XGBoost_Model.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# XGBoost"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load data"
   ]
  },
  {
   "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",
    "from pandas import Series, DataFrame\n",
    "import pickle\n",
    "from xgboost import plot_importance\n",
    "from xgboost.sklearn import XGBRegressor\n",
    "from sklearn.preprocessing import RobustScaler\n",
    "from sklearn.model_selection import KFold, cross_val_score, GridSearchCV\n",
    "from sklearn.metrics import r2_score\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 \"preprocessing\"\n",
    "with open(data_path_out + 'train_pp.obj', 'rb') as file:\n",
    "    train = pickle.load(file)\n",
    "with open(data_path_out + 'test_pp.obj', 'rb') as file:\n",
    "    test = pickle.load(file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(523021, 77)\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>StoreID</th>\n",
       "      <th>Date</th>\n",
       "      <th>IsHoliday</th>\n",
       "      <th>IsOpen</th>\n",
       "      <th>HasPromotions</th>\n",
       "      <th>NearestCompetitor</th>\n",
       "      <th>Region_AreaKM2</th>\n",
       "      <th>Region_GDP</th>\n",
       "      <th>Region_PopulationK</th>\n",
       "      <th>CloudCover</th>\n",
       "      <th>Max_Dew_PointC</th>\n",
       "      <th>Max_Humidity</th>\n",
       "      <th>Max_Sea_Level_PressurehPa</th>\n",
       "      <th>Max_TemperatureC</th>\n",
       "      <th>Max_VisibilityKm</th>\n",
       "      <th>Max_Wind_SpeedKm_h</th>\n",
       "      <th>Mean_Dew_PointC</th>\n",
       "      <th>Mean_Humidity</th>\n",
       "      <th>Mean_Sea_Level_PressurehPa</th>\n",
       "      <th>Mean_TemperatureC</th>\n",
       "      <th>Mean_VisibilityKm</th>\n",
       "      <th>Mean_Wind_SpeedKm_h</th>\n",
       "      <th>Min_Dew_PointC</th>\n",
       "      <th>Min_Humidity</th>\n",
       "      <th>Min_Sea_Level_PressurehPa</th>\n",
       "      <th>Min_TemperatureC</th>\n",
       "      <th>Min_VisibilitykM</th>\n",
       "      <th>Precipitationmm</th>\n",
       "      <th>WindDirDegrees</th>\n",
       "      <th>Hol_and_open</th>\n",
       "      <th>Region_PD</th>\n",
       "      <th>stores_per_region</th>\n",
       "      <th>stores_per_person</th>\n",
       "      <th>stores_per_KM2</th>\n",
       "      <th>GDP_per_capita</th>\n",
       "      <th>real_feel</th>\n",
       "      <th>week_of_month</th>\n",
       "      <th>year</th>\n",
       "      <th>quarter</th>\n",
       "      <th>month</th>\n",
       "      <th>day_of_month</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>day_of_year</th>\n",
       "      <th>WeekOfYear</th>\n",
       "      <th>days_in_month</th>\n",
       "      <th>Events_Hail</th>\n",
       "      <th>Region_0</th>\n",
       "      <th>Region_1</th>\n",
       "      <th>Region_10</th>\n",
       "      <th>Region_2</th>\n",
       "      <th>Region_3</th>\n",
       "      <th>Region_4</th>\n",
       "      <th>Region_5</th>\n",
       "      <th>Region_7</th>\n",
       "      <th>Region_9</th>\n",
       "      <th>Region_6</th>\n",
       "      <th>Region_8</th>\n",
       "      <th>AssortmentType_General</th>\n",
       "      <th>AssortmentType_With Non-Food Department</th>\n",
       "      <th>AssortmentType_With Fish Department</th>\n",
       "      <th>StoreType_Hyper Market</th>\n",
       "      <th>StoreType_Standard Market</th>\n",
       "      <th>StoreType_Super Market</th>\n",
       "      <th>StoreType_Shopping Center</th>\n",
       "      <th>Events_Fog</th>\n",
       "      <th>Events_Rain</th>\n",
       "      <th>Events_Snow</th>\n",
       "      <th>Events_Thunderstorm</th>\n",
       "      <th>Events_Normal</th>\n",
       "      <th>Region</th>\n",
       "      <th>HolidaysWeekBefore</th>\n",
       "      <th>HolidaysWeekCurrent</th>\n",
       "      <th>HolidaysWeekAfter</th>\n",
       "      <th>PromoWeekBefore</th>\n",
       "      <th>PromoWeekCurrent</th>\n",
       "      <th>PromoWeekAfter</th>\n",
       "      <th>NumberOfSales</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",
       "      <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",
       "      <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",
       "      <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",
       "      <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",
       "      <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",
       "      <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",
       "      <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",
