main.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Parkinson's Disease Progression Prediction\n",
    "### Project - SI 618: Data Manipulation and Analysis\n",
    "\n",
    "Authors\n",
    "1. Nowrin Mohamed - nowrin@umich.edu\n",
    "2. Prithvijit Dasgupta - prithvid@umich.edu\n",
    "3. Sachin Salim - sachinks@umich.edu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''\n",
    "Kaggle Competition:\n",
    "@misc{amp-parkinsons-disease-progression-prediction,\n",
    "    author = {Leslie Kirsch, Sohier Dane, Stacey Adam, Victoria Dardov},\n",
    "    title = {AMP®-Parkinson's Disease Progression Prediction},\n",
    "    publisher = {Kaggle},\n",
    "    year = {2023},\n",
    "    url = {https://kaggle.com/competitions/amp-parkinsons-disease-progression-prediction}\n",
    "}\n",
    "'''\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import os\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from pycaret.regression import *\n",
    "\n",
    "%matplotlib inline\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "supplemental_clinical_data = pd.read_csv('data/supplemental_clinical_data.csv')\n",
    "peptides_data = pd.read_csv('data/train_peptides.csv')\n",
    "clinical_data = pd.read_csv('data/train_clinical_data.csv')\n",
    "proteins_data = pd.read_csv('data/train_proteins.csv')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_processed_data(filename):\n",
    "    df_loaded = pd.read_csv(f'data/{filename}')\n",
    "    df_loaded.set_index(\n",
    "        ['visit_id', 'patient_id', 'visit_month'], inplace=True)\n",
    "    return df_loaded\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature extraction"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Data transformation**:<br>\n",
    "The protein data was pivoted to have each protein as a column, resulting in 7.88% of missing values due to not all visits having measurements for all 227 proteins. The same transformation was applied to the peptide data, resulting in 8.87% of missing values.\n",
    "\n",
    "\n",
    "**Filling nan**:<br>\n",
    "To fill the missing protein values, the data was first grouped by patient ID and the missing values were filled with the mean data for each patient. This approach filled around 70% of the missing values. For the remaining missing values, the data was grouped by visit month and filled with the mean data for each month. Finally, any remaining missing values were filled with the mean data for each protein across the entire dataset. The same approach was applied to fill the missing peptide values.\n",
    "\n",
    "**Matching records**:<br>\n",
    "For training purposes, the indices that were common between the protein/peptide data and clinical data were identified, resulting in 1068 matching records. The data was merged based on these indices."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Handling missing data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fill_na(data):\n",
    "    data_1 = data.groupby('patient_id')\\\n",
    "        .transform(lambda x: x.fillna(x.mean()))\n",
    "    data_2 = data_1.groupby('visit_month')\\\n",
    "        .transform(lambda x: x.fillna(x.mean()))\n",
    "    data_3 = data_2.fillna(data_2.mean())\n",
    "    return data_3\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load processed features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def pivot_proteins():\n",
    "    return proteins_data.pivot(\n",
    "        index=[\n",
    "            'visit_id',\n",
    "            'patient_id',\n",
    "            'visit_month'],\n",
    "        columns='UniProt',\n",
    "        values='NPX')\n",
    "\n",
    "\n",
    "def pivot_peptides():\n",
    "    return peptides_data.pivot(\n",
    "        index=[\n",
    "            'visit_id',\n",
    "            'patient_id',\n",
    "            'visit_month'],\n",
    "        columns='Peptide',\n",
    "        values='PeptideAbundance')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_features(type):\n",
    "    '''type: proteins/peptides'''\n",
    "    filename = f'{type}_features.csv'\n",
    "    if os.path.isfile(f'data/{filename}'):\n",
    "        features = read_processed_data(filename)\n",
    "    else:\n",
    "        if type == 'proteins':\n",
    "            features = pivot_proteins()\n",
    "        elif type == 'peptides':\n",
    "            features = pivot_peptides()\n",
    "\n",
    "        features = fill_na(features)\n",
    "\n",
    "        features.to_csv(f'data/{filename}')\n",
    "\n",
    "    return features\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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></th>\n",
       "      <th></th>\n",
       "      <th>O00391</th>\n",
       "      <th>O00533</th>\n",
       "      <th>O00584</th>\n",
       "      <th>O14498</th>\n",
       "      <th>O14773</th>\n",
       "      <th>O14791</th>\n",
       "      <th>O15240</th>\n",
       "      <th>O15394</th>\n",
       "      <th>O43505</th>\n",
       "      <th>O60888</th>\n",
       "      <th>...</th>\n",
       "      <th>Q9HDC9</th>\n",
       "      <th>Q9NQ79</th>\n",
       "      <th>Q9NYU2</th>\n",
       "      <th>Q9UBR2</th>\n",
       "      <th>Q9UBX5</th>\n",
       "      <th>Q9UHG2</th>\n",
       "      <th>Q9UKV8</th>\n",
       "      <th>Q9UNU6</th>\n",
       "      <th>Q9Y646</th>\n",
       "      <th>Q9Y6R7</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>visit_id</th>\n",
       "      <th>patient_id</th>\n",
       "      <th>visit_month</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10053_0</th>\n",
       "      <th>10053</th>\n",
       "      <th>0</th>\n",
       "      <td>9104.27</td>\n",
       "      <td>402321.0</td>\n",
       "      <td>7126.96</td>\n",
       "      <td>24525.7</td>\n",
       "      <td>7150.57</td>\n",
       "      <td>2497.840</td>\n",
       "      <td>83002.9</td>\n",
       "      <td>15113.6</td>\n",
       "      <td>167327.0</td>\n",
       "      <td>129048.0</td>\n",
       "      <td>...</td>\n",
       "      <td>317477.0</td>\n",
       "      <td>9469.45</td>\n",
       "      <td>94237.6</td>\n",
       "      <td>14907.159185</td>\n",
       "      <td>23016.0</td>\n",
       "      <td>177983.0</td>\n",
       "      <td>65900.0</td>\n",
       "      <td>15382.00</td>\n",
       "      <td>9295.65</td>\n",
       "      <td>19017.40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10053_12</th>\n",
       "      <th>10053</th>\n",
       "      <th>12</th>\n",
       "      <td>10464.20</td>\n",
       "      <td>435586.0</td>\n",
       "      <td>7126.96</td>\n",
       "      <td>24525.7</td>\n",
       "      <td>7150.57</td>\n",
       "      <td>2435.275</td>\n",
       "      <td>197117.0</td>\n",
       "      <td>15099.1</td>\n",
       "      <td>164268.0</td>\n",
       "      <td>108114.0</td>\n",
       "      <td>...</td>\n",
       "      <td>317477.0</td>\n",
       "      <td>14408.40</td>\n",
       "      <td>102672.3</td>\n",
       "      <td>14659.712042</td>\n",
       "      <td>28537.0</td>\n",
       "      <td>171733.0</td>\n",
       "      <td>65668.1</td>\n",
       "      <td>13097.65</td>\n",
       "      <td>9295.65</td>\n",
       "      <td>25697.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10053_18</th>\n",
       "      <th>10053</th>\n",
       "      <th>18</th>\n",
       "      <td>13235.70</td>\n",
       "      <td>507386.0</td>\n",
       "      <td>7126.96</td>\n",
       "      <td>24525.7</td>\n",
       "      <td>7150.57</td>\n",
       "      <td>2372.710</td>\n",
       "      <td>126506.0</td>\n",
       "      <td>16289.6</td>\n",
       "      <td>168107.0</td>\n",
       "      <td>163776.0</td>\n",
       "      <td>...</td>\n",
       "      <td>317477.0</td>\n",
       "      <td>38667.20</td>\n",
       "      <td>111107.0</td>\n",
       "      <td>17048.833333</td>\n",
       "      <td>37932.6</td>\n",
       "      <td>245188.0</td>\n",
       "      <td>59986.1</td>\n",
       "      <td>10813.30</td>\n",
       "      <td>9295.65</td>\n",
       "      <td>29102.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10138_12</th>\n",
       "      <th>10138</th>\n",
       "      <th>12</th>\n",
       "      <td>12600.20</td>\n",
       "      <td>494581.0</td>\n",
       "      <td>9165.06</td>\n",
       "      <td>27193.5</td>\n",
       "      <td>22506.10</td>\n",
       "      <td>6015.900</td>\n",
       "      <td>156313.0</td>\n",
       "      <td>54546.4</td>\n",
       "      <td>204013.0</td>\n",
       "      <td>56725.0</td>\n",
       "      <td>...</td>\n",
       "      <td>557904.0</td>\n",
       "      <td>44556.90</td>\n",
       "      <td>155619.0</td>\n",
       "      <td>14647.900000</td>\n",
       "      <td>36927.7</td>\n",
       "      <td>229232.0</td>\n",
       "      <td>106564.0</td>\n",
       "      <td>26077.70</td>\n",
       "      <td>21441.80</td>\n",
       "      <td>7642.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10138_24</th>\n",
       "      <th>10138</th>\n",
       "      <th>24</th>\n",
       "      <td>12003.20</td>\n",
       "      <td>522138.0</td>\n",
       "      <td>4498.51</td>\n",
       "      <td>17189.8</td>\n",
       "      <td>29112.40</td>\n",
       "      <td>2665.150</td>\n",
       "      <td>151169.0</td>\n",
       "      <td>52338.1</td>\n",
       "      <td>240892.0</td>\n",
       "      <td>85767.1</td>\n",
       "      <td>...</td>\n",
       "      <td>495006.0</td>\n",
       "      <td>47836.70</td>\n",
       "      <td>177619.0</td>\n",
       "      <td>17061.100000</td>\n",
       "      <td>25510.4</td>\n",
       "      <td>176722.0</td>\n",
       "      <td>59471.4</td>\n",
       "      <td>12639.20</td>\n",
       "      <td>15091.40</td>\n",
       "      <td>6168.55</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 227 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   O00391    O00533   O00584   O14498  \\\n",
       "visit_id patient_id visit_month                                         \n",
       "10053_0  10053      0             9104.27  402321.0  7126.96  24525.7   \n",
       "10053_12 10053      12           10464.20  435586.0  7126.96  24525.7   \n",
       "10053_18 10053      18           13235.70  507386.0  7126.96  24525.7   \n",
       "10138_12 10138      12           12600.20  494581.0  9165.06  27193.5   \n",
       "10138_24 10138      24           12003.20  522138.0  4498.51  17189.8   \n",
       "\n",
       "                                   O14773    O14791    O15240   O15394  \\\n",
       "visit_id patient_id visit_month                                          \n",
       "10053_0  10053      0             7150.57  2497.840   83002.9  15113.6   \n",
       "10053_12 10053      12            7150.57  2435.275  197117.0  15099.1   \n",
       "10053_18 10053      18            7150.57  2372.710  126506.0  16289.6   \n",
       "10138_12 10138      12           22506.10  6015.900  156313.0  54546.4   \n",
       "10138_24 10138      24           29112.40  2665.150  151169.0  52338.1   \n",
       "\n",
       "                                   O43505    O60888  ...    Q9HDC9    Q9NQ79  \\\n",
       "visit_id patient_id visit_month                      ...                       \n",
       "10053_0  10053      0            167327.0  129048.0  ...  317477.0   9469.45   \n",
       "10053_12 10053      12           164268.0  108114.0  ...  317477.0  14408.40   \n",
       "10053_18 10053      18           168107.0  163776.0  ...  317477.0  38667.20   \n",
       "10138_12 10138      12           204013.0   56725.0  ...  557904.0  44556.90   \n",
       "10138_24 10138      24           240892.0   85767.1  ...  495006.0  47836.70   \n",
       "\n",
       "                                   Q9NYU2        Q9UBR2   Q9UBX5    Q9UHG2  \\\n",
       "visit_id patient_id visit_month                                              \n",
       "10053_0  10053      0             94237.6  14907.159185  23016.0  177983.0   \n",
       "10053_12 10053      12           102672.3  14659.712042  28537.0  171733.0   \n",
       "10053_18 10053      18           111107.0  17048.833333  37932.6  245188.0   \n",
       "10138_12 10138      12           155619.0  14647.900000  36927.7  229232.0   \n",
       "10138_24 10138      24           177619.0  17061.100000  25510.4  176722.0   \n",
       "\n",
       "                                   Q9UKV8    Q9UNU6    Q9Y646    Q9Y6R7  \n",
       "visit_id patient_id visit_month                                          \n",
       "10053_0  10053      0             65900.0  15382.00   9295.65  19017.40  \n",
       "10053_12 10053      12            65668.1  13097.65   9295.65  25697.80  \n",
       "10053_18 10053      18            59986.1  10813.30   9295.65  29102.70  \n",
       "10138_12 10138      12           106564.0  26077.70  21441.80   7642.42  \n",
       "10138_24 10138      24            59471.4  12639.20  15091.40   6168.55  \n",
       "\n",
       "[5 rows x 227 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "proteins_features = load_features('proteins')\n",
    "proteins_features.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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></th>\n",
       "      <th></th>\n",
       "      <th>AADDTWEPFASGK</th>\n",
       "      <th>AAFGQGSGPIMLDEVQC(UniMod_4)TGTEASLADC(UniMod_4)K</th>\n",
       "      <th>AAFTEC(UniMod_4)C(UniMod_4)QAADK</th>\n",
       "      <th>AANEVSSADVK</th>\n",
       "      <th>AATGEC(UniMod_4)TATVGKR</th>\n",
       "      <th>AATVGSLAGQPLQER</th>\n",
       "      <th>AAVYHHFISDGVR</th>\n",
       "      <th>ADDKETC(UniMod_4)FAEEGK</th>\n",
       "      <th>ADDKETC(UniMod_4)FAEEGKK</th>\n",
       "      <th>ADDLGKGGNEESTKTGNAGSR</th>\n",
       "      <th>...</th>\n",
       "      <th>YSLTYIYTGLSK</th>\n",
       "      <th>YTTEIIK</th>\n",
       "      <th>YVGGQEHFAHLLILR</th>\n",
       "      <th>YVM(UniMod_35)LPVADQDQC(UniMod_4)IR</th>\n",
       "      <th>YVMLPVADQDQC(UniMod_4)IR</th>\n",
       "      <th>YVNKEIQNAVNGVK</th>\n",
       "      <th>YWGVASFLQK</th>\n",
       "      <th>YYC(UniMod_4)FQGNQFLR</th>\n",
       "      <th>YYTYLIMNK</th>\n",
       "      <th>YYWGGQYTWDMAK</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>visit_id</th>\n",
       "      <th>patient_id</th>\n",
       "      <th>visit_month</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10053_0</th>\n",
       "      <th>10053</th>\n",
       "      <th>0</th>\n",
       "      <td>6580710.0</td>\n",
       "      <td>31204.4</td>\n",
       "      <td>7735070.0</td>\n",
       "      <td>35984.7</td>\n",
       "      <td>17188.0</td>\n",
       "      <td>19787.3</td>\n",
       "      <td>46620.3</td>\n",
       "      <td>236144.0</td>\n",
       "      <td>2556837.7</td>\n",
       "      <td>30838.2</td>\n",
       "      <td>...</td>\n",
       "      <td>202274.0</td>\n",
       "      <td>8926.159054</td>\n",
       "      <td>4401830.0</td>\n",
       "      <td>77482.6</td>\n",
       "      <td>583075.0</td>\n",
       "      <td>76705.7</td>\n",
       "      <td>104260.0</td>\n",
       "      <td>530223.0</td>\n",
       "      <td>50835.5</td>\n",
       "      <td>7207.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10053_12</th>\n",
       "      <th>10053</th>\n",
       "      <th>12</th>\n",
       "      <td>6333510.0</td>\n",
       "      <td>52277.6</td>\n",
       "      <td>5394390.0</td>\n",
       "      <td>35984.7</td>\n",
       "      <td>17188.0</td>\n",
       "      <td>19787.3</td>\n",
       "      <td>57554.5</td>\n",
       "      <td>108298.0</td>\n",
       "      <td>45885.4</td>\n",
       "      <td>30838.2</td>\n",
       "      <td>...</td>\n",
       "      <td>201009.0</td>\n",
       "      <td>8682.690196</td>\n",
       "      <td>5001750.0</td>\n",
       "      <td>36745.3</td>\n",
       "      <td>355643.0</td>\n",
       "      <td>92078.1</td>\n",
       "      <td>123254.0</td>\n",
       "      <td>453883.0</td>\n",
       "      <td>49281.9</td>\n",
       "      <td>25332.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10053_18</th>\n",
       "      <th>10053</th>\n",
       "      <th>18</th>\n",
       "      <td>7129640.0</td>\n",
       "      <td>61522.0</td>\n",
       "      <td>7011920.0</td>\n",
       "      <td>35984.7</td>\n",
       "      <td>17188.0</td>\n",
       "      <td>19787.3</td>\n",
       "      <td>36029.4</td>\n",
       "      <td>708729.0</td>\n",
       "      <td>5067790.0</td>\n",
       "      <td>30838.2</td>\n",
       "      <td>...</td>\n",
       "      <td>220728.0</td>\n",
       "      <td>10192.130714</td>\n",
       "      <td>5424380.0</td>\n",
       "      <td>39016.0</td>\n",
       "      <td>496021.0</td>\n",
       "      <td>63203.6</td>\n",
       "      <td>128336.0</td>\n",
       "      <td>447505.0</td>\n",
       "      <td>52389.1</td>\n",
       "      <td>21235.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10138_12</th>\n",
       "      <th>10138</th>\n",
       "      <th>12</th>\n",
       "      <td>7404780.0</td>\n",
       "      <td>46107.2</td>\n",
       "      <td>10610900.0</td>\n",
       "      <td>20093.2</td>\n",
       "      <td>20910.2</td>\n",
       "      <td>66662.3</td>\n",
       "      <td>55253.9</td>\n",
       "      <td>79575.5</td>\n",
       "      <td>6201210.0</td>\n",
       "      <td>26720.0</td>\n",
       "      <td>...</td>\n",
       "      <td>188362.0</td>\n",
       "      <td>9433.710000</td>\n",
       "      <td>3900280.0</td>\n",
       "      <td>48210.3</td>\n",
       "      <td>328482.0</td>\n",
       "      <td>89822.1</td>\n",
       "      <td>129964.0</td>\n",
       "      <td>552232.0</td>\n",
       "      <td>65657.8</td>\n",
       "      <td>9876.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10138_24</th>\n",
       "      <th>10138</th>\n",
       "      <th>24</th>\n",
       "      <td>13788300.0</td>\n",
       "      <td>56910.3</td>\n",
       "      <td>6906160.0</td>\n",
       "      <td>13785.5</td>\n",
       "      <td>11004.2</td>\n",
       "      <td>63672.7</td>\n",
       "      <td>36819.8</td>\n",
       "      <td>34160.9</td>\n",
       "      <td>2117430.0</td>\n",
       "      <td>15645.2</td>\n",
       "      <td>...</td>\n",
       "      <td>206187.0</td>\n",
       "      <td>6365.150000</td>\n",
       "      <td>3521800.0</td>\n",
       "      <td>69984.6</td>\n",
       "      <td>496737.0</td>\n",
       "      <td>80919.3</td>\n",
       "      <td>111799.0</td>\n",
       "      <td>568923.5</td>\n",
       "      <td>56977.6</td>\n",
       "      <td>4903.09</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 968 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 AADDTWEPFASGK  \\\n",
       "visit_id patient_id visit_month                  \n",
       "10053_0  10053      0                6580710.0   \n",
       "10053_12 10053      12               6333510.0   \n",
       "10053_18 10053      18               7129640.0   \n",
       "10138_12 10138      12               7404780.0   \n",
       "10138_24 10138      24              13788300.0   \n",
       "\n",
       "                                 AAFGQGSGPIMLDEVQC(UniMod_4)TGTEASLADC(UniMod_4)K  \\\n",
       "visit_id patient_id visit_month                                                     \n",
       "10053_0  10053      0                                                     31204.4   \n",
       "10053_12 10053      12                                                    52277.6   \n",
       "10053_18 10053      18                                                    61522.0   \n",
       "10138_12 10138      12                                                    46107.2   \n",
       "10138_24 10138      24                                                    56910.3   \n",
       "\n",
       "                                 AAFTEC(UniMod_4)C(UniMod_4)QAADK  \\\n",
       "visit_id patient_id visit_month                                     \n",
       "10053_0  10053      0                                   7735070.0   \n",
       "10053_12 10053      12                                  5394390.0   \n",
       "10053_18 10053      18                                  7011920.0   \n",
       "10138_12 10138      12                                 10610900.0   \n",
       "10138_24 10138      24                                  6906160.0   \n",
       "\n",
       "                                 AANEVSSADVK  AATGEC(UniMod_4)TATVGKR  \\\n",
       "visit_id patient_id visit_month                                         \n",
       "10053_0  10053      0                35984.7                  17188.0   \n",
       "10053_12 10053      12               35984.7                  17188.0   \n",
       "10053_18 10053      18               35984.7                  17188.0   \n",
       "10138_12 10138      12               20093.2                  20910.2   \n",
       "10138_24 10138      24               13785.5                  11004.2   \n",
       "\n",
       "                                 AATVGSLAGQPLQER  AAVYHHFISDGVR  \\\n",
       "visit_id patient_id visit_month                                   \n",
       "10053_0  10053      0                    19787.3        46620.3   \n",
       "10053_12 10053      12                   19787.3        57554.5   \n",
       "10053_18 10053      18                   19787.3        36029.4   \n",
       "10138_12 10138      12                   66662.3        55253.9   \n",
       "10138_24 10138      24                   63672.7        36819.8   \n",
       "\n",
       "                                 ADDKETC(UniMod_4)FAEEGK  \\\n",
       "visit_id patient_id visit_month                            \n",
