preprocessing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib \n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "from pandas import Series, DataFrame\n",
    "import pickle\n",
    "import datetime\n",
    "import math\n",
    "from tqdm import tqdm\n",
    "import os\n",
    "import calendar\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.preprocessing import RobustScaler, LabelEncoder\n",
    "from sklearn.model_selection import cross_val_score, KFold\n",
    "from scipy.stats import skew, norm\n",
    "from xgboost.sklearn import XGBRegressor\n",
    "\n",
    "pd.set_option('display.max_columns', None) # no truncate columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Environment settings\n",
    "data_path_out = 'Data/output/'\n",
    "    \n",
    "# Deserialize previously saved data from \"data-visualization\"\n",
    "df = {}\n",
    "for df_name in ['train', 'test', 'all', 'monthly']:\n",
    "    with open(data_path_out+df_name+'_dv.obj', 'rb') as file:\n",
    "        df[df_name] = pickle.load(file)\n",
    "        \n",
    "all_data = df['all']\n",
    "train = df['train']\n",
    "test = df['test']\n",
    "monthly_sales = df['monthly']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Drop features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_data = all_data.drop('Max_Gust_SpeedKm_h', axis=1)\n",
    "# also implicitly drop NumberOfCustomers when spllitting all_data at the end\n",
    "# and combined events when One Hot Encoding is performed"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fill missing values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Events"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_data.Events = all_data.Events.cat.add_categories('Normal')\n",
    "train.Events = train.Events.cat.add_categories('Normal')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 0 missing values of Events left\n",
      "There are 0 missing values of Events left\n"
     ]
    }
   ],
   "source": [
    "all_data.Events = all_data.Events.fillna('Normal')\n",
    "print ('There are {} missing values of Events left'\\\n",
    "       .format(all_data['Events'].isna().sum(),\n",
    "               all_data['Events'].isna().sum()/all_data.shape[0]*100))\n",
    "\n",
    "#  JUST FOR VISUSALIZATION PURPOSES\n",
    "train.Events = train.Events.fillna('Normal')\n",
    "print ('There are {} missing values of Events left'\\\n",
    "       .format(train['Events'].isna().sum(),\n",
    "               train['Events'].isna().sum()/all_data.shape[0]*100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ff850229358>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### ADDED NORMAL\n",
    "f, ax = plt.subplots(figsize=(16, 8))\n",
    "fig = sns.boxplot(x='Events', y=\"NumberOfSales\", data=train[train.IsOpen == True])\n",
    "plt.xticks(rotation=90);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visibility"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since the Visibility has a high negative correlation  with Humidity, a good imputation method could be to group data by Humidity and use the median value to fill in missing values.\n",
    "\n",
    "Intuitively the minimum visibility corresponds to the maximum humidity and viceversa.\n",
    "\n",
    "Some humidity levels could not have visiblity values, so a good solution could be to take the mean between the two adiacent values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "StoreID                       False\n",
       "Date                          False\n",
       "IsHoliday                     False\n",
       "IsOpen                        False\n",
       "HasPromotions                 False\n",
       "StoreType                     False\n",
       "AssortmentType                False\n",
       "NearestCompetitor             False\n",
       "Region                        False\n",
       "Region_AreaKM2                False\n",
       "Region_GDP                    False\n",
       "Region_PopulationK            False\n",
       "CloudCover                     True\n",
       "Events                        False\n",
       "Max_Dew_PointC                False\n",
       "Max_Humidity                  False\n",
       "Max_Sea_Level_PressurehPa     False\n",
       "Max_TemperatureC              False\n",
       "Max_VisibilityKm               True\n",
       "Max_Wind_SpeedKm_h            False\n",
       "Mean_Dew_PointC               False\n",
       "Mean_Humidity                 False\n",
       "Mean_Sea_Level_PressurehPa    False\n",
       "Mean_TemperatureC             False\n",