       "      <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>1000</td>\n",
       "      <td>2016-03-01</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>326</td>\n",
       "      <td>9643</td>\n",
       "      <td>17130</td>\n",
       "      <td>2770</td>\n",
       "      <td>8.0</td>\n",
       "      <td>1</td>\n",
       "      <td>100</td>\n",
       "      <td>1032</td>\n",
       "      <td>2</td>\n",
       "      <td>19.0</td>\n",
       "      <td>21</td>\n",
       "      <td>-1</td>\n",
       "      <td>82</td>\n",
       "      <td>1030</td>\n",
       "      <td>1</td>\n",
       "      <td>11.0</td>\n",
       "      <td>16</td>\n",
       "      <td>-2</td>\n",
       "      <td>70</td>\n",
       "      <td>1029</td>\n",
       "      <td>1</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>23</td>\n",
       "      <td>False</td>\n",
       "      <td>0.287255</td>\n",
       "      <td>62</td>\n",
       "      <td>0.022383</td>\n",
       "      <td>0.00643</td>\n",
       "      <td>0.006184</td>\n",
       "      <td>32.788038</td>\n",
       "      <td>0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>61</td>\n",
       "      <td>9</td>\n",
       "      <td>31</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>5676</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1000</td>\n",
       "      <td>2016-03-02</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>326</td>\n",
       "      <td>9643</td>\n",
       "      <td>17130</td>\n",
       "      <td>2770</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0</td>\n",
       "      <td>87</td>\n",
       "      <td>1030</td>\n",
       "      <td>5</td>\n",
       "      <td>23.0</td>\n",
       "      <td>16</td>\n",
       "      <td>-1</td>\n",
       "      <td>73</td>\n",
       "      <td>1027</td>\n",
       "      <td>3</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10</td>\n",
       "      <td>-2</td>\n",
       "      <td>58</td>\n",
       "      <td>1025</td>\n",
       "      <td>1</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>56</td>\n",
       "      <td>False</td>\n",
       "      <td>0.287255</td>\n",
       "      <td>62</td>\n",
       "      <td>0.022383</td>\n",
       "      <td>0.00643</td>\n",
       "      <td>0.006184</td>\n",
       "      <td>36.601140</td>\n",
       "      <td>0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>62</td>\n",
       "      <td>9</td>\n",
       "      <td>31</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>8111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1000</td>\n",
       "      <td>2016-03-04</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>326</td>\n",
       "      <td>9643</td>\n",
       "      <td>17130</td>\n",
       "      <td>2770</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0</td>\n",
       "      <td>81</td>\n",
       "      <td>1026</td>\n",
       "      <td>4</td>\n",
       "      <td>31.0</td>\n",
       "      <td>23</td>\n",
       "      <td>-1</td>\n",
       "      <td>71</td>\n",
       "      <td>1024</td>\n",
       "      <td>3</td>\n",
       "      <td>11.0</td>\n",
       "      <td>10</td>\n",
       "      <td>-3</td>\n",
       "      <td>55</td>\n",
       "      <td>1023</td>\n",
       "      <td>2</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>22</td>\n",
       "      <td>False</td>\n",
       "      <td>0.287255</td>\n",
       "      <td>62</td>\n",
       "      <td>0.022383</td>\n",
       "      <td>0.00643</td>\n",
       "      <td>0.006184</td>\n",
       "      <td>36.602307</td>\n",
       "      <td>0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>64</td>\n",
       "      <td>9</td>\n",
       "      <td>31</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>8300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1000</td>\n",
       "      <td>2016-03-05</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>326</td>\n",
       "      <td>9643</td>\n",
       "      <td>17130</td>\n",
       "      <td>2770</td>\n",
       "      <td>6.0</td>\n",
       "      <td>-3</td>\n",
       "      <td>80</td>\n",
       "      <td>1027</td>\n",
       "      <td>8</td>\n",
       "      <td>31.0</td>\n",
       "      <td>19</td>\n",
       "      <td>-4</td>\n",
       "      <td>56</td>\n",
       "      <td>1024</td>\n",
       "      <td>3</td>\n",
       "      <td>15.0</td>\n",
       "      <td>10</td>\n",
       "      <td>-6</td>\n",
       "      <td>25</td>\n",
       "      <td>1022</td>\n",
       "      <td>-1</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>108</td>\n",
       "      <td>False</td>\n",
