4 - Spatiotemporal Regression.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Spatiotemporal Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "7fDtK1p7evZT"
   },
   "outputs": [],
   "source": [
    "%%capture\n",
    "!pip3 install pygeohash tensorflow==2.0.0-alpha0 tpot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "wl-nNX21_b98"
   },
   "outputs": [],
   "source": [
    "import math\n",
    "import datetime\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import pygeohash as gh\n",
    "from fbprophet import Prophet\n",
    "from fbprophet.diagnostics import cross_validation, performance_metrics\n",
    "from fbprophet.plot import plot_cross_validation_metric\n",
    "\n",
    "import scipy.fftpack\n",
    "\n",
    "from sklearn.linear_model import LinearRegression, Ridge, Lasso, ElasticNet\n",
    "from sklearn.linear_model import BayesianRidge, ARDRegression\n",
    "from sklearn.pipeline import Pipeline, FeatureUnion\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.preprocessing import PolynomialFeatures, StandardScaler, Normalizer\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "from sklearn.tree import export_graphviz\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "\n",
    "from xgboost import XGBRegressor\n",
    "from xgboost import plot_tree\n",
    "import xgboost\n",
    "\n",
    "from tpot import TPOTRegressor\n",
    "\n",
    "import tensorflow.keras as tfk\n",
    "\n",
    "from sklearn.externals.six import StringIO  \n",
    "from IPython.display import Image  \n",
    "import pydotplus\n",
    "\n",
    "import _pickle as cPickle\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import animation\n",
    "sns.set(style=\"whitegrid\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "rMMIolKw_b-E"
   },
   "source": [
    "## The data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "fbQwPH_s_b-F"
   },
   "source": [
    "### Load the preprocessed data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 11920,
     "status": "ok",
     "timestamp": 1559366995405,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "INzRLvHtfXMX",
    "outputId": "bcfe3b6f-2613-49f0-9e81-b0af385f2909"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Drive already mounted at /content/gdrive; to attempt to forcibly remount, call drive.mount(\"/content/gdrive\", force_remount=True).\n"
     ]
    }
   ],
   "source": [
    "from google.colab import drive\n",
    "drive.mount('/content/gdrive')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 2577,
     "status": "ok",
     "timestamp": 1559367029578,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "G3hWSfLpcRto",
    "outputId": "7ec1ba62-e177-4a02-879a-7eeae8316729"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/content/gdrive/My Drive/Colab Notebooks/aiforsea\n"
     ]
    }
   ],
   "source": [
    "cd '/content/gdrive/My Drive/Colab Notebooks/aiforsea'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 204
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 11960,
     "status": "ok",
     "timestamp": 1559367039706,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "yce--IXleiEN",
    "outputId": "2289da1d-4338-4d03-a03f-2dea65372228"
   },
   "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>geohash6</th>\n",
       "      <th>day</th>\n",
       "      <th>demand</th>\n",
       "      <th>in_raw</th>\n",
       "      <th>weekday</th>\n",
       "      <th>hour</th>\n",
       "      <th>minutes</th>\n",
       "      <th>ds</th>\n",
       "      <th>latitude</th>\n",
       "      <th>longitude</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>qp02yc</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1900-01-01 00:00:00</td>\n",
       "      <td>-5.484924</td>\n",
       "      <td>90.653687</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>qp02z1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1900-01-01 00:00:00</td>\n",
       "      <td>-5.484924</td>\n",
       "      <td>90.664673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>qp02z3</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1900-01-01 00:00:00</td>\n",
       "      <td>-5.484924</td>\n",
       "      <td>90.675659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>qp02z9</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1900-01-01 00:00:00</td>\n",
       "      <td>-5.484924</td>\n",
       "      <td>90.686646</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>qp02zc</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1900-01-01 00:00:00</td>\n",
       "      <td>-5.484924</td>\n",
       "      <td>90.697632</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  geohash6  day  demand  in_raw  weekday  hour  minutes                   ds  \\\n",
       "0   qp02yc    1     0.0   False        1     0        0  1900-01-01 00:00:00   \n",
       "1   qp02z1    1     0.0   False        1     0        0  1900-01-01 00:00:00   \n",
       "2   qp02z3    1     0.0   False        1     0        0  1900-01-01 00:00:00   \n",
       "3   qp02z9    1     0.0   False        1     0        0  1900-01-01 00:00:00   \n",
       "4   qp02zc    1     0.0   False        1     0        0  1900-01-01 00:00:00   \n",
       "\n",
       "   latitude  longitude  \n",
       "0 -5.484924  90.653687  \n",
       "1 -5.484924  90.664673  \n",
       "2 -5.484924  90.675659  \n",
       "3 -5.484924  90.686646  \n",
       "4 -5.484924  90.697632  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('full_data.csv')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "_jC-0pEg_b-m"
   },
   "source": [
    "### Circular embedding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 7656,
     "status": "ok",
     "timestamp": 1559367041844,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "FrITJ1HL_b-o",
    "outputId": "7b3ae2a1-f715-42fa-d33d-92d1049d2594"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 2.1 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df['weekday_sin'] = np.sin(2*np.pi*df['weekday'].values/7)\n",
    "df['weekday_cos'] = np.cos(2*np.pi*df['weekday'].values/7)\n",
    "\n",
    "df['hour_sin']    = np.sin(2*np.pi*df['hour'].values/24)\n",
    "df['hour_cos']    = np.cos(2*np.pi*df['hour'].values/24)\n",
    "\n",
    "df['minutes_sin'] = np.sin(2*np.pi*df['minutes'].values/60)\n",
    "df['minutes_cos'] = np.cos(2*np.pi*df['minutes'].values/60)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 394
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 7129,
     "status": "ok",
     "timestamp": 1559367041846,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "b22sq-pr_b-r",
    "outputId": "b89f7dc6-01e0-4edd-e8dc-75a6bd26c0fc"
   },
   "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>geohash6</th>\n",
       "      <th>day</th>\n",
       "      <th>demand</th>\n",
       "      <th>in_raw</th>\n",
       "      <th>weekday</th>\n",
       "      <th>hour</th>\n",
       "      <th>minutes</th>\n",
       "      <th>ds</th>\n",
       "      <th>latitude</th>\n",
       "      <th>longitude</th>\n",
       "      <th>weekday_sin</th>\n",
       "      <th>weekday_cos</th>\n",
       "      <th>hour_sin</th>\n",
       "      <th>hour_cos</th>\n",
       "      <th>minutes_sin</th>\n",
       "      <th>minutes_cos</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>qp02yc</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1900-01-01 00:00:00</td>\n",
       "      <td>-5.484924</td>\n",
       "      <td>90.653687</td>\n",
       "      <td>0.781831</td>\n",
       "      <td>0.62349</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>qp02z1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1900-01-01 00:00:00</td>\n",
       "      <td>-5.484924</td>\n",
       "      <td>90.664673</td>\n",
       "      <td>0.781831</td>\n",
       "      <td>0.62349</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>qp02z3</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1900-01-01 00:00:00</td>\n",
       "      <td>-5.484924</td>\n",
       "      <td>90.675659</td>\n",
       "      <td>0.781831</td>\n",