       "Mean_VisibilityKm              True\n",
       "Mean_Wind_SpeedKm_h           False\n",
       "Min_Dew_PointC                False\n",
       "Min_Humidity                  False\n",
       "Min_Sea_Level_PressurehPa     False\n",
       "Min_TemperatureC              False\n",
       "Min_VisibilitykM               True\n",
       "Precipitationmm               False\n",
       "WindDirDegrees                False\n",
       "dtype: bool"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_data.isna().any()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_to_predict = ['Mean_VisibilityKm', 'Max_VisibilityKm', 'Min_VisibilitykM', 'CloudCover']\n",
    "\n",
    "cv_nans = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "if cv_nans == True:\n",
    "    for feature in tqdm(feature_to_predict):\n",
    "        drop_features = [f for f in feature_to_predict if f != feature]\n",
    "\n",
    "        df_feature = pd.get_dummies(all_data).drop(['Date'] + drop_features, axis=1)\n",
    "\n",
    "        nan_mask = df_feature[feature].isnull()\n",
    "        feat_test = df_feature[nan_mask]\n",
    "        feat_train = df_feature[-nan_mask]\n",
    "\n",
    "        y_train = feat_train[feature]\n",
    "        X_train = feat_train.drop(feature, axis=1)\n",
    "        X_test = feat_test.drop(feature, axis=1)\n",
    "\n",
    "        model = XGBRegressor(n_jobs=4, max_depth=9)\n",
    "        kfold = KFold(n_splits=5, shuffle=True, random_state=42)\n",
    "        results = cross_val_score(model,X_train,y_train,cv=kfold,verbose=2)\n",
    "\n",
    "        print('Cross-validation for {} -> score: {:.4f} with +/- {:.4f}'\\\n",
    "              .format(feature, results.mean(), results.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4/4 [05:58<00:00, 89.52s/it]\n"
     ]
    }
   ],
   "source": [
    "for feature in tqdm(feature_to_predict):\n",
    "    drop_features = [f for f in feature_to_predict if f != feature]\n",
    "\n",
    "    df_feature = pd.get_dummies(all_data).drop(['Date'] + drop_features, axis=1)\n",
    "    \n",
    "    nan_mask = df_feature[feature].isnull()\n",
    "    feat_test = df_feature[nan_mask]\n",
    "    feat_train = df_feature[-nan_mask]\n",
    "    \n",
    "    y_train = feat_train[feature]\n",
    "    X_train = feat_train.drop(feature, axis=1)\n",
    "    X_test = feat_test.drop(feature, axis=1)\n",
    "    \n",
    "    model = XGBRegressor(n_jobs=4, max_depth=9)\n",
    "    model.fit(X_train,y_train)\n",
    "    predictions = model.predict(X_test)\n",
    "    \n",
    "    nan_mask = all_data[feature].isnull()\n",
    "    all_data.loc[nan_mask,feature] = predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Region 2 missing time window\n",
    "Fill missing data in datetime range 2017-07-04/2018-01-03 (extremes included)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "process_region2 = False"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Window dates filling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "if process_region2:\n",
    "    # create date range dataframe\n",
    "    date_missing_start = '2017-07-04'\n",
    "    date_missing_end = '2018-01-03'\n",
    "\n",
    "    date_range = pd.date_range(date_missing_start, date_missing_end)\n",
    "\n",
    "    # create df and set dtypes\n",
    "    df_region2 = pd.DataFrame(data={'Date': date_range, 'Region': '2'})\n",
    "\n",
    "    df_region2.Date = pd.to_datetime(df_region2.Date)\n",
    "    df_region2.Region = df_region2.Region.astype('category')\n",
    "\n",
    "    print(df_region2.shape)\n",
    "    df_region2.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Replacement by similar store"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "if process_region2:\n",
    "    # HasPromotions (replacement with store 1006 since all are equal)\n",
    "    df_region2['HasPromotions'] = all_data[all_data.StoreID == 1006]\\\n",
    "    .loc[(all_data.Date >= date_missing_start) & (all_data.Date <= date_missing_end)]\\\n",
    "    .HasPromotions.reset_index().drop('index', axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Replacement by most correlated region"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "if process_region2:\n",
    "    # Group all_data by Region\n",
    "    dic_all = {\n",
    "        'IsHoliday': 'mean',\n",
    "        'IsOpen': 'mean',\n",
    "        'HasPromotions': 'mean',\n",
    "        'NearestCompetitor': 'mean',\n",
    "        'Region_AreaKM2': 'mean',\n",
    "        'Region_GDP': 'mean',\n",