       "10053_0  10053      0                           236144.0   \n",
       "10053_12 10053      12                          108298.0   \n",
       "10053_18 10053      18                          708729.0   \n",
       "10138_12 10138      12                           79575.5   \n",
       "10138_24 10138      24                           34160.9   \n",
       "\n",
       "                                 ADDKETC(UniMod_4)FAEEGKK  \\\n",
       "visit_id patient_id visit_month                             \n",
       "10053_0  10053      0                           2556837.7   \n",
       "10053_12 10053      12                            45885.4   \n",
       "10053_18 10053      18                          5067790.0   \n",
       "10138_12 10138      12                          6201210.0   \n",
       "10138_24 10138      24                          2117430.0   \n",
       "\n",
       "                                 ADDLGKGGNEESTKTGNAGSR  ...  YSLTYIYTGLSK  \\\n",
       "visit_id patient_id visit_month                         ...                 \n",
       "10053_0  10053      0                          30838.2  ...      202274.0   \n",
       "10053_12 10053      12                         30838.2  ...      201009.0   \n",
       "10053_18 10053      18                         30838.2  ...      220728.0   \n",
       "10138_12 10138      12                         26720.0  ...      188362.0   \n",
       "10138_24 10138      24                         15645.2  ...      206187.0   \n",
       "\n",
       "                                      YTTEIIK  YVGGQEHFAHLLILR  \\\n",
       "visit_id patient_id visit_month                                  \n",
       "10053_0  10053      0             8926.159054        4401830.0   \n",
       "10053_12 10053      12            8682.690196        5001750.0   \n",
       "10053_18 10053      18           10192.130714        5424380.0   \n",
       "10138_12 10138      12            9433.710000        3900280.0   \n",
       "10138_24 10138      24            6365.150000        3521800.0   \n",
       "\n",
       "                                 YVM(UniMod_35)LPVADQDQC(UniMod_4)IR  \\\n",
       "visit_id patient_id visit_month                                        \n",
       "10053_0  10053      0                                        77482.6   \n",
       "10053_12 10053      12                                       36745.3   \n",
       "10053_18 10053      18                                       39016.0   \n",
       "10138_12 10138      12                                       48210.3   \n",
       "10138_24 10138      24                                       69984.6   \n",
       "\n",
       "                                 YVMLPVADQDQC(UniMod_4)IR  YVNKEIQNAVNGVK  \\\n",
       "visit_id patient_id visit_month                                             \n",
       "10053_0  10053      0                            583075.0         76705.7   \n",
       "10053_12 10053      12                           355643.0         92078.1   \n",
       "10053_18 10053      18                           496021.0         63203.6   \n",
       "10138_12 10138      12                           328482.0         89822.1   \n",
       "10138_24 10138      24                           496737.0         80919.3   \n",
       "\n",
       "                                 YWGVASFLQK  YYC(UniMod_4)FQGNQFLR  YYTYLIMNK  \\\n",
       "visit_id patient_id visit_month                                                 \n",
       "10053_0  10053      0              104260.0               530223.0    50835.5   \n",
       "10053_12 10053      12             123254.0               453883.0    49281.9   \n",
       "10053_18 10053      18             128336.0               447505.0    52389.1   \n",
       "10138_12 10138      12             129964.0               552232.0    65657.8   \n",
       "10138_24 10138      24             111799.0               568923.5    56977.6   \n",
       "\n",
       "                                 YYWGGQYTWDMAK  \n",
       "visit_id patient_id visit_month                 \n",
       "10053_0  10053      0                  7207.30  \n",
       "10053_12 10053      12                25332.80  \n",
       "10053_18 10053      18                21235.70  \n",
       "10138_12 10138      12                 9876.98  \n",
       "10138_24 10138      24                 4903.09  \n",
       "\n",
       "[5 rows x 968 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "peptides_features = load_features('peptides')\n",
    "peptides_features.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2615, 4)\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></th>\n",
       "      <th></th>\n",
       "      <th>updrs_1</th>\n",
       "      <th>updrs_2</th>\n",
       "      <th>updrs_3</th>\n",
       "      <th>updrs_4</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>visit_id</th>\n",
       "      <th>patient_id</th>\n",
       "      <th>visit_month</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>55_0</th>\n",
       "      <th>55</th>\n",
       "      <th>0</th>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55_3</th>\n",
       "      <th>55</th>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>19</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55_6</th>\n",
       "      <th>55</th>\n",
       "      <th>6</th>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "      <td>58</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55_9</th>\n",
       "      <th>55</th>\n",
       "      <th>9</th>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55_12</th>\n",
       "      <th>55</th>\n",
       "      <th>12</th>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 updrs_1  updrs_2  updrs_3  updrs_4\n",
       "visit_id patient_id visit_month                                    \n",
       "55_0     55         0                 10        6        6        0\n",
       "55_3     55         3                 10        7       19        0\n",
       "55_6     55         6                  8       10       58        0\n",
       "55_9     55         9                  8        9       17        0\n",
       "55_12    55         12                10       10        2        0"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clinical_features = clinical_data.drop(\n",
    "    columns=[\"upd23b_clinical_state_on_medication\"]).copy()\n",
    "clinical_features.set_index(\n",
    "    ['visit_id', 'patient_id', 'visit_month'], inplace=True)\n",
    "\n",
    "# nan mask\n",
    "clinical_nan_mask = clinical_features.isna()\n",
    "\n",
    "clinical_features = fill_na(clinical_features)\n",
    "clinical_features = clinical_features.astype(\n",
    "    {'updrs_1': 'int64', 'updrs_2': 'int64', 'updrs_3': 'int64', 'updrs_4': 'int64'})\n",
    "print(clinical_features.shape)\n",
    "clinical_features.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1068 visits are common between input and output features\n"
     ]
    }
   ],
   "source": [
    "common_indices = np.intersect1d(peptides_features.index.values,\n",
    "                                clinical_features.index.values)\n",
    "print(f\"{len(common_indices)} visits are common between input\\\n",
    " and output features\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def smape(a, b):\n",
    "    s = pd.Series(2*abs(a-b)/(abs(a)+abs(b))).fillna(0)\n",
    "    return 100*s.sum()/len(s)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "clinical_data.set_index('visit_id', inplace=True)\n",
    "npx_features = pd.read_csv(\n",
    "    'data/proteins_features.csv').set_index('visit_id')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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>visit_month</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>patient_id</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>942</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1517</th>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3636</th>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3863</th>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62792</th>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63889</th>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64669</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64674</th>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65043</th>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>155 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            visit_month\n",
       "patient_id             \n",
       "55                   10\n",
       "942                   3\n",
       "1517                 10\n",
       "3636                 10\n",
       "3863                  8\n",
       "...                 ...\n",
       "62792                 9\n",
       "63889                 9\n",
       "64669                12\n",
       "64674                 6\n",
       "65043                11\n",
       "\n",
       "[155 rows x 1 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "c = clinical_data.dropna().groupby('patient_id').count()[['visit_month']]\n",
    "c\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n = npx_features.groupby('patient_id').count()[['visit_month']]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "li = []\n",
    "for idx, v in enumerate(clinical_data.index.values):\n",
    "    parts = v.split('_')\n",
    "    if parts[1] == '0':\n",
    "        li.append('00000')\n",
    "    else:\n",
    "        li.append(clinical_data.index.values[idx-1])\n",
    "\n",
    "\n",
    "def offset_updrs2(row):\n",
    "    if (row['offset'] == '00000'):\n",
    "        return 0\n",
    "    else:\n",
    "        return clinical_data.loc[row['offset']]['updrs_2']\n",
    "\n",
    "\n",
    "def offset_updrs1(row):\n",
    "    if (row['offset'] == '00000'):\n",
    "        return 0\n",
    "    else:\n",
    "        return clinical_data.loc[row['offset']]['updrs_1']\n",
    "\n",
    "\n",
    "def offset_updrs3(row):\n",
    "    if (row['offset'] == '00000'):\n",
    "        return 0\n",
    "    else:\n",
    "        return clinical_data.loc[row['offset']]['updrs_3']\n",
    "\n",
    "\n",
    "def offset_updrs4(row):\n",
    "    if (row['offset'] == '00000'):\n",
    "        return 0\n",
    "    else:\n",
    "        return clinical_data.loc[row['offset']]['updrs_4']\n",
    "\n",
    "\n",
    "clinical_data['offset'] = li\n",
    "\n",
    "clinical_data['offset_updrs2_offset'] = clinical_data.apply(\n",
    "    offset_updrs2, axis=1)\n",
    "clinical_data['offset_updrs1_offset'] = clinical_data.apply(\n",
    "    offset_updrs1, axis=1)\n",
    "clinical_data['offset_updrs3_offset'] = clinical_data.apply(\n",
    "    offset_updrs3, axis=1)\n",
    "clinical_data['offset_updrs4_offset'] = clinical_data.apply(\n",
    "    offset_updrs4, axis=1)\n",
    "\n",
    "\n",
    "clinical_data.fillna(0, inplace=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "clinical_data = clinical_data[['updrs_1',\n",
    "                               'updrs_2',\n",
    "                               'offset_updrs2_offset',\n",
    "                               'offset_updrs1_offset',\n",
    "                               'offset_updrs3_offset',\n",
    "                               'offset_updrs4_offset',\n",
    "                               'updrs_3',\n",
    "                               'updrs_4',\n",
    "                               'visit_month',\n",
    "                               'patient_id']]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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>visit_month_c</th>\n",
       "      <th>visit_month_n</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>patient_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>18183</th>\n",
       "      <td>9</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19088</th>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24690</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468</th>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29313</th>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31270</th>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52119</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64674</th>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            visit_month_c  visit_month_n\n",
       "patient_id                              \n",
       "18183                   9              9\n",
       "19088                   6              6\n",
       "24690                   3              3\n",
       "27468                   8              8\n",
       "29313                   5              5\n",
       "31270                   4              4\n",
       "52119                   2              2\n",
       "64674                   6              6"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "t = c.join(n, lsuffix='_c', rsuffix='_n')\n",
    "t[t['visit_month_c'] == t['visit_month_n']]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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>updrs_1</th>\n",
       "      <th>updrs_2</th>\n",
       "      <th>offset_updrs2_offset</th>\n",
       "      <th>offset_updrs1_offset</th>\n",
       "      <th>offset_updrs3_offset</th>\n",
       "      <th>offset_updrs4_offset</th>\n",
       "      <th>updrs_3</th>\n",
       "      <th>updrs_4</th>\n",
       "      <th>visit_month</th>\n",
       "      <th>patient_id</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>visit_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>23391_0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>19</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_6</th>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>19</td>\n",
       "      <td>0.0</td>\n",
       "      <td>58</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_9</th>\n",
       "      <td>2.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>58</td>\n",
       "      <td>0.0</td>\n",
       "      <td>17</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_12</th>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>17</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_18</th>\n",
       "      <td>4.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>18</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_24</th>\n",
       "      <td>4.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_30</th>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>16</td>\n",
       "      <td>0.0</td>\n",
       "      <td>21</td>\n",
       "      <td>0.0</td>\n",
       "      <td>30</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_36</th>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>21</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>36</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_42</th>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>42</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_48</th>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>51</td>\n",
       "      <td>0.0</td>\n",
       "      <td>48</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_54</th>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>51</td>\n",
       "      <td>0.0</td>\n",
       "      <td>55</td>\n",
       "      <td>0.0</td>\n",
       "      <td>54</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_60</th>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>55</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>60</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_72</th>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>58</td>\n",
       "      <td>1.0</td>\n",
       "      <td>72</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23391_84</th>\n",
       "      <td>11.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>58</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>84</td>\n",
       "      <td>23391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_0</th>\n",
       "      <td>9.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_3</th>\n",
       "      <td>12.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>19</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_6</th>\n",
       "      <td>11.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>19</td>\n",
       "      <td>0.0</td>\n",
       "      <td>58</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_9</th>\n",
       "      <td>11.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>58</td>\n",
       "      <td>0.0</td>\n",
       "      <td>17</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_12</th>\n",
       "      <td>12.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>17</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_18</th>\n",
       "      <td>10.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>18</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_24</th>\n",
       "      <td>10.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_30</th>\n",
       "      <td>10.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>16</td>\n",
       "      <td>0.0</td>\n",
       "      <td>21</td>\n",
       "      <td>0.0</td>\n",
       "      <td>30</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_36</th>\n",
       "      <td>16.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>21</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>36</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_42</th>\n",
       "      <td>14.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>42</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_48</th>\n",
       "      <td>15.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>51</td>\n",
       "      <td>3.0</td>\n",
       "      <td>48</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_54</th>\n",
       "      <td>17.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>51</td>\n",
       "      <td>3.0</td>\n",
       "      <td>55</td>\n",
       "      <td>3.0</td>\n",
       "      <td>54</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_60</th>\n",
       "      <td>21.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>55</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>60</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_72</th>\n",
       "      <td>14.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>58</td>\n",
       "      <td>1.0</td>\n",
       "      <td>72</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27468_84</th>\n",
       "      <td>14.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>58</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>84</td>\n",
       "      <td>27468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>19</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_6</th>\n",
       "      <td>0.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>19</td>\n",
       "      <td>0.0</td>\n",
       "      <td>58</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_9</th>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>58</td>\n",
       "      <td>0.0</td>\n",
       "      <td>17</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_12</th>\n",
       "      <td>1.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>17</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_18</th>\n",
       "      <td>0.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>18</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_24</th>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_30</th>\n",
       "      <td>3.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>16</td>\n",
       "      <td>0.0</td>\n",
       "      <td>21</td>\n",
       "      <td>0.0</td>\n",
       "      <td>30</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_36</th>\n",
       "      <td>0.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>21</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>36</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_42</th>\n",
       "      <td>4.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>42</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_48</th>\n",
       "      <td>3.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>51</td>\n",
       "      <td>0.0</td>\n",
       "      <td>48</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_54</th>\n",
       "      <td>5.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>51</td>\n",
       "      <td>0.0</td>\n",
       "      <td>55</td>\n",
       "      <td>0.0</td>\n",
       "      <td>54</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_60</th>\n",
       "      <td>0.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>55</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>60</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_72</th>\n",
       "      <td>1.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>58</td>\n",
       "      <td>0.0</td>\n",
       "      <td>72</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_84</th>\n",
       "      <td>2.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>58</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>84</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40200_96</th>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>96</td>\n",
       "      <td>40200</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          updrs_1  updrs_2  offset_updrs2_offset  offset_updrs1_offset  \\\n",
       "visit_id                                                                 \n",
       "23391_0       0.0      3.0                   0.0                   0.0   \n",
       "23391_3       1.0      4.0                   3.0                   0.0   \n",
       "23391_6       2.0      4.0                   4.0                   1.0   \n",
       "23391_9       2.0      6.0                   4.0                   2.0   \n",
       "23391_12      4.0      5.0                   6.0                   2.0   \n",