       "      <td>0.62349</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>qp02z9</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1900-01-01 00:00:00</td>\n",
       "      <td>-5.484924</td>\n",
       "      <td>90.686646</td>\n",
       "      <td>0.781831</td>\n",
       "      <td>0.62349</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>qp02zc</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1900-01-01 00:00:00</td>\n",
       "      <td>-5.484924</td>\n",
       "      <td>90.697632</td>\n",
       "      <td>0.781831</td>\n",
       "      <td>0.62349</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  geohash6  day  demand  in_raw  weekday  hour  minutes                   ds  \\\n",
       "0   qp02yc    1     0.0   False        1     0        0  1900-01-01 00:00:00   \n",
       "1   qp02z1    1     0.0   False        1     0        0  1900-01-01 00:00:00   \n",
       "2   qp02z3    1     0.0   False        1     0        0  1900-01-01 00:00:00   \n",
       "3   qp02z9    1     0.0   False        1     0        0  1900-01-01 00:00:00   \n",
       "4   qp02zc    1     0.0   False        1     0        0  1900-01-01 00:00:00   \n",
       "\n",
       "   latitude  longitude  weekday_sin  weekday_cos  hour_sin  hour_cos  \\\n",
       "0 -5.484924  90.653687     0.781831      0.62349       0.0       1.0   \n",
       "1 -5.484924  90.664673     0.781831      0.62349       0.0       1.0   \n",
       "2 -5.484924  90.675659     0.781831      0.62349       0.0       1.0   \n",
       "3 -5.484924  90.686646     0.781831      0.62349       0.0       1.0   \n",
       "4 -5.484924  90.697632     0.781831      0.62349       0.0       1.0   \n",
       "\n",
       "   minutes_sin  minutes_cos  \n",
       "0          0.0          1.0  \n",
       "1          0.0          1.0  \n",
       "2          0.0          1.0  \n",
       "3          0.0          1.0  \n",
       "4          0.0          1.0  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "hHgxz9Hy_b_Q"
   },
   "source": [
    "### Collect information for each location"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 6189,
     "status": "ok",
     "timestamp": 1559367042321,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "bXFfPhsD_b_R",
    "outputId": "4815faf0-bb10-4fbe-b0d7-76dfe3c29415",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1329 ['qp02yc' 'qp02z1' 'qp02z3' ... 'qp0djw' 'qp0djy' 'qp0dnn']\n"
     ]
    }
   ],
   "source": [
    "locations = df['geohash6'].unique()\n",
    "print(len(locations), locations)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Ynu8KePA_b_W"
   },
   "outputs": [],
   "source": [
    "df_locinfo = pd.DataFrame()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 235
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 8350,
     "status": "ok",
     "timestamp": 1559367045021,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "nKHDW829_b_Z",
    "outputId": "aebd46a4-fa04-49fd-8aaf-81e11afadfd0"
   },
   "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>median</th>\n",
       "      <th>mean</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>geohash6</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>qp03yb</th>\n",
       "      <td>0.777774</td>\n",
       "      <td>0.694499</td>\n",
       "      <td>5856</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>qp03xw</th>\n",
       "      <td>0.972542</td>\n",
       "      <td>0.745806</td>\n",
       "      <td>5856</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>qp03xx</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.751653</td>\n",
       "      <td>5856</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>qp09e5</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.759626</td>\n",
       "      <td>5856</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>qp09d9</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.867371</td>\n",
       "      <td>5856</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            median      mean  count\n",
       "geohash6                           \n",
       "qp03yb    0.777774  0.694499   5856\n",
       "qp03xw    0.972542  0.745806   5856\n",
       "qp03xx    1.000000  0.751653   5856\n",
       "qp09e5    1.000000  0.759626   5856\n",
       "qp09d9    1.000000  0.867371   5856"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_locinfo['median'] = df.groupby(['geohash6'])['demand'].median()\n",
    "df_locinfo['mean'] = df.groupby(['geohash6'])['demand'].mean()\n",
    "df_locinfo['count'] = df.groupby(['geohash6'])['demand'].count()\n",
    "\n",
    "df_locinfo = df_locinfo.sort_values(by=['mean', 'median'])\n",
    "df_locinfo.tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ez372ca__b_d"
   },
   "source": [
    "### Detrend the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "TPpzob_3_b_i"
   },
   "outputs": [],
   "source": [
    "# median_demand = df['demand'].median()\n",
    "# mean_demand = df['demand'].mean()\n",
    "# \n",
    "# print(\"Median demand:\", median_demand)\n",
    "# print(\"Mean demand:\", mean_demand)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "idC-ztMl_b_q"
   },
   "source": [
    "#### Detrend by mean demand"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "YHgJOS7c_b_r"
   },
   "outputs": [],
   "source": [
    "# df['demand_detrended'] = df.apply(lambda x : x.demand - df_locmedian[x.geohash6], axis=1)\n",
    "# df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "gJkKzB_E_b_v"
   },
   "source": [
    "### Train-test split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 4295,
     "status": "ok",
     "timestamp": 1559367045030,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "YvmHnirQda5p",
    "outputId": "70ef7c26-5ae9-432e-de78-f684e1057a3a"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7782624"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# df = df[df['in_raw']]\n",
    "len(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "WLJ_r1-C_b_x"
   },
   "outputs": [],
   "source": [
    "X_data = df[['latitude', 'longitude',\n",
    "             'day', 'hour', 'minutes',\n",
    "             'weekday_sin', 'weekday_cos',\n",
    "             'hour_sin', 'hour_cos',\n",
    "             'minutes_sin', 'minutes_cos']]\n",
    "# y_data = df['demand_detrended']\n",
    "y_data = df['demand']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 204
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 3267,
     "status": "ok",
     "timestamp": 1559367045034,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "-IFjIDuV_b_0",
    "outputId": "1f19513b-5337-4dc3-9229-85ab45130086"
   },
   "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>latitude</th>\n",
       "      <th>longitude</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "      <th>minutes</th>\n",
       "      <th>weekday_sin</th>\n",
       "      <th>weekday_cos</th>\n",
       "      <th>hour_sin</th>\n",
       "      <th>hour_cos</th>\n",
       "      <th>minutes_sin</th>\n",
       "      <th>minutes_cos</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7782619</th>\n",
       "      <td>-5.237732</td>\n",
       "      <td>90.928345</td>\n",
       "      <td>61</td>\n",
       "      <td>23</td>\n",
       "      <td>45</td>\n",
       "      <td>-0.974928</td>\n",
       "      <td>-0.222521</td>\n",
       "      <td>-0.258819</td>\n",
       "      <td>0.965926</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.836970e-16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7782620</th>\n",
       "      <td>-5.237732</td>\n",
       "      <td>90.939331</td>\n",
       "      <td>61</td>\n",
       "      <td>23</td>\n",
       "      <td>45</td>\n",
       "      <td>-0.974928</td>\n",
       "      <td>-0.222521</td>\n",
       "      <td>-0.258819</td>\n",
       "      <td>0.965926</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.836970e-16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7782621</th>\n",
       "      <td>-5.237732</td>\n",
       "      <td>90.950317</td>\n",
       "      <td>61</td>\n",
       "      <td>23</td>\n",
