day36-imputing-numerical-data/arbitrary-value-imputation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.compose import ColumnTransformer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('titanic_toy.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "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>Age</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Family</th>\n",
       "      <th>Survived</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>22.0</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>38.0</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>26.0</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>35.0</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>35.0</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Age     Fare  Family  Survived\n",
       "0  22.0   7.2500       1         0\n",
       "1  38.0  71.2833       1         1\n",
       "2  26.0   7.9250       0         1\n",
       "3  35.0  53.1000       1         1\n",
       "4  35.0   8.0500       0         0"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Age         0.198653\n",
       "Fare        0.050505\n",
       "Family      0.000000\n",
       "Survived    0.000000\n",
       "dtype: float64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.drop(columns=['Survived'])\n",
    "y = df['Survived']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.2,random_state=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-36-cb3531bd821d>:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  X_train['Age_99'] = X_train['Age'].fillna(99)\n",
      "<ipython-input-36-cb3531bd821d>:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  X_train['Age_minus1'] = X_train['Age'].fillna(-1)\n",
      "<ipython-input-36-cb3531bd821d>:4: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  X_train['Fare_999'] = X_train['Fare'].fillna(999)\n",
      "<ipython-input-36-cb3531bd821d>:5: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  X_train['Fare_minus1'] = X_train['Fare'].fillna(-1)\n"
     ]
    }
   ],
   "source": [
    "X_train['Age_99'] = X_train['Age'].fillna(99)\n",
    "X_train['Age_minus1'] = X_train['Age'].fillna(-1)\n",
    "\n",
    "X_train['Fare_999'] = X_train['Fare'].fillna(999)\n",
    "X_train['Fare_minus1'] = X_train['Fare'].fillna(-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original Age variable variance:  204.3495133904614\n",
      "Age Variance after 99 wala imputation:  951.7275570187172\n",
      "Age Variance after -1 wala imputation:  318.0896202624484\n",
      "Original Fare variable variance:  2448.197913706318\n",
      "Fare Variance after 999 wala imputation:  47219.20265217623\n",
      "Fare Variance after -1 wala imputation:  2378.5676784883503\n"
     ]
    }
   ],
   "source": [
    "print('Original Age variable variance: ', X_train['Age'].var())\n",
    "print('Age Variance after 99 wala imputation: ', X_train['Age_99'].var())\n",
    "print('Age Variance after -1 wala imputation: ', X_train['Age_minus1'].var())\n",
    "\n",
    "print('Original Fare variable variance: ', X_train['Fare'].var())\n",
    "print('Fare Variance after 999 wala imputation: ', X_train['Fare_999'].var())\n",
    "print('Fare Variance after -1 wala imputation: ', X_train['Fare_minus1'].var())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x227a8f2a3a0>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "fig = plt.figure()\n",
    "ax = fig.add_subplot(111)\n",
    "\n",
    "# original variable distribution\n",
    "X_train['Age'].plot(kind='kde', ax=ax)\n",
    "\n",
    "# variable imputed with the median\n",
    "X_train['Age_99'].plot(kind='kde', ax=ax, color='red')\n",
    "\n",
    "# variable imputed with the mean\n",
    "X_train['Age_minus1'].plot(kind='kde', ax=ax, color='green')\n",
    "\n",
    "# add legends\n",
    "lines, labels = ax.get_legend_handles_labels()\n",
    "ax.legend(lines, labels, loc='best')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x227a8bb0430>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "fig = plt.figure()\n",
    "ax = fig.add_subplot(111)\n",
    "\n",
    "# original variable distribution\n",
    "X_train['Fare'].plot(kind='kde', ax=ax)\n",
    "\n",
    "# variable imputed with the median\n",
    "X_train['Fare_999'].plot(kind='kde', ax=ax, color='red')\n",
    "\n",
    "# variable imputed with the mean\n",
    "X_train['Fare_minus1'].plot(kind='kde', ax=ax, color='green')\n",
    "\n",
    "# add legends\n",
    "lines, labels = ax.get_legend_handles_labels()\n",
    "ax.legend(lines, labels, loc='best')"
   ]
  },
  {