    "        'Region_PopulationK': 'mean',\n",
    "        'CloudCover': 'mean',\n",
    "        'Max_Humidity': 'mean',\n",
    "        'Max_Dew_PointC': 'mean',\n",
    "        'Max_Sea_Level_PressurehPa': 'mean',\n",
    "        'Max_TemperatureC': 'mean',\n",
    "        'Max_VisibilityKm': 'mean',\n",
    "        'Max_Wind_SpeedKm_h': 'mean',\n",
    "        'Mean_Dew_PointC': 'mean',\n",
    "        'Mean_Humidity': 'mean',\n",
    "        'Mean_Sea_Level_PressurehPa': 'mean',\n",
    "        'Mean_TemperatureC': 'mean',\n",
    "        'Mean_VisibilityKm': 'mean',\n",
    "        'Mean_Wind_SpeedKm_h': 'mean',\n",
    "        'Min_Dew_PointC': 'mean',\n",
    "        'Min_Humidity': 'mean',\n",
    "        'Min_Sea_Level_PressurehPa': 'mean',\n",
    "        'Min_TemperatureC': 'mean',\n",
    "        'Min_VisibilitykM': 'mean',\n",
    "        'Precipitationmm': 'mean',\n",
    "        'WindDirDegrees': 'mean'}\n",
    "\n",
    "\n",
    "    # Define train_data grouped by Region in common date range\n",
    "    all_data_byRegion = all_data.groupby(['Region', pd.Grouper(key='Date', freq='D')])\\\n",
    "    .agg(dic_all).reset_index()\n",
    "\n",
    "    all_data_common_byRegion =\\\n",
    "    all_data_byRegion.loc[(all_data_byRegion.Date >= '2016-03-01') &\n",
    "                          (all_data_byRegion.Date <= '2017-07-03')]\n",
    "\n",
    "    all_data_missing_byRegion =\\\n",
    "    all_data_byRegion[all_data_byRegion.Region != '2']\\\n",
    "    .loc[(all_data_byRegion.Date >= date_missing_start) &\n",
    "         (all_data_byRegion.Date <= date_missing_end)]\n",
    "\n",
    "    # Define train_data grouped by Region in common date range\n",
    "    all_data_byRegion_except_region2 = all_data_common_byRegion[all_data_common_byRegion.Region != '2']\n",
    "    all_data_byRegion_except_region2.Region.cat.remove_unused_categories(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "if process_region2:\n",
    "    # Replace by most correlated region\n",
    "    median_correction = ['Max_TemperatureC', 'CloudCover', 'Max_Sea_Level_PressurehPa',\n",
    "                         'Max_Wind_SpeedKm_h', 'Max_Dew_PointC', 'WindDirDegrees',\n",
    "                         'Mean_Dew_PointC', 'Mean_Sea_Level_PressurehPa', 'Mean_TemperatureC',\n",
    "                         'Mean_Wind_SpeedKm_h', 'Min_Dew_PointC', 'Min_Sea_Level_PressurehPa',\n",
    "                         'Min_VisibilitykM']\n",
    "\n",
    "    print('Most correlated region for the given feature')\n",
    "    print('{:<30}{:10}{:>10}'.format('FEATURE', 'REGION', 'CORRELATION'))\n",
    "    print('-' * (30+10+10))\n",
    "\n",
    "    # get the date range\n",
    "    for feature_name in median_correction:\n",
    "        # compute correlation for each region wrt region 2 for current feature\n",
    "        corr_byRegion = {}\n",
    "        for region_name, region in all_data_byRegion_except_region2.groupby('Region'):\n",
    "            region = region[feature_name]\n",
    "            region_2 = all_data_common_byRegion[feature_name][all_data_common_byRegion.Region == '2']\n",
    "\n",
    "            corr_byRegion[region_name] = pd.np.corrcoef(region_2, region)[1,0]\n",
    "\n",
    "        # get region with max correlation for the current feature\n",
    "        max_corr_region = max(corr_byRegion, key=corr_byRegion.get)\n",
    "\n",
    "        print('{:<30}{:10}{:10.4f}'.format(feature_name, max_corr_region, corr_byRegion[max_corr_region]))\n",
    "\n",
    "        # replace missing feature for date window with most correlated region feature \n",
    "        feature_column = all_data_missing_byRegion[all_data_missing_byRegion.Region == max_corr_region]\\\n",
    "        [feature_name].reset_index().drop('index', axis=1)\n",
    "\n",
    "        df_region2 = pd.concat([df_region2, feature_column], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Replacement by region 2 itself"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "if process_region2:\n",
    "    # copy from Region 2: Region_AreaKM2, Region_GDP, Region_PopulationK\n",
    "    for feature_name in ['Region_AreaKM2', 'Region_GDP', 'Region_PopulationK']:\n",
    "        df_region2[feature_name] = all_data[all_data.Region == '2'][feature_name].unique()[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Replacement by single store from Region 2 and concatenation with dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "if process_region2:\n",
    "    # Add missing dates window to dataset\n",