       "      <td>45</td>\n",
       "      <td>-0.974928</td>\n",
       "      <td>-0.222521</td>\n",
       "      <td>-0.258819</td>\n",
       "      <td>0.965926</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.836970e-16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7782622</th>\n",
       "      <td>-5.237732</td>\n",
       "      <td>90.961304</td>\n",
       "      <td>61</td>\n",
       "      <td>23</td>\n",
       "      <td>45</td>\n",
       "      <td>-0.974928</td>\n",
       "      <td>-0.222521</td>\n",
       "      <td>-0.258819</td>\n",
       "      <td>0.965926</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.836970e-16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7782623</th>\n",
       "      <td>-5.237732</td>\n",
       "      <td>90.972290</td>\n",
       "      <td>61</td>\n",
       "      <td>23</td>\n",
       "      <td>45</td>\n",
       "      <td>-0.974928</td>\n",
       "      <td>-0.222521</td>\n",
       "      <td>-0.258819</td>\n",
       "      <td>0.965926</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.836970e-16</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         latitude  longitude  day  hour  minutes  weekday_sin  weekday_cos  \\\n",
       "7782619 -5.237732  90.928345   61    23       45    -0.974928    -0.222521   \n",
       "7782620 -5.237732  90.939331   61    23       45    -0.974928    -0.222521   \n",
       "7782621 -5.237732  90.950317   61    23       45    -0.974928    -0.222521   \n",
       "7782622 -5.237732  90.961304   61    23       45    -0.974928    -0.222521   \n",
       "7782623 -5.237732  90.972290   61    23       45    -0.974928    -0.222521   \n",
       "\n",
       "         hour_sin  hour_cos  minutes_sin   minutes_cos  \n",
       "7782619 -0.258819  0.965926         -1.0 -1.836970e-16  \n",
       "7782620 -0.258819  0.965926         -1.0 -1.836970e-16  \n",
       "7782621 -0.258819  0.965926         -1.0 -1.836970e-16  \n",
       "7782622 -0.258819  0.965926         -1.0 -1.836970e-16  \n",
       "7782623 -0.258819  0.965926         -1.0 -1.836970e-16  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_data.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 119
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 3484,
     "status": "ok",
     "timestamp": 1559367045600,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "_m3XQDIe_b_6",
    "outputId": "392afc3a-3431-408f-e993-7248fd579c98"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7782619    0.0\n",
       "7782620    0.0\n",
       "7782621    0.0\n",
       "7782622    0.0\n",
       "7782623    0.0\n",
       "Name: demand, dtype: float64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_data.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "rQiTOifzeiFV"
   },
   "outputs": [],
   "source": [
    "test_index = (X_data['day'] == 61) & ((X_data['hour'] == 23) | ((X_data['hour'] == 22) & (X_data['minutes'] == 45)))\n",
    "train_index = (~test_index) & (X_data['day'] >= 61-14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Z_r9GOAk_b_-"
   },
   "outputs": [],
   "source": [
    "X_train, X_test = X_data[train_index], X_data[test_index]\n",
    "y_train, y_test = y_data[train_index], y_data[test_index]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "HWlLxh2reiFf"
   },
   "outputs": [],
   "source": [
    "X_train = X_train.drop(['hour', 'minutes'], axis=1)\n",
    "X_test  = X_test.drop(['hour', 'minutes'], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "3ASENZv5_cAB"
   },
   "source": [
    "## Polynomial Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "RetKvAzr_cAD"
   },
   "outputs": [],
   "source": [
    "class ColumnExtractor(object):\n",
    "    def __init__(self, columns=None):\n",
    "        self.columns = columns\n",
    "\n",
    "    def fit(self, X, y=None):\n",
    "        return self\n",
    "\n",
    "    def transform(self, X):\n",
    "        X_cols = X[self.columns]\n",
    "        return X_cols"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "DySzkUjfeiFs"
   },
   "source": [
    "### Train polynomial regressors"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "m2hWJw4idhRt"
   },
   "source": [
    "Loss with full data:\n",
    "- degree 1 polynomial regression loss: 0.048594363494371254\n",
    "- degree 2 polynomial regression loss: 0.04912314854368942\n",
    "- degree 3 polynomial regression loss: 0.0495815584173599\n",
    "\n",
    "Loss with partial data:\n",
    "- degree 1 polynomial regression loss: 0.0665519070143781\n",
    "- degree 2 polynomial regression loss: 0.06692959565100215\n",
    "- degree 3 polynomial regression loss: 0.06690783480405857"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 68
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 6047,
     "status": "ok",
     "timestamp": 1559367667019,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "UgKwPT7__cAG",
    "outputId": "573320b4-5e7f-4a12-a961-871cb0ef15cb"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "degree 1 polynomial regression loss: 0.048594363494371254\n",
      "degree 2 polynomial regression loss: 0.04912314854368942\n",
      "degree 3 polynomial regression loss: 0.0495815584173599\n"
     ]
    }
   ],
   "source": [
    "poly_regressors = [None]\n",
    "for deg in range(1, 3+1):\n",
    "  poly_reg = Pipeline([\n",
    "    ('features', FeatureUnion([\n",
    "        ('spatial', Pipeline([\n",
    "            ('extract', ColumnExtractor(columns=['latitude', 'longitude'])),\n",
    "            ('poly', PolynomialFeatures(degree=deg))\n",
    "        ])),\n",
    "        ('temporal', ColumnExtractor(columns=['day',\n",
    "                                             'weekday_sin', 'weekday_cos',\n",
    "                                             'hour_sin', 'hour_cos',\n",
    "                                             'minutes_sin', 'minutes_cos']))\n",
    "    ])),\n",
    "    ('std_scaler', StandardScaler()),\n",
    "    ('lr', Ridge())\n",
    "  ])\n",
    "  poly_reg = poly_reg.fit(X_train, y_train)\n",
    "  poly_regressors.append(poly_reg)\n",
    "  \n",
    "  y_pred = poly_reg.predict(X_test)\n",
    "  print(\"degree\", deg, \"polynomial regression loss:\", math.sqrt(mean_squared_error(y_test, y_pred)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('models/geotemporal/poly_regressors.pkl', 'wb') as fid:\n",
    "    cPickle.dump(poly_regressors, fid)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "RW0P48BdeiFx"
   },
   "source": [
    "### Features of 1st degree polynomial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "jSf1kbLAeiFy",
    "outputId": "2d1efa48-6d76-4c33-b741-30fc4a83b4dc"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['1',\n",
       " 'lat',\n",
       " 'long',\n",
       " 'day',\n",
       " 'weekday_sin',\n",
       " 'weekday_cos',\n",
       " 'hour_sin',\n",
       " 'hour_cos',\n",
       " 'minutes_sin',\n",
       " 'minutes_cos']"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "poly_features = poly_regressors[1]['features'].transformer_list[0][1]['poly'].get_feature_names(['lat', 'long'])\n",
    "poly_features += ['day', 'weekday_sin', 'weekday_cos', 'hour_sin', 'hour_cos', 'minutes_sin', 'minutes_cos']\n",
    "poly_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ASqnpBs0eiF3",
    "outputId": "7449f6b7-7992-4b3e-d905-ddaae4cf8b5f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x16c9fa121d0>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_coefs = pd.DataFrame(np.expand_dims(poly_regressors[1]['lr'].coef_, axis=0),\n",
    "                        columns=poly_features).T\n",
    "df_coefs.sort_values(by=0).plot.barh(figsize=(4, 8))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "qZrvavkWeiF7"
   },
   "source": [
    "### Features of 3rd degree polynomial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 306
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1358,
     "status": "ok",
     "timestamp": 1559367846475,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "RCGvUc8d_cAL",
