model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "2ec1fbd8",
   "metadata": {},
   "outputs": [],
   "source": [
    "from datetime import datetime\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import openpyxl\n",
    "import xlsxwriter\n",
    "from pathlib import Path\n",
    "from pandas import DataFrame\n",
    "import re\n",
    "import os\n",
    "from datetime import timedelta\n",
    "from sklearn.impute import SimpleImputer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "bd959270",
   "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>Date</th>\n",
       "      <th>Country</th>\n",
       "      <th>NumDay</th>\n",
       "      <th>LP</th>\n",
       "      <th>GDP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1996-06-30</td>\n",
       "      <td>DE</td>\n",
       "      <td>13</td>\n",
       "      <td>87.9</td>\n",
       "      <td>77.663</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1996-09-30</td>\n",
       "      <td>DE</td>\n",
       "      <td>8</td>\n",
       "      <td>88.3</td>\n",
       "      <td>77.934</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1997-06-30</td>\n",
       "      <td>DE</td>\n",
       "      <td>0</td>\n",
       "      <td>89.7</td>\n",
       "      <td>79.098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1997-09-30</td>\n",
       "      <td>DE</td>\n",
       "      <td>19</td>\n",
       "      <td>89.9</td>\n",
       "      <td>79.368</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1998-06-30</td>\n",
       "      <td>DE</td>\n",
       "      <td>8</td>\n",
       "      <td>90.3</td>\n",
       "      <td>80.354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>2017-09-30</td>\n",
       "      <td>NL</td>\n",
       "      <td>16</td>\n",
       "      <td>101.3</td>\n",
       "      <td>105.622</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>2018-06-30</td>\n",
       "      <td>NL</td>\n",
       "      <td>22</td>\n",
       "      <td>101.1</td>\n",
       "      <td>107.590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>2018-09-30</td>\n",
       "      <td>NL</td>\n",
       "      <td>41</td>\n",
       "      <td>100.7</td>\n",
       "      <td>107.830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>2019-06-30</td>\n",
       "      <td>NL</td>\n",
       "      <td>14</td>\n",
       "      <td>100.7</td>\n",
       "      <td>109.203</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>2019-09-30</td>\n",
       "      <td>NL</td>\n",
       "      <td>37</td>\n",
       "      <td>100.8</td>\n",
       "      <td>109.569</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>144 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          Date Country  NumDay     LP      GDP\n",
       "0   1996-06-30      DE      13   87.9   77.663\n",
       "1   1996-09-30      DE       8   88.3   77.934\n",
       "2   1997-06-30      DE       0   89.7   79.098\n",
       "3   1997-09-30      DE      19   89.9   79.368\n",
       "4   1998-06-30      DE       8   90.3   80.354\n",
       "..         ...     ...     ...    ...      ...\n",
       "139 2017-09-30      NL      16  101.3  105.622\n",
       "140 2018-06-30      NL      22  101.1  107.590\n",
       "141 2018-09-30      NL      41  100.7  107.830\n",
       "142 2019-06-30      NL      14  100.7  109.203\n",
       "143 2019-09-30      NL      37  100.8  109.569\n",
       "\n",
       "[144 rows x 5 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_excel(\"/Users/zhengxin/Downloads/NumDay_quarter.xlsx\",engine='openpyxl')\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "1b71f9f7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 144 entries, 0 to 143\n",
      "Data columns (total 7 columns):\n",
      " #   Column   Non-Null Count  Dtype  \n",
      "---  ------   --------------  -----  \n",
      " 0   Date     144 non-null    object \n",
      " 1   Country  144 non-null    object \n",
      " 2   NumDay   144 non-null    int64  \n",
      " 3   LP       144 non-null    float64\n",
      " 4   GDP      144 non-null    float64\n",
      " 5   Year     144 non-null    int64  \n",
      " 6   Quarter  144 non-null    int64  \n",
      "dtypes: float64(2), int64(3), object(2)\n",
      "memory usage: 8.0+ KB\n",
      "None\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>Date</th>\n",
       "      <th>Country</th>\n",
       "      <th>NumDay</th>\n",
       "      <th>LP</th>\n",
       "      <th>GDP</th>\n",
       "      <th>Year</th>\n",
       "      <th>Quarter</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>30.06.1996</td>\n",
       "      <td>DE</td>\n",
       "      <td>13</td>\n",
       "      <td>87.9</td>\n",
       "      <td>77.663</td>\n",
       "      <td>1996</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>30.09.1996</td>\n",
       "      <td>DE</td>\n",
       "      <td>8</td>\n",
       "      <td>88.3</td>\n",
       "      <td>77.934</td>\n",