       "      <td>0.287255</td>\n",
       "      <td>62</td>\n",
       "      <td>0.022383</td>\n",
       "      <td>0.00643</td>\n",
       "      <td>0.006184</td>\n",
       "      <td>36.602307</td>\n",
       "      <td>0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>65</td>\n",
       "      <td>9</td>\n",
       "      <td>31</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>7154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1000</td>\n",
       "      <td>2016-03-06</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>326</td>\n",
       "      <td>9643</td>\n",
       "      <td>17130</td>\n",
       "      <td>2770</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "      <td>93</td>\n",
       "      <td>1025</td>\n",
       "      <td>7</td>\n",
       "      <td>31.0</td>\n",
       "      <td>16</td>\n",
       "      <td>-3</td>\n",
       "      <td>75</td>\n",
       "      <td>1023</td>\n",
       "      <td>1</td>\n",
       "      <td>12.0</td>\n",
       "      <td>5</td>\n",
       "      <td>-6</td>\n",
       "      <td>48</td>\n",
       "      <td>1022</td>\n",
       "      <td>-5</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>46</td>\n",
       "      <td>False</td>\n",
       "      <td>0.287255</td>\n",
       "      <td>62</td>\n",
       "      <td>0.022383</td>\n",
       "      <td>0.00643</td>\n",
       "      <td>0.006184</td>\n",
       "      <td>33.236222</td>\n",
       "      <td>0</td>\n",
       "      <td>2016</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "      <td>66</td>\n",
       "      <td>9</td>\n",
       "      <td>31</td>\n",
       "      <td>False</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       StoreID       Date  IsHoliday  IsOpen  HasPromotions  \\\n",
       "index                                                         \n",
       "0         1000 2016-03-01      False    True          False   \n",
       "1         1000 2016-03-02      False    True          False   \n",
       "2         1000 2016-03-04      False    True          False   \n",
       "3         1000 2016-03-05      False    True          False   \n",
       "4         1000 2016-03-06      False   False          False   \n",
       "\n",
       "       NearestCompetitor  Region_AreaKM2  Region_GDP  Region_PopulationK  \\\n",
       "index                                                                      \n",
       "0                    326            9643       17130                2770   \n",
       "1                    326            9643       17130                2770   \n",
       "2                    326            9643       17130                2770   \n",
       "3                    326            9643       17130                2770   \n",
       "4                    326            9643       17130                2770   \n",
       "\n",
       "       CloudCover  Max_Dew_PointC  Max_Humidity  Max_Sea_Level_PressurehPa  \\\n",
       "index                                                                        \n",
       "0             8.0               1           100                       1032   \n",
       "1             8.0               0            87                       1030   \n",
       "2             8.0               0            81                       1026   \n",
       "3             6.0              -3            80                       1027   \n",
       "4             6.0               0            93                       1025   \n",
       "\n",
       "       Max_TemperatureC  Max_VisibilityKm  Max_Wind_SpeedKm_h  \\\n",
       "index                                                           \n",
       "0                     2              19.0                  21   \n",
       "1                     5              23.0                  16   \n",
       "2                     4              31.0                  23   \n",
       "3                     8              31.0                  19   \n",
       "4                     7              31.0                  16   \n",
       "\n",
       "       Mean_Dew_PointC  Mean_Humidity  Mean_Sea_Level_PressurehPa  \\\n",
       "index                                                               \n",
       "0                   -1             82                        1030   \n",
       "1                   -1             73                        1027   \n",
       "2                   -1             71                        1024   \n",
       "3                   -4             56                        1024   \n",
       "4                   -3             75                        1023   \n",
       "\n",
       "       Mean_TemperatureC  Mean_VisibilityKm  Mean_Wind_SpeedKm_h  \\\n",