    "    print(all_data.shape)\n",
    "\n",
    "    # add to all_data missing region 2 values\n",
    "    for storeid in all_data[all_data.Region == '2'].StoreID.unique():    \n",
    "        # copy from same stores from Region 2: StoreID, AssortmentType, NearestCompetitor, StoreType\n",
    "        store_data = {\n",
    "            'StoreID': np.array([storeid], dtype='uint16'),\n",
    "            'AssortmentType': all_data[all_data.StoreID == storeid].AssortmentType.unique()[0],\n",
    "            'NearestCompetitor': all_data[all_data.StoreID == storeid].NearestCompetitor.unique()[0],\n",
    "            'StoreType': all_data[all_data.StoreID == storeid].StoreType.unique()[0]\n",
    "        }\n",
    "\n",
    "        # create 'storeid' column\n",
    "        df_storeid = pd.DataFrame(data=store_data, index=np.arange(0, df_region2.shape[0]))\n",
    "\n",
    "        df_storeid = pd.concat([df_region2, df_storeid], axis=1)\n",
    "\n",
    "        # add to all_data df the missing values range for 'storeid'\n",
    "        all_data = pd.concat([all_data, df_storeid], axis=0, ignore_index=True, sort=True)\n",
    "\n",
    "    all_data = all_data\\\n",
    "    .sort_values(by=['Region', 'StoreID', 'Date'], ascending=[False, False, True])\\\n",
    "    .reset_index().drop('index', axis=1)\n",
    "\n",
    "    print(all_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "if process_region2:\n",
    "    all_data['IsHoliday']=all_data['IsHoliday'].astype('bool')\n",
    "    all_data['IsOpen']=all_data['IsOpen'].astype('bool')\n",
    "    all_data['AssortmentType']=all_data['AssortmentType'].astype('category')\n",
    "    all_data['StoreType']=all_data['StoreType'].astype('category')\n",
    "    all_data['Region']=all_data['Region'].astype('category')\n",
    "    all_data['Events']=all_data['Events'].astype('category')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "if process_region2:\n",
    "    print('TODO:')\n",
    "    all_data[all_data.Region == '2'].isna().any()\n",
    "    all_data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## New Features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Store is open and is holiday\n",
    "This feature has been added for those stores that are open during holidays: we want to inspect if the sales are higher in those days"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "is_open_holiday = all_data.IsHoliday & all_data.IsOpen\n",
    "all_data['Hol_and_open'] = is_open_holiday"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ff80c861b70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# For plotting reasons\n",
    "holidays_open_mask = train.IsHoliday & train.IsOpen\n",
    "open_mask = train['IsOpen'] == True\n",
    "not_holidays_mask =  train['IsHoliday'] == False\n",
    "not_holidays_open_mask = not_holidays_mask & open_mask\n",
    "\n",
    "df_holidays_open = train[holidays_open_mask].groupby(\"StoreID\")\n",
    "avg_sales_holidays = df_holidays_open.NumberOfSales.mean()\n",
    "avg_sales_holidays.plot(figsize=(16,8),legend=True,marker='o')\n",
    "\n",
    "df_not_holidays_open = train[not_holidays_open_mask].groupby(\"StoreID\")\n",
    "avg_sales_not_holidays = df_not_holidays_open.NumberOfSales.mean()\n",
    "avg_sales_not_holidays = avg_sales_not_holidays.filter(items = avg_sales_holidays.index)\n",
    "avg_sales_not_holidays.plot(figsize=(16,8),legend=True,marker='o');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Region's population density"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_data['Region_PD'] = all_data.Region_PopulationK.div(all_data.Region_AreaKM2, axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ff80c86dd30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_PD = all_data.groupby(\"Region\").Region_PD.mean().sort_values(ascending=False)\\\n",
    "        .plot(kind='bar', title='Population Density', figsize=(16,8));\n",
    "plot_PD.set_ylabel(\"K people / squared Km\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Number of stores per region"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "stores_per_region = all_data[['Region', 'StoreID']].drop_duplicates().groupby(['Region'])\\\n",
    ".count().rename(columns={'StoreID': 'stores_per_region'})\n",
    "all_data = all_data.reset_index().merge(stores_per_region, on='Region', how='left').set_index('index')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Number of stores per person"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# compute stores_per_region and persons_per_region\n",