    "outputId": "fe7a0997-8388-4bf0-8587-1566294c845b"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['1',\n",
       " 'lat',\n",
       " 'long',\n",
       " 'lat^2',\n",
       " 'lat long',\n",
       " 'long^2',\n",
       " 'lat^3',\n",
       " 'lat^2 long',\n",
       " 'lat long^2',\n",
       " 'long^3',\n",
       " 'day',\n",
       " 'weekday_sin',\n",
       " 'weekday_cos',\n",
       " 'hour_sin',\n",
       " 'hour_cos',\n",
       " 'minutes_sin',\n",
       " 'minutes_cos']"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "poly_features = poly_regressors[3]['features'].transformer_list[0][1]['poly'].get_feature_names(['lat', 'long'])\n",
    "poly_features += ['day', 'weekday_sin', 'weekday_cos', 'hour_sin', 'hour_cos', 'minutes_sin', 'minutes_cos']\n",
    "poly_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 508
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 1312,
     "status": "ok",
     "timestamp": 1559367847141,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "tt3_VCkY_cAT",
    "outputId": "23151af4-fe64-4d90-dd4e-2a24b308d666"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x16c9fa0b550>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAHUCAYAAABBO1+tAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAgAElEQVR4nO3deXxOd97/8deVjUYUI9VS6Vhqq71KLFOxTaqN2KKi2tCpm5qboX6dEEtJM6habircVBdpo7G0TUW4SVpSnaLoQlASd1ETIS0mRDSL5Pz+8Mh1SyWRkOTKda7385+Zazvn871O8/Y951znfCyGYRiIiJiUk60LEBGpSAo5ETE1hZyImJpCTkRMzcXWBZhFfn4+mZmZuLq6YrFYbF2OiEMxDIPc3Fxq1KiBk1PhuZtCrpxkZmaSnJxs6zJEHFrz5s2pWbNmoecUcuXE1dUVuPklu7m52biau3P06FHatGlj6zIqnSOO22xjzsnJITk52fp3eCuFXDkp2EV1c3OjWrVqNq7m7tlz7ffCEcdtxjEXdahIJx5ExNQUciJiago5ETE1HZOrBLm5uaSkpJCVlWXrUkrk4uLC8ePHcXZ2pnbt2nh6et52Ol7E3ijkKkFKSgo1a9akUaNGVfo3dJmZmbi7u5Obm0taWhopKSk88sgjti6rwrVs1drWJVS6Tp062bqEUsnJzcPN1fmelqGQqwRZWVlVPuAKWCwW3NzcePjhh0lKSrJ1OZWihnt1/F+NsXUZUoTYJYPueRnaF6kk9hBwt9JuqpiFaf9L3r9/P0FBQbYuQ0RszLQhV5Xl5ObZfLmxsbE888wz+Pr68tFHH1VIPSJVgemPyZ0+fZrZs2eTnp6Ou7s7M2fOpF27doSEhODh4cGxY8dIS0tjwoQJBAQEkJGRwdSpUzl79ixeXl5cuHCBFStW0LBhw3Kryc3VuUKOAZX2+EVaWhpLly4lOjoaNzc3RowYgbe3N/Xr1y/3mkRszfQhFxwczLhx4/D19eXQoUNMnjyZuLg4AC5cuEBUVBTJycmMGjWKgIAAVq5cSePGjVm1ahVHjhwhMDCwTOs7evTobc+5uLiQmZlpfVyjRo17G1QJbl1PcRISEnjiiSdwdXXFMAz69OlDbGws48aNK/T5nJwcvvvuuwqrtaqwlzONjupe/xs0dchlZmaSkpKCr68vAB06dKBWrVqcOnUKgB49emCxWGjevDnp6ekA7Nmzh8WLFwPQtm1bmjdvXqZ1tmnT5rZrAo8fP16hwXar0qznypUr1K9f3/rehx9+mMTExNs+7+bmRvv27SumUJFSKs0/QtnZ2UVOMMDkx+SK6tFjGAZ5eTePXRWE0a1nPp2dnYv8nJnk5+cXGrNhGHZ39lektEwdch4eHjRs2JD4+HgADh06xMWLF2nWrFmxn+nWrRuxsbEAJCUlcfLkSdMFwEMPPcSvv/5qffzrr79Sr149G1YkUnFMHXIAixYtIjIyEn9/f8LCwggPDy/xfm8TJkzg7Nmz+Pv7s3z5cjw9PalevXolVlzxunfvzr59+7h8+TK//fYb8fHx9OzZ09ZliVQI0x6T8/b2xtvbG4DIyMjbXl+wYEGhxwW/7t+5cycvvvginTp1IjU1lRdeeIE6deqUa205uXnl8kvuopZbmktgHnzwQaZMmcKoUaPIzc1l2LBhtGvXrlQnLUTsjWlD7m41adKEOXPmkJ+fj5OTE2FhYeX+6/97vRavPJbr7++Pv79/hdQhUpUo5H6nbdu2REdH27oMqUSZ17MqZGYt9648LtA3/TE5kTs5cfyYrUuodPby+8fy2OtRyImIqSnkKom9/fYuPz/f1iWIlAuFXCWoXr06ly5dsougMwyDnJwczp07V2lXaYhUJJ14qAQNGzYkJSWl0A9wq6KcnBzc3NxwcXGhVq1aeHp62rokkXumkKsErq6uNG7c2NZl3NF3332na1XFdLS7KiKmppATEVNTyImIqSnkRMTUFHIiYmoKOXF4Vbm5dEU1PXIk+gmJOLyq3FxaNw64d5rJiYipOXzInTlzhq1bt9q6DBGpIJUacqXpap+QkMDatWtvez48PJzw8PByr+mdd95hwYIFZGRkADc7fE2ePNl6U8lt27aV+zpFpPJUuZnc0aNHuXbtWqWs68qVKyQmJjJ69Gg+/fRTANasWUODBg2IjY0lIiKCN954g4sXL1ZKPSJS/mx24uHAgQMsXbqUrKwsrl69yvTp02nUqBEbNmwAoEGDBgQEBBT52YSEBJYtW0Z+fj5eXl6EhYXh6elJnz59GDhwIF9//TW//fYbb775Jm3atCE5OZmQkBDy8vJ44okn+Oqrr/j888/59NNPeeaZZxg6dCijRo1i9OjRdOnSxXqdad26dalduzYXL17UxeoidspmIbdu3Trmzp1L06ZN2bdvH/Pnzyc2NpYRI0YAFBtwly5dYvbs2axfv56GDRvy7rvvEhYWxvLlywGoXbs2n3zyCZGRkbz99tuEh4cTEhLC5MmT8fHxISIigry8PAzDYOPGjXzwwQfUrVuXpk2b8tVXX+Hj42Nd1//8z/+Qk5PDo48+WupxFdfg1l7Yyx1jy1NpmhfbUkVtE0fZ1jYLuUWLFpGQkMCOHTs4fPhwqTtFJSYm0q5dOxo2bAhAYGAga9assb7+5JNPAtCsWTPi4+NJT0/n3Llz1vAKCAjgww8/xGKxEBcXZ/1cQUgW2L59O/Pnz+fdd9/FxaX0X1ObNm2sTavtzXfffVfl/+AdUUVsE7Nt6+zs7GInGDY7Jjdy5EgSExNp06YN48ePL/Xnfn/HWsMwuHHjhvVxQcAUNIR2dnYu880qIyMjefPNN3nvvfdo2bJlmT4rIlWLTUIuPT2dM2fOMHnyZHr27MnOnTvJy7v5y25nZ+dCofV77du35/Dhw6SkpACwceNGa3/VotSsWRMvLy92794NQGxsbIm1ffHFF0RERLB+/XpatGhR1qGJSBVjk93V2rVrM2zYMPz8/HBxcaFr165kZWVx/fp1OnfuzLRp0/D09Czy5yaenp6EhYUxceJEcnNzadCgAfPmzStxfQsXLmTGjBksW7aMFi1aUL169WLfu3z5crKzswvNLufOnUvbtm3vfsAiYjMWwx4aD9yjFStWMHz4cOrVq0d8fDyxsbHl/pu7gmMCOiZnnxztsi6zbeuS/v4c4trVBg0a8NJLL+Hi4sL9999/x5mfOJaq3Fy6PJorOzqHCLmhQ4cydOhQW5chVdSJ48eq7KxGAXfvqtwVDyIi5UkhJyKmppATEVNTyImIqSnkRMTUFHIiYmoKORExNYWciJiaQk5ETE0hJyKmppATEVNTyInDa9mqdbkvMyc3r9yXKXfHIS7QFylJDffq5X6rpap6VxNH5PAzucuXLxMVFWXrMkSkgpgu5O7UwPr3zas3bNjAkiVLOHPmDAB5eXnMnj2bAQMG4O/vf8fbpYtI1Wa6kLuTW5tX5+XlsXXrVmbMmMFHH30EwJYtW7h27Rpbt27lgw8+YO7cuZXW7FpEyp9pj8mVpnm1h4cHjz/+OP7+/rz99tu88sorDBkyBH9/fwB++eUXXF1dcXV1teVQROQemDbkStO8etSoUUyZMgU3Nzf69u1LTEwMI0eOxMXFhZkzZxITE8O4cePK1LNBzaXtT0XdFbiqf5dVvb7yYtqQK03z6g8//ND6/6dNm1botXnz5vH3v/+doKAgHn/8cf70pz+Var1qZCMFqvJ3abZtXSWbS1e0u21effToUetJiDp16vDkk0+SlJRUQVWKSEUzZcjdS/Pqw4cPs2jRIvLz87l27Rpff/01jz/+eGWVLiLlzJS7q7Vr16Z79+531bx6xIgRJCUl4e/vj5OTE88//zwdO3a0wShEpDw4RHPpyqDm0vbN0a54MNu2Lunvz5S7qyIiBUy5uypSFpnXs8p95pWTq873VYVmcuLwThw/Vu7LVMBVHQo5ETE1hZyImJpCTkRMTSEnIqamkBMRU1PIiYipKeRExNQUciJiago5ETE1hZyImJpCTkRMTSEnDq9lq9Zl/kxObl4FVCIVQXchEYdXw716me8nV9XvFyf/x+FncpcvXyYqKsrWZYhIBTFdyO3fv7/I25oXSEhIYO3atdbHGzZsYMmSJdbmNXl5ecyZM4cBAwbg5+dHREREBVcsIhXJdCF3J0ePHuXatWvAzUDbunUrM2bM4KOPPgIgOjqa9PR0tmzZwscff8ymTZs4dqz87zcmIpXDtMfkDhw4wNKlS8nKyuLq1atMnz6dRo0asWHDBgAaNGiAh4cHjz/+OP7+/rz99tu88sorNGvWjPbt2+Pk5IS7uzteXl6cP3+e1q3LfnBaRGzPtCG3bt065s6dS9OmTdm3bx/z588nNjaWESNGABAQEMCoUaOYMmUKbm5u9O3bl5iYGEaOHGldxvfff09iYiILFy4s9XqLa3BrLxylq/qt7rahi71/V/Zef2mZNuQWLVpEQkICO3bs4PDhw2RmZt72ng8//ND6/6dNm1botYMHDzJlyhQWL15MrVq1Sr1edetyHPb8XZltWxd06yqKaY/JjRw5ksTERNq0acP48ePL9Nn4+HgmT57MkiVL6NGjRwVVKCKVwZQzufT0dFJTU4mKisLNzY3FixeTl3fzx5vOzs5kZ2cX+9nExERCQ0N5//33admyZWWVLCIVxJQhV7t2bbp3746fnx8uLi507dqVrKwsrl+/TufOnZk2bRqenp5F/tRk1apV5OXlFdp9nTRpEn379q3MIYhIOTFdyHl7e+Pt7Q3A9OnTrc+HhoYC0LlzZ3bt2lXs51etWlWh9YlI5TJdyImU1d00l1bzaPth2hMPIqV1N82lFXD2QyEnIqamkBMRU1PIiYipKeRExNQUciJiago5ETE1hZyImJpCTkRMTSEnIqamkBMRU1PIiYipKeTE4RU0l1bDaHNSyInDK2gurYvuzUkhJyKm5tAhl5OTw5o1a2xdhohUILsJuf379xd5u/J7sX37dlatWsXevXutz61YsQI/Pz/8/PzK1IpQRKomuwm5irB+/XrmzZvHunXrANi7dy9ff/01n332GZs3b+bYsWN8/vnnNq5SRO6F3d3+/PTp08yePZv09HTc3d2ZOXMm7dq1IyQkBA8PD44dO0ZaWhoTJkwgICCAjIwMpk6dytmzZ/Hy8uLChQusWLGCK1eucN999/HMM8/w7rvvcu7cOR544AFCQkJwc3MDoGnTpqSmptp4xCJyL+wu5IKDgxk3bhy+vr4cOnSIyZMnExcXB8CFCxeIiooiOTmZUaNGERAQwMqVK2ncuDGrVq3iyJEjBAYGArBu3ToCAgIACAgIICoqiuDgYOt6zpw5w/bt21m/fn2Z6iuuwa29cJSu6re6tcmyI43fUcZqVyGXmZlJSkoKvr6+AHTo0IFatWpx6tQpAHr06IHFYqF58+akp6cDsGfPHhYvXgxA27Ztad68OQBvvPGGdbnPP/98ofWcPHmSl19+malTp9KoUaMy1dimTRuqVat2V+OzNbN1Vb8bjjJ+s23r7OzsYicYdnVMzjCMIp8raBxdEC4Wi8X6urOzc5GfK853333Hiy++yKuvvsqQIUPusWIRsTW7CjkPDw8aNmxIfHw8AIcOHeLixYs0a9as2M9069aN2NhYAJKSkjh58mShELzV+fPnmTBhAosXL8bPz6/8ByAilc6udlcBFi1aRGhoKOHh4bi6uhIeHm49UVCUCRMmMH36dPz9/XnkkUfw9PSkevXqRb73vffeIzs7mwULFlifGzFiBM8991y5j0NEKofFKMu+nB2KiYmhYcOGdOrUidTUVF544QW++OILnJzKdxJbcExAx+Tsk/+rMWVuMG3PzLatS/r7s7uZXFk1adKEOXPmkJ+fj5OTE2FhYeUecGLfMq9nEbtkEDm5ebp+1YRMH3Jt27YlOjra1mVIFXbi+DE6deqkgDMpTWlExNQUciJiago5ETE1hZyImJpCTkRMTSEnIqamkBMRU1PIiYipKeRExNQUciJiago5ETE1hZw4vPr169u6BKlACjlxeA0aNLB1CVKBFHIiYmoKORExNZuH3P79+wkKCrJpDW+99RY7d+60aQ0iUjFMf9PM0pg8ebKtSxCRClIlQu7y5cuMHTuWs2fP0rhxY5YvX05sbCxr167FYrHQunVrXnvtNWrUqEGLFi1ISkoCIDo6mgMHDrBgwQL69OlDu3btOH78OFFRUdStW/e29eTm5jJjxgxOnjwJwMiRIxk+fDghISF06dKFLl26MHHiRJo1a8bx48epW7cub731FrVr1y71WNRc2v506tTJIcftKGOuEiGXmprK6tWrefjhhxk+fDjr169n3bp1bNq0iTp16vD666+zYsUKpk2bVuJyevbsybJly4p9/YcffuDKlSts3ryZtLQ0lixZwvDhwwu958SJE8yfP5/HHnuMv/3tb8TGxpZpd1qNbOyTo43bbNu6yjeXbtmyJV5eXjg5OdG0aVMyMjLo3bs3derUASAwMJBvvvnmjstp3759ia83a9aM06dPM2bMGHbs2MHUqVNve0/dunV57LHHrO+/cuXKXYxIRKqKKhFyLi7/N6G0WCzcf//9hV43DIMbN24UegwUeg644wyqTp06bNu2jRdeeIHTp08zZMgQrl69WuwyLBYLJu/YKGJ6VSLkirJr1y7S09MB2LRpE97e3sDNoDp58iSGYbBr164yLXPnzp0EBwfTq1cvZs2ahbu7O+fPny/32kWk6qgSx+R+z8PDg5dffpmgoCByc3Np3bo1r7/+OgCvvvoq48ePx9PTk06dOvHvf/+71Mvt2bMn8fHx+Pn5Ua1aNQYOHEiLFi0qahgiUgVYDO2PlYuSOnjbC7MdjJbimW1bl/T3VyVncvciKyuLwMDAIl+bNGkSffv2reSKpKpLTU3V9asmZrqQq169OjExMbYuQ+zI+fPnFXImVmVPPIiIlAeFnIiYmkJORExNIScipqaQExFTU8iJiKkp5ETE1BRyImJqCjkRMTWFnIiYmkJOHJ6aS5ubQk4cnq5bNTeFnIiYms1CrjQ3q+zTpw8pKSkVWseRI0eYOXNmha5DRGzHdLdaKqu2bdvStm1bW5chIhWk1DM5f39/fvrpJ+DmLcjnzJkD3GzzN27cONasWcOQIUMYOHAgCxcutDaA2bx5M0OGDGHQoEHMmDGD7OzsQsv9/vvv8fX15eeffyY9PZ2xY8fi7+/PK6+8Yn3vtWvXmDRpEoGBgfTu3ZsZM2ZgGAbBwcFs2rTJuqygoCAOHz5c7BhiY2MZNGgQQ4cOZdKkSWRnZ7N//35ry8GgoCAWLlxIYGAgf/7zn9m9e3dpvx4RqaJKHXI+Pj7s27cPgOTkZL7//nsA/vnPf9KrVy+OHj3KJ598Yu1pumXLFk6ePMmmTZvYsGEDMTEx1K1bl/fee8+6zBMnTjBz5kxWr17NH//4R5YvX85jjz1GbGwszz//PBcvXgTgyy+/pFWrVmzcuJG4uDgOHjzIsWPHCAgIsN4g89y5c1y+fLnEtoTLli3j/fffJzo6mocffphTp07d9p7c3Fw2btzI9OnTeeutt0r79YhIFVXq3VUfHx8iIiLo2rUrjz76KKdOneLSpUt89dVXNGvWjMTERIYOHQrcvAV5gwYNyMjI4Oeff7Y2cM7NzbX2NAUYM2YM/fv3p0mTJgAcOHCAJUuWANC5c2e8vLwAGDBgAImJiURERHDq1CnS09O5fv063t7evPbaa6SkpBATE8OgQYNKHEPv3r157rnn6NevH0899RStWrVi//79hd7z5JNPAjd7rhZ0CyuL4hrc2gtH6ap+q06dOjnkuB1lzKUOuY4dOxISEsLevXvp0qULdevWZceOHdy4cYOaNWsyevRo/vKXvwBw9epVnJ2d+eSTT3j66aeZNWsWAJmZmeTl5VmXuXjxYqZOncqzzz5Ly5Ytb+tz6uzsDEBkZCRxcXEMHz6c7t27k5ycjGEYWCwWBg8ezLZt29i+fXuhWWJRZs2axYkTJ9i9ezfBwcFMnDiRhx56qNB7CppgWCyW0n41haiRjX1ytHGbbVsXNLIpSql3V11cXGjXrh2RkZF06dKFrl27snr1anx8fOjatSsxMTFkZmZy48YNJkyYQFxcHN7e3nz++edcunQJwzAIDQ3lgw8+sC6zW7duvPrqq8yaNYv8/Hy6detm3f1MTEzk7NmzAOzZs4fAwEAGDhxIdnY2J06cID8/H4ChQ4eyYcMG6tevz4MPPlhs/Tdu3MDX15c6derw8ssvM2jQII4fP17a4YuInSrT2VUfHx8OHjxI06ZNeeCBB7h06RK9evWiY8eOnDhxguHDh5OXl8eTTz7JkCFDsFgsTJw4kdGjR5Ofn0+rVq0YN25coWUOHjyY6OhoIiMjmTRpEiEhIfj5+dGkSRPr7uro0aMJDQ1lzZo1eHh40LFjR+tPS+rXr0/9+vUZMmRIyQN1cWHSpEm89NJLVKtWjbp167JgwQL+93//tyxfgYjYGbvuu2oYBr/88gtBQUFs3boVNzc3m9WivqtiT8y2rU3bdzUuLo7Q0FBCQ0OtAffmm2+yd+/e297bpk0b5s2bV9klioiN2XXI9e/fn/79+xd6btq0aTaqRuyVmkubm65dFYd3/vx5W5cgFUghJyKmppATEVNTyImIqSnkRMTUFHIiYmoKORExNYWciJiaQk5ETE0hJyKmppATEVNTyImIqSnkxOHVr1/f1iVIBVLIicPTHUjMTSEnIqbmcCEXEhJCdHS0rcsQkUricCEnIo7Fru8MXBqGYbBgwQK+/PJL6tWrR15eHl26dGHp0qXs27ePK1euUK9ePZYuXUpCQgLffPONtfdreHg41apVu635jojYD9OHXFxcHD/++CNbt24lIyODgQMHkpeXx6lTp9iwYQNOTk5MnTqVLVu2EBgYyNKlS7l27RoeHh5s3bqVDz/8sEzrU3Np+6Pm0uZm+pA7cOAAvr6+uLq68oc//IGePXvi7OzMtGnT+Pjjjzl9+jSHDh3ikUceoUaNGvj4+PD555/j5eWFl5dXib1ci6JuXfbJ0cZttm1dLs2l7ZXFYuHWrosuLi6kp6czZswY8vPzeeqpp+jXr5/1PQEBAWzdupXY2FiGDh1qq7JFpJyYPuS6devG9u3bycnJ4cqVK/zzn//EYrHQpUsXnnvuORo1asSXX35JXl4eAE888QQXLlxg//799OvXz8bVi8i9Mv3uar9+/Thy5AgDBgzA09OTpk2bkpWVxYkTJ/D39wdu7mKmpKRYP/PnP/+Z9PR0mzarFpHyYfqQA5gyZQpTpkwp9Nxf//rX295nGAa5ubkcPHiQGTNmVFZ5IlKBTL+7Wha//vorPXr0oH379rRu3drW5YhIOXCImVxp1atXj4MHD9q6DKlkqampun7VxDSTE4d3/vx5W5cgFUghJyKmppATEVNTyImIqSnkRMTUFHIiYmoKORExNYWciJiaQk5ETE0hJyKmppATEVNTyInDU3Npc1PIicPTxfnmppATEVNTyImIqSnkRMTUKizkBg0adFefy8jIYMKECeVcjYg4qgoLuZiYmLv63JUrVzh+/Hg5VyMijspi3NqUtJT279/P6tWrcXV1JSUlhT59+uDu7s4XX3wBwJo1a+jRowdJSUmEh4eTlpbGzz//zLlz53j22Wf561//SnR0NAcOHGDBggUABAUFMXHiRNauXcvXX3+Nj48PK1euZPPmzXzwwQfk5+fTunVr5syZg5OTEzNmzODkyZMAjBw5kuHDhxdbb3p6OjNnzuTUqVO4ubkREhJCt27dSEhIYNmyZeTn5+Pl5UVYWBienp68+eab7NmzBycnJ/r168fEiRPv+J2U1NxWqrZOnTo5TDd5syuyubtxF7755hujY8eORmpqqnH9+nWjQ4cOxvr16w3DMIyQkBAjIiLCaN68uWEYhrF8+XJj2LBhRnZ2tnHx4kWjQ4cOxpUrV4xPP/3UmDZtmnWZL7zwgvHNN98Y//rXv4zevXsbhmEYycnJxnPPPWdkZWUZhmEYixcvNlauXGns37/fGDt2rGEYhnHhwgUjODi4xHpDQ0ONBQsWGIZhGCdOnDCGDx9uXLx40fjTn/5k/Otf/zIMwzDeeecd429/+5uRkpJiPPPMM4ZhGMb169eNyZMnW9dfkqysLOPbb78t1Xurqm+//dbWJUglMdu2Lunv764b2TRv3tz6I8o6derQrVs34OZvjq5evVrovd7e3ri5uVG3bl1q165NRkZGqdaxf/9+fv75Z+ssLTc3l8cee4znnnuO06dPM2bMGHr27MnUqVNLXM7BgwdZvHgxAC1atGDjxo0kJCTQrl07GjZsCEBgYCBr1qzhwQcfpFq1aowYMYLevXvz97///fZ/GUTEbtx1yLm6uhZ67OzsXOx7bw0Ji8WCYRjW/y2Qm5t72+fy8vJ4+umnmTVrFgCZmZnk5eVx//33s23bNvbs2cPu3bsZMmQI27Zt4/777y9y/S4uLlgsFuvjn376ifz8/ELvMQyDGzdu4OLiwscff8yBAwf46quvGDFiBJGRkTRu3LiEb0NEqiqb/YSkTp06/PTTTxiGwb/+9S+SkpKAm4F048YN4OYM8PPPP+fSpUsYhkFoaCgffPABO3fuJDg4mF69ejFr1izc3d1L7Lj0xBNPsG3bNuBmwI0dO5b27dtz+PBhUlJSANi4cSPe3t78+OOPvPDCC3Tu3Jlp06bRtGlTTp8+XcHfhohUFJv1Xe3evTuffvop/fv3p3HjxnTq1AmAunXr0qBBA4KCgoiMjGTixImMHj2a/Px8WrVqxbhx43ByciI+Ph4/Pz+qVavGwIEDadGiRbHrmjRpErNmzWLgwIG4uLiwcOFCPD09CQsLY+LEieTm5tKgQQPmzZtHvXr16NChAwMGDOC+++7j8ccfp2fPnpX1tYhIOburs6tyu4Kzq0We3bET3333nfUfG0fiiM2lzbatS/r7s9lMrrxFRETw2Wef3fZ8vXr1eOedd2xQkdiL8+fPO1zIORLThNyLL77Iiy++aOsyRKSK0bWrImJqCjkRMTWFnIiYmkJORExNIScipqaQE51gvs4AABoJSURBVBFTU8iJiKkp5ETE1BRyImJqCjkRMTWFnDi8gpu/ijkp5MTh6eJ8c1PIiYipKeRExNQqJeSqeqPptLQ0xo4dW+HrEZHKVykhV9UbTT/44IO6saaISd3z7c/trdH0vn37WLRoEQC1atViyZIlXL9+nVGjRrFr1y5CQkLw8PDg2LFjpKWlMWHCBAICAu74Pai5tP1Sc2nzqLDbnx8+fJht27ZRu3ZtunfvzrRp04iOjmb69OnWLlkFkpKS+Oijj8jIyKBfv348//zzxS531qxZjBo1ipUrV3Ly5Ek2bdrEhg0bqFatGkuWLOG9997jiSee4MqVK2zevJm0tDSWLFlSYsj993//N6GhobRr14533nmHH3/8kUaNGhV6z4ULF4iKiiI5OZlRo0aVKuQKqMeDfXK0cZttW5c0ySiXkLOnRtN9+/Zl4sSJ9OvXj759+9KjRw9rW8ICPXr0wGKx0Lx5c9LT00tVn4hUTeUScvbUaPrFF1+kd+/eJCQksGjRIhITE/H39y+yxlsbUouIfaoSPyGpzEbTzz77LJmZmdbGNz/++GOljFFEbKNKdOuqzEbT/+///T9CQkJwcXHB3d2duXPnVtYwRcQG1Fy6nKi5tNgTs21rh2gufSs1mpaySE1N1fWrJmbKkFOjaSmL8+fPK+RMrEqceBARqSgKORExNYWciJiaQk5ETE0hJyKmppATEVNTyImIqSnkRMTUFHIiYmoKORExNYWciJiaQk4cXsFdrcWcFHLi8HRxvrkp5ETE1BRyxbh27RoDBgy4rcmNiNgXhVwRDh8+zHPPPceZM2dsXYqI3COFXBE2bdrEnDlzqFevnq1LEZF7ZMo7A9+refPm3fVni2tway8csZN8p06dHHLcjjJmhVw5UyMb++Ro4zbbti5oZFMU7a6KiKkp5ETE1BRyImJqOiZXgl27dtm6BBG5R5rJicNLTU21dQlSgRRy4vDOnz9v6xKkAinkRMTUFHIiYmoKORExNYWciJiaQk5ETE0hJyKmppATEVNTyImIqSnkRMTUFHIiYmoKORExNYWcODw1lzY3hZw4PDWXNjeFnIiYWqWGXIsWLe74nj59+qihs4iUG83kRMTUSgw5f39/fvrpJwBeffVV5syZA8APP/zAuHHjWLNmDUOGDGHgwIEsXLgQwzAA2Lx5M0OGDGHQoEHMmDGD7OzsQsv9/vvv8fX15eeffyY9PZ2xY8fi7+/PK6+8Yn3vtWvXmDRpEoGBgfTu3ZsZM2ZgGAbBwcFs2rTJuqygoCAOHz5c7BjOnTvHqFGjGDBgAMOGDePEiRMAfPrppwwYMAB/f39CQkLIzMwkNzeX4OBgBg8ezODBgwutR0TsU4kh5+Pjw759+wBITk7m+++/B+Cf//wnvXr14ujRo3zyySds3ryZtLQ0tmzZwsmTJ9m0aRMbNmwgJiaGunXr8t5771mXeeLECWbOnMnq1av54x//yPLly3nssceIjY3l+eef5+LFiwB8+eWXtGrVio0bNxIXF8fBgwc5duwYAQEBxMTEADcD7PLly7Rv377YMbz++us89dRTbN26lb/97W+sWrWKpKQkVq9eTWRkJLGxsdx3332sWLGCH374gStXrrB582befvttvv3223v7dkXE5kpsZOPj40NERARdu3bl0Ucf5dSpU1y6dImvvvqKZs2akZiYyNChQwHIysqiQYMGZGRk8PPPPzN8+HAAcnNzeeyxx6zLHDNmDP3796dJkyYAHDhwgCVLlgDQuXNnvLy8ABgwYACJiYlERERw6tQp0tPTuX79Ot7e3rz22mukpKQQExPDoEGDShzgwYMH+a//+i/reHx8fFi3bh29e/emTp06AAQGBjJ9+nTGjRvH6dOnGTNmDD179mTq1Kll/kKLa3BrLxylq/qtOnXq5JDjdpQxlxhyHTt2JCQkhL1799KlSxfq1q3Ljh07uHHjBjVr1mT06NH85S9/AeDq1as4OzvzySef8PTTTzNr1iwAMjMzycvLsy5z8eLFTJ06lWeffZaWLVtisVisu7kAzs7OAERGRhIXF8fw4cPp3r07ycnJGIaBxWJh8ODBbNu2je3btxeaJRY5QJf/G6JhGPz000/k5+cXeo9hGNy4cYM6deqwbds29uzZw+7duxkyZAjbtm3j/vvvL813CUCbNm2oVq1aqd9flZitq3pZONq4zbats7Ozi51glLi76uLiQrt27YiMjKRLly507dqV1atX4+PjQ9euXYmJiSEzM5MbN24wYcIE4uLi8Pb25vPPP+fSpUsYhkFoaCgffPCBdZndunXj1VdfZdasWeTn59OtWzfr7mdiYiJnz54FYM+ePQQGBjJw4ECys7M5ceKENZyGDh3Khg0bqF+/Pg8++GCJg3/iiSfYtm0bAHv37uW1116jS5cu7Nq1i/T0dAA2bdqEt7c3O3fuJDg4mF69ejFr1izc3d3V5ETEzt2x76qPjw8HDx6kadOmPPDAA1y6dIlevXrRsWNHTpw4wfDhw8nLy+PJJ59kyJAhWCwWJk6cyOjRo8nPz6dVq1aMGzeu0DIHDx5MdHQ0kZGRTJo0iZCQEPz8/GjSpIl1d3X06NGEhoayZs0aPDw86Nixo/WnJfXr16d+/foMGTLkjgOcPXs2s2bNIioqivvuu4+5c+fy6KOP8vLLLxMUFERubi6tW7fm9ddfp1q1asTHx+Pn50e1atUYOHBgqX72IiJVl8W4dV/RDhiGwS+//EJQUBBbt27Fzc3N1iUB/zdd1u6q2AOzbeuS/v7uOJOrauLi4ggNDSU0NNQacG+++SZ79+697b1t2rRh3rx5lV2iiFQhdhdy/fv3p3///oWemzZtmo2qETNITU3V9asmpisexOHp5JK5KeRExNQUciJiago5ETE1hZyImJpCTkRMTSEnIqamkBMRU1PIiYipKeRExNQUciJiago5ETE1hZw4tJzcPFq2am3rMqQC2d1dSETKk5urM26uzrYuQyqQZnIiYmoKORExNZuE3P79+wkKCrLFqkXEwWgmJyKmZrMTD5cvX2bs2LGcPXuWxo0bs3z5cmJjY1m7di0Wi4XWrVvz2muvUaNGDVq0aEFSUhIA0dHRHDhwgAULFtCnTx/atWvH8ePHiYqKom7dukWuKzY2llWrVmGxWGjbti3/+Mc/uHHjBrNmzSIpKQmLxcKYMWMYPHgwJ06cYPbs2dy4cYNq1arxxhtv0KhRo1KPS82l7UtBMxdHGzc4zphtFnKpqamsXr2ahx9+mOHDh7N+/XrWrVvHpk2bqFOnDq+//jorVqy4Y/+Gnj17smzZsmJfT0tL44033iA6OpqHHnqI4OBgdu/ezffff0+dOnXYunUrly9ftja7/uCDD/jLX/7C008/zWeffcahQ4fKFHLq1mWfHG3cZtvWd91cuiK1bNkSLy8vnJycaNq0KRkZGfTu3Zs6deoAEBgYyDfffHPH5bRv377E13/44Qcef/xxHnroIQAWLVpEv379+Oabbxg2bBgAf/jDH+jbty8HDhzAx8eHf/zjH8yYMYOaNWvi7+9/jyMVEVuyWci5uPzfJNJisXD//fcXet0wDG7cuFHoMVDoOeCOsyYXFxcsFov18eXLl7l8+TK/bzdrGAZ5eXn079+fzz77jHbt2hEREcGcOXPKNjARqVKq1ImHXbt2kZ6eDsCmTZvw9vYGoE6dOpw8eRLDMNi1a1eZltm2bVsOHTrEr7/+CsD8+fPZuXMnXbt25ZNPPgFuBt/OnTvp0qULr7zyCkeOHGHEiBFMnjyZH3/8sRxHKCKVrcpc8eDh4cHLL79MUFAQubm5tG7dmtdffx2AV199lfHjx+Pp6UmnTp3497//XerlPvjgg8ycOZMxY8aQn59Phw4dGDp0KL/99huhoaH4+/uTl5fH+PHjad26NePHj2fmzJmsXLkSV1dXQkNDK2jEIlIZLMbv99vkrhQc+NSJB7EHZtvWJf39VZmZ3L3IysoiMDCwyNcmTZpE3759K7kisRc5uXnk5uZSw726rUuRCmKKkKtevToxMTG2LkPskJurM0cSD5lqViOFVakTDyIi5U0hJyKmppATEVNTyImIqSnkRMTUFHIiYmoKORExNYWciJiaQk5ETE0hJyKmppATu5OTm1