       "index                                                              \n",
       "0                      1               11.0                   16   \n",
       "1                      3               13.0                   10   \n",
       "2                      3               11.0                   10   \n",
       "3                      3               15.0                   10   \n",
       "4                      1               12.0                    5   \n",
       "\n",
       "       Min_Dew_PointC  Min_Humidity  Min_Sea_Level_PressurehPa  \\\n",
       "index                                                            \n",
       "0                  -2            70                       1029   \n",
       "1                  -2            58                       1025   \n",
       "2                  -3            55                       1023   \n",
       "3                  -6            25                       1022   \n",
       "4                  -6            48                       1022   \n",
       "\n",
       "       Min_TemperatureC  Min_VisibilitykM  Precipitationmm  WindDirDegrees  \\\n",
       "index                                                                        \n",
       "0                     1               6.0              0.0              23   \n",
       "1                     1              10.0              0.0              56   \n",
       "2                     2               8.0              0.0              22   \n",
       "3                    -1              10.0              0.0             108   \n",
       "4                    -5               5.0              0.0              46   \n",
       "\n",
       "       Hol_and_open  Region_PD  stores_per_region  stores_per_person  \\\n",
       "index                                                                  \n",
       "0             False   0.287255                 62           0.022383   \n",
       "1             False   0.287255                 62           0.022383   \n",
       "2             False   0.287255                 62           0.022383   \n",
       "3             False   0.287255                 62           0.022383   \n",
       "4             False   0.287255                 62           0.022383   \n",
       "\n",
       "       stores_per_KM2  GDP_per_capita  real_feel  week_of_month  year  \\\n",
       "index                                                                   \n",
       "0             0.00643        0.006184  32.788038              0  2016   \n",
       "1             0.00643        0.006184  36.601140              0  2016   \n",
       "2             0.00643        0.006184  36.602307              0  2016   \n",
       "3             0.00643        0.006184  36.602307              0  2016   \n",
       "4             0.00643        0.006184  33.236222              0  2016   \n",
       "\n",
       "       quarter  month  day_of_month  day_of_week  day_of_year  WeekOfYear  \\\n",
       "index                                                                       \n",
       "0            1      3             1            1           61           9   \n",
       "1            1      3             2            2           62           9   \n",
       "2            1      3             4            4           64           9   \n",
       "3            1      3             5            5           65           9   \n",
       "4            1      3             6            6           66           9   \n",
       "\n",
       "       days_in_month  Events_Hail  Region_0  Region_1  Region_10  Region_2  \\\n",
       "index                                                                        \n",
       "0                 31        False         0         0          0         0   \n",
       "1                 31        False         0         0          0         0   \n",
       "2                 31        False         0         0          0         0   \n",
       "3                 31        False         0         0          0         0   \n",
       "4                 31        False         0         0          0         0   \n",
       "\n",
       "       Region_3  Region_4  Region_5  Region_7  Region_9  Region_6  Region_8  \\\n",
       "index                                                                         \n",
       "0             0         0         0         1         0         0         0   \n",
       "1             0         0         0         1         0         0         0   \n",
       "2             0         0         0         1         0         0         0   \n",