    "stores_per_region = all_data[['Region', 'StoreID']].drop_duplicates().groupby(['Region'])\\\n",
    ".count().rename(columns={'StoreID': 'stores_per_region'})\n",
    "persons_per_region = all_data[['Region', 'Region_PopulationK']]\\\n",
    ".drop_duplicates().set_index('Region')\n",
    "\n",
    "# compute stores_per_person\n",
    "stores_per_person = pd.merge(stores_per_region, persons_per_region, on='Region')\n",
    "stores_per_person['stores_per_person'] = stores_per_person.apply(lambda x: x[0]/x[1], axis=1)\n",
    "stores_per_person = stores_per_person.drop(['Region_PopulationK', 'stores_per_region'], axis=1)\n",
    "\n",
    "# merge stores_per_person to all_data dataframe\n",
    "all_data = all_data.reset_index().merge(stores_per_person, on='Region', how='left').set_index('index')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Number of Stores per Km^2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# compute stores_per_region and KM2_per_region\n",
    "stores_per_region = all_data[['Region', 'StoreID']].drop_duplicates()\\\n",
    ".groupby(['Region']).count().rename(columns={'StoreID': 'stores_per_region'})\n",
    "KM2_per_region = all_data[['Region', 'Region_AreaKM2']].drop_duplicates()\\\n",
    ".set_index('Region')\n",
    "\n",
    "# compute stores_per_KM2\n",
    "stores_per_KM2 = pd.merge(stores_per_region, KM2_per_region, on='Region')\n",
    "stores_per_KM2['stores_per_KM2'] = stores_per_KM2.apply(lambda x: x[0]/x[1], axis=1)\n",
    "stores_per_KM2 = stores_per_KM2.drop(['Region_AreaKM2', 'stores_per_region'], axis=1)\n",
    "\n",
    "# merge stores_per_KM2 to all_data dataframe\n",
    "all_data = all_data.reset_index().merge(stores_per_KM2, on='Region', how='left').set_index('index')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### GDP per capita"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get Region_GDP and Region_PopulationK\n",
    "GDP_per_capita = all_data[['Region', 'Region_GDP', 'Region_PopulationK']].drop_duplicates()\\\n",
    ".set_index('Region')\n",
    "\n",
    "# compute GDP_per_capita\n",
    "GDP_per_capita['GDP_per_capita'] = GDP_per_capita.apply(lambda x: x[0]/(x[1]*1000), axis=1)\n",
    "GDP_per_capita = GDP_per_capita.drop(['Region_GDP', 'Region_PopulationK'], axis=1)\n",
    "\n",
    "# merge GDP_per_capita to all_data dataframe\n",
    "all_data = all_data.reset_index().merge(GDP_per_capita, on='Region', how='left').set_index('index')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### RealFeel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from https://gist.githubusercontent.com/mumblepins/d4e6f2189200cbd0914e217a2e4adde9/raw/05bebe8eea8c482da0a080d6e030ce015ebae6de/realfeel.py\n",
    "# Based on patent https://www.google.com/patents/US7251579\n",
    "# Not responsible if you use this for anything other than personal use\n",
    "from math import sqrt\n",
    "\n",
    "def realfeel(W, #windspeed mph\n",
    "            A, #pressure mb\n",
    "            T, # temperature F\n",
    "            UV, # UV Index\n",
    "            D, # Dew Point F\n",
    "            P2, # preciptation Factor from 0-5\n",
    "        ):\n",
    "    if W<4:\n",
    "        Wa=W/2+2\n",
    "    elif W<56:\n",
    "        Wa=W\n",
    "    else:\n",
    "        Wa=56\n",
    "\n",
    "    WSP2=(80-T)*(0.566+0.25*sqrt(Wa)-0.0166*Wa)*((sqrt(A/10))/10)\n",
    "    WSP1=sqrt(W)*((sqrt(A/10))/10)\n",
    "\n",
    "    SI2 = UV\n",
    "\n",
    "    if D >= (55+sqrt(W)):\n",
    "        Da=D\n",
    "    else:\n",
    "        Da=55+sqrt(W)\n",
    "\n",
    "    H2=(Da-55-sqrt(W))**2/30\n",
    "\n",
    "    if T>= 65:\n",
    "        MFT=80-WSP2+SI2+H2-P2\n",
    "    else:\n",
    "        MFT=T-WSP1+SI2+H2-P2\n",
    "    \n",
    "    return MFT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "real_feel = all_data[['Precipitationmm', 'Mean_Dew_PointC', 'Mean_TemperatureC', 'Mean_Sea_Level_PressurehPa', 'Mean_Wind_SpeedKm_h', 'CloudCover', 'Date']].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# compute uv_index estimation\n",
    "# based on https://science.howstuffworks.com/nature/climate-weather/atmospheric/question160.htm\n",
    "real_feel['season_weight'] = real_feel.Date.dt.quarter.apply(lambda q: 2 if q==4 else q)\n",
    "real_feel['uv_index'] = real_feel[['season_weight', 'CloudCover']].apply(lambda x: (x[0]+1)/(x[1]+1), axis=1)\n",
    "\n",
    "# scale to max 5 uv index as nordic countries have low uv index\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "mms = MinMaxScaler(feature_range=(0, 5))\n",