       "      <td>1996</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>30.06.1997</td>\n",
       "      <td>DE</td>\n",
       "      <td>0</td>\n",
       "      <td>89.7</td>\n",
       "      <td>79.098</td>\n",
       "      <td>1997</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>30.09.1997</td>\n",
       "      <td>DE</td>\n",
       "      <td>19</td>\n",
       "      <td>89.9</td>\n",
       "      <td>79.368</td>\n",
       "      <td>1997</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>30.06.1998</td>\n",
       "      <td>DE</td>\n",
       "      <td>8</td>\n",
       "      <td>90.3</td>\n",
       "      <td>80.354</td>\n",
       "      <td>1998</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>30.09.2017</td>\n",
       "      <td>NL</td>\n",
       "      <td>16</td>\n",
       "      <td>101.3</td>\n",
       "      <td>105.622</td>\n",
       "      <td>2017</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>30.06.2018</td>\n",
       "      <td>NL</td>\n",
       "      <td>22</td>\n",
       "      <td>101.1</td>\n",
       "      <td>107.590</td>\n",
       "      <td>2018</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>30.09.2018</td>\n",
       "      <td>NL</td>\n",
       "      <td>41</td>\n",
       "      <td>100.7</td>\n",
       "      <td>107.830</td>\n",
       "      <td>2018</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>30.06.2019</td>\n",
       "      <td>NL</td>\n",
       "      <td>14</td>\n",
       "      <td>100.7</td>\n",
       "      <td>109.203</td>\n",
       "      <td>2019</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>30.09.2019</td>\n",
       "      <td>NL</td>\n",
       "      <td>37</td>\n",
       "      <td>100.8</td>\n",
       "      <td>109.569</td>\n",
       "      <td>2019</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>144 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           Date Country  NumDay     LP      GDP  Year  Quarter\n",
       "0    30.06.1996      DE      13   87.9   77.663  1996        2\n",
       "1    30.09.1996      DE       8   88.3   77.934  1996        3\n",
       "2    30.06.1997      DE       0   89.7   79.098  1997        2\n",
       "3    30.09.1997      DE      19   89.9   79.368  1997        3\n",
       "4    30.06.1998      DE       8   90.3   80.354  1998        2\n",
       "..          ...     ...     ...    ...      ...   ...      ...\n",
       "139  30.09.2017      NL      16  101.3  105.622  2017        3\n",
       "140  30.06.2018      NL      22  101.1  107.590  2018        2\n",
       "141  30.09.2018      NL      41  100.7  107.830  2018        3\n",
       "142  30.06.2019      NL      14  100.7  109.203  2019        2\n",
       "143  30.09.2019      NL      37  100.8  109.569  2019        3\n",
       "\n",
       "[144 rows x 7 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Date'] = pd.to_datetime(df['Date'])\n",
    "df['Date'] = df['Date'].dt.strftime('%d.%m.%Y')\n",
    "df['Year'] = pd.DatetimeIndex(df['Date']).year\n",
    "df['Quarter'] = pd.DatetimeIndex(df['Date']).quarter\n",
    "print(df.info())\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "91dbbe38",
   "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>Date</th>\n",
       "      <th>Country</th>\n",
       "      <th>NumDay</th>\n",
       "      <th>LP</th>\n",
       "      <th>GDP</th>\n",
       "      <th>Year</th>\n",
       "      <th>Quarter</th>\n",
       "      <th>Country_DE</th>\n",
       "      <th>Country_ES</th>\n",
       "      <th>Country_NL</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>30.06.1996</td>\n",
       "      <td>DE</td>\n",
       "      <td>13</td>\n",
       "      <td>87.9</td>\n",
       "      <td>77.663</td>\n",
       "      <td>1996</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>30.09.1996</td>\n",
       "      <td>DE</td>\n",
       "      <td>8</td>\n",
       "      <td>88.3</td>\n",
       "      <td>77.934</td>\n",
       "      <td>1996</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>30.06.1997</td>\n",
       "      <td>DE</td>\n",
       "      <td>0</td>\n",
       "      <td>89.7</td>\n",
       "      <td>79.098</td>\n",
       "      <td>1997</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>30.09.1997</td>\n",
       "      <td>DE</td>\n",
       "      <td>19</td>\n",
       "      <td>89.9</td>\n",
       "      <td>79.368</td>\n",
       "      <td>1997</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>30.06.1998</td>\n",
       "      <td>DE</td>\n",
       "      <td>8</td>\n",
       "      <td>90.3</td>\n",
       "      <td>80.354</td>\n",
       "      <td>1998</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>30.09.2017</td>\n",
       "      <td>NL</td>\n",
       "      <td>16</td>\n",
       "      <td>101.3</td>\n",
       "      <td>105.622</td>\n",
       "      <td>2017</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>30.06.2018</td>\n",
       "      <td>NL</td>\n",
       "      <td>22</td>\n",