       "3             0         0         0         1         0         0         0   \n",
       "4             0         0         0         1         0         0         0   \n",
       "\n",
       "       AssortmentType_General  AssortmentType_With Non-Food Department  \\\n",
       "index                                                                    \n",
       "0                           1                                        0   \n",
       "1                           1                                        0   \n",
       "2                           1                                        0   \n",
       "3                           1                                        0   \n",
       "4                           1                                        0   \n",
       "\n",
       "       AssortmentType_With Fish Department  StoreType_Hyper Market  \\\n",
       "index                                                                \n",
       "0                                        0                       1   \n",
       "1                                        0                       1   \n",
       "2                                        0                       1   \n",
       "3                                        0                       1   \n",
       "4                                        0                       1   \n",
       "\n",
       "       StoreType_Standard Market  StoreType_Super Market  \\\n",
       "index                                                      \n",
       "0                              0                       0   \n",
       "1                              0                       0   \n",
       "2                              0                       0   \n",
       "3                              0                       0   \n",
       "4                              0                       0   \n",
       "\n",
       "       StoreType_Shopping Center  Events_Fog  Events_Rain  Events_Snow  \\\n",
       "index                                                                    \n",
       "0                              0           0            0            0   \n",
       "1                              0           0            0            1   \n",
       "2                              0           0            1            0   \n",
       "3                              0           0            0            0   \n",
       "4                              0           0            0            0   \n",
       "\n",
       "       Events_Thunderstorm  Events_Normal Region  HolidaysWeekBefore  \\\n",
       "index                                                                  \n",
       "0                        0              0      7                   0   \n",
       "1                        0              0      7                   0   \n",
       "2                        0              0      7                   0   \n",
       "3                        0              1      7                   0   \n",
       "4                        0              1      7                   0   \n",
       "\n",
       "       HolidaysWeekCurrent  HolidaysWeekAfter  PromoWeekBefore  \\\n",
       "index                                                            \n",
       "0                        0                  0                0   \n",
       "1                        0                  0                0   \n",
       "2                        0                  0                0   \n",
       "3                        0                  0                0   \n",
       "4                        0                  0                0   \n",
       "\n",
       "       PromoWeekCurrent  PromoWeekAfter  NumberOfSales  \n",
       "index                                                   \n",
       "0                     0               5           5676  \n",
       "1                     0               5           8111  \n",
       "2                     0               5           8300  \n",
       "3                     0               5           7154  \n",
       "4                     0               5              0  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(train.shape)\n",
    "train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Utilities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# categorized feature lists to ease the preprocessing\n",
    "filter_startswith = lambda l,s: [x for x in l if x.startswith(s)]\n",
    "\n",
    "# regional features\n",
    "r_regions = ['Region_'+str(i) for i in range(11)]\n",
    "r_assortments = filter_startswith(train.columns, 'AssortmentType_')\n",
    "r_storetype = filter_startswith(train.columns, 'StoreType_')\n",