    "real_feel['uv_index'] = mms.fit_transform(real_feel[['uv_index']].values).astype(int)"
   ]
  },
  {
   "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>Precipitationmm</th>\n",
       "      <th>Mean_Dew_PointC</th>\n",
       "      <th>Mean_TemperatureC</th>\n",
       "      <th>Mean_Sea_Level_PressurehPa</th>\n",
       "      <th>Mean_Wind_SpeedKm_h</th>\n",
       "      <th>CloudCover</th>\n",
       "      <th>Date</th>\n",
       "      <th>season_weight</th>\n",
       "      <th>uv_index</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>index</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1030</td>\n",
       "      <td>16</td>\n",
       "      <td>8.0</td>\n",
       "      <td>2016-03-01</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>-1</td>\n",
       "      <td>3</td>\n",
       "      <td>1027</td>\n",
       "      <td>10</td>\n",
       "      <td>8.0</td>\n",
       "      <td>2016-03-02</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>-1</td>\n",
       "      <td>3</td>\n",
       "      <td>1024</td>\n",
       "      <td>10</td>\n",
       "      <td>8.0</td>\n",
       "      <td>2016-03-04</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>-4</td>\n",
       "      <td>3</td>\n",
       "      <td>1024</td>\n",
       "      <td>10</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2016-03-05</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>-3</td>\n",
       "      <td>1</td>\n",
       "      <td>1023</td>\n",
       "      <td>5</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2016-03-06</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Precipitationmm  Mean_Dew_PointC  Mean_TemperatureC  \\\n",
       "index                                                        \n",
       "0                  0.0               -1                  1   \n",
       "1                  0.0               -1                  3   \n",
       "2                  0.0               -1                  3   \n",
       "3                  0.0               -4                  3   \n",
       "4                  0.0               -3                  1   \n",
       "\n",
       "       Mean_Sea_Level_PressurehPa  Mean_Wind_SpeedKm_h  CloudCover       Date  \\\n",
       "index                                                                           \n",
       "0                            1030                   16         8.0 2016-03-01   \n",
       "1                            1027                   10         8.0 2016-03-02   \n",
       "2                            1024                   10         8.0 2016-03-04   \n",
       "3                            1024                   10         6.0 2016-03-05   \n",
       "4                            1023                    5         6.0 2016-03-06   \n",
       "\n",
       "       season_weight  uv_index  \n",
       "index                           \n",
       "0                  1         0  \n",
       "1                  1         0  \n",
       "2                  1         0  \n",
       "3                  1         0  \n",
       "4                  1         0  "
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "real_feel.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# temperatures to fahreneit\n",
    "to_fahreneit = lambda c: c * 9/5 + 32\n",
    "\n",
    "real_feel['Mean_Dew_PointC'] = real_feel['Mean_Dew_PointC'].apply(to_fahreneit)\n",
    "real_feel['Mean_TemperatureC'] = real_feel['Mean_TemperatureC'].apply(to_fahreneit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# speed to mph\n",
    "to_mph = lambda kmh: 0.6214 * kmh\n",
    "\n",
    "real_feel['Mean_Wind_SpeedKm_h'] = real_feel['Mean_Wind_SpeedKm_h'].apply(to_mph)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# pa to mb\n",
    "to_mb = lambda pa: pa / 10\n",
    "\n",
    "real_feel['Mean_Sea_Level_PressurehPa'] = real_feel['Mean_Sea_Level_PressurehPa'].apply(to_mb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# rescale precipitation\n",
    "mms = MinMaxScaler(feature_range=(0, 5))\n",
    "real_feel['Precipitationmm'] = mms.fit_transform(real_feel[['Precipitationmm']].values).astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# compute RealFeel\n",
    "real_feel['real_feel'] =\\\n",
    "real_feel.apply(lambda x: realfeel(x['Mean_Wind_SpeedKm_h'], x['Mean_Sea_Level_PressurehPa'],\n",
    "                                   x['Mean_TemperatureC'], x['uv_index'],\n",
    "                                   x['Mean_Dew_PointC'], x['Precipitationmm']), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# add real feel to dataset\n",