euy1NzaXMzxbWr4ljcXJ3xf7X8rlWOXTKo3JYlVY9mciJiago5ETG1KhtypWlAnZCQwNq1a297Pjw8nPDw8IoqTUTsiF0fk7P3HqciUvHsIuQOHDjA0qVLycrK4urVq0yfPp1GjRqxYcMGABo0aEBAQECRn01ISGDZsmXk5+fj5eVFWFgYnp6e9OnTh4EDB/L111/z22+/8eabb9KmTRuSk5MJCQkhLy+PJ554gq+++orPP/+8MocrIuXILkJu3bp1zJ07l6ZNm7Jv3z7mz59PbGwsI0aMACg24C5dusTs2bNZv349DRs25N133yUsLIzly5cDULt2bT755BMiIyN5++23CQ8PJyQkhMmTJ+Pj40NERAR5eWX7uYK9zy7toat6RdzF1x7GXd4cZcx2EXKLFi0iISGBHTt2cPjwYTIzM0v1ucTERNq1a0fDhg2Bmw2r16xZY339ySefBKBZs2bEx8eTnp7OuXPn8PHxAW6G54cfflimWtXIxj452rjNtq0LGtkUpcqeeLjVyJEjSUxMpE2bNowfP77Un8vPzy/0+PcNqwvCqKD5tLOz821Np0XEvlX5kEtPT+fMmTNMnjyZnj17snPnTusupLOzc6HQ+r327dtz+PBhUlJSANi4caO1YXVRatasiZeXF7t37wYgNja2HEciIrZQ5XdXa9euzbBhw/Dz88PFxYWuXbuSlZXF9evX6dy5M9OmTcPT07PIn5t4enoSFhbGxIkTyc3NpUGDBsybN6/E9S1cuJAZM2awbNkyWrRoQfXqalUnYs/UXPp3VqxYwfDhw6lXrx7x8fHExsaW6jd3ai5duXRZ172xp21dGqZvLl2eGjRowEsvvYSLiwv333//HWd+UvlycvPKNZgyr2epubSJKeR+Z+jQoQwdOtTWZUgJ3Fydy3V5J44fM9WsRgqr8iceRETuhUJORExNIScipqaQExFTU8iJiKkp5ETE1BRyImJqCjkRMTWFnIiYmkJORExNIScipqaQkyojJ7dst5ovLy1btbbJeqVy6AJ9qTLcXJ3L9RZKpeWIt1pyJFV+Jnf58mWioqJsXYaI2KlKDbm7aRi9YcMGlixZwpkzZwDIy8tjzpw5DBgwAD8/PyIiIu5qPSLiGKrcTO7o0aNcu3YNuBloW7duZcaMGXz00UcAREdHk56ezpYtW/j444/ZtGkTx44ds2XJIlKF2eyYXGkaRnt4ePD444/j7+/P22+/zSuvvEKzZs1o3749Tk5OuLu74+Xlxfnz52nduuiDx6dPn2b27Nmkp6fj7u7OzJkzadeuHSEhIXh4eHDs2DHS0tKYMGECAQEBZGRkMHXqVM6ePYuXlxcXLlxgxYoV1raGImJfbBZypWkYPWrUKKZMmYKbmxt9+/YlJiaGkSNHWpfx/fffk5iYyMKFC4tdT3BwMOPGjcPX15dDhw4xefJk4uLiALhw4QJRUVEkJyczatQoAgICWLlyJY0bN2bVqlUcOXKEwMDAMo1LzaXvni3vzusojZZv5ShjtlnIlaZh9K2NnadNm1botYMHDzJlyhQWL15MrVq1ilxHZmYmZ8+exdfXF4AOHTpQq1YtTp06BUCPHj2wWCw0b96c9PR0APbs2cPixYsBaNu2Lc2bNy/TuNTIxj452rjNtq2rZHPpu20YDRAfH8/kyZNZsmQJPXr0KPZ9RTUiMwzD2rf1982lQQ2mRczGJiF3Lw2jExMTCQ0N5f333y+xUTSAh4cHDRs2JD4+HoBDhw5x8eJFmjVrVuxnunXrZm0qnZSUxMmTJwuFoIjYF5vsrt5Lw+hVq1aRl5dXaPd10qRJ9O3bt8h1LVq0iNDQUMLDw3F1dSU8PBw3N7dia5swYQLTp0/H39+fRx55BE9PTzWYFrFjai79OzExMTRs2JBOnTqRmprKCy+8wBdffIGTU8mTXjWXLh+64qFyVIVtXZ7UXLoMmjRpwpw5c8jPz8fJyYmwsLA7BpyIVF0Kud9p27Yt0dHRti7DIeXk5tlkVpV5PYsa7jokYVaaokiV4ebqbJP1njiuK2bMTCEnIqamkBMRU1PIiYipKeRExNQUciJiago5ETE1hZyImJpCTkRMTSEnIqamkBMRU1PIiYipKeSkWLbqaF/ZWrYqugmSmIPuQiLFslVH+8rmiPeTcySayYmIqTl0yO3fv7/IW6wXSEhIYO3atZVYkYiUN+2ulsDee6iKiEIOgAMHDrB06VKysrK4evUq06dPp1GjRmzYsAGABg0aEBAQYOMqReRuKOSAdevWMXfuXJo2bcq+ffuYP38+sbGxjBgxAqBMAWfvs79bu6qbqdHJnThKN/lbOcqYFXLcbFuYkJDAjh07OHz4MJmZmXe9LHXrsk+ONm6zbeuCbl1FcegTDwVGjhxJYmIibdq0Yfz48bYuR0TKkcPP5NLT00lNTSUqKgo3NzcWL15MXt7NH8E6OzuTnZ1t4wpF5F44/Eyudu3aDBs2DD8/P55++mkyMzPJysri+vXrdO7cmdjYWCIjI21dpojcJYeeyXl7e+Pt7Q3A9OnTrc+HhoYC0LlzZ3bt2mWL0kSknDh0yEnJbNXsubKpubS5OfzuqhTPVs2eK5uaS5ubQk5ETE0hJyKmppATEVNTyImIqSnkRMTUFHIiYmoKORExNYWciJiaQk5ETE0hJyKmppATEVNTyNlAVW3abKY7xZaFmkubm+5CYgOO0rTZXjjCnVYcmWZyImJqDh1yOTk5rFmzxtZliEgFspuQu1O3+7uxfft2Vq1axd69e63PvfXWWzzzzDP4+fmxdu3acl2fiFQ+uwm5irB+/XrmzZvHunXrgJtNpr/55hu2bNnCp59+SmRkJKdOnbJxlSJyL+zuxMPp06eZPXs26enpuLu7M3PmTNq1a0dISAgeHh4cO3aMtLQ0JkyYQEBAABkZGUydOpWzZ8/i5eXFhQsXWLFiBVeuXOG+++7jmWee4d133+XcuXN06dKFDz/8EBcXF9LS0sjLy8Pd3d3WQxaRe2B3IRccHMy4cePw9fXl0KFDTJ48mbi4OAAuXLhAVFQUycnJjBo1ioCAAFauXEnjxo1ZtWoVR44cITAwEIB169YREBAAQEBAAFFRUQQHB+Pq6sry5ct5//336d+/Pw8++GCZ6iuuwe2tHPWnGlWZo3STv5WjjNmuQi4zM5OUlBR8fX0B6NChA7Vq1bLuUvbo0QOLxULz5s1JT08HYM+ePSxevBiAtm3b0rx5cwDeeOMN63Kff/75QuuZNGkSY8eOZfz48WzatMkajKXRpk0bqlWrdveDFJtwtH94vvvuO1ONOTs7u9gJhl0dkzMMo8jnCppBF4SLxWKxvu7s7Fzk54ry008/cfz4cQDuu+8+fH19SUpKuteyRcSG7CrkPDw8aNiwIfHx8QAcOnSIixcv0qxZs2I/061bN2JjYwFISkri5MmThULwVikpKcyaNYucnBxycnLYuXOnqf61E3FEdrW7CrBo0SJCQ0MJDw/H1dWV8PBw3Nzcin3/hAkTmD59Ov7+/jzyyCN4enpSvXrRPTZ9fHxITExk8ODBODs74+vri5+fX0UNRUQqgcUo7b6cnYqJiaFhw4Z06tSJ1NRUXnjhBb744gucnMp3EltwTKC0x+R0WVfV4YiXdZn1mFxRf392N5MrqyZNmjBnzhzy8/NxcnIiLCys3AOurBylM729yLyeRQ33omf3Yv9MH3Jt27YlOjra1mUUUlU705vtX/fSOnH8mEOO21HY1YkHEZGyUsiJiKkp5ETE1Ex/TK6yFJykzsnJsXEl9yY7O9vWJdiEI47bTGMu+Lsr6scipv8JSWXJyMggOTnZ1mWIOLTmzZtTs2bNQs8p5MpJfn4+mZmZuLq6FntFhYhUDMMwyM3NpUaNGrf9REwhJyKmphMPImJqCjkRMTWFnIiYmkJORExNIScipqaQExFTU8iJiKkp5ETE1HTtqlh99tlnLFmyhLp16wLQq1cvpkyZYuOqKk5sbCyrVq3ixo0bjB49+raubWYVFBTE5cuXcXG5+ecfFhZG+/btbVxVxVHIidXRo0cJCQlhwIABti6lwqWlpbF06VKio6Nxc3NjxIgReHt78+ijj9q6tAplGAZnzpwhISHBGnJmp91VsTpy5AifffYZ/v7+/P3vf+fKlSu2LqnC7N27l65du1K7dm3c3d156qmn2LFjh63LqnAFPYpfeuklBg4cyLp162xcUcVTyInVAw88wH/+53+yZcsW6tevT1hYmK1LqjC//PILDzzwgPVxvXr1SEtLs2FFlePq1at069aNlStXEhERwYYNG9izZ4+ty6pQjjFflUK2b9/OG2+8Uei5Jk2aEBERYX38H//xH/z5z3+u5MoqT35+fqG7xRiG4RB3j+nYsSMdO3a0Ph42bBi7d++mR48eNqyqYinkHNDTTz/N008/Xei5jIwMIiIiePHFF4Gbf/TOzlWz4U55eOihh/j222+tj3/99Vfq1atnw4oqx7fffktubi7dunUDbm5nsx+b0+6qAODu7s67777L4cOHAVi3bp2pZ3Ldu3dn3759XL58md9++434+Hh69uxp67IqXEZGBgsXLiQ7O5tr167x2WefmXo7g+4nJ7f49ttvmTdvHllZWTRq1IiFCxfedpdVM4mNjeXtt98mNzeXYcOGMXbsWFuXVCmWLVtGXFwc+fn5jBw5ktGjR9u6pAqlkBMRU9PuqoiYmkJORExNIScipqaQExFTU8iJiKkp5ETE1BRyImJq/x9TkLMmP9/OPwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 288x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_coefs = pd.DataFrame(np.expand_dims(poly_regressors[3]['lr'].coef_, axis=0),\n",
    "                        columns=poly_features).T\n",
    "df_coefs.sort_values(by=0).plot.barh(figsize=(4, 8))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "gxrFkfv0_cAX"
   },
   "source": [
    "## Random Forest"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "umNSn9r-eKOb"
   },
   "source": [
    "Loss with full data: 0.09902381906309178\n",
    "\n",
    "Loss with partial data: 0.09902381906309177\n",
    "\n",
    "Loss with parital data and dropped 'day' column: 0.09902381906309199"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "tbVExEMQf2Dg"
   },
   "outputs": [],
   "source": [
    "X_train = X_train.drop('day', axis=1)\n",
    "X_test  = X_test.drop('day', axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train random forest model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 187
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 21096,
     "status": "ok",
     "timestamp": 1559366751592,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "lArZ2RrU_cAY",
    "outputId": "0e5212c9-7cdc-4050-ed9e-f8ccc2e41812"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend ThreadingBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  42 tasks      | elapsed:   31.3s\n",
      "[Parallel(n_jobs=-1)]: Done  50 out of  50 | elapsed:   36.1s finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=3,\n",
       "                      max_features='auto', max_leaf_nodes=None,\n",
       "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                      min_samples_leaf=1, min_samples_split=2,\n",
       "                      min_weight_fraction_leaf=0.0, n_estimators=50, n_jobs=-1,\n",
       "                      oob_score=False, random_state=0, verbose=1,\n",
       "                      warm_start=True)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc = RandomForestRegressor(n_estimators=50, max_depth=3, n_jobs=-1,\n",
    "                            random_state=0, warm_start=True, verbose=1)\n",
    "rfc.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 85
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 19916,
     "status": "ok",
     "timestamp": 1559366752601,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "LTBpOWa-_cAo",
    "outputId": "5fccfb32-5f38-4594-9c55-dc5790192433"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend ThreadingBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done  42 tasks      | elapsed:    0.0s\n",
      "[Parallel(n_jobs=4)]: Done  50 out of  50 | elapsed:    0.0s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest loss: 0.07614693023190335\n"
     ]
    }
   ],
   "source": [
    "y_pred = rfc.predict(X_test)\n",
    "print(\"Random Forest loss:\", math.sqrt(mean_squared_error(y_test, y_pred)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('models/geotemporal/random_forest.pkl', 'wb') as fid:\n",
    "    cPickle.dump(rfc, fid)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Inspect the first tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 492
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 21456,
     "status": "ok",
     "timestamp": 1559366752595,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "ULpyupZ2_cAg",