    "r_general = ['NearestCompetitor', 'Region_AreaKM2', 'Region_GDP', 'Region_PopulationK', 'Region_PD',\n",
    "             'stores_per_region', 'stores_per_person', 'stores_per_KM2', 'GDP_per_capita']\n",
    "regional_features =  r_regions + r_assortments + r_storetype + r_general\n",
    "\n",
    "# meteo features\n",
    "m_events = filter_startswith(train.columns, 'Events_')\n",
    "m_general = ['CloudCover', 'Max_Dew_PointC', 'Max_Humidity', 'Max_Sea_Level_PressurehPa',\n",
    "             'Max_TemperatureC', 'Max_VisibilityKm', 'Max_Wind_SpeedKm_h',\n",
    "             'Mean_Dew_PointC', 'Mean_Humidity', 'Mean_Sea_Level_PressurehPa',\n",
    "             'Mean_TemperatureC', 'Mean_VisibilityKm', 'Mean_Wind_SpeedKm_h',\n",
    "             'Min_Dew_PointC', 'Min_Humidity', 'Min_Sea_Level_PressurehPa',\n",
    "             'Min_TemperatureC', 'Min_VisibilitykM', 'Precipitationmm',\n",
    "             'WindDirDegrees', 'real_feel']\n",
    "meteo_features =  m_events + m_general\n",
    "\n",
    "# store features\n",
    "store_features = ['IsHoliday', 'IsOpen', 'HasPromotions', 'Hol_and_open']\n",
    "\n",
    "# seasonal features\n",
    "seasonal_features = ['week_of_month', 'year', 'quarter', 'month', 'day_of_month',\n",
    "                     'day_of_week', 'day_of_year', 'WeekOfYear', 'days_in_month',\n",
    "                     'HolidaysWeekBefore', 'HolidaysWeekCurrent','HolidaysWeekAfter',\n",
    "                     'PromoWeekBefore', 'PromoWeekCurrent', 'PromoWeekAfter']\n",
    "\n",
    "# x and y specific\n",
    "x_features = ['StoreID']\n",
    "y_features = ['NumberOfSales']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# split train set in x and y\n",
    "def split_dataset(dataset):\n",
    "    x = dataset.drop('NumberOfSales', axis=1)\n",
    "    y = dataset.NumberOfSales\n",
    "    \n",
    "    return (x, y)\n",
    "\n",
    "# split train set by months\n",
    "def split_dataset_bymonth(test_year, test_months, train_set):\n",
    "    test_mask = (train_set.year == test_year) & train_set.month.isin(test_months)\n",
    "\n",
    "    # define the train set\n",
    "    train_dataset = train_set[~test_mask]\n",
    "    x_train, y_train = split_dataset(train_dataset)\n",
    "    \n",
    "    # define the test set\n",
    "    test_dataset = train_set[test_mask]\n",
    "    x_test, y_test = split_dataset(test_dataset)\n",
    "    \n",
    "    return (x_train, y_train, x_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set to 0 the NumberOfSales of closed stores\n",
    "def nos_correction(dataset, results):\n",
    "    dataset['results'] = results\n",
    "    \n",
    "    mask_close = (dataset.IsOpen == False)\n",
    "    \n",
    "    dataset.loc[mask_close, 'results'] = 0\n",
    "    \n",
    "    return dataset['results']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# reorder results monthly wrt regions, stores pairing predictions with actual values\n",
    "def get_regional_results(results, x, y=0):\n",
    "    # create the dataframe and round predicted values to int\n",
    "    df_results = x[['StoreID', 'month'] + r_regions].copy()\n",
    "    df_results['predicted'] = results.astype(int)\n",
    "    \n",
    "    # create 'actual' values column if passed\n",
    "    if isinstance(y, int):\n",
    "        agg_dict = {'predicted': 'sum'}\n",
    "    else:\n",
    "        agg_dict = {'predicted': 'sum', 'actual': 'sum'}\n",
    "        df_results['actual'] = y.astype(int)\n",
    "    \n",
    "    # aggregate prediction monthly\n",
    "    df_results = df_results.groupby(['StoreID', 'month']+r_regions)\\\n",
    "    .agg(agg_dict).reset_index()\n",
    "    \n",
    "    # reverse region one-hot-encoding\n",
    "    df_results['region'] = df_results[r_regions].idxmax(axis=1)\\\n",
    "    .map(lambda x: str(x).replace('Region_', ''))\n",
    "    df_results = df_results.drop(r_regions, axis=1)\n",
    "    \n",
    "    df_results = df_results.sort_values(by=['region', 'StoreID', 'month'])\\\n",
    "    .reset_index().drop('index', axis=1)    \n",
    "    \n",
    "    return df_results\n",
    "\n",
    "# format format suitable for the submission of the project\n",
    "def get_submission_format(regional_results):\n",
    "    submit_results = regional_results.copy()\n",
    "    \n",
    "    # drop region column and reorder wrt storeid and month\n",
    "    submit_results = submit_results.drop('region', axis=1)\\\n",
    "    .sort_values(by=['StoreID', 'month'])\\\n",
    "    .reset_index().drop('index', axis=1)\n",
    "    \n",
    "    return submit_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Lanzi error\n",