    "all_data['real_feel'] = real_feel['real_feel']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ff80b8f8a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot real feel\n",
    "real_feel_monthly = real_feel.groupby([pd.Grouper(key='Date',freq='M')]).mean().reset_index()\n",
    "\n",
    "pd.Series(real_feel_monthly['real_feel'].values, index=real_feel_monthly['Date'])\\\n",
    ".plot(figsize=(20,6))\n",
    "pd.Series(real_feel_monthly['Mean_TemperatureC'].values, index=real_feel_monthly['Date'])\\\n",
    ".plot(figsize=(20,6), color='red')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Seasonality features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "def week_of_month(tgtdate):\n",
    "    tgtdate = pd.to_datetime(tgtdate)\n",
    "\n",
    "    days_this_month = calendar.mdays[tgtdate.month]\n",
    "    for i in range(1, days_this_month):\n",
    "        d = datetime.datetime(tgtdate.year, tgtdate.month, i)\n",
    "        if d.day - d.weekday() > 0:\n",
    "            startdate = d\n",
    "            break\n",
    "    return (tgtdate - startdate).days //7 + 1\n",
    "\n",
    "all_data['week_of_month'] = all_data.Date.apply(week_of_month)\n",
    "all_data['year'] = all_data.Date.dt.year\n",
    "all_data['quarter'] = all_data.Date.dt.quarter\n",
    "all_data['month'] = all_data.Date.dt.month\n",
    "all_data['day_of_month'] = all_data.Date.dt.day\n",
    "all_data['day_of_week'] = all_data.Date.dt.dayofweek\n",
    "all_data['day_of_year'] = all_data.Date.dt.dayofyear\n",
    "all_data['WeekOfYear'] = all_data.Date.dt.weekofyear\n",
    "all_data['days_in_month'] = all_data.Date.dt.daysinmonth"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## One-Hot encoding"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Add Hail single event"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "# all features which contain 'Hails'\n",
    "event_list = all_data.Events.unique()\n",
    "hail_list = [feature for feature in event_list if 'Hail' in feature.split('-')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create Hails feature and concatenate it to the dataset\n",
    "hail_feature = all_data.Events.isin(hail_list).rename('Events_Hail')\n",
    "all_data = pd.concat([all_data, hail_feature], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Perform One Hot Encoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_data_reg_temp = all_data['Region']\n",
    "all_data = pd.get_dummies(all_data,columns=['Region','AssortmentType', 'StoreType', 'Events'])\n",
    "all_data['Region'] = all_data_reg_temp"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Drop combined Events"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "events_to_drop = ['Events_'+feature for feature in event_list if '-' in feature]\n",
    "all_data = all_data.drop(events_to_drop, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## New Features Holidays"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "def num_of_features(regional_week_year, week, year, column_of_interest):\n",
    "    num_holidays = 0\n",
    "\n",
    "    if week == 1:\n",
    "        week_before = [52,year-1]\n",
    "        week_after = [2,year]\n",
    "    if week == 52:\n",
    "        week_after = [1,year+1]\n",
    "        week_before = [51,year]\n",
    "    if (week>1) & (week<52):\n",
    "        week_before = [week-1,year]\n",
    "        week_after = [week+1,year]\n",
    "    \n",
    "    before_df = regional_week_year[(regional_week_year['WeekOfYear']==week_before[0])&\n",
    "                                  (regional_week_year['year']==week_before[1])]\n",
    "    \n",
    "    after_df = regional_week_year[(regional_week_year['WeekOfYear']==week_after[0])&\n",
    "                                  (regional_week_year['year']==week_after[1])]\n",
    "    this_df = regional_week_year[(regional_week_year['WeekOfYear']==week)&\n",
    "                                 (regional_week_year['year']==year)]\n",
    "    \n",
    "    num_ago = before_df[before_df[column_of_interest]][column_of_interest].sum()\n",
    "    num_next = after_df[after_df[column_of_interest]][column_of_interest].sum()\n",
    "    num_this = this_df[this_df[column_of_interest]][column_of_interest].sum()\n",
    "    \n",
    "    return num_ago, num_next, num_this\n",
    "\n",
    "def create_new_features(before,current,after,column_of_interest):\n",
    "    all_data[before] = 0\n",
    "    all_data[current] = 0\n",
    "    all_data[after] = 0\n",
    "\n",
    "    region_list= [\"Region_\"+str(d) for d in range(0,11)]\n",
