      "No    0.944388\n",
      "Yes   0.363636\n"
     ]
    }
   ],
   "source": [
    "for i in [1,3,5]:\n",
    "    pred, score, classes = KNN(i)\n",
    "    cm = confusion_matrix(y_test, pred)\n",
    "    cm_normalized = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "    plot_confusion_matrix(cm_normalized.T, classes, n_neighbors=i)\n",
    "    cm_df = pd.DataFrame(cm.T, index=classes, columns=classes)\n",
    "    cm_df.index.name = 'Predicted'\n",
    "    cm_df.columns.name = 'True'\n",
    "    print(cm_df)    \n",
    "    print(pd.DataFrame(precision_score(y_test, pred, average=None),\n",
    "                       index=classes, columns=['Precision']))        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "####  Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
       "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
       "          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
       "          verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regr = skl_lm.LogisticRegression()\n",
    "regr.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True        No  Yes\n",
      "Predicted          \n",
      "No         935   59\n",
      "Yes          6    0\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "         No       0.94      0.99      0.97       941\n",
      "        Yes       0.00      0.00      0.00        59\n",
      "\n",
      "avg / total       0.89      0.94      0.91      1000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "pred = regr.predict(X_test)\n",
    "cm_df = pd.DataFrame(confusion_matrix(y_test, pred).T, index=regr.classes_,\n",
    "                     columns=regr.classes_)\n",
    "cm_df.index.name = 'Predicted'\n",
    "cm_df.columns.name = 'True'\n",
    "print(cm_df)\n",
    "print(classification_report(y_test, pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Pred   No  Yes\n",
      "True          \n",
      "No    919   22\n",
      "Yes    48   11\n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "         No       0.95      0.98      0.96       941\n",
      "        Yes       0.33      0.19      0.24        59\n",
      "\n",
      "avg / total       0.91      0.93      0.92      1000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "pred_p = regr.predict_proba(X_test)\n",
    "cm_df = pd.DataFrame({'True': y_test, 'Pred': pred_p[:,1] > .25})\n",
    "cm_df.Pred.replace(to_replace={True:'Yes', False:'No'}, inplace=True)\n",
    "print(cm_df.groupby(['True', 'Pred']).size().unstack('True').T)\n",
    "print(classification_report(y_test, cm_df.Pred))"
   ]
  }
 ],
 "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}