   "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></th>\n",
       "      <th>Age</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Family</th>\n",
       "      <th>Age_99</th>\n",
       "      <th>Age_minus1</th>\n",
       "      <th>Fare_999</th>\n",
       "      <th>Fare_minus1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Age</th>\n",
       "      <td>204.349513</td>\n",
       "      <td>70.719262</td>\n",
       "      <td>-6.498901</td>\n",
       "      <td>204.349513</td>\n",
       "      <td>204.349513</td>\n",
       "      <td>162.793430</td>\n",
       "      <td>63.321188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fare</th>\n",
       "      <td>70.719262</td>\n",
       "      <td>2448.197914</td>\n",
       "      <td>17.258917</td>\n",
       "      <td>-101.671097</td>\n",
       "      <td>125.558364</td>\n",
       "      <td>2448.197914</td>\n",
       "      <td>2448.197914</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Family</th>\n",
       "      <td>-6.498901</td>\n",
       "      <td>17.258917</td>\n",
       "      <td>2.735252</td>\n",
       "      <td>-7.387287</td>\n",
       "      <td>-4.149246</td>\n",
       "      <td>11.528625</td>\n",
       "      <td>16.553989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Age_99</th>\n",
       "      <td>204.349513</td>\n",
       "      <td>-101.671097</td>\n",
       "      <td>-7.387287</td>\n",
       "      <td>951.727557</td>\n",
       "      <td>-189.535540</td>\n",
       "      <td>-159.931663</td>\n",
       "      <td>-94.317400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Age_minus1</th>\n",
       "      <td>204.349513</td>\n",
       "      <td>125.558364</td>\n",
       "      <td>-4.149246</td>\n",
       "      <td>-189.535540</td>\n",
       "      <td>318.089620</td>\n",
       "      <td>257.379887</td>\n",
       "      <td>114.394141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fare_999</th>\n",
       "      <td>162.793430</td>\n",
       "      <td>2448.197914</td>\n",
       "      <td>11.528625</td>\n",
       "      <td>-159.931663</td>\n",
       "      <td>257.379887</td>\n",
       "      <td>47219.202652</td>\n",
       "      <td>762.474982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fare_minus1</th>\n",
       "      <td>63.321188</td>\n",
       "      <td>2448.197914</td>\n",
       "      <td>16.553989</td>\n",
       "      <td>-94.317400</td>\n",
       "      <td>114.394141</td>\n",
       "      <td>762.474982</td>\n",
       "      <td>2378.567678</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    Age         Fare     Family      Age_99  Age_minus1  \\\n",
       "Age          204.349513    70.719262  -6.498901  204.349513  204.349513   \n",
       "Fare          70.719262  2448.197914  17.258917 -101.671097  125.558364   \n",
       "Family        -6.498901    17.258917   2.735252   -7.387287   -4.149246   \n",
       "Age_99       204.349513  -101.671097  -7.387287  951.727557 -189.535540   \n",
       "Age_minus1   204.349513   125.558364  -4.149246 -189.535540  318.089620   \n",
       "Fare_999     162.793430  2448.197914  11.528625 -159.931663  257.379887   \n",
       "Fare_minus1   63.321188  2448.197914  16.553989  -94.317400  114.394141   \n",
       "\n",
       "                 Fare_999  Fare_minus1  \n",
       "Age            162.793430    63.321188  \n",
       "Fare          2448.197914  2448.197914  \n",
       "Family          11.528625    16.553989  \n",
       "Age_99        -159.931663   -94.317400  \n",
       "Age_minus1     257.379887   114.394141  \n",
       "Fare_999     47219.202652   762.474982  \n",
       "Fare_minus1    762.474982  2378.567678  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.cov()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>Age</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Family</th>\n",
       "      <th>Age_99</th>\n",
       "      <th>Age_minus1</th>\n",
       "      <th>Fare_999</th>\n",
       "      <th>Fare_minus1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Age</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.092644</td>\n",
       "      <td>-0.299113</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.051179</td>\n",
       "      <td>0.084585</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fare</th>\n",
       "      <td>0.092644</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.208268</td>\n",
       "      <td>-0.066273</td>\n",
       "      <td>0.142022</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Family</th>\n",
       "      <td>-0.299113</td>\n",