       "      <td>101.1</td>\n",
       "      <td>107.590</td>\n",
       "      <td>2018</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>30.09.2018</td>\n",
       "      <td>NL</td>\n",
       "      <td>41</td>\n",
       "      <td>100.7</td>\n",
       "      <td>107.830</td>\n",
       "      <td>2018</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>30.06.2019</td>\n",
       "      <td>NL</td>\n",
       "      <td>14</td>\n",
       "      <td>100.7</td>\n",
       "      <td>109.203</td>\n",
       "      <td>2019</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>30.09.2019</td>\n",
       "      <td>NL</td>\n",
       "      <td>37</td>\n",
       "      <td>100.8</td>\n",
       "      <td>109.569</td>\n",
       "      <td>2019</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>144 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           Date Country  NumDay     LP      GDP  Year  Quarter  Country_DE  \\\n",
       "0    30.06.1996      DE      13   87.9   77.663  1996        2           1   \n",
       "1    30.09.1996      DE       8   88.3   77.934  1996        3           1   \n",
       "2    30.06.1997      DE       0   89.7   79.098  1997        2           1   \n",
       "3    30.09.1997      DE      19   89.9   79.368  1997        3           1   \n",
       "4    30.06.1998      DE       8   90.3   80.354  1998        2           1   \n",
       "..          ...     ...     ...    ...      ...   ...      ...         ...   \n",
       "139  30.09.2017      NL      16  101.3  105.622  2017        3           0   \n",
       "140  30.06.2018      NL      22  101.1  107.590  2018        2           0   \n",
       "141  30.09.2018      NL      41  100.7  107.830  2018        3           0   \n",
       "142  30.06.2019      NL      14  100.7  109.203  2019        2           0   \n",
       "143  30.09.2019      NL      37  100.8  109.569  2019        3           0   \n",
       "\n",
       "     Country_ES  Country_NL  \n",
       "0             0           0  \n",
       "1             0           0  \n",
       "2             0           0  \n",
       "3             0           0  \n",
       "4             0           0  \n",
       "..          ...         ...  \n",
       "139           0           1  \n",
       "140           0           1  \n",
       "141           0           1  \n",
       "142           0           1  \n",
       "143           0           1  \n",
       "\n",
       "[144 rows x 10 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# dummy encoding\n",
    "df['Country_'] = df['Country']\n",
    "df = pd.get_dummies(df, columns=['Country_'], drop_first=False, prefix='Country')\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52ca09c8",
   "metadata": {},
   "source": [
    "## Time series"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "1710d9b1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 3600x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(50, 8))\n",
    "plt.plot(df.loc[df['Country'] == 'DE']['Date'], df.loc[df['Country'] == 'DE']['LP'], \n",
    "         'b-', label = 'DE')\n",
    "plt.plot(df.loc[df['Country'] == 'NL']['Date'], df.loc[df['Country'] == 'NL']['LP'], \n",
    "         'r-', label = 'NL')\n",
    "plt.plot(df.loc[df['Country'] == 'ES']['Date'], df.loc[df['Country'] == 'ES']['LP'], \n",
    "         'g-', label = 'ES')\n",
    "plt.xlabel('Date'); \n",
    "plt.ylabel('LP'); \n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3b3b3edc",
   "metadata": {},
   "source": [
    "## Machine learning models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "378c7b00",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import model_selection\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from math import sqrt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "d4c3f106",
   "metadata": {},
   "outputs": [],
   "source": [
    "# from sklearn.model_selection import train_test_split\n",
    "target_column_test = ['LP'] \n",
    "predictors_test = ['NumDay', 'Year', 'Quarter',\n",
    "                  'Country_DE', 'Country_ES', 'Country_NL']\n",
    "X = df[predictors_test]\n",
    "y = df[target_column_test]\n",
    "# X_train, X_test, y_train, y_test = train_test_split(\n",
    "#     df[predictors_test], df[target_column_test], test_size=0.20, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "21dc465b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TimeSeriesSplit(gap=0, max_train_size=None, n_splits=5, test_size=None)\n",
      "TRAIN: [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23] TEST: [24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47]\n",
      "TRAIN: [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23\n",
      " 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47] TEST: [48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71]\n",
      "TRAIN: [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23\n",
      " 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47\n",