    "\n",
    "# Regional error\n",
    "# DataFrame: 'StoreID' 'month' 'predicted' 'actual' 'region'\n",
    "def lanzi_region_error(region):\n",
    "    num = region[['actual', 'predicted']].apply(lambda x: abs(x[0] - x[1]), axis=1).sum()\n",
    "    den = region['actual'].sum()\n",
    "    \n",
    "    return num/den\n",
    "    \n",
    "# Total error across all regions\n",
    "# region_errors = [0.3, 0.5, ... ]\n",
    "def lanzi_total_error(region_errors):\n",
    "    return sum(region_errors)/len(region_errors)\n",
    "\n",
    "def lanzi_error(regional_results):\n",
    "    region_errors = []\n",
    "    for regionid, region in regional_results.groupby('region'):\n",
    "        region_errors.append(lanzi_region_error(region))\n",
    "    \n",
    "    return (region_errors, lanzi_total_error(region_errors))\n",
    "\n",
    "# Print results of Lanzi error\n",
    "def lanzi_print(r_error, t_error):\n",
    "    print('LANZI ERROR')\n",
    "    print('-' * 20) \n",
    "    print('{:<10}{:>10}'.format('REGION', 'ERROR'))\n",
    "    \n",
    "    for regionid, error in enumerate(r_error):\n",
    "        print('{:<10}{:>10.4f}'.format(regionid, error))\n",
    "    print('-' * 20)\n",
    "    print('{:<10}{:>10.4f}'.format('TOTAL', t_error))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Robust Scaler\n",
    "rs = RobustScaler()\n",
    "\n",
    "features_to_scale = r_general + m_general\n",
    "\n",
    "train[features_to_scale] = rs.fit_transform(train[features_to_scale])\n",
    "test[features_to_scale] = rs.fit_transform(test[features_to_scale])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# drop useless features\n",
    "m_general.remove('real_feel')\n",
    "\n",
    "features_to_drop = ['Region', 'Date'] + m_general\n",
    "\n",
    "train = train.drop(labels=features_to_drop, axis=1)\n",
    "test = test.drop(labels=features_to_drop, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train and predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# settings\n",
    "xgboost_params = {    \n",
    "    'max_depth': 12,\n",
    "    'learning_rate': 0.2,\n",
    "    'n_jobs': 4,\n",
    "    'silent': False\n",
    "}\n",
    "\n",
    "kfold_params = {\n",
    "    'n_splits': 2,\n",
    "    'shuffle': True,\n",
    "    'random_state': 42\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  ................................................................\n",
      "[CV] ................................................. , total= 1.1min\n",
      "[CV]  ................................................................\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:  1.1min remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] ................................................. , total= 1.1min\n",
      "Cross-validation score: 0.9632 with +/- 0.0001\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:  2.2min finished\n"
     ]
    }
   ],
   "source": [
    "# Cross validation score\n",
    "x_train, y_train = split_dataset(train)\n",
    "\n",
    "model = XGBRegressor(**xgboost_params)\n",
    "kfold = KFold(**kfold_params)\n",
    "results = cross_val_score(model, x_train, y_train, scoring='r2', cv=kfold, verbose=2)\n",
    "\n",
    "print('Cross-validation score: {:.4f} with +/- {:.4f}'\\\n",
    "      .format(results.mean(),results.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x864 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Feature importance plot\n",
    "model.fit(x_train, y_train)\n",
    "fig, ax = plt.subplots(1, 1, figsize=(20, 12))\n",
    "plot_importance(model, ax=ax)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# results = {}\n",
    "# for storeid in all_train.StoreID.unique():\n",
    "#     train = all_train[all_train.StoreID == storeid]\n",
    "#     y_train = train.NumberOfSales\n",
    "#     x_train = train.drop('NumberOfSales',axis = 1)\n",
    "    \n",
    "#     model = XGBRegressor(silent=False)\n",
    "#     kfold = KFold(n_splits=10, shuffle = True, random_state=7)\n",
    "#     results[storeid] = cross_val_score(model, x_train, y_train, scoring='r2', cv=kfold)\n",
    "#     print('Cross-validation for {} -> score: {:.4f} with +/- {:.4f}'\\\n",
    "#           .format(storeid,results[storeid].mean(),results[storeid].std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_results = pd.DataFrame.from_dict(results).T\n",
    "# df_results_mean = df_results.mean(axis=1)\n",
    "# df_results_mean[df_results_mean < 0.8]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Lanzi Error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split train set and predict the test set\n",