    "outputId": "ed71a158-8360-485c-b84a-e2093fc4b36d"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dot_data = StringIO()\n",
    "export_graphviz(rfc.estimators_[0], out_file=dot_data,  \n",
    "                filled=True, rounded=True,\n",
    "                special_characters=True, \n",
    "                feature_names=X_train.columns,\n",
    "                max_depth=2)\n",
    "graph = pydotplus.graph_from_dot_data(dot_data.getvalue())  \n",
    "Image(graph.create_png())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ULxFNg7r_cAu"
   },
   "source": [
    "## XGBoost"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "1tj2LiCSfVcK"
   },
   "source": [
    "Loss with partial data: 0.07101681478370973\n",
    "\n",
    "Loss with full data: 0.07101681478370973\n",
    "\n",
    "Loss with partial data and dropped 'day' column: 0.06948751120370361"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 221
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 29885,
     "status": "ok",
     "timestamp": 1559366835783,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "TriHXxwr_cAv",
    "outputId": "b578752d-cf35-4690-dd7c-3159aefde59a"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\admin\\.conda\\envs\\py36\\lib\\site-packages\\xgboost\\core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
      "  if getattr(data, 'base', None) is not None and \\\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[01:09:53] WARNING: C:/Jenkins/workspace/xgboost-win64_release_0.90/src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "XGBRegressor(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "             colsample_bynode=1, colsample_bytree=1, eval_metric='rmse',\n",
       "             gamma=0, importance_type='gain', learning_rate=0.1,\n",
       "             max_delta_step=0, max_depth=3, min_child_weight=1, missing=None,\n",
       "             n_estimators=100, n_jobs=-1, nthread=None, objective='reg:linear',\n",
       "             random_state=0, reg_alpha=0, reg_lambda=1, scale_pos_weight=1,\n",
       "             seed=None, silent=None, subsample=1, verbosity=1)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xgb = XGBRegressor(random_state=0, n_jobs=-1, eval_metric='rmse')\n",
    "xgb.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 29412,
     "status": "ok",
     "timestamp": 1559366835785,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "10P3nWx-_cA5",
    "outputId": "d21be5d4-7b82-41e2-af15-63a7329b1691"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "XGBoost loss: 0.04855022108270493\n"
     ]
    }
   ],
   "source": [
    "y_pred = xgb.predict(X_test)\n",
    "print(\"XGBoost loss:\", math.sqrt(mean_squared_error(y_test, y_pred)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('models/geotemporal/xgboost.pkl', 'wb') as fid:\n",
    "    cPickle.dump(xgb, fid)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "nKNNGOwQepSF"
   },
   "source": [
    "## Plain Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "zJABwdsweiG3"
   },
   "outputs": [],
   "source": [
    "X_train_means = X_train.mean()\n",
    "X_train_maxs  = X_train.max()\n",
    "X_train_mins  = X_train.min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "LpViEUMTeiG6"
   },
   "outputs": [],
   "source": [
    "X_train_normed = (X_train - X_train_means) / (X_train_maxs - X_train_mins)\n",
    "X_test_normed = (X_test - X_train_means) / (X_train_maxs - X_train_mins)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "uJbVmqV5eiG8"
   },
   "outputs": [],
   "source": [
    "def build_model():\n",
    "    model = tfk.Sequential([\n",
    "        tfk.layers.Input((9,)),\n",
    "        tfk.layers.Dense(32, activation='relu'),\n",
    "        tfk.layers.Dense(16, activation='relu'),\n",
    "        tfk.layers.Dense(8, activation='relu'),\n",
    "        tfk.layers.Dense(1)\n",
    "    ], name='nn_spatiotemporal')\n",
    "\n",
    "    model.compile(loss='mean_squared_error',\n",
    "                  optimizer='adam',\n",
    "                  metrics=['mean_absolute_error'])\n",
    "\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "JRqplcP3eiG_"
   },
   "outputs": [],
   "source": [
    "model = build_model()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 289
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 540,
     "status": "ok",
     "timestamp": 1559333155072,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "_2FeH9NzeiHC",
    "outputId": "81469180-4750-4765-e14f-57c55b80a57c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"nn_spatiotemporal\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense (Dense)                (None, 32)                320       \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 16)                528       \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 8)                 136       \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 1)                 9         \n",
      "=================================================================\n",
      "Total params: 993\n",
      "Trainable params: 993\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 887
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 876014,
     "status": "error",
     "timestamp": 1559334869135,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "_I-GAAREeiHK",
    "outputId": "e018255d-591f-42b3-ebab-3801f40fb97a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 1525692 samples, validate on 381423 samples\n",
      "Epoch 1/1000\n",
      "1525692/1525692 [==============================] - 62s 41us/sample - loss: 0.0054 - mean_absolute_error: 0.0424 - val_loss: 0.0064 - val_mean_absolute_error: 0.0464\n",
      "Epoch 2/1000\n",
      "1525692/1525692 [==============================] - 64s 42us/sample - loss: 0.0053 - mean_absolute_error: 0.0422 - val_loss: 0.0060 - val_mean_absolute_error: 0.0441\n",
      "Epoch 3/1000\n",
      "1525692/1525692 [==============================] - 62s 40us/sample - loss: 0.0053 - mean_absolute_error: 0.0422 - val_loss: 0.0058 - val_mean_absolute_error: 0.0447\n",
      "Epoch 4/1000\n",
      "1525692/1525692 [==============================] - 61s 40us/sample - loss: 0.0053 - mean_absolute_error: 0.0422 - val_loss: 0.0061 - val_mean_absolute_error: 0.0448\n",
      "Epoch 5/1000\n",
      "1525692/1525692 [==============================] - 61s 40us/sample - loss: 0.0052 - mean_absolute_error: 0.0420 - val_loss: 0.0059 - val_mean_absolute_error: 0.0437\n",
      "Epoch 6/1000\n",
      "1525692/1525692 [==============================] - 61s 40us/sample - loss: 0.0052 - mean_absolute_error: 0.0418 - val_loss: 0.0059 - val_mean_absolute_error: 0.0436\n",
      "Epoch 7/1000\n",
      "1525692/1525692 [==============================] - 61s 40us/sample - loss: 0.0052 - mean_absolute_error: 0.0417 - val_loss: 0.0058 - val_mean_absolute_error: 0.0444\n",
      "Epoch 8/1000\n",
      "1525692/1525692 [==============================] - 61s 40us/sample - loss: 0.0051 - mean_absolute_error: 0.0416 - val_loss: 0.0061 - val_mean_absolute_error: 0.0454\n",
      "Epoch 9/1000\n",
      "1525692/1525692 [==============================] - 60s 39us/sample - loss: 0.0051 - mean_absolute_error: 0.0415 - val_loss: 0.0063 - val_mean_absolute_error: 0.0450\n",
      "Epoch 10/1000\n",
      "1525692/1525692 [==============================] - 61s 40us/sample - loss: 0.0051 - mean_absolute_error: 0.0414 - val_loss: 0.0057 - val_mean_absolute_error: 0.0446\n",
      "Epoch 11/1000\n",
      "1525692/1525692 [==============================] - 61s 40us/sample - loss: 0.0051 - mean_absolute_error: 0.0414 - val_loss: 0.0060 - val_mean_absolute_error: 0.0432\n",
      "Epoch 12/1000\n",
      "1525692/1525692 [==============================] - 58s 38us/sample - loss: 0.0051 - mean_absolute_error: 0.0413 - val_loss: 0.0059 - val_mean_absolute_error: 0.0471\n",
      "Epoch 13/1000\n",
      "1525692/1525692 [==============================] - 59s 39us/sample - loss: 0.0050 - mean_absolute_error: 0.0412 - val_loss: 0.0061 - val_mean_absolute_error: 0.0450\n",
      "Epoch 14/1000\n",