      " 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71] TEST: [72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95]\n",
      "TRAIN: [ 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22 23\n",
      " 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47\n",
      " 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71\n",
      " 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95] TEST: [ 96  97  98  99 100 101 102 103 104 105 106 107 108 109 110 111 112 113\n",
      " 114 115 116 117 118 119]\n",
      "TRAIN: [  0   1   2   3   4   5   6   7   8   9  10  11  12  13  14  15  16  17\n",
      "  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35\n",
      "  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53\n",
      "  54  55  56  57  58  59  60  61  62  63  64  65  66  67  68  69  70  71\n",
      "  72  73  74  75  76  77  78  79  80  81  82  83  84  85  86  87  88  89\n",
      "  90  91  92  93  94  95  96  97  98  99 100 101 102 103 104 105 106 107\n",
      " 108 109 110 111 112 113 114 115 116 117 118 119] TEST: [120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137\n",
      " 138 139 140 141 142 143]\n"
     ]
    }
   ],
   "source": [
    "# generate pairs of train/test sets\n",
    "from sklearn.model_selection import TimeSeriesSplit\n",
    "tscv = TimeSeriesSplit()\n",
    "print(tscv)\n",
    "for train_index, test_index in tscv.split(X):\n",
    "    print(\"TRAIN:\", train_index, \"TEST:\", test_index) \n",
    "    X_train, X_test = X.iloc[train_index], X.iloc[test_index]\n",
    "    y_train, y_test = y.iloc[train_index], y.iloc[test_index]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d25c5125",
   "metadata": {},
   "source": [
    "### Decision trees"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "fe4fec3a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE(train): 0.0\n",
      "R2(train): 1.0\n",
      "RMSE(test): 1.4002975874196644\n",
      "R2(test): 0.5318577468291467\n"
     ]
    }
   ],
   "source": [
    "# timeseries split cross validation\n",
    "from sklearn.model_selection import TimeSeriesSplit\n",
    "from sklearn.metrics import mean_squared_error\n",
    "import statsmodels.api as sm\n",
    "tscv = TimeSeriesSplit()\n",
    "for train_index, test_index in tscv.split(X):\n",
    "    X_train, X_test = X.iloc[train_index], X.iloc[test_index]\n",
    "    y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "    \n",
    "    model = DecisionTreeRegressor().fit(X_train, y_train)\n",
    "    pred_test_tree = model.predict(X_test)\n",
    "    pred_train_tree= model.predict(X_train)\n",
    "    \n",
    "print(\"RMSE(train):\",np.sqrt(mean_squared_error(y_train,pred_train_tree)))\n",
    "print(\"R2(train):\", r2_score(y_train, pred_train_tree))\n",
    "print(\"RMSE(test):\", np.sqrt(mean_squared_error(y_test,pred_test_tree))) \n",
    "print(\"R2(test):\", r2_score(y_test, pred_test_tree))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "2bad02e3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 5-fold cross-validation\n",
    "# from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "# model = DecisionTreeRegressor()\n",
    "# gs = GridSearchCV(model,\n",
    "#                   param_grid = {'max_depth': range(1, 11),\n",
    "#                                 'min_samples_split': range(10, 60, 10)},\n",
    "#                   cv=5,\n",
    "#                   n_jobs=1,\n",
    "#                   scoring='neg_mean_squared_error')\n",
    "\n",
    "# gs.fit(X_train, y_train)\n",
    "\n",
    "# print(gs.best_params_)\n",
    "# print(-gs.best_score_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "6790a850",
   "metadata": {},
   "outputs": [],
   "source": [
    "# new_model = DecisionTreeRegressor(max_depth=4,\n",
    "#                                   min_samples_split=10)\n",
    "# #or new_model = gs.best_estimator_\n",
    "# new_model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "9a85a407",
   "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>Variable</th>\n",
       "      <th>Importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Year</td>\n",
       "      <td>0.891037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Country_ES</td>\n",
       "      <td>0.069658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Country_NL</td>\n",
       "      <td>0.031450</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Country_DE</td>\n",
       "      <td>0.005474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NumDay</td>\n",
       "      <td>0.001359</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Quarter</td>\n",
       "      <td>0.001022</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Variable  Importance\n",
       "1        Year    0.891037\n",
       "4  Country_ES    0.069658\n",
       "5  Country_NL    0.031450\n",
       "3  Country_DE    0.005474\n",
       "0      NumDay    0.001359\n",
       "2     Quarter    0.001022"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# the more importance, the more relevant the variable is, according to the model.\n",
    "pd.DataFrame({'Variable':X_train.columns,\n",
    "              'Importance':model.feature_importances_}).sort_values('Importance', \n",
    "                                                                        ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "dcfebae2",
   "metadata": {},
   "outputs": [],
   "source": [
    "# pred_train_tree= new_model.predict(X_train)\n",
    "# print(np.sqrt(mean_squared_error(y_train,pred_train_tree)))\n",
    "# print(r2_score(y_train, pred_train_tree))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "6201ef04",
   "metadata": {},
   "outputs": [],
   "source": [
    "# pred_test_tree= new_model.predict(X_test)\n",
    "# print(np.sqrt(mean_squared_error(y_test,pred_test_tree))) \n",
    "# print(r2_score(y_test, pred_test_tree))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2aada012",
   "metadata": {},
   "source": [
    "### Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "3c640aa7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE(train): 0.29990886115617366\n",
      "R2(train): 0.9962056229878642\n",
      "RMSE(test): 0.804283117647839\n",
      "R2(test): 0.8455616811738386\n"
     ]
    }
   ],
   "source": [
    "# timeseries split cross validation\n",
    "from sklearn.model_selection import TimeSeriesSplit\n",
    "from sklearn.metrics import mean_squared_error\n",
    "import statsmodels.api as sm\n",
    "tscv = TimeSeriesSplit()\n",
    "for train_index, test_index in tscv.split(X):\n",
    "    X_train, X_test = X.iloc[train_index], X.iloc[test_index]\n",
    "    y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "    \n",
    "    rf_model = RandomForestRegressor().fit(X_train, y_train.values.ravel())\n",
    "    pred_test_tree = rf_model.predict(X_test)\n",
    "    pred_train_tree= rf_model.predict(X_train)\n",
    "    \n",
    "print(\"RMSE(train):\",np.sqrt(mean_squared_error(y_train,pred_train_tree)))\n",
    "print(\"R2(train):\", r2_score(y_train, pred_train_tree))\n",
    "print(\"RMSE(test):\", np.sqrt(mean_squared_error(y_test,pred_test_tree))) \n",
    "print(\"R2(test):\", r2_score(y_test, pred_test_tree))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "992d20a9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# # 5-fold cross-validation\n",
    "# from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "# model = RandomForestRegressor()\n",
    "# gs = GridSearchCV(model,\n",
    "#                   param_grid = {'max_depth': range(1, 11),\n",
    "#                                 'min_samples_split': range(10, 60, 10)},\n",
    "#                   cv=5,\n",
    "#                   n_jobs=1,\n",
    "#                   scoring='neg_mean_squared_error')\n",
    "\n",
    "# gs.fit(X_train, y_train.values.ravel())\n",
    "\n",
    "# print(gs.best_params_)\n",
    "# print(-gs.best_score_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "71d855c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# new_model = RandomForestRegressor(max_depth=5,\n",
    "#                                   min_samples_split=10)\n",
    "# #or new_model = gs.best_estimator_\n",
    "# new_model.fit(X_train, y_train.values.ravel())\n",
    "# pred_train_tree= new_model.predict(X_train)\n",
    "# print(np.sqrt(mean_squared_error(y_train,pred_train_tree)))\n",
    "# print(r2_score(y_train, pred_train_tree))\n",
    "# pred_test_tree= new_model.predict(X_test)\n",
    "# print(np.sqrt(mean_squared_error(y_test,pred_test_tree))) \n",
    "# print(r2_score(y_test, pred_test_tree))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "99e67d73",
   "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>Variable</th>\n",
       "      <th>Importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Year</td>\n",
       "      <td>0.878767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Country_ES</td>\n",
       "      <td>0.054093</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Country_NL</td>\n",
       "      <td>0.028707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NumDay</td>\n",
       "      <td>0.019941</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Country_DE</td>\n",
       "      <td>0.017378</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Quarter</td>\n",
       "      <td>0.001115</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Variable  Importance\n",
       "1        Year    0.878767\n",
       "4  Country_ES    0.054093\n",
       "5  Country_NL    0.028707\n",
       "0      NumDay    0.019941\n",
       "3  Country_DE    0.017378\n",
       "2     Quarter    0.001115"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# the more importance, the more relevant the variable is, according to the model.\n",
    "pd.DataFrame({'Variable':X_train.columns,\n",
    "              'Importance':rf_model.feature_importances_}).sort_values('Importance', \n",
    "                                                                        ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "da7567f3",