       "23391_18      4.0      6.0                   5.0                   4.0   \n",
       "23391_24      4.0      7.0                   6.0                   4.0   \n",
       "23391_30      5.0      7.0                   7.0                   4.0   \n",
       "23391_36      6.0      7.0                   7.0                   5.0   \n",
       "23391_42      5.0      6.0                   7.0                   6.0   \n",
       "23391_48      5.0      6.0                   6.0                   5.0   \n",
       "23391_54      1.0      5.0                   6.0                   5.0   \n",
       "23391_60      4.0      5.0                   5.0                   1.0   \n",
       "23391_72      8.0      6.0                   5.0                   4.0   \n",
       "23391_84     11.0     15.0                   6.0                   8.0   \n",
       "27468_0       9.0      8.0                   0.0                   0.0   \n",
       "27468_3      12.0      9.0                   8.0                   9.0   \n",
       "27468_6      11.0     12.0                   9.0                  12.0   \n",
       "27468_9      11.0      7.0                  12.0                  11.0   \n",
       "27468_12     12.0      7.0                   7.0                  11.0   \n",
       "27468_18     10.0      8.0                   7.0                  12.0   \n",
       "27468_24     10.0     12.0                   8.0                  10.0   \n",
       "27468_30     10.0     14.0                  12.0                  10.0   \n",
       "27468_36     16.0     13.0                  14.0                  10.0   \n",
       "27468_42     14.0     11.0                  13.0                  16.0   \n",
       "27468_48     15.0     14.0                  11.0                  14.0   \n",
       "27468_54     17.0     16.0                  14.0                  15.0   \n",
       "27468_60     21.0     15.0                  16.0                  17.0   \n",
       "27468_72     14.0     14.0                  15.0                  21.0   \n",
       "27468_84     14.0     16.0                  14.0                  14.0   \n",
       "40200_0       0.0      4.0                   0.0                   0.0   \n",
       "40200_3       0.0      4.0                   4.0                   0.0   \n",
       "40200_6       0.0      6.0                   4.0                   0.0   \n",
       "40200_9       0.0      5.0                   6.0                   0.0   \n",
       "40200_12      1.0     10.0                   5.0                   0.0   \n",
       "40200_18      0.0      9.0                  10.0                   1.0   \n",
       "40200_24      2.0      3.0                   9.0                   0.0   \n",
       "40200_30      3.0      9.0                   3.0                   2.0   \n",
       "40200_36      0.0      7.0                   9.0                   3.0   \n",
       "40200_42      4.0     15.0                   7.0                   0.0   \n",
       "40200_48      3.0     11.0                  15.0                   4.0   \n",
       "40200_54      5.0     14.0                  11.0                   3.0   \n",
       "40200_60      0.0     14.0                  14.0                   5.0   \n",
       "40200_72      1.0      7.0                  14.0                   0.0   \n",
       "40200_84      2.0     10.0                   7.0                   1.0   \n",
       "40200_96      2.0     15.0                  10.0                   2.0   \n",
       "\n",
       "          offset_updrs3_offset  offset_updrs4_offset  updrs_3  updrs_4  \\\n",
       "visit_id                                                                 \n",
       "23391_0                      0                   0.0        6      0.0   \n",
       "23391_3                      6                   0.0       19      0.0   \n",
       "23391_6                     19                   0.0       58      0.0   \n",
       "23391_9                     58                   0.0       17      0.0   \n",
       "23391_12                    17                   0.0        2      0.0   \n",
       "23391_18                     2                   0.0        7      0.0   \n",
       "23391_24                     7                   0.0       16      0.0   \n",
       "23391_30                    16                   0.0       21      0.0   \n",
       "23391_36                    21                   0.0        9      0.0   \n",
       "23391_42                     9                   0.0        4      0.0   \n",
       "23391_48                     4                   0.0       51      0.0   \n",
       "23391_54                    51                   0.0       55      0.0   \n",
       "23391_60                    55                   0.0        0      0.0   \n",
       "23391_72                     0                   0.0       58      1.0   \n",
       "23391_84                    58                   1.0        0      0.0   \n",
       "27468_0                      0                   0.0        6      0.0   \n",
       "27468_3                      6                   0.0       19      0.0   \n",
       "27468_6                     19                   0.0       58      0.0   \n",
       "27468_9                     58                   0.0       17      0.0   \n",
       "27468_12                    17                   0.0        2      0.0   \n",
       "27468_18                     2                   0.0        7      0.0   \n",
       "27468_24                     7                   0.0       16      0.0   \n",
       "27468_30                    16                   0.0       21      0.0   \n",
       "27468_36                    21                   0.0        9      0.0   \n",
       "27468_42                     9                   0.0        4      0.0   \n",
       "27468_48                     4                   0.0       51      3.0   \n",
       "27468_54                    51                   3.0       55      3.0   \n",
       "27468_60                    55                   3.0        0      7.0   \n",
       "27468_72                     0                   7.0       58      1.0   \n",
       "27468_84                    58                   1.0        0      0.0   \n",
       "40200_0                      0                   0.0        6      0.0   \n",
       "40200_3                      6                   0.0       19      0.0   \n",
       "40200_6                     19                   0.0       58      0.0   \n",
       "40200_9                     58                   0.0       17      0.0   \n",
       "40200_12                    17                   0.0        2      0.0   \n",
       "40200_18                     2                   0.0        7      0.0   \n",
       "40200_24                     7                   0.0       16      0.0   \n",
       "40200_30                    16                   0.0       21      0.0   \n",
       "40200_36                    21                   0.0        9      0.0   \n",
       "40200_42                     9                   0.0        4      0.0   \n",
       "40200_48                     4                   0.0       51      0.0   \n",
       "40200_54                    51                   0.0       55      0.0   \n",
       "40200_60                    55                   0.0        0      0.0   \n",
       "40200_72                     0                   0.0       58      0.0   \n",
       "40200_84                    58                   0.0        0      0.0   \n",
       "40200_96                     0                   0.0       50      0.0   \n",
       "\n",
       "          visit_month  patient_id  \n",
       "visit_id                           \n",
       "23391_0             0       23391  \n",
       "23391_3             3       23391  \n",
       "23391_6             6       23391  \n",
       "23391_9             9       23391  \n",
       "23391_12           12       23391  \n",
       "23391_18           18       23391  \n",
       "23391_24           24       23391  \n",
       "23391_30           30       23391  \n",
       "23391_36           36       23391  \n",
       "23391_42           42       23391  \n",
       "23391_48           48       23391  \n",
       "23391_54           54       23391  \n",
       "23391_60           60       23391  \n",
       "23391_72           72       23391  \n",
       "23391_84           84       23391  \n",
       "27468_0             0       27468  \n",
       "27468_3             3       27468  \n",
       "27468_6             6       27468  \n",
       "27468_9             9       27468  \n",
       "27468_12           12       27468  \n",
       "27468_18           18       27468  \n",
       "27468_24           24       27468  \n",
       "27468_30           30       27468  \n",
       "27468_36           36       27468  \n",
       "27468_42           42       27468  \n",
       "27468_48           48       27468  \n",
       "27468_54           54       27468  \n",
       "27468_60           60       27468  \n",
       "27468_72           72       27468  \n",
       "27468_84           84       27468  \n",
       "40200_0             0       40200  \n",
       "40200_3             3       40200  \n",
       "40200_6             6       40200  \n",
       "40200_9             9       40200  \n",
       "40200_12           12       40200  \n",
       "40200_18           18       40200  \n",
       "40200_24           24       40200  \n",
       "40200_30           30       40200  \n",
       "40200_36           36       40200  \n",
       "40200_42           42       40200  \n",
       "40200_48           48       40200  \n",
       "40200_54           54       40200  \n",
       "40200_60           60       40200  \n",
       "40200_72           72       40200  \n",
       "40200_84           84       40200  \n",
       "40200_96           96       40200  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "patient_visits = []\n",
    "for vi in clinical_data.index:\n",
    "    if vi.startswith('40200') or vi.startswith(\n",
    "            '27468') or vi.startswith('23391'):\n",
    "        patient_visits.append(vi)\n",
    "clinical_data_holdout = clinical_data[clinical_data.index.isin(patient_visits)]\n",
    "clinical_data_train = clinical_data[~clinical_data.index.isin(patient_visits)]\n",
    "clinical_data_holdout\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# UPDRS 1 Predictions with just protein data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_a054c_row8_col1 {\n",
       "  background-color: lightgreen;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_a054c\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_a054c_level0_col0\" class=\"col_heading level0 col0\" >Description</th>\n",
       "      <th id=\"T_a054c_level0_col1\" class=\"col_heading level0 col1\" >Value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_a054c_row0_col0\" class=\"data row0 col0\" >Session id</td>\n",
       "      <td id=\"T_a054c_row0_col1\" class=\"data row0 col1\" >5597</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_a054c_row1_col0\" class=\"data row1 col0\" >Target</td>\n",
       "      <td id=\"T_a054c_row1_col1\" class=\"data row1 col1\" >target</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_a054c_row2_col0\" class=\"data row2 col0\" >Target type</td>\n",
       "      <td id=\"T_a054c_row2_col1\" class=\"data row2 col1\" >Regression</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_a054c_row3_col0\" class=\"data row3 col0\" >Original data shape</td>\n",
       "      <td id=\"T_a054c_row3_col1\" class=\"data row3 col1\" >(1047, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_a054c_row4_col0\" class=\"data row4 col0\" >Transformed data shape</td>\n",
       "      <td id=\"T_a054c_row4_col1\" class=\"data row4 col1\" >(1047, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_a054c_row5_col0\" class=\"data row5 col0\" >Transformed train set shape</td>\n",
       "      <td id=\"T_a054c_row5_col1\" class=\"data row5 col1\" >(732, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_a054c_row6_col0\" class=\"data row6 col0\" >Transformed test set shape</td>\n",
       "      <td id=\"T_a054c_row6_col1\" class=\"data row6 col1\" >(315, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_a054c_row7_col0\" class=\"data row7 col0\" >Numeric features</td>\n",
       "      <td id=\"T_a054c_row7_col1\" class=\"data row7 col1\" >234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_a054c_row8_col0\" class=\"data row8 col0\" >Preprocess</td>\n",
       "      <td id=\"T_a054c_row8_col1\" class=\"data row8 col1\" >True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_a054c_row9_col0\" class=\"data row9 col0\" >Imputation type</td>\n",
       "      <td id=\"T_a054c_row9_col1\" class=\"data row9 col1\" >simple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row10\" class=\"row_heading level0 row10\" >10</th>\n",
       "      <td id=\"T_a054c_row10_col0\" class=\"data row10 col0\" >Numeric imputation</td>\n",
       "      <td id=\"T_a054c_row10_col1\" class=\"data row10 col1\" >mean</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row11\" class=\"row_heading level0 row11\" >11</th>\n",
       "      <td id=\"T_a054c_row11_col0\" class=\"data row11 col0\" >Categorical imputation</td>\n",
       "      <td id=\"T_a054c_row11_col1\" class=\"data row11 col1\" >mode</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row12\" class=\"row_heading level0 row12\" >12</th>\n",
       "      <td id=\"T_a054c_row12_col0\" class=\"data row12 col0\" >Fold Generator</td>\n",
       "      <td id=\"T_a054c_row12_col1\" class=\"data row12 col1\" >KFold</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row13\" class=\"row_heading level0 row13\" >13</th>\n",
       "      <td id=\"T_a054c_row13_col0\" class=\"data row13 col0\" >Fold Number</td>\n",
       "      <td id=\"T_a054c_row13_col1\" class=\"data row13 col1\" >10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row14\" class=\"row_heading level0 row14\" >14</th>\n",
       "      <td id=\"T_a054c_row14_col0\" class=\"data row14 col0\" >CPU Jobs</td>\n",
       "      <td id=\"T_a054c_row14_col1\" class=\"data row14 col1\" >-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row15\" class=\"row_heading level0 row15\" >15</th>\n",
       "      <td id=\"T_a054c_row15_col0\" class=\"data row15 col0\" >Use GPU</td>\n",
       "      <td id=\"T_a054c_row15_col1\" class=\"data row15 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row16\" class=\"row_heading level0 row16\" >16</th>\n",
       "      <td id=\"T_a054c_row16_col0\" class=\"data row16 col0\" >Log Experiment</td>\n",
       "      <td id=\"T_a054c_row16_col1\" class=\"data row16 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row17\" class=\"row_heading level0 row17\" >17</th>\n",
       "      <td id=\"T_a054c_row17_col0\" class=\"data row17 col0\" >Experiment Name</td>\n",
       "      <td id=\"T_a054c_row17_col1\" class=\"data row17 col1\" >reg-default-name</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_a054c_level0_row18\" class=\"row_heading level0 row18\" >18</th>\n",
       "      <td id=\"T_a054c_row18_col0\" class=\"data row18 col0\" >USI</td>\n",
       "      <td id=\"T_a054c_row18_col1\" class=\"data row18 col1\" >8d9d</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x275395eaee0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_c6c97_row10_col0, #T_c6c97_row10_col1, #T_c6c97_row10_col2, #T_c6c97_row10_col3, #T_c6c97_row10_col4, #T_c6c97_row10_col5, #T_c6c97_row10_col6 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_c6c97\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_c6c97_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_c6c97_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_c6c97_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_c6c97_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_c6c97_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_c6c97_level0_col5\" class=\"col_heading level0 col5\" >MAPE</th>\n",
       "      <th id=\"T_c6c97_level0_col6\" class=\"col_heading level0 col6\" >SMAPE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >Fold</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "      <th class=\"blank col6\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_c6c97_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_c6c97_row0_col0\" class=\"data row0 col0\" >3.0954</td>\n",
       "      <td id=\"T_c6c97_row0_col1\" class=\"data row0 col1\" >16.0217</td>\n",
       "      <td id=\"T_c6c97_row0_col2\" class=\"data row0 col2\" >4.0027</td>\n",
       "      <td id=\"T_c6c97_row0_col3\" class=\"data row0 col3\" >0.4470</td>\n",
       "      <td id=\"T_c6c97_row0_col4\" class=\"data row0 col4\" >0.6256</td>\n",
       "      <td id=\"T_c6c97_row0_col5\" class=\"data row0 col5\" >0.7975</td>\n",
       "      <td id=\"T_c6c97_row0_col6\" class=\"data row0 col6\" >61.3100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c6c97_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_c6c97_row1_col0\" class=\"data row1 col0\" >2.0699</td>\n",
       "      <td id=\"T_c6c97_row1_col1\" class=\"data row1 col1\" >7.7146</td>\n",
       "      <td id=\"T_c6c97_row1_col2\" class=\"data row1 col2\" >2.7775</td>\n",
       "      <td id=\"T_c6c97_row1_col3\" class=\"data row1 col3\" >0.6601</td>\n",
       "      <td id=\"T_c6c97_row1_col4\" class=\"data row1 col4\" >0.5199</td>\n",
       "      <td id=\"T_c6c97_row1_col5\" class=\"data row1 col5\" >0.5347</td>\n",
       "      <td id=\"T_c6c97_row1_col6\" class=\"data row1 col6\" >46.5653</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c6c97_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_c6c97_row2_col0\" class=\"data row2 col0\" >2.6015</td>\n",
       "      <td id=\"T_c6c97_row2_col1\" class=\"data row2 col1\" >12.2183</td>\n",
       "      <td id=\"T_c6c97_row2_col2\" class=\"data row2 col2\" >3.4955</td>\n",
       "      <td id=\"T_c6c97_row2_col3\" class=\"data row2 col3\" >0.6571</td>\n",
       "      <td id=\"T_c6c97_row2_col4\" class=\"data row2 col4\" >0.4950</td>\n",
       "      <td id=\"T_c6c97_row2_col5\" class=\"data row2 col5\" >0.5346</td>\n",
       "      <td id=\"T_c6c97_row2_col6\" class=\"data row2 col6\" >46.4948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c6c97_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_c6c97_row3_col0\" class=\"data row3 col0\" >2.6370</td>\n",
       "      <td id=\"T_c6c97_row3_col1\" class=\"data row3 col1\" >14.0089</td>\n",
       "      <td id=\"T_c6c97_row3_col2\" class=\"data row3 col2\" >3.7428</td>\n",
       "      <td id=\"T_c6c97_row3_col3\" class=\"data row3 col3\" >0.4643</td>\n",
       "      <td id=\"T_c6c97_row3_col4\" class=\"data row3 col4\" >0.5632</td>\n",
       "      <td id=\"T_c6c97_row3_col5\" class=\"data row3 col5\" >0.7103</td>\n",
       "      <td id=\"T_c6c97_row3_col6\" class=\"data row3 col6\" >52.0997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c6c97_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_c6c97_row4_col0\" class=\"data row4 col0\" >2.6971</td>\n",
       "      <td id=\"T_c6c97_row4_col1\" class=\"data row4 col1\" >12.5559</td>\n",
       "      <td id=\"T_c6c97_row4_col2\" class=\"data row4 col2\" >3.5434</td>\n",
       "      <td id=\"T_c6c97_row4_col3\" class=\"data row4 col3\" >0.4742</td>\n",
       "      <td id=\"T_c6c97_row4_col4\" class=\"data row4 col4\" >0.6518</td>\n",
       "      <td id=\"T_c6c97_row4_col5\" class=\"data row4 col5\" >0.9844</td>\n",
       "      <td id=\"T_c6c97_row4_col6\" class=\"data row4 col6\" >59.3582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c6c97_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_c6c97_row5_col0\" class=\"data row5 col0\" >2.4730</td>\n",
       "      <td id=\"T_c6c97_row5_col1\" class=\"data row5 col1\" >10.8286</td>\n",
       "      <td id=\"T_c6c97_row5_col2\" class=\"data row5 col2\" >3.2907</td>\n",
       "      <td id=\"T_c6c97_row5_col3\" class=\"data row5 col3\" >0.6270</td>\n",
       "      <td id=\"T_c6c97_row5_col4\" class=\"data row5 col4\" >0.5892</td>\n",
       "      <td id=\"T_c6c97_row5_col5\" class=\"data row5 col5\" >0.5716</td>\n",
       "      <td id=\"T_c6c97_row5_col6\" class=\"data row5 col6\" >53.5641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c6c97_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_c6c97_row6_col0\" class=\"data row6 col0\" >2.7421</td>\n",
       "      <td id=\"T_c6c97_row6_col1\" class=\"data row6 col1\" >12.8090</td>\n",
       "      <td id=\"T_c6c97_row6_col2\" class=\"data row6 col2\" >3.5790</td>\n",
       "      <td id=\"T_c6c97_row6_col3\" class=\"data row6 col3\" >0.5524</td>\n",
       "      <td id=\"T_c6c97_row6_col4\" class=\"data row6 col4\" >0.6498</td>\n",
       "      <td id=\"T_c6c97_row6_col5\" class=\"data row6 col5\" >0.7454</td>\n",
       "      <td id=\"T_c6c97_row6_col6\" class=\"data row6 col6\" >60.2849</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c6c97_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_c6c97_row7_col0\" class=\"data row7 col0\" >2.0725</td>\n",
       "      <td id=\"T_c6c97_row7_col1\" class=\"data row7 col1\" >8.7264</td>\n",
       "      <td id=\"T_c6c97_row7_col2\" class=\"data row7 col2\" >2.9540</td>\n",
       "      <td id=\"T_c6c97_row7_col3\" class=\"data row7 col3\" >0.7396</td>\n",
       "      <td id=\"T_c6c97_row7_col4\" class=\"data row7 col4\" >0.5046</td>\n",
       "      <td id=\"T_c6c97_row7_col5\" class=\"data row7 col5\" >0.3986</td>\n",
       "      <td id=\"T_c6c97_row7_col6\" class=\"data row7 col6\" >42.9123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c6c97_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_c6c97_row8_col0\" class=\"data row8 col0\" >2.5404</td>\n",