    "x_train, y_train, x_test, y_test =\\\n",
    "    split_dataset_bymonth(2017, [3, 4], train)\n",
    "\n",
    "model = XGBRegressor(**xgboost_params)\n",
    "model.fit(x_train, y_train)\n",
    "\n",
    "results = model.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "results = nos_correction(x_test, results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "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>StoreID</th>\n",
       "      <th>month</th>\n",
       "      <th>predicted</th>\n",
       "      <th>actual</th>\n",
       "      <th>region</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1001</td>\n",
       "      <td>3</td>\n",
       "      <td>75421</td>\n",
       "      <td>70708</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1001</td>\n",
       "      <td>4</td>\n",
       "      <td>73826</td>\n",
       "      <td>67093</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1009</td>\n",
       "      <td>3</td>\n",
       "      <td>136429</td>\n",
       "      <td>130890</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1009</td>\n",
       "      <td>4</td>\n",
       "      <td>115192</td>\n",
       "      <td>116915</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1019</td>\n",
       "      <td>3</td>\n",
       "      <td>114703</td>\n",
       "      <td>113892</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   StoreID  month  predicted  actual region\n",
       "0     1001      3      75421   70708      0\n",
       "1     1001      4      73826   67093      0\n",
       "2     1009      3     136429  130890      0\n",
       "3     1009      4     115192  116915      0\n",
       "4     1019      3     114703  113892      0"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get monthly results ordered by region, store and month\n",
    "regional_results = get_regional_results(results, x_test, y_test)\n",
    "regional_results.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LANZI ERROR\n",
      "--------------------\n",
      "REGION         ERROR\n",
      "0             0.0349\n",
      "1             0.0378\n",
      "2             0.0391\n",
      "3             0.0348\n",
      "4             0.0360\n",
      "5             0.0334\n",
      "6             0.0337\n",
      "7             0.0314\n",
      "8             0.0344\n",
      "9             0.0287\n",
      "10            0.0339\n",
      "--------------------\n",
      "TOTAL         0.0344\n"
     ]
    }
   ],
   "source": [
    "# get lanzi error and print results\n",
    "r_error, t_error = lanzi_error(regional_results)\n",
    "lanzi_print(r_error, t_error)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Score submit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split train set and predict the submission test set\n",
    "x_train, y_train = split_dataset(train)\n",
    "\n",
    "model = XGBRegressor(**xgboost_params)\n",
    "model.fit(x_train, y_train)\n",
    "\n",
    "submit_results = model.predict(test)\n",
    "results = nos_correction(x_test, results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1498, 3)\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>StoreID</th>\n",
       "      <th>month</th>\n",
       "      <th>predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1000</td>\n",
       "      <td>3</td>\n",
       "      <td>169917</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1000</td>\n",
       "      <td>4</td>\n",
       "      <td>162199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1001</td>\n",
       "      <td>3</td>\n",
       "      <td>94528</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1001</td>\n",
       "      <td>4</td>\n",
       "      <td>81944</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1002</td>\n",
       "      <td>3</td>\n",
       "      <td>122412</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   StoreID  month  predicted\n",
       "0     1000      3     169917\n",
       "1     1000      4     162199\n",
       "2     1001      3      94528\n",
       "3     1001      4      81944\n",
       "4     1002      3     122412"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get monthly results and format it for submission\n",
    "submit_regional_results = get_regional_results(submit_results, test)\n",
    "df_submission = get_submission_format(submit_regional_results)\n",
    "\n",
    "print(df_submission.shape)\n",
    "df_submission.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# write submission csv\n",
    "df_submission.to_csv(data_path_out+'submission.csv', index=False)"
   ]
  }
 ],
 "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}