   "metadata": {},
   "source": [
    "### Linear Mixed-Effects Models Using lme4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "d596dd06",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "           LP    NumDay      Year  Quarter  Country_DE  Country_ES  \\\n",
      "0    0.233161  0.228070  0.000000      0.0         1.0         0.0   \n",
      "1    0.253886  0.140351  0.000000      1.0         1.0         0.0   \n",
      "2    0.326425  0.000000  0.043478      0.0         1.0         0.0   \n",
      "3    0.336788  0.333333  0.043478      1.0         1.0         0.0   \n",
      "4    0.357513  0.140351  0.086957      0.0         1.0         0.0   \n",
      "..        ...       ...       ...      ...         ...         ...   \n",
      "139  0.927461  0.280702  0.913043      1.0         0.0         0.0   \n",
      "140  0.917098  0.385965  0.956522      0.0         0.0         0.0   \n",
      "141  0.896373  0.719298  0.956522      1.0         0.0         0.0   \n",
      "142  0.896373  0.245614  1.000000      0.0         0.0         0.0   \n",
      "143  0.901554  0.649123  1.000000      1.0         0.0         0.0   \n",
      "\n",
      "     Country_NL Country  \n",
      "0           0.0      DE  \n",
      "1           0.0      DE  \n",
      "2           0.0      DE  \n",
      "3           0.0      DE  \n",
      "4           0.0      DE  \n",
      "..          ...     ...  \n",
      "139         1.0      NL  \n",
      "140         1.0      NL  \n",
      "141         1.0      NL  \n",
      "142         1.0      NL  \n",
      "143         1.0      NL  \n",
      "\n",
      "[144 rows x 8 columns]\n"
     ]
    }
   ],
   "source": [
    "# Normalize time series data\n",
    "from sklearn import preprocessing\n",
    "columns = ['LP','NumDay', 'Year', 'Quarter','Country_DE', 'Country_ES', 'Country_NL']\n",
    "\n",
    "x = df[columns].values #returns a numpy array\n",
    "min_max_scaler = preprocessing.MinMaxScaler()\n",
    "x_scaled = min_max_scaler.fit_transform(x)\n",
    "df_norm = pd.DataFrame(x_scaled)\n",
    "df_norm.columns = columns\n",
    "df_norm[\"Country\"] = df[\"Country\"]\n",
    "print(df_norm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "e34a6035",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "boundary (singular) fit: see ?isSingular \n",
      "\n",
      "Formula: LP~Year+Quarter+NumDay+(Year+Quarter|Country)\n",
      "\n",
      "Family: gaussian\t Inference: parametric\n",
      "\n",
      "Number of observations: 144\t Groups: {'Country': 3.0}\n",
      "\n",
      "Log-likelihood: 131.448 \t AIC: -262.895\n",
      "\n",
      "Random effects:\n",
      "\n",
      "                 Name    Var    Std\n",
      "Country   (Intercept)  0.008  0.087\n",
      "Country          Year  0.005  0.069\n",
      "Country       Quarter  0.000  0.003\n",
      "Residual               0.008  0.088\n",
      "\n",
      "                 IV1      IV2   Corr\n",
      "Country  (Intercept)     Year -0.834\n",
      "Country  (Intercept)  Quarter -0.897\n",
      "Country         Year  Quarter  0.992\n",
      "\n",
      "Fixed effects:\n",
      "\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>Estimate</th>\n",
       "      <th>2.5_ci</th>\n",
       "      <th>97.5_ci</th>\n",
       "      <th>SE</th>\n",
       "      <th>DF</th>\n",
       "      <th>T-stat</th>\n",
       "      <th>P-val</th>\n",
       "      <th>Sig</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>(Intercept)</th>\n",
       "      <td>0.216</td>\n",
       "      <td>0.113</td>\n",
       "      <td>0.319</td>\n",
       "      <td>0.053</td>\n",
       "      <td>2.015</td>\n",
       "      <td>4.098</td>\n",
       "      <td>0.054</td>\n",
       "      <td>.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Year</th>\n",
       "      <td>0.760</td>\n",
       "      <td>0.668</td>\n",
       "      <td>0.853</td>\n",
       "      <td>0.047</td>\n",
       "      <td>2.058</td>\n",
       "      <td>16.112</td>\n",
       "      <td>0.003</td>\n",
       "      <td>**</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Quarter</th>\n",
       "      <td>0.016</td>\n",
       "      <td>-0.018</td>\n",
       "      <td>0.050</td>\n",
       "      <td>0.018</td>\n",
       "      <td>81.896</td>\n",
       "      <td>0.914</td>\n",
       "      <td>0.363</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NumDay</th>\n",
       "      <td>-0.020</td>\n",
       "      <td>-0.085</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.033</td>\n",
       "      <td>137.303</td>\n",
       "      <td>-0.598</td>\n",
       "      <td>0.551</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Estimate  2.5_ci  97.5_ci     SE       DF  T-stat  P-val Sig\n",
       "(Intercept)     0.216   0.113    0.319  0.053    2.015   4.098  0.054   .\n",
       "Year            0.760   0.668    0.853  0.047    2.058  16.112  0.003  **\n",