    "    for region in tqdm(region_list):\n",
    "        curr_region = all_data[all_data[region]==1]\n",
    "\n",
    "        #get all valid dates of that region\n",
    "        regional_week_year=curr_region[['WeekOfYear','year',column_of_interest]]\n",
    "\n",
    "        #get all store ids of that region\n",
    "        regional_stores=len(curr_region['StoreID'].unique())\n",
    "\n",
    "        week_year_list =regional_week_year[['WeekOfYear','year']].drop_duplicates().values.tolist()\n",
    "\n",
    "        for date in week_year_list:\n",
    "            num_ago,num_next,num_this = num_of_features(regional_week_year,\n",
    "                                                        date[0],\n",
    "                                                        date[1],\n",
    "                                                        column_of_interest)\n",
    "\n",
    "            all_data.at[((all_data['WeekOfYear']==date[0])&\n",
    "                         (all_data['year']==date[1])),before]=int(num_ago/regional_stores)\n",
    "            all_data.at[((all_data['WeekOfYear']==date[0])&\n",
    "                         (all_data['year']==date[1])),after]=int(num_next/regional_stores)\n",
    "            all_data.at[((all_data['WeekOfYear']==date[0])&\n",
    "                         (all_data['year']==date[1])),current]=int(num_this/regional_stores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 11/11 [01:03<00:00,  5.74s/it]\n",
      "100%|██████████| 11/11 [01:03<00:00,  5.75s/it]\n"
     ]
    }
   ],
   "source": [
    "create_new_features('HolidaysWeekBefore', 'HolidaysWeekCurrent', 'HolidaysWeekAfter','IsHoliday')\n",
    "create_new_features('PromoWeekBefore', 'PromoWeekCurrent', 'PromoWeekAfter', 'HasPromotions')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Split train and test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_final = all_data[:train.shape[0]].copy()\n",
    "train_final['NumberOfSales'] = train.NumberOfSales\n",
    "\n",
    "# extract the test examples (we don't have the class value for this)\n",
    "test_final = all_data[train.shape[0]:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Add differential feature (on train set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Activate Differential\n",
    "act_differential = False\n",
    "\n",
    "if act_differential == True:\n",
    "    march_2016  = datetime.date(2016,3,31)\n",
    "    monthly_sales['Differential'] = 0        \n",
    "\n",
    "    for storeID, store in tqdm(monthly_sales.groupby('StoreID')):\n",
    "        monthly_sales.loc[store.index,'one_step'] = store['NumberOfSales'].shift(1).fillna(0)\n",
    "        monthly_sales.loc[store.index,'two_steps'] = store['NumberOfSales'].shift(2).fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "if act_differential == True:\n",
    "    train_final['Differential'] = 0\n",
    "\n",
    "    #correct one step in april\n",
    "    april_mask = monthly_sales.Date.dt.month == 4\n",
    "    year_mask = monthly_sales.Date.dt.year == 2016\n",
    "    monthly_sales.loc[year_mask & april_mask,'one_step'] = 0\n",
    "\n",
    "    #compute differential\n",
    "    monthly_sales['Differential'] = monthly_sales['one_step'] - monthly_sales['two_steps']\n",
    "    pippo = monthly_sales[['StoreID','Date','Differential']]\n",
    "\n",
    "    #longest for ever - more than 10 mins - there must be a better way..\n",
    "    for index,row in tqdm(pippo.iterrows()):\n",
    "        differential = row.Differential\n",
    "        month = row.Date.month\n",
    "        year = row.Date.year\n",
    "        store = row.StoreID\n",
    "        month_mask = train_final.Date.dt.month == month\n",
    "        year_mask = train_final.Date.dt.year == year\n",
    "        store_mask = train_final.StoreID == store\n",
    "\n",
    "        train_final.loc[store_mask & year_mask & month_mask,'Differential'] = differential\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Save Train and Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "# train_final.to_csv(data_path_out+'train_final.csv')\n",
    "# test_final.to_csv(data_path_out+'test_final.csv')\n",
    "\n",
    "# Serialize dataframes for later use in implementations\n",
    "with open(data_path_out+'train_pp.obj', 'wb') as file:\n",
    "        pickle.dump(train_final, file)\n",
    "with open(data_path_out+'test_pp.obj', 'wb') as file:\n",
    "        pickle.dump(test_final, file)"
   ]
  },
  {
   "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}