       "      <td id=\"T_c6c97_row8_col1\" class=\"data row8 col1\" >9.9943</td>\n",
       "      <td id=\"T_c6c97_row8_col2\" class=\"data row8 col2\" >3.1614</td>\n",
       "      <td id=\"T_c6c97_row8_col3\" class=\"data row8 col3\" >0.6125</td>\n",
       "      <td id=\"T_c6c97_row8_col4\" class=\"data row8 col4\" >0.5062</td>\n",
       "      <td id=\"T_c6c97_row8_col5\" class=\"data row8 col5\" >0.6224</td>\n",
       "      <td id=\"T_c6c97_row8_col6\" class=\"data row8 col6\" >45.0658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c6c97_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_c6c97_row9_col0\" class=\"data row9 col0\" >2.9825</td>\n",
       "      <td id=\"T_c6c97_row9_col1\" class=\"data row9 col1\" >17.2046</td>\n",
       "      <td id=\"T_c6c97_row9_col2\" class=\"data row9 col2\" >4.1478</td>\n",
       "      <td id=\"T_c6c97_row9_col3\" class=\"data row9 col3\" >0.4874</td>\n",
       "      <td id=\"T_c6c97_row9_col4\" class=\"data row9 col4\" >0.7146</td>\n",
       "      <td id=\"T_c6c97_row9_col5\" class=\"data row9 col5\" >0.6870</td>\n",
       "      <td id=\"T_c6c97_row9_col6\" class=\"data row9 col6\" >61.8737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c6c97_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_c6c97_row10_col0\" class=\"data row10 col0\" >2.5911</td>\n",
       "      <td id=\"T_c6c97_row10_col1\" class=\"data row10 col1\" >12.2082</td>\n",
       "      <td id=\"T_c6c97_row10_col2\" class=\"data row10 col2\" >3.4695</td>\n",
       "      <td id=\"T_c6c97_row10_col3\" class=\"data row10 col3\" >0.5721</td>\n",
       "      <td id=\"T_c6c97_row10_col4\" class=\"data row10 col4\" >0.5820</td>\n",
       "      <td id=\"T_c6c97_row10_col5\" class=\"data row10 col5\" >0.6587</td>\n",
       "      <td id=\"T_c6c97_row10_col6\" class=\"data row10 col6\" >52.9529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c6c97_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_c6c97_row11_col0\" class=\"data row11 col0\" >0.3164</td>\n",
       "      <td id=\"T_c6c97_row11_col1\" class=\"data row11 col1\" >2.8697</td>\n",
       "      <td id=\"T_c6c97_row11_col2\" class=\"data row11 col2\" >0.4134</td>\n",
       "      <td id=\"T_c6c97_row11_col3\" class=\"data row11 col3\" >0.0959</td>\n",
       "      <td id=\"T_c6c97_row11_col4\" class=\"data row11 col4\" >0.0725</td>\n",
       "      <td id=\"T_c6c97_row11_col5\" class=\"data row11 col5\" >0.1561</td>\n",
       "      <td id=\"T_c6c97_row11_col6\" class=\"data row11 col6\" >7.0060</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x275385f5af0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_c1972_row10_col0, #T_c1972_row10_col1, #T_c1972_row10_col2, #T_c1972_row10_col3, #T_c1972_row10_col4, #T_c1972_row10_col5, #T_c1972_row10_col6 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_c1972\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_c1972_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_c1972_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_c1972_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_c1972_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_c1972_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_c1972_level0_col5\" class=\"col_heading level0 col5\" >MAPE</th>\n",
       "      <th id=\"T_c1972_level0_col6\" class=\"col_heading level0 col6\" >SMAPE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >Fold</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "      <th class=\"blank col6\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_c1972_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_c1972_row0_col0\" class=\"data row0 col0\" >3.1629</td>\n",
       "      <td id=\"T_c1972_row0_col1\" class=\"data row0 col1\" >17.5933</td>\n",
       "      <td id=\"T_c1972_row0_col2\" class=\"data row0 col2\" >4.1944</td>\n",
       "      <td id=\"T_c1972_row0_col3\" class=\"data row0 col3\" >0.3927</td>\n",
       "      <td id=\"T_c1972_row0_col4\" class=\"data row0 col4\" >0.6409</td>\n",
       "      <td id=\"T_c1972_row0_col5\" class=\"data row0 col5\" >0.7579</td>\n",
       "      <td id=\"T_c1972_row0_col6\" class=\"data row0 col6\" >64.9219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c1972_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_c1972_row1_col0\" class=\"data row1 col0\" >1.9677</td>\n",
       "      <td id=\"T_c1972_row1_col1\" class=\"data row1 col1\" >7.3461</td>\n",
       "      <td id=\"T_c1972_row1_col2\" class=\"data row1 col2\" >2.7104</td>\n",
       "      <td id=\"T_c1972_row1_col3\" class=\"data row1 col3\" >0.6763</td>\n",
       "      <td id=\"T_c1972_row1_col4\" class=\"data row1 col4\" >0.5431</td>\n",
       "      <td id=\"T_c1972_row1_col5\" class=\"data row1 col5\" >0.4873</td>\n",
       "      <td id=\"T_c1972_row1_col6\" class=\"data row1 col6\" >46.5516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c1972_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_c1972_row2_col0\" class=\"data row2 col0\" >2.6929</td>\n",
       "      <td id=\"T_c1972_row2_col1\" class=\"data row2 col1\" >13.2543</td>\n",
       "      <td id=\"T_c1972_row2_col2\" class=\"data row2 col2\" >3.6406</td>\n",
       "      <td id=\"T_c1972_row2_col3\" class=\"data row2 col3\" >0.6280</td>\n",
       "      <td id=\"T_c1972_row2_col4\" class=\"data row2 col4\" >0.4999</td>\n",
       "      <td id=\"T_c1972_row2_col5\" class=\"data row2 col5\" >0.4962</td>\n",
       "      <td id=\"T_c1972_row2_col6\" class=\"data row2 col6\" >47.4530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c1972_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_c1972_row3_col0\" class=\"data row3 col0\" >2.5962</td>\n",
       "      <td id=\"T_c1972_row3_col1\" class=\"data row3 col1\" >13.8326</td>\n",
       "      <td id=\"T_c1972_row3_col2\" class=\"data row3 col2\" >3.7192</td>\n",
       "      <td id=\"T_c1972_row3_col3\" class=\"data row3 col3\" >0.4710</td>\n",
       "      <td id=\"T_c1972_row3_col4\" class=\"data row3 col4\" >0.5621</td>\n",
       "      <td id=\"T_c1972_row3_col5\" class=\"data row3 col5\" >0.6819</td>\n",
       "      <td id=\"T_c1972_row3_col6\" class=\"data row3 col6\" >52.2211</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c1972_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_c1972_row4_col0\" class=\"data row4 col0\" >2.6047</td>\n",
       "      <td id=\"T_c1972_row4_col1\" class=\"data row4 col1\" >12.3966</td>\n",
       "      <td id=\"T_c1972_row4_col2\" class=\"data row4 col2\" >3.5209</td>\n",
       "      <td id=\"T_c1972_row4_col3\" class=\"data row4 col3\" >0.4809</td>\n",
       "      <td id=\"T_c1972_row4_col4\" class=\"data row4 col4\" >0.6318</td>\n",
       "      <td id=\"T_c1972_row4_col5\" class=\"data row4 col5\" >0.8953</td>\n",
       "      <td id=\"T_c1972_row4_col6\" class=\"data row4 col6\" >57.1597</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c1972_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_c1972_row5_col0\" class=\"data row5 col0\" >2.5273</td>\n",
       "      <td id=\"T_c1972_row5_col1\" class=\"data row5 col1\" >10.5646</td>\n",
       "      <td id=\"T_c1972_row5_col2\" class=\"data row5 col2\" >3.2503</td>\n",
       "      <td id=\"T_c1972_row5_col3\" class=\"data row5 col3\" >0.6360</td>\n",
       "      <td id=\"T_c1972_row5_col4\" class=\"data row5 col4\" >0.5879</td>\n",
       "      <td id=\"T_c1972_row5_col5\" class=\"data row5 col5\" >0.5912</td>\n",
       "      <td id=\"T_c1972_row5_col6\" class=\"data row5 col6\" >55.3868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c1972_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_c1972_row6_col0\" class=\"data row6 col0\" >2.6550</td>\n",
       "      <td id=\"T_c1972_row6_col1\" class=\"data row6 col1\" >12.9968</td>\n",
       "      <td id=\"T_c1972_row6_col2\" class=\"data row6 col2\" >3.6051</td>\n",
       "      <td id=\"T_c1972_row6_col3\" class=\"data row6 col3\" >0.5458</td>\n",
       "      <td id=\"T_c1972_row6_col4\" class=\"data row6 col4\" >0.6644</td>\n",
       "      <td id=\"T_c1972_row6_col5\" class=\"data row6 col5\" >0.6651</td>\n",
       "      <td id=\"T_c1972_row6_col6\" class=\"data row6 col6\" >60.2813</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c1972_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_c1972_row7_col0\" class=\"data row7 col0\" >2.1350</td>\n",
       "      <td id=\"T_c1972_row7_col1\" class=\"data row7 col1\" >9.1743</td>\n",
       "      <td id=\"T_c1972_row7_col2\" class=\"data row7 col2\" >3.0289</td>\n",
       "      <td id=\"T_c1972_row7_col3\" class=\"data row7 col3\" >0.7262</td>\n",
       "      <td id=\"T_c1972_row7_col4\" class=\"data row7 col4\" >0.5079</td>\n",
       "      <td id=\"T_c1972_row7_col5\" class=\"data row7 col5\" >0.3914</td>\n",
       "      <td id=\"T_c1972_row7_col6\" class=\"data row7 col6\" >44.7149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c1972_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_c1972_row8_col0\" class=\"data row8 col0\" >2.4336</td>\n",
       "      <td id=\"T_c1972_row8_col1\" class=\"data row8 col1\" >9.3595</td>\n",
       "      <td id=\"T_c1972_row8_col2\" class=\"data row8 col2\" >3.0593</td>\n",
       "      <td id=\"T_c1972_row8_col3\" class=\"data row8 col3\" >0.6371</td>\n",
       "      <td id=\"T_c1972_row8_col4\" class=\"data row8 col4\" >0.4808</td>\n",
       "      <td id=\"T_c1972_row8_col5\" class=\"data row8 col5\" >0.5450</td>\n",
       "      <td id=\"T_c1972_row8_col6\" class=\"data row8 col6\" >44.1198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c1972_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_c1972_row9_col0\" class=\"data row9 col0\" >2.8647</td>\n",
       "      <td id=\"T_c1972_row9_col1\" class=\"data row9 col1\" >15.0992</td>\n",
       "      <td id=\"T_c1972_row9_col2\" class=\"data row9 col2\" >3.8858</td>\n",
       "      <td id=\"T_c1972_row9_col3\" class=\"data row9 col3\" >0.5502</td>\n",
       "      <td id=\"T_c1972_row9_col4\" class=\"data row9 col4\" >0.6744</td>\n",
       "      <td id=\"T_c1972_row9_col5\" class=\"data row9 col5\" >0.6735</td>\n",
       "      <td id=\"T_c1972_row9_col6\" class=\"data row9 col6\" >61.4463</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c1972_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_c1972_row10_col0\" class=\"data row10 col0\" >2.5640</td>\n",
       "      <td id=\"T_c1972_row10_col1\" class=\"data row10 col1\" >12.1617</td>\n",
       "      <td id=\"T_c1972_row10_col2\" class=\"data row10 col2\" >3.4615</td>\n",
       "      <td id=\"T_c1972_row10_col3\" class=\"data row10 col3\" >0.5744</td>\n",
       "      <td id=\"T_c1972_row10_col4\" class=\"data row10 col4\" >0.5793</td>\n",
       "      <td id=\"T_c1972_row10_col5\" class=\"data row10 col5\" >0.6185</td>\n",
       "      <td id=\"T_c1972_row10_col6\" class=\"data row10 col6\" >53.4256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c1972_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_c1972_row11_col0\" class=\"data row11 col0\" >0.3217</td>\n",
       "      <td id=\"T_c1972_row11_col1\" class=\"data row11 col1\" >2.9221</td>\n",
       "      <td id=\"T_c1972_row11_col2\" class=\"data row11 col2\" >0.4240</td>\n",
       "      <td id=\"T_c1972_row11_col3\" class=\"data row11 col3\" >0.0989</td>\n",
       "      <td id=\"T_c1972_row11_col4\" class=\"data row11 col4\" >0.0676</td>\n",
       "      <td id=\"T_c1972_row11_col5\" class=\"data row11 col5\" >0.1397</td>\n",
       "      <td id=\"T_c1972_row11_col6\" class=\"data row11 col6\" >7.1284</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x27538575580>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 10 candidates, totalling 100 fits\n",
      "Original model was better than the tuned model, hence it will be returned. NOTE: The display metrics are for the tuned model (not the original one).\n"
     ]
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_013ff_row10_col0, #T_013ff_row10_col1, #T_013ff_row10_col2, #T_013ff_row10_col3, #T_013ff_row10_col4, #T_013ff_row10_col5, #T_013ff_row10_col6 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_013ff\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_013ff_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_013ff_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_013ff_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_013ff_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_013ff_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_013ff_level0_col5\" class=\"col_heading level0 col5\" >MAPE</th>\n",
       "      <th id=\"T_013ff_level0_col6\" class=\"col_heading level0 col6\" >SMAPE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >Fold</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "      <th class=\"blank col6\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_013ff_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_013ff_row0_col0\" class=\"data row0 col0\" >3.2085</td>\n",
       "      <td id=\"T_013ff_row0_col1\" class=\"data row0 col1\" >16.3920</td>\n",
       "      <td id=\"T_013ff_row0_col2\" class=\"data row0 col2\" >4.0487</td>\n",
       "      <td id=\"T_013ff_row0_col3\" class=\"data row0 col3\" >0.4342</td>\n",
       "      <td id=\"T_013ff_row0_col4\" class=\"data row0 col4\" >0.6609</td>\n",
       "      <td id=\"T_013ff_row0_col5\" class=\"data row0 col5\" >0.9231</td>\n",
       "      <td id=\"T_013ff_row0_col6\" class=\"data row0 col6\" >65.9575</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_013ff_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_013ff_row1_col0\" class=\"data row1 col0\" >2.2053</td>\n",
       "      <td id=\"T_013ff_row1_col1\" class=\"data row1 col1\" >7.9106</td>\n",
       "      <td id=\"T_013ff_row1_col2\" class=\"data row1 col2\" >2.8126</td>\n",
       "      <td id=\"T_013ff_row1_col3\" class=\"data row1 col3\" >0.6514</td>\n",
       "      <td id=\"T_013ff_row1_col4\" class=\"data row1 col4\" >0.5897</td>\n",
       "      <td id=\"T_013ff_row1_col5\" class=\"data row1 col5\" >0.6146</td>\n",
       "      <td id=\"T_013ff_row1_col6\" class=\"data row1 col6\" >49.8593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_013ff_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_013ff_row2_col0\" class=\"data row2 col0\" >2.6364</td>\n",
       "      <td id=\"T_013ff_row2_col1\" class=\"data row2 col1\" >11.8302</td>\n",
       "      <td id=\"T_013ff_row2_col2\" class=\"data row2 col2\" >3.4395</td>\n",
       "      <td id=\"T_013ff_row2_col3\" class=\"data row2 col3\" >0.6680</td>\n",
       "      <td id=\"T_013ff_row2_col4\" class=\"data row2 col4\" >0.5219</td>\n",
       "      <td id=\"T_013ff_row2_col5\" class=\"data row2 col5\" >0.5358</td>\n",
       "      <td id=\"T_013ff_row2_col6\" class=\"data row2 col6\" >46.7434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_013ff_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_013ff_row3_col0\" class=\"data row3 col0\" >2.7438</td>\n",
       "      <td id=\"T_013ff_row3_col1\" class=\"data row3 col1\" >14.3004</td>\n",
       "      <td id=\"T_013ff_row3_col2\" class=\"data row3 col2\" >3.7816</td>\n",
       "      <td id=\"T_013ff_row3_col3\" class=\"data row3 col3\" >0.4532</td>\n",
       "      <td id=\"T_013ff_row3_col4\" class=\"data row3 col4\" >0.6020</td>\n",
       "      <td id=\"T_013ff_row3_col5\" class=\"data row3 col5\" >0.8203</td>\n",
       "      <td id=\"T_013ff_row3_col6\" class=\"data row3 col6\" >54.2314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_013ff_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_013ff_row4_col0\" class=\"data row4 col0\" >2.7458</td>\n",
       "      <td id=\"T_013ff_row4_col1\" class=\"data row4 col1\" >12.5027</td>\n",
       "      <td id=\"T_013ff_row4_col2\" class=\"data row4 col2\" >3.5359</td>\n",
       "      <td id=\"T_013ff_row4_col3\" class=\"data row4 col3\" >0.4764</td>\n",
       "      <td id=\"T_013ff_row4_col4\" class=\"data row4 col4\" >0.6683</td>\n",
       "      <td id=\"T_013ff_row4_col5\" class=\"data row4 col5\" >1.0152</td>\n",
       "      <td id=\"T_013ff_row4_col6\" class=\"data row4 col6\" >59.7003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_013ff_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_013ff_row5_col0\" class=\"data row5 col0\" >2.5797</td>\n",
       "      <td id=\"T_013ff_row5_col1\" class=\"data row5 col1\" >10.4306</td>\n",
       "      <td id=\"T_013ff_row5_col2\" class=\"data row5 col2\" >3.2296</td>\n",
       "      <td id=\"T_013ff_row5_col3\" class=\"data row5 col3\" >0.6407</td>\n",
       "      <td id=\"T_013ff_row5_col4\" class=\"data row5 col4\" >0.6303</td>\n",
       "      <td id=\"T_013ff_row5_col5\" class=\"data row5 col5\" >0.6971</td>\n",
       "      <td id=\"T_013ff_row5_col6\" class=\"data row5 col6\" >55.3701</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_013ff_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_013ff_row6_col0\" class=\"data row6 col0\" >2.8226</td>\n",
       "      <td id=\"T_013ff_row6_col1\" class=\"data row6 col1\" >12.9993</td>\n",
       "      <td id=\"T_013ff_row6_col2\" class=\"data row6 col2\" >3.6055</td>\n",
       "      <td id=\"T_013ff_row6_col3\" class=\"data row6 col3\" >0.5457</td>\n",
       "      <td id=\"T_013ff_row6_col4\" class=\"data row6 col4\" >0.7091</td>\n",
       "      <td id=\"T_013ff_row6_col5\" class=\"data row6 col5\" >0.7905</td>\n",
       "      <td id=\"T_013ff_row6_col6\" class=\"data row6 col6\" >61.8698</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_013ff_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_013ff_row7_col0\" class=\"data row7 col0\" >2.2030</td>\n",
       "      <td id=\"T_013ff_row7_col1\" class=\"data row7 col1\" >9.4104</td>\n",
       "      <td id=\"T_013ff_row7_col2\" class=\"data row7 col2\" >3.0676</td>\n",
       "      <td id=\"T_013ff_row7_col3\" class=\"data row7 col3\" >0.7192</td>\n",
       "      <td id=\"T_013ff_row7_col4\" class=\"data row7 col4\" >0.5615</td>\n",
       "      <td id=\"T_013ff_row7_col5\" class=\"data row7 col5\" >0.4731</td>\n",
       "      <td id=\"T_013ff_row7_col6\" class=\"data row7 col6\" >46.1942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_013ff_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_013ff_row8_col0\" class=\"data row8 col0\" >2.5368</td>\n",
       "      <td id=\"T_013ff_row8_col1\" class=\"data row8 col1\" >9.7984</td>\n",
       "      <td id=\"T_013ff_row8_col2\" class=\"data row8 col2\" >3.1302</td>\n",
       "      <td id=\"T_013ff_row8_col3\" class=\"data row8 col3\" >0.6201</td>\n",
       "      <td id=\"T_013ff_row8_col4\" class=\"data row8 col4\" >0.5201</td>\n",
       "      <td id=\"T_013ff_row8_col5\" class=\"data row8 col5\" >0.6498</td>\n",
       "      <td id=\"T_013ff_row8_col6\" class=\"data row8 col6\" >45.4628</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_013ff_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_013ff_row9_col0\" class=\"data row9 col0\" >3.0422</td>\n",
       "      <td id=\"T_013ff_row9_col1\" class=\"data row9 col1\" >16.5612</td>\n",
       "      <td id=\"T_013ff_row9_col2\" class=\"data row9 col2\" >4.0695</td>\n",
       "      <td id=\"T_013ff_row9_col3\" class=\"data row9 col3\" >0.5066</td>\n",
       "      <td id=\"T_013ff_row9_col4\" class=\"data row9 col4\" >0.7244</td>\n",
       "      <td id=\"T_013ff_row9_col5\" class=\"data row9 col5\" >0.7859</td>\n",
       "      <td id=\"T_013ff_row9_col6\" class=\"data row9 col6\" >62.6314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_013ff_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_013ff_row10_col0\" class=\"data row10 col0\" >2.6724</td>\n",
       "      <td id=\"T_013ff_row10_col1\" class=\"data row10 col1\" >12.2136</td>\n",
       "      <td id=\"T_013ff_row10_col2\" class=\"data row10 col2\" >3.4721</td>\n",
       "      <td id=\"T_013ff_row10_col3\" class=\"data row10 col3\" >0.5715</td>\n",
       "      <td id=\"T_013ff_row10_col4\" class=\"data row10 col4\" >0.6188</td>\n",
       "      <td id=\"T_013ff_row10_col5\" class=\"data row10 col5\" >0.7305</td>\n",
       "      <td id=\"T_013ff_row10_col6\" class=\"data row10 col6\" >54.8020</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_013ff_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_013ff_row11_col0\" class=\"data row11 col0\" >0.3040</td>\n",
       "      <td id=\"T_013ff_row11_col1\" class=\"data row11 col1\" >2.7740</td>\n",
       "      <td id=\"T_013ff_row11_col2\" class=\"data row11 col2\" >0.3978</td>\n",
       "      <td id=\"T_013ff_row11_col3\" class=\"data row11 col3\" >0.0956</td>\n",
       "      <td id=\"T_013ff_row11_col4\" class=\"data row11 col4\" >0.0687</td>\n",
       "      <td id=\"T_013ff_row11_col5\" class=\"data row11 col5\" >0.1606</td>\n",
       "      <td id=\"T_013ff_row11_col6\" class=\"data row11 col6\" >7.1486</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2753854a970>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original model was better than the ensembled model, hence it will be returned. NOTE: The display metrics are for the ensembled model (not the original one).\n"
     ]
    }
   ],
   "source": [
    "dataset = pd.merge(clinical_data_train[['updrs_1',\n",
    "                                        'offset_updrs1_offset',\n",
    "                                        'offset_updrs2_offset',\n",
    "                                        'offset_updrs3_offset',\n",
    "                                        'offset_updrs4_offset',\n",
    "                                        'visit_month']],\n",
    "                   npx_features,\n",
    "                   how='inner',\n",
    "                   left_on=['visit_id'],\n",
    "                   right_on=['visit_id'])\n",
    "dataset.head()\n",
    "Y = dataset[dataset.columns[0]].values\n",