       "Quarter         0.016  -0.018    0.050  0.018   81.896   0.914  0.363    \n",
       "NumDay         -0.020  -0.085    0.045  0.033  137.303  -0.598  0.551    "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SS Type III Analysis of Variance Table with Satterthwaite approximated degrees of freedom:\n",
      "(NOTE: Using original model contrasts, orthogonality not guaranteed)\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>SS</th>\n",
       "      <th>MS</th>\n",
       "      <th>NumDF</th>\n",
       "      <th>DenomDF</th>\n",
       "      <th>F-stat</th>\n",
       "      <th>P-val</th>\n",
       "      <th>Sig</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.994837</td>\n",
       "      <td>1.994837</td>\n",
       "      <td>1</td>\n",
       "      <td>2.057824</td>\n",
       "      <td>259.597910</td>\n",
       "      <td>0.003387</td>\n",
       "      <td>**</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.006419</td>\n",
       "      <td>0.006419</td>\n",
       "      <td>1</td>\n",
       "      <td>81.896164</td>\n",
       "      <td>0.835365</td>\n",
       "      <td>0.363409</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.002749</td>\n",
       "      <td>0.002749</td>\n",
       "      <td>1</td>\n",
       "      <td>137.303031</td>\n",
       "      <td>0.357700</td>\n",
       "      <td>0.550772</td>\n",
       "      <td></td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         SS        MS  NumDF     DenomDF      F-stat     P-val Sig\n",
       "0  1.994837  1.994837      1    2.057824  259.597910  0.003387  **\n",
       "1  0.006419  0.006419      1   81.896164    0.835365  0.363409    \n",
       "2  0.002749  0.002749      1  137.303031    0.357700  0.550772    "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Install pymer4\n",
    "# !pip install -q pymer4\n",
    "# load pymer4\n",
    "from pymer4.models import Lmer\n",
    "model = Lmer('LP ~ Year + Quarter + NumDay + (Year + Quarter|Country)', data=df_norm)\n",
    "display(model.fit())\n",
    "# ANOVA results from fitted model\n",
    "display(model.anova())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e9cf5a37",
   "metadata": {},
   "source": [
    "### regression (gradient descent)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "cf127c24",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index(['Date', 'Country', 'NumDay', 'LP', 'GDP', 'Year', 'Quarter',\n",
      "       'Country_DE', 'Country_ES', 'Country_NL'],\n",
      "      dtype='object')\n",
      "(144, 10)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Date          False\n",
       "Country       False\n",
       "NumDay        False\n",
       "LP            False\n",
       "GDP           False\n",
       "Year          False\n",
       "Quarter       False\n",
       "Country_DE    False\n",
       "Country_ES    False\n",
       "Country_NL    False\n",
       "dtype: bool"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(df.columns)\n",
    "print(df.shape)\n",
    "pd.isna(df).any()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "dade8933",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(144, 6)\n",
      "144\n",
      "(144, 1)\n"
     ]
    }
   ],
   "source": [
    "x = df[predictors_test].values\n",
    "print(x.shape)\n",
    "y = df[target_column_test].values.reshape(-1,1)\n",
    "print(len(y))\n",
    "print(y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "243f543d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Correlation between Numday and y = 0.19\n",
      "Correlation between Year and y = 0.91\n",
      "Correlation between Quarter and y = 0.02\n"
     ]
    }
   ],
   "source": [
    "from scipy.stats import pearsonr\n",
    "print(f'Correlation between Numday and y = {pearsonr(x[:,0],y[:,0])[0]:.2f}')\n",
    "print(f'Correlation between Year and y = {pearsonr(x[:,1],y[:,0])[0]:.2f}')\n",
    "print(f'Correlation between Quarter and y = {pearsonr(x[:,2],y[:,0])[0]:.2f}')\n",
    "# Correlation between year and LP = 0.91"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "f693d386",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(144, 7)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add a bias to the input vector\n",
    "X = np.concatenate((np.ones((len(x),1)),x),axis=1)\n",
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "a1a79ce2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.00000000e+00, 6.43883110e-03, 9.88608222e-01, 9.90589401e-04,\n",
       "        4.95294700e-04, 0.00000000e+00, 0.00000000e+00],\n",
       "       [1.00000000e+00, 3.96235760e-03, 9.88608222e-01, 1.48588410e-03,\n",
       "        4.95294700e-04, 0.00000000e+00, 0.00000000e+00],\n",
       "       [1.00000000e+00, 0.00000000e+00, 9.89103517e-01, 9.90589401e-04,\n",
       "        4.95294700e-04, 0.00000000e+00, 0.00000000e+00],\n",
       "       [1.00000000e+00, 9.41059931e-03, 9.89103517e-01, 1.48588410e-03,\n",
       "        4.95294700e-04, 0.00000000e+00, 0.00000000e+00],\n",