       "      <td>0.208268</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.144787</td>\n",
       "      <td>-0.140668</td>\n",
       "      <td>0.032079</td>\n",
       "      <td>0.205233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Age_99</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.066273</td>\n",
       "      <td>-0.144787</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.344476</td>\n",
       "      <td>-0.023857</td>\n",
       "      <td>-0.062687</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Age_minus1</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.142022</td>\n",
       "      <td>-0.140668</td>\n",
       "      <td>-0.344476</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.066411</td>\n",
       "      <td>0.131514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fare_999</th>\n",
       "      <td>0.051179</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.032079</td>\n",
       "      <td>-0.023857</td>\n",
       "      <td>0.066411</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.071946</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fare_minus1</th>\n",
       "      <td>0.084585</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.205233</td>\n",
       "      <td>-0.062687</td>\n",
       "      <td>0.131514</td>\n",
       "      <td>0.071946</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Age      Fare    Family    Age_99  Age_minus1  Fare_999  \\\n",
       "Age          1.000000  0.092644 -0.299113  1.000000    1.000000  0.051179   \n",
       "Fare         0.092644  1.000000  0.208268 -0.066273    0.142022  1.000000   \n",
       "Family      -0.299113  0.208268  1.000000 -0.144787   -0.140668  0.032079   \n",
       "Age_99       1.000000 -0.066273 -0.144787  1.000000   -0.344476 -0.023857   \n",
       "Age_minus1   1.000000  0.142022 -0.140668 -0.344476    1.000000  0.066411   \n",
       "Fare_999     0.051179  1.000000  0.032079 -0.023857    0.066411  1.000000   \n",
       "Fare_minus1  0.084585  1.000000  0.205233 -0.062687    0.131514  0.071946   \n",
       "\n",
       "             Fare_minus1  \n",
       "Age             0.084585  \n",
       "Fare            1.000000  \n",
       "Family          0.205233  \n",
       "Age_99         -0.062687  \n",
       "Age_minus1      0.131514  \n",
       "Fare_999        0.071946  \n",
       "Fare_minus1     1.000000  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.corr()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Using Sklearn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.2,random_state=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "imputer1 = SimpleImputer(strategy='constant',fill_value=99)\n",
    "imputer2 = SimpleImputer(strategy='constant',fill_value=999)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "trf = ColumnTransformer([\n",
    "    ('imputer1',imputer1,['Age']),\n",
    "    ('imputer2',imputer2,['Fare'])\n",
    "],remainder='passthrough')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ColumnTransformer(remainder='passthrough',\n",
       "                  transformers=[('imputer1',\n",
       "                                 SimpleImputer(fill_value=99,\n",
       "                                               strategy='constant'),\n",
       "                                 ['Age']),\n",
       "                                ('imputer2',\n",
       "                                 SimpleImputer(fill_value=999,\n",
       "                                               strategy='constant'),\n",
       "                                 ['Fare'])])"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trf.fit(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([99.])"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trf.named_transformers_['imputer1'].statistics_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([999.])"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trf.named_transformers_['imputer2'].statistics_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = trf.transform(X_train)\n",
    "X_test = trf.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 40.    ,  27.7208,   0.    ],\n",
       "       [  4.    ,  16.7   ,   2.    ],\n",
       "       [ 47.    ,   9.    ,   0.    ],\n",
       "       ...,\n",
       "       [ 71.    ,  49.5042,   0.    ],\n",
       "       [ 99.    , 221.7792,   0.    ],\n",
       "       [ 99.    ,  25.925 ,   0.    ]])"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train"
   ]
  },
  {
   "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}