    "X = dataset[dataset.columns[6:]].values\n",
    "X_noscale = dataset[dataset.columns[1:6]].values\n",
    "X_scaled = StandardScaler().fit_transform(X)\n",
    "final_X = np.hstack((X_noscale, X_scaled))\n",
    "\n",
    "s = setup(data=final_X, target=Y)\n",
    "add_metric('SMAPE', 'SMAPE', smape, greater_is_better=False)\n",
    "b1 = create_model('et')\n",
    "t1 = tune_model(b1, choose_better=True, optimize='SMAPE')\n",
    "ensemble_u1 = ensemble_model(\n",
    "    t1,\n",
    "    choose_better=True,\n",
    "    n_estimators=20,\n",
    "    method='Boosting')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_86238\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_86238_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_86238_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_86238_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_86238_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_86238_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_86238_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_86238_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_86238_level0_col7\" class=\"col_heading level0 col7\" >SMAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_86238_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_86238_row0_col0\" class=\"data row0 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_86238_row0_col1\" class=\"data row0 col1\" >2.2361</td>\n",
       "      <td id=\"T_86238_row0_col2\" class=\"data row0 col2\" >9.2379</td>\n",
       "      <td id=\"T_86238_row0_col3\" class=\"data row0 col3\" >3.0394</td>\n",
       "      <td id=\"T_86238_row0_col4\" class=\"data row0 col4\" >0.6439</td>\n",
       "      <td id=\"T_86238_row0_col5\" class=\"data row0 col5\" >0.5285</td>\n",
       "      <td id=\"T_86238_row0_col6\" class=\"data row0 col6\" >0.6583</td>\n",
       "      <td id=\"T_86238_row0_col7\" class=\"data row0 col7\" >48.3763</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x275385e0520>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_9e754\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_9e754_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_9e754_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_9e754_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_9e754_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_9e754_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_9e754_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_9e754_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_9e754_level0_col7\" class=\"col_heading level0 col7\" >SMAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_9e754_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_9e754_row0_col0\" class=\"data row0 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_9e754_row0_col1\" class=\"data row0 col1\" >2.2361</td>\n",
       "      <td id=\"T_9e754_row0_col2\" class=\"data row0 col2\" >9.2379</td>\n",
       "      <td id=\"T_9e754_row0_col3\" class=\"data row0 col3\" >3.0394</td>\n",
       "      <td id=\"T_9e754_row0_col4\" class=\"data row0 col4\" >0.6439</td>\n",
       "      <td id=\"T_9e754_row0_col5\" class=\"data row0 col5\" >0.5285</td>\n",
       "      <td id=\"T_9e754_row0_col6\" class=\"data row0 col6\" >0.6583</td>\n",
       "      <td id=\"T_9e754_row0_col7\" class=\"data row0 col7\" >48.3763</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2753946b4c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_2b3c3\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_2b3c3_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_2b3c3_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_2b3c3_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_2b3c3_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_2b3c3_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_2b3c3_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_2b3c3_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_2b3c3_level0_col7\" class=\"col_heading level0 col7\" >SMAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_2b3c3_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_2b3c3_row0_col0\" class=\"data row0 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_2b3c3_row0_col1\" class=\"data row0 col1\" >2.2361</td>\n",
       "      <td id=\"T_2b3c3_row0_col2\" class=\"data row0 col2\" >9.2379</td>\n",
       "      <td id=\"T_2b3c3_row0_col3\" class=\"data row0 col3\" >3.0394</td>\n",
       "      <td id=\"T_2b3c3_row0_col4\" class=\"data row0 col4\" >0.6439</td>\n",
       "      <td id=\"T_2b3c3_row0_col5\" class=\"data row0 col5\" >0.5285</td>\n",
       "      <td id=\"T_2b3c3_row0_col6\" class=\"data row0 col6\" >0.6583</td>\n",
       "      <td id=\"T_2b3c3_row0_col7\" class=\"data row0 col7\" >48.3763</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x27539214070>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(memory=FastMemory(location=C:\\Users\\prith\\AppData\\Local\\Temp\\joblib),\n",
       "         steps=[(&#x27;numerical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[&#x27;feature_1&#x27;, &#x27;feature_2&#x27;,\n",
       "                                             &#x27;feature_3&#x27;, &#x27;feature_4&#x27;,\n",
       "                                             &#x27;feature_5&#x27;, &#x27;feature_6&#x27;,\n",
       "                                             &#x27;feature_7&#x27;, &#x27;feature_8&#x27;,\n",
       "                                             &#x27;feature_9&#x27;, &#x27;feature_10&#x27;,\n",
       "                                             &#x27;feature_11&#x27;, &#x27;feature_12&#x27;,\n",
       "                                             &#x27;feature_13&#x27;, &#x27;feature_14&#x27;,\n",
       "                                             &#x27;feature_15&#x27;, &#x27;feature_16&#x27;,\n",
       "                                             &#x27;feature_17&#x27;, &#x27;fea...\n",
       "                                             &#x27;feature_19&#x27;, &#x27;feature_20&#x27;,\n",
       "                                             &#x27;feature_21&#x27;, &#x27;feature_22&#x27;,\n",
       "                                             &#x27;feature_23&#x27;, &#x27;feature_24&#x27;,\n",
       "                                             &#x27;feature_25&#x27;, &#x27;feature_26&#x27;,\n",
       "                                             &#x27;feature_27&#x27;, &#x27;feature_28&#x27;,\n",
       "                                             &#x27;feature_29&#x27;, &#x27;feature_30&#x27;, ...],\n",
       "                                    transformer=SimpleImputer())),\n",
       "                (&#x27;categorical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[],\n",
       "                                    transformer=SimpleImputer(strategy=&#x27;most_frequent&#x27;))),\n",
       "                (&#x27;actual_estimator&#x27;,\n",
       "                 ExtraTreesRegressor(n_jobs=-1, random_state=5597))])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" ><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(memory=FastMemory(location=C:\\Users\\prith\\AppData\\Local\\Temp\\joblib),\n",
       "         steps=[(&#x27;numerical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[&#x27;feature_1&#x27;, &#x27;feature_2&#x27;,\n",
       "                                             &#x27;feature_3&#x27;, &#x27;feature_4&#x27;,\n",
       "                                             &#x27;feature_5&#x27;, &#x27;feature_6&#x27;,\n",
       "                                             &#x27;feature_7&#x27;, &#x27;feature_8&#x27;,\n",
       "                                             &#x27;feature_9&#x27;, &#x27;feature_10&#x27;,\n",
       "                                             &#x27;feature_11&#x27;, &#x27;feature_12&#x27;,\n",
       "                                             &#x27;feature_13&#x27;, &#x27;feature_14&#x27;,\n",
       "                                             &#x27;feature_15&#x27;, &#x27;feature_16&#x27;,\n",
       "                                             &#x27;feature_17&#x27;, &#x27;fea...\n",
       "                                             &#x27;feature_19&#x27;, &#x27;feature_20&#x27;,\n",
       "                                             &#x27;feature_21&#x27;, &#x27;feature_22&#x27;,\n",
       "                                             &#x27;feature_23&#x27;, &#x27;feature_24&#x27;,\n",
       "                                             &#x27;feature_25&#x27;, &#x27;feature_26&#x27;,\n",
       "                                             &#x27;feature_27&#x27;, &#x27;feature_28&#x27;,\n",
       "                                             &#x27;feature_29&#x27;, &#x27;feature_30&#x27;, ...],\n",
       "                                    transformer=SimpleImputer())),\n",
       "                (&#x27;categorical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[],\n",
       "                                    transformer=SimpleImputer(strategy=&#x27;most_frequent&#x27;))),\n",
       "                (&#x27;actual_estimator&#x27;,\n",
       "                 ExtraTreesRegressor(n_jobs=-1, random_state=5597))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">numerical_imputer: TransformerWrapper</label><div class=\"sk-toggleable__content\"><pre>TransformerWrapper(include=[&#x27;feature_1&#x27;, &#x27;feature_2&#x27;, &#x27;feature_3&#x27;, &#x27;feature_4&#x27;,\n",
       "                            &#x27;feature_5&#x27;, &#x27;feature_6&#x27;, &#x27;feature_7&#x27;, &#x27;feature_8&#x27;,\n",
       "                            &#x27;feature_9&#x27;, &#x27;feature_10&#x27;, &#x27;feature_11&#x27;,\n",
       "                            &#x27;feature_12&#x27;, &#x27;feature_13&#x27;, &#x27;feature_14&#x27;,\n",
       "                            &#x27;feature_15&#x27;, &#x27;feature_16&#x27;, &#x27;feature_17&#x27;,\n",
       "                            &#x27;feature_18&#x27;, &#x27;feature_19&#x27;, &#x27;feature_20&#x27;,\n",
       "                            &#x27;feature_21&#x27;, &#x27;feature_22&#x27;, &#x27;feature_23&#x27;,\n",
       "                            &#x27;feature_24&#x27;, &#x27;feature_25&#x27;, &#x27;feature_26&#x27;,\n",
       "                            &#x27;feature_27&#x27;, &#x27;feature_28&#x27;, &#x27;feature_29&#x27;,\n",
       "                            &#x27;feature_30&#x27;, ...],\n",
       "                   transformer=SimpleImputer())</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">transformer: SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer()</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" ><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-5\" type=\"checkbox\" ><label for=\"sk-estimator-id-5\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">categorical_imputer: TransformerWrapper</label><div class=\"sk-toggleable__content\"><pre>TransformerWrapper(include=[],\n",
       "                   transformer=SimpleImputer(strategy=&#x27;most_frequent&#x27;))</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-6\" type=\"checkbox\" ><label for=\"sk-estimator-id-6\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">transformer: SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(strategy=&#x27;most_frequent&#x27;)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-7\" type=\"checkbox\" ><label for=\"sk-estimator-id-7\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(strategy=&#x27;most_frequent&#x27;)</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-8\" type=\"checkbox\" ><label for=\"sk-estimator-id-8\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ExtraTreesRegressor</label><div class=\"sk-toggleable__content\"><pre>ExtraTreesRegressor(n_jobs=-1, random_state=5597)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(memory=FastMemory(location=C:\\Users\\prith\\AppData\\Local\\Temp\\joblib),\n",
       "         steps=[('numerical_imputer',\n",
       "                 TransformerWrapper(include=['feature_1', 'feature_2',\n",
       "                                             'feature_3', 'feature_4',\n",
       "                                             'feature_5', 'feature_6',\n",
       "                                             'feature_7', 'feature_8',\n",
       "                                             'feature_9', 'feature_10',\n",
       "                                             'feature_11', 'feature_12',\n",
       "                                             'feature_13', 'feature_14',\n",
       "                                             'feature_15', 'feature_16',\n",
       "                                             'feature_17', 'fea...\n",
       "                                             'feature_19', 'feature_20',\n",
       "                                             'feature_21', 'feature_22',\n",
       "                                             'feature_23', 'feature_24',\n",
       "                                             'feature_25', 'feature_26',\n",
       "                                             'feature_27', 'feature_28',\n",
       "                                             'feature_29', 'feature_30', ...],\n",
       "                                    transformer=SimpleImputer())),\n",
       "                ('categorical_imputer',\n",
       "                 TransformerWrapper(include=[],\n",
       "                                    transformer=SimpleImputer(strategy='most_frequent'))),\n",
       "                ('actual_estimator',\n",
       "                 ExtraTreesRegressor(n_jobs=-1, random_state=5597))])"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "predict_model(b1)\n",
    "predict_model(t1)\n",
    "predict_model(ensemble_u1)\n",
    "finalize_model(ensemble_u1)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# UPDRS 2 Predictions with just Protein Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_c8f61_row8_col1 {\n",
       "  background-color: lightgreen;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_c8f61\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_c8f61_level0_col0\" class=\"col_heading level0 col0\" >Description</th>\n",
       "      <th id=\"T_c8f61_level0_col1\" class=\"col_heading level0 col1\" >Value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_c8f61_row0_col0\" class=\"data row0 col0\" >Session id</td>\n",
       "      <td id=\"T_c8f61_row0_col1\" class=\"data row0 col1\" >4350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_c8f61_row1_col0\" class=\"data row1 col0\" >Target</td>\n",
       "      <td id=\"T_c8f61_row1_col1\" class=\"data row1 col1\" >target</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_c8f61_row2_col0\" class=\"data row2 col0\" >Target type</td>\n",
       "      <td id=\"T_c8f61_row2_col1\" class=\"data row2 col1\" >Regression</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_c8f61_row3_col0\" class=\"data row3 col0\" >Original data shape</td>\n",
       "      <td id=\"T_c8f61_row3_col1\" class=\"data row3 col1\" >(1047, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_c8f61_row4_col0\" class=\"data row4 col0\" >Transformed data shape</td>\n",
       "      <td id=\"T_c8f61_row4_col1\" class=\"data row4 col1\" >(1047, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_c8f61_row5_col0\" class=\"data row5 col0\" >Transformed train set shape</td>\n",
       "      <td id=\"T_c8f61_row5_col1\" class=\"data row5 col1\" >(732, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_c8f61_row6_col0\" class=\"data row6 col0\" >Transformed test set shape</td>\n",
       "      <td id=\"T_c8f61_row6_col1\" class=\"data row6 col1\" >(315, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_c8f61_row7_col0\" class=\"data row7 col0\" >Numeric features</td>\n",
       "      <td id=\"T_c8f61_row7_col1\" class=\"data row7 col1\" >234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_c8f61_row8_col0\" class=\"data row8 col0\" >Preprocess</td>\n",
       "      <td id=\"T_c8f61_row8_col1\" class=\"data row8 col1\" >True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_c8f61_row9_col0\" class=\"data row9 col0\" >Imputation type</td>\n",
       "      <td id=\"T_c8f61_row9_col1\" class=\"data row9 col1\" >simple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row10\" class=\"row_heading level0 row10\" >10</th>\n",
       "      <td id=\"T_c8f61_row10_col0\" class=\"data row10 col0\" >Numeric imputation</td>\n",
       "      <td id=\"T_c8f61_row10_col1\" class=\"data row10 col1\" >mean</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row11\" class=\"row_heading level0 row11\" >11</th>\n",
       "      <td id=\"T_c8f61_row11_col0\" class=\"data row11 col0\" >Categorical imputation</td>\n",
       "      <td id=\"T_c8f61_row11_col1\" class=\"data row11 col1\" >mode</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row12\" class=\"row_heading level0 row12\" >12</th>\n",
       "      <td id=\"T_c8f61_row12_col0\" class=\"data row12 col0\" >Fold Generator</td>\n",
       "      <td id=\"T_c8f61_row12_col1\" class=\"data row12 col1\" >KFold</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row13\" class=\"row_heading level0 row13\" >13</th>\n",
       "      <td id=\"T_c8f61_row13_col0\" class=\"data row13 col0\" >Fold Number</td>\n",
       "      <td id=\"T_c8f61_row13_col1\" class=\"data row13 col1\" >10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row14\" class=\"row_heading level0 row14\" >14</th>\n",
       "      <td id=\"T_c8f61_row14_col0\" class=\"data row14 col0\" >CPU Jobs</td>\n",
       "      <td id=\"T_c8f61_row14_col1\" class=\"data row14 col1\" >-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row15\" class=\"row_heading level0 row15\" >15</th>\n",
       "      <td id=\"T_c8f61_row15_col0\" class=\"data row15 col0\" >Use GPU</td>\n",
       "      <td id=\"T_c8f61_row15_col1\" class=\"data row15 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row16\" class=\"row_heading level0 row16\" >16</th>\n",
       "      <td id=\"T_c8f61_row16_col0\" class=\"data row16 col0\" >Log Experiment</td>\n",
       "      <td id=\"T_c8f61_row16_col1\" class=\"data row16 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row17\" class=\"row_heading level0 row17\" >17</th>\n",
       "      <td id=\"T_c8f61_row17_col0\" class=\"data row17 col0\" >Experiment Name</td>\n",
       "      <td id=\"T_c8f61_row17_col1\" class=\"data row17 col1\" >reg-default-name</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_c8f61_level0_row18\" class=\"row_heading level0 row18\" >18</th>\n",
       "      <td id=\"T_c8f61_row18_col0\" class=\"data row18 col0\" >USI</td>\n",
       "      <td id=\"T_c8f61_row18_col1\" class=\"data row18 col1\" >a71c</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x275385d1400>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_ee657_row10_col0, #T_ee657_row10_col1, #T_ee657_row10_col2, #T_ee657_row10_col3, #T_ee657_row10_col4, #T_ee657_row10_col5, #T_ee657_row10_col6 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_ee657\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_ee657_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_ee657_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_ee657_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_ee657_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_ee657_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_ee657_level0_col5\" class=\"col_heading level0 col5\" >MAPE</th>\n",
       "      <th id=\"T_ee657_level0_col6\" class=\"col_heading level0 col6\" >SMAPE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >Fold</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "      <th class=\"blank col6\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_ee657_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_ee657_row0_col0\" class=\"data row0 col0\" >2.1728</td>\n",
       "      <td id=\"T_ee657_row0_col1\" class=\"data row0 col1\" >8.3400</td>\n",
       "      <td id=\"T_ee657_row0_col2\" class=\"data row0 col2\" >2.8879</td>\n",
       "      <td id=\"T_ee657_row0_col3\" class=\"data row0 col3\" >0.7383</td>\n",
       "      <td id=\"T_ee657_row0_col4\" class=\"data row0 col4\" >0.6216</td>\n",
       "      <td id=\"T_ee657_row0_col5\" class=\"data row0 col5\" >0.5153</td>\n",
       "      <td id=\"T_ee657_row0_col6\" class=\"data row0 col6\" >82.3297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ee657_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_ee657_row1_col0\" class=\"data row1 col0\" >2.1980</td>\n",
       "      <td id=\"T_ee657_row1_col1\" class=\"data row1 col1\" >11.9131</td>\n",
       "      <td id=\"T_ee657_row1_col2\" class=\"data row1 col2\" >3.4515</td>\n",
       "      <td id=\"T_ee657_row1_col3\" class=\"data row1 col3\" >0.6706</td>\n",
       "      <td id=\"T_ee657_row1_col4\" class=\"data row1 col4\" >0.7638</td>\n",
       "      <td id=\"T_ee657_row1_col5\" class=\"data row1 col5\" >0.4986</td>\n",
       "      <td id=\"T_ee657_row1_col6\" class=\"data row1 col6\" >91.4793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ee657_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_ee657_row2_col0\" class=\"data row2 col0\" >2.3996</td>\n",
       "      <td id=\"T_ee657_row2_col1\" class=\"data row2 col1\" >11.2053</td>\n",
       "      <td id=\"T_ee657_row2_col2\" class=\"data row2 col2\" >3.3474</td>\n",
       "      <td id=\"T_ee657_row2_col3\" class=\"data row2 col3\" >0.7124</td>\n",
       "      <td id=\"T_ee657_row2_col4\" class=\"data row2 col4\" >0.6138</td>\n",
       "      <td id=\"T_ee657_row2_col5\" class=\"data row2 col5\" >0.6924</td>\n",
       "      <td id=\"T_ee657_row2_col6\" class=\"data row2 col6\" >85.4547</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ee657_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_ee657_row3_col0\" class=\"data row3 col0\" >2.4033</td>\n",