       "       [1.00000000e+00, 3.96235760e-03, 9.89598811e-01, 9.90589401e-04,\n",
       "        4.95294700e-04, 0.00000000e+00, 0.00000000e+00],\n",
       "       [1.00000000e+00, 8.42000991e-03, 9.89598811e-01, 1.48588410e-03,\n",
       "        4.95294700e-04, 0.00000000e+00, 0.00000000e+00],\n",
       "       [1.00000000e+00, 0.00000000e+00, 9.90094106e-01, 9.90589401e-04,\n",
       "        4.95294700e-04, 0.00000000e+00, 0.00000000e+00],\n",
       "       [1.00000000e+00, 1.78306092e-02, 9.90094106e-01, 1.48588410e-03,\n",
       "        4.95294700e-04, 0.00000000e+00, 0.00000000e+00],\n",
       "       [1.00000000e+00, 8.42000991e-03, 9.90589401e-01, 9.90589401e-04,\n",
       "        4.95294700e-04, 0.00000000e+00, 0.00000000e+00],\n",
       "       [1.00000000e+00, 0.00000000e+00, 9.90589401e-01, 1.48588410e-03,\n",
       "        4.95294700e-04, 0.00000000e+00, 0.00000000e+00]])"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# data normalization\n",
    "Xnorm = X.copy()\n",
    "minx = np.min(X[:,1:])\n",
    "maxx = np.max(X[:,1:])\n",
    "Xnorm[:,1:] = (X[:,1:]-minx)/(maxx-minx)\n",
    "Xnorm[:10,:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "040ceac9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(144, 1)\n"
     ]
    }
   ],
   "source": [
    "ynorm = y.copy()\n",
    "maxy = np.max(y)\n",
    "miny = np.min(y)\n",
    "ynorm = (y-miny)/(maxy - miny) \n",
    "ynorm[:10,0]\n",
    "print(ynorm.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "e5877e16",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fad439f1460>]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Initial estimate of parameters\n",
    "theta0 = np.zeros((X.shape[1],1))+0.4\n",
    "#theta0 = np.array([[0],[0.5],[2],[0.5]])\n",
    "\n",
    "ypred = Xnorm.dot(theta0)\n",
    "\n",
    "sortidx = np.argsort(ynorm[:,0]) # sort the values for better visualization\n",
    "plt.plot(ynorm[sortidx,0],'o')\n",
    "plt.plot(ypred[sortidx,0],'--')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "e3218781",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[2.04492907e-01],\n",
       "       [1.52270031e-03],\n",
       "       [2.02554561e-01],\n",
       "       [2.52363656e-04],\n",
       "       [2.24915133e-05],\n",
       "       [4.19925144e-05],\n",
       "       [3.68002253e-05]])"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# calculate gradient\n",
    "m=144\n",
    "def grad(theta):\n",
    "    dJ = 1/m*np.sum((Xnorm.dot(theta)-ynorm)*Xnorm,axis=0).reshape(-1,1)\n",
    "    return dJ\n",
    "grad(theta0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "7c0cece3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "14.824883386088173"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def cost(theta):\n",
    "    J = np.sum((Xnorm.dot(theta)-ynorm)**2,axis=0)[0]\n",
    "    return J\n",
    "cost(theta0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "bd014195",
   "metadata": {},
   "outputs": [],
   "source": [
    "def GD(theta0,learning_rate = 0.5,epochs=1000,TOL=1e-7):\n",
    "    \n",
    "    theta_history = [theta0]\n",
    "    J_history = [cost(theta0)]\n",
    "    \n",
    "    thetanew = theta0*10000\n",
    "    print(f'epoch \\t Cost(J) \\t')\n",
    "    for epoch in range(epochs):\n",
    "        if epoch%100 == 0:\n",
    "            print(f'{epoch:5d}\\t{J_history[-1]:7.4f}\\t')\n",
    "        dJ = grad(theta0)\n",
    "        J = cost(theta0)\n",
    "        \n",
    "        thetanew = theta0 - learning_rate*dJ\n",
    "        theta_history.append(thetanew)\n",
    "        J_history.append(J)\n",
    "        \n",
    "        if np.sum((thetanew - theta0)**2) < TOL:\n",
    "            print('Convergence achieved.')\n",
    "            break\n",
    "        theta0 = thetanew\n",
    "\n",
    "    return thetanew,theta_history,J_history"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "07f31ba7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch \t Cost(J) \t\n",
      "    0\t14.8249\t\n",
      "  100\t 8.8202\t\n",
      "  200\t 8.8144\t\n",
      "  300\t 8.8085\t\n",
      "  400\t 8.8027\t\n",
      "  500\t 8.7968\t\n",
      "  600\t 8.7910\t\n",
      "  700\t 8.7852\t\n",
      "  800\t 8.7794\t\n",
      "  900\t 8.7736\t\n",
      "Convergence achieved.\n"
     ]
    }
   ],
   "source": [
    "theta,theta_history,J_history = GD(theta0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "298bee4a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fad463ec9a0>]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(J_history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "8e601a00",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fad4d1720a0>]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "yprednorm = Xnorm.dot(theta)\n",
    "\n",
    "ypred = yprednorm*(maxy-miny) + miny\n",
    "plt.plot(y[sortidx,0],'o')\n",
    "plt.plot(ypred[sortidx,0],'--')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}