      "1525692/1525692 [==============================] - 58s 38us/sample - loss: 0.0050 - mean_absolute_error: 0.0410 - val_loss: 0.0057 - val_mean_absolute_error: 0.0466\n",
      "Epoch 15/1000\n",
      " 814464/1525692 [===============>..............] - ETA: 22s - loss: 0.0049 - mean_absolute_error: 0.0409"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "ignored",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-36-6c007d287178>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     10\u001b[0m   \u001b[0mX_train_normed\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_train\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     11\u001b[0m   \u001b[0mepochs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1000\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalidation_split\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0.2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mverbose\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 12\u001b[0;31m   callbacks=[early_stop])  # PrintDot()\n\u001b[0m",
      "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_freq, max_queue_size, workers, use_multiprocessing, **kwargs)\u001b[0m\n\u001b[1;32m    871\u001b[0m           \u001b[0mvalidation_steps\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mvalidation_steps\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    872\u001b[0m           \u001b[0mvalidation_freq\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mvalidation_freq\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 873\u001b[0;31m           steps_name='steps_per_epoch')\n\u001b[0m\u001b[1;32m    874\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    875\u001b[0m   def evaluate(self,\n",
      "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training_arrays.py\u001b[0m in \u001b[0;36mmodel_iteration\u001b[0;34m(model, inputs, targets, sample_weights, batch_size, epochs, verbose, callbacks, val_inputs, val_targets, val_sample_weights, shuffle, initial_epoch, steps_per_epoch, validation_steps, validation_freq, mode, validation_in_fit, prepared_feed_values_from_dataset, steps_name, **kwargs)\u001b[0m\n\u001b[1;32m    350\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    351\u001b[0m         \u001b[0;31m# Get outputs.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 352\u001b[0;31m         \u001b[0mbatch_outs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mins_batch\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    353\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch_outs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    354\u001b[0m           \u001b[0mbatch_outs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mbatch_outs\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/backend.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, inputs)\u001b[0m\n\u001b[1;32m   3215\u001b[0m         \u001b[0mvalue\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmath_ops\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcast\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtensor\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3216\u001b[0m       \u001b[0mconverted_inputs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 3217\u001b[0;31m     \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_graph_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mconverted_inputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   3218\u001b[0m     return nest.pack_sequence_as(self._outputs_structure,\n\u001b[1;32m   3219\u001b[0m                                  [x.numpy() for x in outputs])\n",
      "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m    556\u001b[0m       raise TypeError(\"Keyword arguments {} unknown. Expected {}.\".format(\n\u001b[1;32m    557\u001b[0m           list(kwargs.keys()), list(self._arg_keywords)))\n\u001b[0;32m--> 558\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_call_flat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    559\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    560\u001b[0m   \u001b[0;32mdef\u001b[0m \u001b[0m_filtered_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36m_call_flat\u001b[0;34m(self, args)\u001b[0m\n\u001b[1;32m    625\u001b[0m     \u001b[0;31m# Only need to override the gradient in graph mode and when we have outputs.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    626\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mcontext\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mexecuting_eagerly\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 627\u001b[0;31m       \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_inference_function\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcall\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mctx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    628\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    629\u001b[0m       \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_register_gradient\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36mcall\u001b[0;34m(self, ctx, args)\u001b[0m\n\u001b[1;32m    413\u001b[0m             attrs=(\"executor_type\", executor_type,\n\u001b[1;32m    414\u001b[0m                    \"config_proto\", config),\n\u001b[0;32m--> 415\u001b[0;31m             ctx=ctx)\n\u001b[0m\u001b[1;32m    416\u001b[0m       \u001b[0;31m# Replace empty list with None\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    417\u001b[0m       \u001b[0moutputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0moutputs\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/tensorflow/python/eager/execute.py\u001b[0m in \u001b[0;36mquick_execute\u001b[0;34m(op_name, num_outputs, inputs, attrs, ctx, name)\u001b[0m\n\u001b[1;32m     58\u001b[0m     tensors = pywrap_tensorflow.TFE_Py_Execute(ctx._handle, device_name,\n\u001b[1;32m     59\u001b[0m                                                \u001b[0mop_name\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minputs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mattrs\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 60\u001b[0;31m                                                num_outputs)\n\u001b[0m\u001b[1;32m     61\u001b[0m   \u001b[0;32mexcept\u001b[0m \u001b[0mcore\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_NotOkStatusException\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     62\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mname\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "class PrintDot(tfk.callbacks.Callback):\n",
    "  def on_epoch_end(self, epoch, logs):\n",
    "    if epoch % 10 == 0:\n",
    "      print('')\n",
    "    print('.', end='')\n",
    "\n",
    "early_stop = tfk.callbacks.EarlyStopping(monitor='val_loss', patience=5)\n",
    "\n",
    "history = model.fit(\n",
    "  X_train_normed, y_train,\n",
    "  epochs=100, validation_split=0.2, verbose=1,\n",
    "  callbacks=[early_stop])  # PrintDot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "executionInfo": {
     "elapsed": 659,
     "status": "ok",
     "timestamp": 1559335007324,
     "user": {
      "displayName": "Franz Louis Cesista",
      "photoUrl": "https://lh3.googleusercontent.com/-NquKeLt1770/AAAAAAAAAAI/AAAAAAAAAMo/RHnx0T1cGIk/s64/photo.jpg",
      "userId": "08273419339046628946"
     },
     "user_tz": -480
    },
    "id": "XRG5bAyFeiHO",
    "outputId": "b8ee2667-4439-48f7-8b40-c7a466199edc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing set Root Mean Squared Error: 0.049747\n"
     ]
    }
   ],
   "source": [
    "loss, mae = model.evaluate(X_test_normed, y_test, verbose=0)\n",
    "\n",
    "print(\"Testing set Root Mean Squared Error: {:f}\".format(math.sqrt(loss)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "XGh2yHjtmh33"
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "name": "4 - Spatiotemporal Regression.ipynb",
   "provenance": [],
   "version": "0.3.2"
  },
  "kernelspec": {
   "display_name": "Python [conda env:py36]",
   "language": "python",
   "name": "conda-env-py36-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}