       "      <td id=\"T_ee657_row3_col1\" class=\"data row3 col1\" >10.8463</td>\n",
       "      <td id=\"T_ee657_row3_col2\" class=\"data row3 col2\" >3.2934</td>\n",
       "      <td id=\"T_ee657_row3_col3\" class=\"data row3 col3\" >0.6065</td>\n",
       "      <td id=\"T_ee657_row3_col4\" class=\"data row3 col4\" >0.6284</td>\n",
       "      <td id=\"T_ee657_row3_col5\" class=\"data row3 col5\" >0.5604</td>\n",
       "      <td id=\"T_ee657_row3_col6\" class=\"data row3 col6\" >78.9644</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ee657_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_ee657_row4_col0\" class=\"data row4 col0\" >2.5562</td>\n",
       "      <td id=\"T_ee657_row4_col1\" class=\"data row4 col1\" >12.7411</td>\n",
       "      <td id=\"T_ee657_row4_col2\" class=\"data row4 col2\" >3.5695</td>\n",
       "      <td id=\"T_ee657_row4_col3\" class=\"data row4 col3\" >0.6757</td>\n",
       "      <td id=\"T_ee657_row4_col4\" class=\"data row4 col4\" >0.7050</td>\n",
       "      <td id=\"T_ee657_row4_col5\" class=\"data row4 col5\" >0.8831</td>\n",
       "      <td id=\"T_ee657_row4_col6\" class=\"data row4 col6\" >94.7960</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ee657_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_ee657_row5_col0\" class=\"data row5 col0\" >2.7192</td>\n",
       "      <td id=\"T_ee657_row5_col1\" class=\"data row5 col1\" >13.4031</td>\n",
       "      <td id=\"T_ee657_row5_col2\" class=\"data row5 col2\" >3.6610</td>\n",
       "      <td id=\"T_ee657_row5_col3\" class=\"data row5 col3\" >0.6271</td>\n",
       "      <td id=\"T_ee657_row5_col4\" class=\"data row5 col4\" >0.6394</td>\n",
       "      <td id=\"T_ee657_row5_col5\" class=\"data row5 col5\" >0.5507</td>\n",
       "      <td id=\"T_ee657_row5_col6\" class=\"data row5 col6\" >85.6433</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ee657_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_ee657_row6_col0\" class=\"data row6 col0\" >2.2278</td>\n",
       "      <td id=\"T_ee657_row6_col1\" class=\"data row6 col1\" >10.2845</td>\n",
       "      <td id=\"T_ee657_row6_col2\" class=\"data row6 col2\" >3.2070</td>\n",
       "      <td id=\"T_ee657_row6_col3\" class=\"data row6 col3\" >0.6423</td>\n",
       "      <td id=\"T_ee657_row6_col4\" class=\"data row6 col4\" >0.6027</td>\n",
       "      <td id=\"T_ee657_row6_col5\" class=\"data row6 col5\" >0.6207</td>\n",
       "      <td id=\"T_ee657_row6_col6\" class=\"data row6 col6\" >75.3185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ee657_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_ee657_row7_col0\" class=\"data row7 col0\" >1.8282</td>\n",
       "      <td id=\"T_ee657_row7_col1\" class=\"data row7 col1\" >6.7561</td>\n",
       "      <td id=\"T_ee657_row7_col2\" class=\"data row7 col2\" >2.5992</td>\n",
       "      <td id=\"T_ee657_row7_col3\" class=\"data row7 col3\" >0.8075</td>\n",
       "      <td id=\"T_ee657_row7_col4\" class=\"data row7 col4\" >0.4270</td>\n",
       "      <td id=\"T_ee657_row7_col5\" class=\"data row7 col5\" >0.4498</td>\n",
       "      <td id=\"T_ee657_row7_col6\" class=\"data row7 col6\" >81.5777</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ee657_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_ee657_row8_col0\" class=\"data row8 col0\" >2.4492</td>\n",
       "      <td id=\"T_ee657_row8_col1\" class=\"data row8 col1\" >12.7253</td>\n",
       "      <td id=\"T_ee657_row8_col2\" class=\"data row8 col2\" >3.5672</td>\n",
       "      <td id=\"T_ee657_row8_col3\" class=\"data row8 col3\" >0.7394</td>\n",
       "      <td id=\"T_ee657_row8_col4\" class=\"data row8 col4\" >0.5191</td>\n",
       "      <td id=\"T_ee657_row8_col5\" class=\"data row8 col5\" >0.5300</td>\n",
       "      <td id=\"T_ee657_row8_col6\" class=\"data row8 col6\" >68.2006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ee657_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_ee657_row9_col0\" class=\"data row9 col0\" >1.8563</td>\n",
       "      <td id=\"T_ee657_row9_col1\" class=\"data row9 col1\" >5.7468</td>\n",
       "      <td id=\"T_ee657_row9_col2\" class=\"data row9 col2\" >2.3972</td>\n",
       "      <td id=\"T_ee657_row9_col3\" class=\"data row9 col3\" >0.8229</td>\n",
       "      <td id=\"T_ee657_row9_col4\" class=\"data row9 col4\" >0.5535</td>\n",
       "      <td id=\"T_ee657_row9_col5\" class=\"data row9 col5\" >0.6310</td>\n",
       "      <td id=\"T_ee657_row9_col6\" class=\"data row9 col6\" >75.6818</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ee657_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_ee657_row10_col0\" class=\"data row10 col0\" >2.2811</td>\n",
       "      <td id=\"T_ee657_row10_col1\" class=\"data row10 col1\" >10.3961</td>\n",
       "      <td id=\"T_ee657_row10_col2\" class=\"data row10 col2\" >3.1981</td>\n",
       "      <td id=\"T_ee657_row10_col3\" class=\"data row10 col3\" >0.7043</td>\n",
       "      <td id=\"T_ee657_row10_col4\" class=\"data row10 col4\" >0.6074</td>\n",
       "      <td id=\"T_ee657_row10_col5\" class=\"data row10 col5\" >0.5932</td>\n",
       "      <td id=\"T_ee657_row10_col6\" class=\"data row10 col6\" >81.9446</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ee657_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_ee657_row11_col0\" class=\"data row11 col0\" >0.2705</td>\n",
       "      <td id=\"T_ee657_row11_col1\" class=\"data row11 col1\" >2.4946</td>\n",
       "      <td id=\"T_ee657_row11_col2\" class=\"data row11 col2\" >0.4099</td>\n",
       "      <td id=\"T_ee657_row11_col3\" class=\"data row11 col3\" >0.0697</td>\n",
       "      <td id=\"T_ee657_row11_col4\" class=\"data row11 col4\" >0.0889</td>\n",
       "      <td id=\"T_ee657_row11_col5\" class=\"data row11 col5\" >0.1177</td>\n",
       "      <td id=\"T_ee657_row11_col6\" class=\"data row11 col6\" >7.5051</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x27539721520>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_b27f3_row10_col0, #T_b27f3_row10_col1, #T_b27f3_row10_col2, #T_b27f3_row10_col3, #T_b27f3_row10_col4, #T_b27f3_row10_col5, #T_b27f3_row10_col6 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_b27f3\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_b27f3_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_b27f3_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_b27f3_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_b27f3_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_b27f3_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_b27f3_level0_col5\" class=\"col_heading level0 col5\" >MAPE</th>\n",
       "      <th id=\"T_b27f3_level0_col6\" class=\"col_heading level0 col6\" >SMAPE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >Fold</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "      <th class=\"blank col6\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_b27f3_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_b27f3_row0_col0\" class=\"data row0 col0\" >2.1715</td>\n",
       "      <td id=\"T_b27f3_row0_col1\" class=\"data row0 col1\" >8.2885</td>\n",
       "      <td id=\"T_b27f3_row0_col2\" class=\"data row0 col2\" >2.8790</td>\n",
       "      <td id=\"T_b27f3_row0_col3\" class=\"data row0 col3\" >0.7399</td>\n",
       "      <td id=\"T_b27f3_row0_col4\" class=\"data row0 col4\" >0.6452</td>\n",
       "      <td id=\"T_b27f3_row0_col5\" class=\"data row0 col5\" >0.4561</td>\n",
       "      <td id=\"T_b27f3_row0_col6\" class=\"data row0 col6\" >77.3723</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b27f3_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_b27f3_row1_col0\" class=\"data row1 col0\" >2.2090</td>\n",
       "      <td id=\"T_b27f3_row1_col1\" class=\"data row1 col1\" >9.8870</td>\n",
       "      <td id=\"T_b27f3_row1_col2\" class=\"data row1 col2\" >3.1444</td>\n",
       "      <td id=\"T_b27f3_row1_col3\" class=\"data row1 col3\" >0.7266</td>\n",
       "      <td id=\"T_b27f3_row1_col4\" class=\"data row1 col4\" >0.7378</td>\n",
       "      <td id=\"T_b27f3_row1_col5\" class=\"data row1 col5\" >0.4937</td>\n",
       "      <td id=\"T_b27f3_row1_col6\" class=\"data row1 col6\" >91.9422</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b27f3_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_b27f3_row2_col0\" class=\"data row2 col0\" >2.3592</td>\n",
       "      <td id=\"T_b27f3_row2_col1\" class=\"data row2 col1\" >10.8221</td>\n",
       "      <td id=\"T_b27f3_row2_col2\" class=\"data row2 col2\" >3.2897</td>\n",
       "      <td id=\"T_b27f3_row2_col3\" class=\"data row2 col3\" >0.7222</td>\n",
       "      <td id=\"T_b27f3_row2_col4\" class=\"data row2 col4\" >0.6408</td>\n",
       "      <td id=\"T_b27f3_row2_col5\" class=\"data row2 col5\" >0.5853</td>\n",
       "      <td id=\"T_b27f3_row2_col6\" class=\"data row2 col6\" >78.0634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b27f3_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_b27f3_row3_col0\" class=\"data row3 col0\" >2.5918</td>\n",
       "      <td id=\"T_b27f3_row3_col1\" class=\"data row3 col1\" >11.1839</td>\n",
       "      <td id=\"T_b27f3_row3_col2\" class=\"data row3 col2\" >3.3442</td>\n",
       "      <td id=\"T_b27f3_row3_col3\" class=\"data row3 col3\" >0.5942</td>\n",
       "      <td id=\"T_b27f3_row3_col4\" class=\"data row3 col4\" >0.6622</td>\n",
       "      <td id=\"T_b27f3_row3_col5\" class=\"data row3 col5\" >0.5560</td>\n",
       "      <td id=\"T_b27f3_row3_col6\" class=\"data row3 col6\" >79.8231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b27f3_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_b27f3_row4_col0\" class=\"data row4 col0\" >2.4902</td>\n",
       "      <td id=\"T_b27f3_row4_col1\" class=\"data row4 col1\" >12.2851</td>\n",
       "      <td id=\"T_b27f3_row4_col2\" class=\"data row4 col2\" >3.5050</td>\n",
       "      <td id=\"T_b27f3_row4_col3\" class=\"data row4 col3\" >0.6873</td>\n",
       "      <td id=\"T_b27f3_row4_col4\" class=\"data row4 col4\" >0.7401</td>\n",
       "      <td id=\"T_b27f3_row4_col5\" class=\"data row4 col5\" >0.7689</td>\n",
       "      <td id=\"T_b27f3_row4_col6\" class=\"data row4 col6\" >88.3986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b27f3_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_b27f3_row5_col0\" class=\"data row5 col0\" >2.6749</td>\n",
       "      <td id=\"T_b27f3_row5_col1\" class=\"data row5 col1\" >13.3083</td>\n",
       "      <td id=\"T_b27f3_row5_col2\" class=\"data row5 col2\" >3.6481</td>\n",
       "      <td id=\"T_b27f3_row5_col3\" class=\"data row5 col3\" >0.6297</td>\n",
       "      <td id=\"T_b27f3_row5_col4\" class=\"data row5 col4\" >0.6619</td>\n",
       "      <td id=\"T_b27f3_row5_col5\" class=\"data row5 col5\" >0.5081</td>\n",
       "      <td id=\"T_b27f3_row5_col6\" class=\"data row5 col6\" >80.7748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b27f3_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_b27f3_row6_col0\" class=\"data row6 col0\" >2.2903</td>\n",
       "      <td id=\"T_b27f3_row6_col1\" class=\"data row6 col1\" >8.9903</td>\n",
       "      <td id=\"T_b27f3_row6_col2\" class=\"data row6 col2\" >2.9984</td>\n",
       "      <td id=\"T_b27f3_row6_col3\" class=\"data row6 col3\" >0.6873</td>\n",
       "      <td id=\"T_b27f3_row6_col4\" class=\"data row6 col4\" >0.6332</td>\n",
       "      <td id=\"T_b27f3_row6_col5\" class=\"data row6 col5\" >0.5390</td>\n",
       "      <td id=\"T_b27f3_row6_col6\" class=\"data row6 col6\" >75.4658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b27f3_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_b27f3_row7_col0\" class=\"data row7 col0\" >2.1658</td>\n",
       "      <td id=\"T_b27f3_row7_col1\" class=\"data row7 col1\" >7.6867</td>\n",
       "      <td id=\"T_b27f3_row7_col2\" class=\"data row7 col2\" >2.7725</td>\n",
       "      <td id=\"T_b27f3_row7_col3\" class=\"data row7 col3\" >0.7810</td>\n",
       "      <td id=\"T_b27f3_row7_col4\" class=\"data row7 col4\" >0.6018</td>\n",
       "      <td id=\"T_b27f3_row7_col5\" class=\"data row7 col5\" >0.4817</td>\n",
       "      <td id=\"T_b27f3_row7_col6\" class=\"data row7 col6\" >81.5664</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b27f3_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_b27f3_row8_col0\" class=\"data row8 col0\" >2.7333</td>\n",
       "      <td id=\"T_b27f3_row8_col1\" class=\"data row8 col1\" >15.2333</td>\n",
       "      <td id=\"T_b27f3_row8_col2\" class=\"data row8 col2\" >3.9030</td>\n",
       "      <td id=\"T_b27f3_row8_col3\" class=\"data row8 col3\" >0.6881</td>\n",
       "      <td id=\"T_b27f3_row8_col4\" class=\"data row8 col4\" >0.5883</td>\n",
       "      <td id=\"T_b27f3_row8_col5\" class=\"data row8 col5\" >0.4995</td>\n",
       "      <td id=\"T_b27f3_row8_col6\" class=\"data row8 col6\" >67.4565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b27f3_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_b27f3_row9_col0\" class=\"data row9 col0\" >1.8696</td>\n",
       "      <td id=\"T_b27f3_row9_col1\" class=\"data row9 col1\" >5.3561</td>\n",
       "      <td id=\"T_b27f3_row9_col2\" class=\"data row9 col2\" >2.3143</td>\n",
       "      <td id=\"T_b27f3_row9_col3\" class=\"data row9 col3\" >0.8350</td>\n",
       "      <td id=\"T_b27f3_row9_col4\" class=\"data row9 col4\" >0.5976</td>\n",
       "      <td id=\"T_b27f3_row9_col5\" class=\"data row9 col5\" >0.5430</td>\n",
       "      <td id=\"T_b27f3_row9_col6\" class=\"data row9 col6\" >73.3666</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b27f3_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_b27f3_row10_col0\" class=\"data row10 col0\" >2.3556</td>\n",
       "      <td id=\"T_b27f3_row10_col1\" class=\"data row10 col1\" >10.3041</td>\n",
       "      <td id=\"T_b27f3_row10_col2\" class=\"data row10 col2\" >3.1799</td>\n",
       "      <td id=\"T_b27f3_row10_col3\" class=\"data row10 col3\" >0.7091</td>\n",
       "      <td id=\"T_b27f3_row10_col4\" class=\"data row10 col4\" >0.6509</td>\n",
       "      <td id=\"T_b27f3_row10_col5\" class=\"data row10 col5\" >0.5431</td>\n",
       "      <td id=\"T_b27f3_row10_col6\" class=\"data row10 col6\" >79.4230</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b27f3_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_b27f3_row11_col0\" class=\"data row11 col0\" >0.2550</td>\n",
       "      <td id=\"T_b27f3_row11_col1\" class=\"data row11 col1\" >2.7493</td>\n",
       "      <td id=\"T_b27f3_row11_col2\" class=\"data row11 col2\" >0.4389</td>\n",
       "      <td id=\"T_b27f3_row11_col3\" class=\"data row11 col3\" >0.0658</td>\n",
       "      <td id=\"T_b27f3_row11_col4\" class=\"data row11 col4\" >0.0504</td>\n",
       "      <td id=\"T_b27f3_row11_col5\" class=\"data row11 col5\" >0.0836</td>\n",
       "      <td id=\"T_b27f3_row11_col6\" class=\"data row11 col6\" >6.6743</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x27538537910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 10 candidates, totalling 100 fits\n"
     ]
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_81e86_row10_col0, #T_81e86_row10_col1, #T_81e86_row10_col2, #T_81e86_row10_col3, #T_81e86_row10_col4, #T_81e86_row10_col5, #T_81e86_row10_col6 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_81e86\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_81e86_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_81e86_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_81e86_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_81e86_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_81e86_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_81e86_level0_col5\" class=\"col_heading level0 col5\" >MAPE</th>\n",
       "      <th id=\"T_81e86_level0_col6\" class=\"col_heading level0 col6\" >SMAPE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >Fold</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "      <th class=\"blank col6\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_81e86_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_81e86_row0_col0\" class=\"data row0 col0\" >2.2501</td>\n",
       "      <td id=\"T_81e86_row0_col1\" class=\"data row0 col1\" >8.6878</td>\n",
       "      <td id=\"T_81e86_row0_col2\" class=\"data row0 col2\" >2.9475</td>\n",
       "      <td id=\"T_81e86_row0_col3\" class=\"data row0 col3\" >0.7274</td>\n",
       "      <td id=\"T_81e86_row0_col4\" class=\"data row0 col4\" >0.7384</td>\n",
       "      <td id=\"T_81e86_row0_col5\" class=\"data row0 col5\" >0.4425</td>\n",
       "      <td id=\"T_81e86_row0_col6\" class=\"data row0 col6\" >75.6133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_81e86_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_81e86_row1_col0\" class=\"data row1 col0\" >2.7049</td>\n",
       "      <td id=\"T_81e86_row1_col1\" class=\"data row1 col1\" >12.2398</td>\n",
       "      <td id=\"T_81e86_row1_col2\" class=\"data row1 col2\" >3.4985</td>\n",
       "      <td id=\"T_81e86_row1_col3\" class=\"data row1 col3\" >0.6615</td>\n",
       "      <td id=\"T_81e86_row1_col4\" class=\"data row1 col4\" >0.9259</td>\n",
       "      <td id=\"T_81e86_row1_col5\" class=\"data row1 col5\" >0.5871</td>\n",
       "      <td id=\"T_81e86_row1_col6\" class=\"data row1 col6\" >92.9019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_81e86_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_81e86_row2_col0\" class=\"data row2 col0\" >2.9788</td>\n",
       "      <td id=\"T_81e86_row2_col1\" class=\"data row2 col1\" >13.0663</td>\n",
       "      <td id=\"T_81e86_row2_col2\" class=\"data row2 col2\" >3.6147</td>\n",
       "      <td id=\"T_81e86_row2_col3\" class=\"data row2 col3\" >0.6646</td>\n",
       "      <td id=\"T_81e86_row2_col4\" class=\"data row2 col4\" >0.8603</td>\n",
       "      <td id=\"T_81e86_row2_col5\" class=\"data row2 col5\" >0.7848</td>\n",
       "      <td id=\"T_81e86_row2_col6\" class=\"data row2 col6\" >83.1692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_81e86_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_81e86_row3_col0\" class=\"data row3 col0\" >2.7420</td>\n",
       "      <td id=\"T_81e86_row3_col1\" class=\"data row3 col1\" >10.9417</td>\n",
       "      <td id=\"T_81e86_row3_col2\" class=\"data row3 col2\" >3.3078</td>\n",
       "      <td id=\"T_81e86_row3_col3\" class=\"data row3 col3\" >0.6030</td>\n",
       "      <td id=\"T_81e86_row3_col4\" class=\"data row3 col4\" >0.7658</td>\n",
       "      <td id=\"T_81e86_row3_col5\" class=\"data row3 col5\" >0.5902</td>\n",
       "      <td id=\"T_81e86_row3_col6\" class=\"data row3 col6\" >76.6801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_81e86_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_81e86_row4_col0\" class=\"data row4 col0\" >2.7458</td>\n",
       "      <td id=\"T_81e86_row4_col1\" class=\"data row4 col1\" >12.6783</td>\n",
       "      <td id=\"T_81e86_row4_col2\" class=\"data row4 col2\" >3.5607</td>\n",
       "      <td id=\"T_81e86_row4_col3\" class=\"data row4 col3\" >0.6773</td>\n",
       "      <td id=\"T_81e86_row4_col4\" class=\"data row4 col4\" >0.8372</td>\n",
       "      <td id=\"T_81e86_row4_col5\" class=\"data row4 col5\" >0.8067</td>\n",
       "      <td id=\"T_81e86_row4_col6\" class=\"data row4 col6\" >88.7642</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_81e86_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_81e86_row5_col0\" class=\"data row5 col0\" >2.8383</td>\n",
       "      <td id=\"T_81e86_row5_col1\" class=\"data row5 col1\" >13.4247</td>\n",
       "      <td id=\"T_81e86_row5_col2\" class=\"data row5 col2\" >3.6640</td>\n",
       "      <td id=\"T_81e86_row5_col3\" class=\"data row5 col3\" >0.6265</td>\n",
       "      <td id=\"T_81e86_row5_col4\" class=\"data row5 col4\" >0.7421</td>\n",
       "      <td id=\"T_81e86_row5_col5\" class=\"data row5 col5\" >0.5358</td>\n",
       "      <td id=\"T_81e86_row5_col6\" class=\"data row5 col6\" >79.0338</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_81e86_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_81e86_row6_col0\" class=\"data row6 col0\" >2.6586</td>\n",
       "      <td id=\"T_81e86_row6_col1\" class=\"data row6 col1\" >11.3268</td>\n",
       "      <td id=\"T_81e86_row6_col2\" class=\"data row6 col2\" >3.3655</td>\n",
       "      <td id=\"T_81e86_row6_col3\" class=\"data row6 col3\" >0.6060</td>\n",
       "      <td id=\"T_81e86_row6_col4\" class=\"data row6 col4\" >0.8016</td>\n",
       "      <td id=\"T_81e86_row6_col5\" class=\"data row6 col5\" >0.6315</td>\n",
       "      <td id=\"T_81e86_row6_col6\" class=\"data row6 col6\" >75.5150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_81e86_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_81e86_row7_col0\" class=\"data row7 col0\" >2.6148</td>\n",
       "      <td id=\"T_81e86_row7_col1\" class=\"data row7 col1\" >9.1811</td>\n",
       "      <td id=\"T_81e86_row7_col2\" class=\"data row7 col2\" >3.0300</td>\n",
       "      <td id=\"T_81e86_row7_col3\" class=\"data row7 col3\" >0.7385</td>\n",
       "      <td id=\"T_81e86_row7_col4\" class=\"data row7 col4\" >0.8343</td>\n",
       "      <td id=\"T_81e86_row7_col5\" class=\"data row7 col5\" >0.5249</td>\n",
       "      <td id=\"T_81e86_row7_col6\" class=\"data row7 col6\" >81.4771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_81e86_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_81e86_row8_col0\" class=\"data row8 col0\" >2.9796</td>\n",
       "      <td id=\"T_81e86_row8_col1\" class=\"data row8 col1\" >15.1008</td>\n",
       "      <td id=\"T_81e86_row8_col2\" class=\"data row8 col2\" >3.8860</td>\n",
       "      <td id=\"T_81e86_row8_col3\" class=\"data row8 col3\" >0.6908</td>\n",
       "      <td id=\"T_81e86_row8_col4\" class=\"data row8 col4\" >0.7035</td>\n",
       "      <td id=\"T_81e86_row8_col5\" class=\"data row8 col5\" >0.5865</td>\n",
       "      <td id=\"T_81e86_row8_col6\" class=\"data row8 col6\" >67.5533</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_81e86_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_81e86_row9_col0\" class=\"data row9 col0\" >2.3254</td>\n",
       "      <td id=\"T_81e86_row9_col1\" class=\"data row9 col1\" >7.2601</td>\n",
       "      <td id=\"T_81e86_row9_col2\" class=\"data row9 col2\" >2.6945</td>\n",
       "      <td id=\"T_81e86_row9_col3\" class=\"data row9 col3\" >0.7763</td>\n",
       "      <td id=\"T_81e86_row9_col4\" class=\"data row9 col4\" >0.8029</td>\n",
       "      <td id=\"T_81e86_row9_col5\" class=\"data row9 col5\" >0.7096</td>\n",
       "      <td id=\"T_81e86_row9_col6\" class=\"data row9 col6\" >75.5810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_81e86_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_81e86_row10_col0\" class=\"data row10 col0\" >2.6838</td>\n",
       "      <td id=\"T_81e86_row10_col1\" class=\"data row10 col1\" >11.3907</td>\n",
       "      <td id=\"T_81e86_row10_col2\" class=\"data row10 col2\" >3.3569</td>\n",
       "      <td id=\"T_81e86_row10_col3\" class=\"data row10 col3\" >0.6772</td>\n",
       "      <td id=\"T_81e86_row10_col4\" class=\"data row10 col4\" >0.8012</td>\n",
       "      <td id=\"T_81e86_row10_col5\" class=\"data row10 col5\" >0.6200</td>\n",
       "      <td id=\"T_81e86_row10_col6\" class=\"data row10 col6\" >79.6289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_81e86_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_81e86_row11_col0\" class=\"data row11 col0\" >0.2298</td>\n",
       "      <td id=\"T_81e86_row11_col1\" class=\"data row11 col1\" >2.2923</td>\n",
       "      <td id=\"T_81e86_row11_col2\" class=\"data row11 col2\" >0.3490</td>\n",
       "      <td id=\"T_81e86_row11_col3\" class=\"data row11 col3\" >0.0546</td>\n",
       "      <td id=\"T_81e86_row11_col4\" class=\"data row11 col4\" >0.0630</td>\n",
       "      <td id=\"T_81e86_row11_col5\" class=\"data row11 col5\" >0.1100</td>\n",
       "      <td id=\"T_81e86_row11_col6\" class=\"data row11 col6\" >6.9273</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x275394b6250>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original model was better than the ensembled model, hence it will be returned. NOTE: The display metrics are for the ensembled model (not the original one).\n"
     ]
    }
   ],
   "source": [
    "dataset = pd.merge(clinical_data_train[['updrs_2',\n",
    "                                        'offset_updrs1_offset',\n",
    "                                        'offset_updrs2_offset',\n",
    "                                        'offset_updrs3_offset',\n",
    "                                        'offset_updrs4_offset',\n",
    "                                        'visit_month']],\n",
    "                   npx_features,\n",
    "                   how='inner',\n",
    "                   left_on=['visit_id'],\n",
    "                   right_on=['visit_id'])\n",
    "dataset.head()\n",
    "Y = dataset[dataset.columns[0]].values\n",
    "X = dataset[dataset.columns[6:]].values\n",
    "X_noscale = dataset[dataset.columns[1:6]].values\n",
    "X_scaled = StandardScaler().fit_transform(X)\n",
    "final_X = np.hstack((X_noscale, X_scaled))\n",
    "\n",
    "s = setup(data=final_X, target=Y)\n",
    "add_metric('SMAPE', 'SMAPE', smape, greater_is_better=False)\n",
    "b2 = create_model('et')\n",
    "t2 = tune_model(b2, choose_better=True, optimize='SMAPE')\n",
    "ensemble_u2 = ensemble_model(\n",
    "    t2,\n",
    "    choose_better=True,\n",
    "    n_estimators=20,\n",
    "    method='Boosting')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_5ff10\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_5ff10_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_5ff10_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_5ff10_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_5ff10_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_5ff10_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_5ff10_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_5ff10_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_5ff10_level0_col7\" class=\"col_heading level0 col7\" >SMAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_5ff10_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_5ff10_row0_col0\" class=\"data row0 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_5ff10_row0_col1\" class=\"data row0 col1\" >2.2543</td>\n",
       "      <td id=\"T_5ff10_row0_col2\" class=\"data row0 col2\" >9.9189</td>\n",
       "      <td id=\"T_5ff10_row0_col3\" class=\"data row0 col3\" >3.1494</td>\n",
       "      <td id=\"T_5ff10_row0_col4\" class=\"data row0 col4\" >0.7102</td>\n",
       "      <td id=\"T_5ff10_row0_col5\" class=\"data row0 col5\" >0.6573</td>\n",
       "      <td id=\"T_5ff10_row0_col6\" class=\"data row0 col6\" >0.5158</td>\n",
       "      <td id=\"T_5ff10_row0_col7\" class=\"data row0 col7\" >84.0014</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x27538581160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_9567f\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_9567f_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_9567f_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_9567f_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_9567f_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_9567f_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_9567f_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_9567f_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_9567f_level0_col7\" class=\"col_heading level0 col7\" >SMAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_9567f_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_9567f_row0_col0\" class=\"data row0 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_9567f_row0_col1\" class=\"data row0 col1\" >2.3281</td>\n",
       "      <td id=\"T_9567f_row0_col2\" class=\"data row0 col2\" >9.9262</td>\n",
       "      <td id=\"T_9567f_row0_col3\" class=\"data row0 col3\" >3.1506</td>\n",
       "      <td id=\"T_9567f_row0_col4\" class=\"data row0 col4\" >0.7100</td>\n",
       "      <td id=\"T_9567f_row0_col5\" class=\"data row0 col5\" >0.6819</td>\n",
       "      <td id=\"T_9567f_row0_col6\" class=\"data row0 col6\" >0.4535</td>\n",
       "      <td id=\"T_9567f_row0_col7\" class=\"data row0 col7\" >84.4329</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x275385c03d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_31662\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_31662_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_31662_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_31662_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_31662_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_31662_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_31662_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_31662_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_31662_level0_col7\" class=\"col_heading level0 col7\" >SMAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_31662_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_31662_row0_col0\" class=\"data row0 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_31662_row0_col1\" class=\"data row0 col1\" >2.3281</td>\n",
       "      <td id=\"T_31662_row0_col2\" class=\"data row0 col2\" >9.9262</td>\n",
       "      <td id=\"T_31662_row0_col3\" class=\"data row0 col3\" >3.1506</td>\n",
       "      <td id=\"T_31662_row0_col4\" class=\"data row0 col4\" >0.7100</td>\n",
       "      <td id=\"T_31662_row0_col5\" class=\"data row0 col5\" >0.6819</td>\n",
       "      <td id=\"T_31662_row0_col6\" class=\"data row0 col6\" >0.4535</td>\n",
       "      <td id=\"T_31662_row0_col7\" class=\"data row0 col7\" >84.4329</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x27538581b20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-2 {color: black;background-color: white;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(memory=FastMemory(location=C:\\Users\\prith\\AppData\\Local\\Temp\\joblib),\n",
       "         steps=[(&#x27;numerical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[&#x27;feature_1&#x27;, &#x27;feature_2&#x27;,\n",
       "                                             &#x27;feature_3&#x27;, &#x27;feature_4&#x27;,\n",
       "                                             &#x27;feature_5&#x27;, &#x27;feature_6&#x27;,\n",
       "                                             &#x27;feature_7&#x27;, &#x27;feature_8&#x27;,\n",
       "                                             &#x27;feature_9&#x27;, &#x27;feature_10&#x27;,\n",
       "                                             &#x27;feature_11&#x27;, &#x27;feature_12&#x27;,\n",
       "                                             &#x27;feature_13&#x27;, &#x27;feature_14&#x27;,\n",
       "                                             &#x27;feature_15&#x27;, &#x27;feature_16&#x27;,\n",
       "                                             &#x27;feature_17&#x27;, &#x27;fea...\n",
       "                                             &#x27;feature_29&#x27;, &#x27;feature_30&#x27;, ...],\n",
       "                                    transformer=SimpleImputer())),\n",
       "                (&#x27;categorical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[],\n",
       "                                    transformer=SimpleImputer(strategy=&#x27;most_frequent&#x27;))),\n",
       "                (&#x27;actual_estimator&#x27;,\n",
       "                 ExtraTreesRegressor(bootstrap=True, max_depth=4,\n",
       "                                     min_impurity_decrease=0.0005,\n",
       "                                     min_samples_leaf=3, min_samples_split=5,\n",
       "                                     n_estimators=30, n_jobs=-1,\n",
       "                                     random_state=4350))])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-9\" type=\"checkbox\" ><label for=\"sk-estimator-id-9\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(memory=FastMemory(location=C:\\Users\\prith\\AppData\\Local\\Temp\\joblib),\n",
       "         steps=[(&#x27;numerical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[&#x27;feature_1&#x27;, &#x27;feature_2&#x27;,\n",
       "                                             &#x27;feature_3&#x27;, &#x27;feature_4&#x27;,\n",
       "                                             &#x27;feature_5&#x27;, &#x27;feature_6&#x27;,\n",
       "                                             &#x27;feature_7&#x27;, &#x27;feature_8&#x27;,\n",
       "                                             &#x27;feature_9&#x27;, &#x27;feature_10&#x27;,\n",
       "                                             &#x27;feature_11&#x27;, &#x27;feature_12&#x27;,\n",
       "                                             &#x27;feature_13&#x27;, &#x27;feature_14&#x27;,\n",
       "                                             &#x27;feature_15&#x27;, &#x27;feature_16&#x27;,\n",
       "                                             &#x27;feature_17&#x27;, &#x27;fea...\n",
       "                                             &#x27;feature_29&#x27;, &#x27;feature_30&#x27;, ...],\n",
       "                                    transformer=SimpleImputer())),\n",
       "                (&#x27;categorical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[],\n",
       "                                    transformer=SimpleImputer(strategy=&#x27;most_frequent&#x27;))),\n",
       "                (&#x27;actual_estimator&#x27;,\n",
       "                 ExtraTreesRegressor(bootstrap=True, max_depth=4,\n",
       "                                     min_impurity_decrease=0.0005,\n",
       "                                     min_samples_leaf=3, min_samples_split=5,\n",
       "                                     n_estimators=30, n_jobs=-1,\n",
       "                                     random_state=4350))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-10\" type=\"checkbox\" ><label for=\"sk-estimator-id-10\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">numerical_imputer: TransformerWrapper</label><div class=\"sk-toggleable__content\"><pre>TransformerWrapper(include=[&#x27;feature_1&#x27;, &#x27;feature_2&#x27;, &#x27;feature_3&#x27;, &#x27;feature_4&#x27;,\n",
       "                            &#x27;feature_5&#x27;, &#x27;feature_6&#x27;, &#x27;feature_7&#x27;, &#x27;feature_8&#x27;,\n",
       "                            &#x27;feature_9&#x27;, &#x27;feature_10&#x27;, &#x27;feature_11&#x27;,\n",
       "                            &#x27;feature_12&#x27;, &#x27;feature_13&#x27;, &#x27;feature_14&#x27;,\n",
       "                            &#x27;feature_15&#x27;, &#x27;feature_16&#x27;, &#x27;feature_17&#x27;,\n",
       "                            &#x27;feature_18&#x27;, &#x27;feature_19&#x27;, &#x27;feature_20&#x27;,\n",
       "                            &#x27;feature_21&#x27;, &#x27;feature_22&#x27;, &#x27;feature_23&#x27;,\n",
       "                            &#x27;feature_24&#x27;, &#x27;feature_25&#x27;, &#x27;feature_26&#x27;,\n",
       "                            &#x27;feature_27&#x27;, &#x27;feature_28&#x27;, &#x27;feature_29&#x27;,\n",
       "                            &#x27;feature_30&#x27;, ...],\n",
       "                   transformer=SimpleImputer())</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-11\" type=\"checkbox\" ><label for=\"sk-estimator-id-11\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">transformer: SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer()</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-12\" type=\"checkbox\" ><label for=\"sk-estimator-id-12\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-13\" type=\"checkbox\" ><label for=\"sk-estimator-id-13\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">categorical_imputer: TransformerWrapper</label><div class=\"sk-toggleable__content\"><pre>TransformerWrapper(include=[],\n",
       "                   transformer=SimpleImputer(strategy=&#x27;most_frequent&#x27;))</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-14\" type=\"checkbox\" ><label for=\"sk-estimator-id-14\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">transformer: SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(strategy=&#x27;most_frequent&#x27;)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-15\" type=\"checkbox\" ><label for=\"sk-estimator-id-15\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(strategy=&#x27;most_frequent&#x27;)</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-16\" type=\"checkbox\" ><label for=\"sk-estimator-id-16\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ExtraTreesRegressor</label><div class=\"sk-toggleable__content\"><pre>ExtraTreesRegressor(bootstrap=True, max_depth=4, min_impurity_decrease=0.0005,\n",
       "                    min_samples_leaf=3, min_samples_split=5, n_estimators=30,\n",
       "                    n_jobs=-1, random_state=4350)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(memory=FastMemory(location=C:\\Users\\prith\\AppData\\Local\\Temp\\joblib),\n",
       "         steps=[('numerical_imputer',\n",
       "                 TransformerWrapper(include=['feature_1', 'feature_2',\n",
       "                                             'feature_3', 'feature_4',\n",
       "                                             'feature_5', 'feature_6',\n",
       "                                             'feature_7', 'feature_8',\n",
       "                                             'feature_9', 'feature_10',\n",
       "                                             'feature_11', 'feature_12',\n",
       "                                             'feature_13', 'feature_14',\n",
       "                                             'feature_15', 'feature_16',\n",
       "                                             'feature_17', 'fea...\n",
       "                                             'feature_29', 'feature_30', ...],\n",
       "                                    transformer=SimpleImputer())),\n",
       "                ('categorical_imputer',\n",
       "                 TransformerWrapper(include=[],\n",
       "                                    transformer=SimpleImputer(strategy='most_frequent'))),\n",
       "                ('actual_estimator',\n",
       "                 ExtraTreesRegressor(bootstrap=True, max_depth=4,\n",
       "                                     min_impurity_decrease=0.0005,\n",
       "                                     min_samples_leaf=3, min_samples_split=5,\n",
       "                                     n_estimators=30, n_jobs=-1,\n",
       "                                     random_state=4350))])"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "predict_model(b2)\n",
    "predict_model(t2)\n",
    "predict_model(ensemble_u2)\n",
    "finalize_model(ensemble_u2)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# UPDRS 3 Predictions with Proteins Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_ca0dd_row8_col1 {\n",
       "  background-color: lightgreen;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_ca0dd\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_ca0dd_level0_col0\" class=\"col_heading level0 col0\" >Description</th>\n",
       "      <th id=\"T_ca0dd_level0_col1\" class=\"col_heading level0 col1\" >Value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_ca0dd_row0_col0\" class=\"data row0 col0\" >Session id</td>\n",
       "      <td id=\"T_ca0dd_row0_col1\" class=\"data row0 col1\" >7050</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_ca0dd_row1_col0\" class=\"data row1 col0\" >Target</td>\n",
       "      <td id=\"T_ca0dd_row1_col1\" class=\"data row1 col1\" >target</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_ca0dd_row2_col0\" class=\"data row2 col0\" >Target type</td>\n",
       "      <td id=\"T_ca0dd_row2_col1\" class=\"data row2 col1\" >Regression</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_ca0dd_row3_col0\" class=\"data row3 col0\" >Original data shape</td>\n",
       "      <td id=\"T_ca0dd_row3_col1\" class=\"data row3 col1\" >(1047, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_ca0dd_row4_col0\" class=\"data row4 col0\" >Transformed data shape</td>\n",
       "      <td id=\"T_ca0dd_row4_col1\" class=\"data row4 col1\" >(1047, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_ca0dd_row5_col0\" class=\"data row5 col0\" >Transformed train set shape</td>\n",
       "      <td id=\"T_ca0dd_row5_col1\" class=\"data row5 col1\" >(732, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_ca0dd_row6_col0\" class=\"data row6 col0\" >Transformed test set shape</td>\n",
       "      <td id=\"T_ca0dd_row6_col1\" class=\"data row6 col1\" >(315, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_ca0dd_row7_col0\" class=\"data row7 col0\" >Numeric features</td>\n",
       "      <td id=\"T_ca0dd_row7_col1\" class=\"data row7 col1\" >234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_ca0dd_row8_col0\" class=\"data row8 col0\" >Preprocess</td>\n",
       "      <td id=\"T_ca0dd_row8_col1\" class=\"data row8 col1\" >True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_ca0dd_row9_col0\" class=\"data row9 col0\" >Imputation type</td>\n",
       "      <td id=\"T_ca0dd_row9_col1\" class=\"data row9 col1\" >simple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row10\" class=\"row_heading level0 row10\" >10</th>\n",
       "      <td id=\"T_ca0dd_row10_col0\" class=\"data row10 col0\" >Numeric imputation</td>\n",
       "      <td id=\"T_ca0dd_row10_col1\" class=\"data row10 col1\" >mean</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row11\" class=\"row_heading level0 row11\" >11</th>\n",
       "      <td id=\"T_ca0dd_row11_col0\" class=\"data row11 col0\" >Categorical imputation</td>\n",
       "      <td id=\"T_ca0dd_row11_col1\" class=\"data row11 col1\" >mode</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row12\" class=\"row_heading level0 row12\" >12</th>\n",
       "      <td id=\"T_ca0dd_row12_col0\" class=\"data row12 col0\" >Fold Generator</td>\n",
       "      <td id=\"T_ca0dd_row12_col1\" class=\"data row12 col1\" >KFold</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row13\" class=\"row_heading level0 row13\" >13</th>\n",
       "      <td id=\"T_ca0dd_row13_col0\" class=\"data row13 col0\" >Fold Number</td>\n",
       "      <td id=\"T_ca0dd_row13_col1\" class=\"data row13 col1\" >10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row14\" class=\"row_heading level0 row14\" >14</th>\n",
       "      <td id=\"T_ca0dd_row14_col0\" class=\"data row14 col0\" >CPU Jobs</td>\n",
       "      <td id=\"T_ca0dd_row14_col1\" class=\"data row14 col1\" >-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row15\" class=\"row_heading level0 row15\" >15</th>\n",
       "      <td id=\"T_ca0dd_row15_col0\" class=\"data row15 col0\" >Use GPU</td>\n",
       "      <td id=\"T_ca0dd_row15_col1\" class=\"data row15 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row16\" class=\"row_heading level0 row16\" >16</th>\n",
       "      <td id=\"T_ca0dd_row16_col0\" class=\"data row16 col0\" >Log Experiment</td>\n",
       "      <td id=\"T_ca0dd_row16_col1\" class=\"data row16 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row17\" class=\"row_heading level0 row17\" >17</th>\n",
       "      <td id=\"T_ca0dd_row17_col0\" class=\"data row17 col0\" >Experiment Name</td>\n",
       "      <td id=\"T_ca0dd_row17_col1\" class=\"data row17 col1\" >reg-default-name</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ca0dd_level0_row18\" class=\"row_heading level0 row18\" >18</th>\n",
       "      <td id=\"T_ca0dd_row18_col0\" class=\"data row18 col0\" >USI</td>\n",
       "      <td id=\"T_ca0dd_row18_col1\" class=\"data row18 col1\" >2b0a</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x275385c4640>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_9b99c_row10_col0, #T_9b99c_row10_col1, #T_9b99c_row10_col2, #T_9b99c_row10_col3, #T_9b99c_row10_col4, #T_9b99c_row10_col5, #T_9b99c_row10_col6 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_9b99c\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_9b99c_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_9b99c_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_9b99c_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_9b99c_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_9b99c_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_9b99c_level0_col5\" class=\"col_heading level0 col5\" >MAPE</th>\n",
       "      <th id=\"T_9b99c_level0_col6\" class=\"col_heading level0 col6\" >SMAPE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >Fold</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "      <th class=\"blank col6\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_9b99c_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_9b99c_row0_col0\" class=\"data row0 col0\" >0.4166</td>\n",
       "      <td id=\"T_9b99c_row0_col1\" class=\"data row0 col1\" >2.8177</td>\n",
       "      <td id=\"T_9b99c_row0_col2\" class=\"data row0 col2\" >1.6786</td>\n",
       "      <td id=\"T_9b99c_row0_col3\" class=\"data row0 col3\" >0.9934</td>\n",
       "      <td id=\"T_9b99c_row0_col4\" class=\"data row0 col4\" >0.3332</td>\n",
       "      <td id=\"T_9b99c_row0_col5\" class=\"data row0 col5\" >0.0309</td>\n",
       "      <td id=\"T_9b99c_row0_col6\" class=\"data row0 col6\" >15.9021</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9b99c_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_9b99c_row1_col0\" class=\"data row1 col0\" >0.2631</td>\n",
       "      <td id=\"T_9b99c_row1_col1\" class=\"data row1 col1\" >0.5625</td>\n",
       "      <td id=\"T_9b99c_row1_col2\" class=\"data row1 col2\" >0.7500</td>\n",
       "      <td id=\"T_9b99c_row1_col3\" class=\"data row1 col3\" >0.9986</td>\n",
       "      <td id=\"T_9b99c_row1_col4\" class=\"data row1 col4\" >0.1130</td>\n",
       "      <td id=\"T_9b99c_row1_col5\" class=\"data row1 col5\" >0.0629</td>\n",
       "      <td id=\"T_9b99c_row1_col6\" class=\"data row1 col6\" >4.8346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9b99c_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_9b99c_row2_col0\" class=\"data row2 col0\" >0.5374</td>\n",
       "      <td id=\"T_9b99c_row2_col1\" class=\"data row2 col1\" >0.8693</td>\n",
       "      <td id=\"T_9b99c_row2_col2\" class=\"data row2 col2\" >0.9324</td>\n",
       "      <td id=\"T_9b99c_row2_col3\" class=\"data row2 col3\" >0.9974</td>\n",
       "      <td id=\"T_9b99c_row2_col4\" class=\"data row2 col4\" >0.3280</td>\n",
       "      <td id=\"T_9b99c_row2_col5\" class=\"data row2 col5\" >0.0690</td>\n",
       "      <td id=\"T_9b99c_row2_col6\" class=\"data row2 col6\" >26.8993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9b99c_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_9b99c_row3_col0\" class=\"data row3 col0\" >0.5626</td>\n",
       "      <td id=\"T_9b99c_row3_col1\" class=\"data row3 col1\" >4.1042</td>\n",
       "      <td id=\"T_9b99c_row3_col2\" class=\"data row3 col2\" >2.0259</td>\n",
       "      <td id=\"T_9b99c_row3_col3\" class=\"data row3 col3\" >0.9871</td>\n",
       "      <td id=\"T_9b99c_row3_col4\" class=\"data row3 col4\" >0.4737</td>\n",
       "      <td id=\"T_9b99c_row3_col5\" class=\"data row3 col5\" >0.0288</td>\n",
       "      <td id=\"T_9b99c_row3_col6\" class=\"data row3 col6\" >29.3939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9b99c_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_9b99c_row4_col0\" class=\"data row4 col0\" >1.1432</td>\n",
       "      <td id=\"T_9b99c_row4_col1\" class=\"data row4 col1\" >27.7781</td>\n",
       "      <td id=\"T_9b99c_row4_col2\" class=\"data row4 col2\" >5.2705</td>\n",
       "      <td id=\"T_9b99c_row4_col3\" class=\"data row4 col3\" >0.9319</td>\n",
       "      <td id=\"T_9b99c_row4_col4\" class=\"data row4 col4\" >0.3969</td>\n",
       "      <td id=\"T_9b99c_row4_col5\" class=\"data row4 col5\" >0.0663</td>\n",
       "      <td id=\"T_9b99c_row4_col6\" class=\"data row4 col6\" >19.6261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9b99c_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_9b99c_row5_col0\" class=\"data row5 col0\" >0.4047</td>\n",
       "      <td id=\"T_9b99c_row5_col1\" class=\"data row5 col1\" >1.0506</td>\n",
       "      <td id=\"T_9b99c_row5_col2\" class=\"data row5 col2\" >1.0250</td>\n",
       "      <td id=\"T_9b99c_row5_col3\" class=\"data row5 col3\" >0.9977</td>\n",
       "      <td id=\"T_9b99c_row5_col4\" class=\"data row5 col4\" >0.2172</td>\n",
       "      <td id=\"T_9b99c_row5_col5\" class=\"data row5 col5\" >0.0448</td>\n",
       "      <td id=\"T_9b99c_row5_col6\" class=\"data row5 col6\" >11.7563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9b99c_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_9b99c_row6_col0\" class=\"data row6 col0\" >1.9127</td>\n",
       "      <td id=\"T_9b99c_row6_col1\" class=\"data row6 col1\" >64.4337</td>\n",
       "      <td id=\"T_9b99c_row6_col2\" class=\"data row6 col2\" >8.0271</td>\n",
       "      <td id=\"T_9b99c_row6_col3\" class=\"data row6 col3\" >0.8757</td>\n",
       "      <td id=\"T_9b99c_row6_col4\" class=\"data row6 col4\" >0.4325</td>\n",
       "      <td id=\"T_9b99c_row6_col5\" class=\"data row6 col5\" >0.0538</td>\n",
       "      <td id=\"T_9b99c_row6_col6\" class=\"data row6 col6\" >11.9860</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9b99c_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_9b99c_row7_col0\" class=\"data row7 col0\" >0.7219</td>\n",
       "      <td id=\"T_9b99c_row7_col1\" class=\"data row7 col1\" >3.0194</td>\n",
       "      <td id=\"T_9b99c_row7_col2\" class=\"data row7 col2\" >1.7376</td>\n",
       "      <td id=\"T_9b99c_row7_col3\" class=\"data row7 col3\" >0.9923</td>\n",
       "      <td id=\"T_9b99c_row7_col4\" class=\"data row7 col4\" >0.3941</td>\n",
       "      <td id=\"T_9b99c_row7_col5\" class=\"data row7 col5\" >0.1249</td>\n",
       "      <td id=\"T_9b99c_row7_col6\" class=\"data row7 col6\" >26.1619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9b99c_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_9b99c_row8_col0\" class=\"data row8 col0\" >0.8581</td>\n",
       "      <td id=\"T_9b99c_row8_col1\" class=\"data row8 col1\" >9.0486</td>\n",
       "      <td id=\"T_9b99c_row8_col2\" class=\"data row8 col2\" >3.0081</td>\n",
       "      <td id=\"T_9b99c_row8_col3\" class=\"data row8 col3\" >0.9769</td>\n",
       "      <td id=\"T_9b99c_row8_col4\" class=\"data row8 col4\" >0.5310</td>\n",
       "      <td id=\"T_9b99c_row8_col5\" class=\"data row8 col5\" >0.0262</td>\n",
       "      <td id=\"T_9b99c_row8_col6\" class=\"data row8 col6\" >18.5613</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9b99c_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_9b99c_row9_col0\" class=\"data row9 col0\" >0.8588</td>\n",
       "      <td id=\"T_9b99c_row9_col1\" class=\"data row9 col1\" >11.9397</td>\n",
       "      <td id=\"T_9b99c_row9_col2\" class=\"data row9 col2\" >3.4554</td>\n",
       "      <td id=\"T_9b99c_row9_col3\" class=\"data row9 col3\" >0.9756</td>\n",
       "      <td id=\"T_9b99c_row9_col4\" class=\"data row9 col4\" >0.3236</td>\n",
       "      <td id=\"T_9b99c_row9_col5\" class=\"data row9 col5\" >0.0392</td>\n",
       "      <td id=\"T_9b99c_row9_col6\" class=\"data row9 col6\" >22.5309</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9b99c_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_9b99c_row10_col0\" class=\"data row10 col0\" >0.7679</td>\n",
       "      <td id=\"T_9b99c_row10_col1\" class=\"data row10 col1\" >12.5624</td>\n",
       "      <td id=\"T_9b99c_row10_col2\" class=\"data row10 col2\" >2.7911</td>\n",
       "      <td id=\"T_9b99c_row10_col3\" class=\"data row10 col3\" >0.9727</td>\n",
       "      <td id=\"T_9b99c_row10_col4\" class=\"data row10 col4\" >0.3543</td>\n",
       "      <td id=\"T_9b99c_row10_col5\" class=\"data row10 col5\" >0.0547</td>\n",
       "      <td id=\"T_9b99c_row10_col6\" class=\"data row10 col6\" >18.7652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9b99c_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_9b99c_row11_col0\" class=\"data row11 col0\" >0.4556</td>\n",
       "      <td id=\"T_9b99c_row11_col1\" class=\"data row11 col1\" >18.9887</td>\n",
       "      <td id=\"T_9b99c_row11_col2\" class=\"data row11 col2\" >2.1846</td>\n",
       "      <td id=\"T_9b99c_row11_col3\" class=\"data row11 col3\" >0.0375</td>\n",
       "      <td id=\"T_9b99c_row11_col4\" class=\"data row11 col4\" >0.1158</td>\n",
       "      <td id=\"T_9b99c_row11_col5\" class=\"data row11 col5\" >0.0278</td>\n",
       "      <td id=\"T_9b99c_row11_col6\" class=\"data row11 col6\" >7.3902</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2753905b490>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_56bb2_row10_col0, #T_56bb2_row10_col1, #T_56bb2_row10_col2, #T_56bb2_row10_col3, #T_56bb2_row10_col4, #T_56bb2_row10_col5, #T_56bb2_row10_col6 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_56bb2\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_56bb2_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_56bb2_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_56bb2_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_56bb2_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_56bb2_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_56bb2_level0_col5\" class=\"col_heading level0 col5\" >MAPE</th>\n",
       "      <th id=\"T_56bb2_level0_col6\" class=\"col_heading level0 col6\" >SMAPE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >Fold</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "      <th class=\"blank col6\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_56bb2_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_56bb2_row0_col0\" class=\"data row0 col0\" >0.3280</td>\n",
       "      <td id=\"T_56bb2_row0_col1\" class=\"data row0 col1\" >4.2554</td>\n",
       "      <td id=\"T_56bb2_row0_col2\" class=\"data row0 col2\" >2.0629</td>\n",
       "      <td id=\"T_56bb2_row0_col3\" class=\"data row0 col3\" >0.9901</td>\n",
       "      <td id=\"T_56bb2_row0_col4\" class=\"data row0 col4\" >0.3423</td>\n",
       "      <td id=\"T_56bb2_row0_col5\" class=\"data row0 col5\" >0.0132</td>\n",
       "      <td id=\"T_56bb2_row0_col6\" class=\"data row0 col6\" >14.6722</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_56bb2_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_56bb2_row1_col0\" class=\"data row1 col0\" >0.2948</td>\n",
       "      <td id=\"T_56bb2_row1_col1\" class=\"data row1 col1\" >0.6712</td>\n",
       "      <td id=\"T_56bb2_row1_col2\" class=\"data row1 col2\" >0.8193</td>\n",
       "      <td id=\"T_56bb2_row1_col3\" class=\"data row1 col3\" >0.9983</td>\n",
       "      <td id=\"T_56bb2_row1_col4\" class=\"data row1 col4\" >0.1036</td>\n",
       "      <td id=\"T_56bb2_row1_col5\" class=\"data row1 col5\" >0.0599</td>\n",
       "      <td id=\"T_56bb2_row1_col6\" class=\"data row1 col6\" >4.8399</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_56bb2_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_56bb2_row2_col0\" class=\"data row2 col0\" >0.2515</td>\n",
       "      <td id=\"T_56bb2_row2_col1\" class=\"data row2 col1\" >0.5377</td>\n",
       "      <td id=\"T_56bb2_row2_col2\" class=\"data row2 col2\" >0.7333</td>\n",
       "      <td id=\"T_56bb2_row2_col3\" class=\"data row2 col3\" >0.9984</td>\n",
       "      <td id=\"T_56bb2_row2_col4\" class=\"data row2 col4\" >0.2999</td>\n",
       "      <td id=\"T_56bb2_row2_col5\" class=\"data row2 col5\" >0.0176</td>\n",
       "      <td id=\"T_56bb2_row2_col6\" class=\"data row2 col6\" >17.9005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_56bb2_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_56bb2_row3_col0\" class=\"data row3 col0\" >0.7693</td>\n",
       "      <td id=\"T_56bb2_row3_col1\" class=\"data row3 col1\" >10.8986</td>\n",
       "      <td id=\"T_56bb2_row3_col2\" class=\"data row3 col2\" >3.3013</td>\n",
       "      <td id=\"T_56bb2_row3_col3\" class=\"data row3 col3\" >0.9656</td>\n",
       "      <td id=\"T_56bb2_row3_col4\" class=\"data row3 col4\" >0.5142</td>\n",
       "      <td id=\"T_56bb2_row3_col5\" class=\"data row3 col5\" >0.0226</td>\n",
       "      <td id=\"T_56bb2_row3_col6\" class=\"data row3 col6\" >29.0818</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_56bb2_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_56bb2_row4_col0\" class=\"data row4 col0\" >1.0096</td>\n",
       "      <td id=\"T_56bb2_row4_col1\" class=\"data row4 col1\" >37.3102</td>\n",
       "      <td id=\"T_56bb2_row4_col2\" class=\"data row4 col2\" >6.1082</td>\n",
       "      <td id=\"T_56bb2_row4_col3\" class=\"data row4 col3\" >0.9085</td>\n",
       "      <td id=\"T_56bb2_row4_col4\" class=\"data row4 col4\" >0.4184</td>\n",
       "      <td id=\"T_56bb2_row4_col5\" class=\"data row4 col5\" >0.0205</td>\n",
       "      <td id=\"T_56bb2_row4_col6\" class=\"data row4 col6\" >11.0708</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_56bb2_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_56bb2_row5_col0\" class=\"data row5 col0\" >0.4012</td>\n",
       "      <td id=\"T_56bb2_row5_col1\" class=\"data row5 col1\" >1.3619</td>\n",
       "      <td id=\"T_56bb2_row5_col2\" class=\"data row5 col2\" >1.1670</td>\n",
       "      <td id=\"T_56bb2_row5_col3\" class=\"data row5 col3\" >0.9970</td>\n",
       "      <td id=\"T_56bb2_row5_col4\" class=\"data row5 col4\" >0.2256</td>\n",
       "      <td id=\"T_56bb2_row5_col5\" class=\"data row5 col5\" >0.0639</td>\n",
       "      <td id=\"T_56bb2_row5_col6\" class=\"data row5 col6\" >14.0589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_56bb2_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_56bb2_row6_col0\" class=\"data row6 col0\" >1.7470</td>\n",
       "      <td id=\"T_56bb2_row6_col1\" class=\"data row6 col1\" >67.9880</td>\n",
       "      <td id=\"T_56bb2_row6_col2\" class=\"data row6 col2\" >8.2455</td>\n",
       "      <td id=\"T_56bb2_row6_col3\" class=\"data row6 col3\" >0.8689</td>\n",
       "      <td id=\"T_56bb2_row6_col4\" class=\"data row6 col4\" >0.4284</td>\n",
       "      <td id=\"T_56bb2_row6_col5\" class=\"data row6 col5\" >0.0516</td>\n",
       "      <td id=\"T_56bb2_row6_col6\" class=\"data row6 col6\" >15.0521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_56bb2_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_56bb2_row7_col0\" class=\"data row7 col0\" >0.4934</td>\n",
       "      <td id=\"T_56bb2_row7_col1\" class=\"data row7 col1\" >3.0643</td>\n",
       "      <td id=\"T_56bb2_row7_col2\" class=\"data row7 col2\" >1.7505</td>\n",
       "      <td id=\"T_56bb2_row7_col3\" class=\"data row7 col3\" >0.9922</td>\n",
       "      <td id=\"T_56bb2_row7_col4\" class=\"data row7 col4\" >0.2481</td>\n",
       "      <td id=\"T_56bb2_row7_col5\" class=\"data row7 col5\" >0.0339</td>\n",
       "      <td id=\"T_56bb2_row7_col6\" class=\"data row7 col6\" >22.4406</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_56bb2_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_56bb2_row8_col0\" class=\"data row8 col0\" >0.3809</td>\n",
       "      <td id=\"T_56bb2_row8_col1\" class=\"data row8 col1\" >2.7903</td>\n",
       "      <td id=\"T_56bb2_row8_col2\" class=\"data row8 col2\" >1.6704</td>\n",
       "      <td id=\"T_56bb2_row8_col3\" class=\"data row8 col3\" >0.9929</td>\n",
       "      <td id=\"T_56bb2_row8_col4\" class=\"data row8 col4\" >0.3306</td>\n",
       "      <td id=\"T_56bb2_row8_col5\" class=\"data row8 col5\" >0.0173</td>\n",
       "      <td id=\"T_56bb2_row8_col6\" class=\"data row8 col6\" >17.7580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_56bb2_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_56bb2_row9_col0\" class=\"data row9 col0\" >0.7967</td>\n",
       "      <td id=\"T_56bb2_row9_col1\" class=\"data row9 col1\" >22.9852</td>\n",
       "      <td id=\"T_56bb2_row9_col2\" class=\"data row9 col2\" >4.7943</td>\n",
       "      <td id=\"T_56bb2_row9_col3\" class=\"data row9 col3\" >0.9531</td>\n",
       "      <td id=\"T_56bb2_row9_col4\" class=\"data row9 col4\" >0.2856</td>\n",
       "      <td id=\"T_56bb2_row9_col5\" class=\"data row9 col5\" >0.0207</td>\n",
       "      <td id=\"T_56bb2_row9_col6\" class=\"data row9 col6\" >16.2953</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_56bb2_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_56bb2_row10_col0\" class=\"data row10 col0\" >0.6472</td>\n",
       "      <td id=\"T_56bb2_row10_col1\" class=\"data row10 col1\" >15.1863</td>\n",
       "      <td id=\"T_56bb2_row10_col2\" class=\"data row10 col2\" >3.0653</td>\n",
       "      <td id=\"T_56bb2_row10_col3\" class=\"data row10 col3\" >0.9665</td>\n",
       "      <td id=\"T_56bb2_row10_col4\" class=\"data row10 col4\" >0.3197</td>\n",
       "      <td id=\"T_56bb2_row10_col5\" class=\"data row10 col5\" >0.0321</td>\n",
       "      <td id=\"T_56bb2_row10_col6\" class=\"data row10 col6\" >16.3170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_56bb2_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_56bb2_row11_col0\" class=\"data row11 col0\" >0.4377</td>\n",
       "      <td id=\"T_56bb2_row11_col1\" class=\"data row11 col1\" >20.9599</td>\n",
       "      <td id=\"T_56bb2_row11_col2\" class=\"data row11 col2\" >2.4063</td>\n",
       "      <td id=\"T_56bb2_row11_col3\" class=\"data row11 col3\" >0.0424</td>\n",
       "      <td id=\"T_56bb2_row11_col4\" class=\"data row11 col4\" >0.1106</td>\n",
       "      <td id=\"T_56bb2_row11_col5\" class=\"data row11 col5\" >0.0182</td>\n",
       "      <td id=\"T_56bb2_row11_col6\" class=\"data row11 col6\" >6.1129</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2753855d2b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 10 candidates, totalling 100 fits\n"
     ]
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_4430c_row10_col0, #T_4430c_row10_col1, #T_4430c_row10_col2, #T_4430c_row10_col3, #T_4430c_row10_col4, #T_4430c_row10_col5, #T_4430c_row10_col6 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_4430c\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_4430c_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_4430c_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_4430c_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_4430c_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_4430c_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_4430c_level0_col5\" class=\"col_heading level0 col5\" >MAPE</th>\n",
       "      <th id=\"T_4430c_level0_col6\" class=\"col_heading level0 col6\" >SMAPE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >Fold</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "      <th class=\"blank col6\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_4430c_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_4430c_row0_col0\" class=\"data row0 col0\" >3.1919</td>\n",
       "      <td id=\"T_4430c_row0_col1\" class=\"data row0 col1\" >16.5896</td>\n",
       "      <td id=\"T_4430c_row0_col2\" class=\"data row0 col2\" >4.0730</td>\n",
       "      <td id=\"T_4430c_row0_col3\" class=\"data row0 col3\" >0.9614</td>\n",
       "      <td id=\"T_4430c_row0_col4\" class=\"data row0 col4\" >0.6294</td>\n",
       "      <td id=\"T_4430c_row0_col5\" class=\"data row0 col5\" >0.4627</td>\n",
       "      <td id=\"T_4430c_row0_col6\" class=\"data row0 col6\" >44.3853</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_4430c_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_4430c_row1_col0\" class=\"data row1 col0\" >2.9814</td>\n",
       "      <td id=\"T_4430c_row1_col1\" class=\"data row1 col1\" >14.9018</td>\n",
       "      <td id=\"T_4430c_row1_col2\" class=\"data row1 col2\" >3.8603</td>\n",
       "      <td id=\"T_4430c_row1_col3\" class=\"data row1 col3\" >0.9618</td>\n",
       "      <td id=\"T_4430c_row1_col4\" class=\"data row1 col4\" >0.4149</td>\n",
       "      <td id=\"T_4430c_row1_col5\" class=\"data row1 col5\" >0.5536</td>\n",
       "      <td id=\"T_4430c_row1_col6\" class=\"data row1 col6\" >33.9739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_4430c_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_4430c_row2_col0\" class=\"data row2 col0\" >3.0575</td>\n",
       "      <td id=\"T_4430c_row2_col1\" class=\"data row2 col1\" >13.0401</td>\n",
       "      <td id=\"T_4430c_row2_col2\" class=\"data row2 col2\" >3.6111</td>\n",
       "      <td id=\"T_4430c_row2_col3\" class=\"data row2 col3\" >0.9606</td>\n",
       "      <td id=\"T_4430c_row2_col4\" class=\"data row2 col4\" >0.6843</td>\n",
       "      <td id=\"T_4430c_row2_col5\" class=\"data row2 col5\" >0.5107</td>\n",
       "      <td id=\"T_4430c_row2_col6\" class=\"data row2 col6\" >51.9528</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_4430c_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_4430c_row3_col0\" class=\"data row3 col0\" >3.1074</td>\n",
       "      <td id=\"T_4430c_row3_col1\" class=\"data row3 col1\" >17.9758</td>\n",
       "      <td id=\"T_4430c_row3_col2\" class=\"data row3 col2\" >4.2398</td>\n",
       "      <td id=\"T_4430c_row3_col3\" class=\"data row3 col3\" >0.9433</td>\n",
       "      <td id=\"T_4430c_row3_col4\" class=\"data row3 col4\" >0.8607</td>\n",
       "      <td id=\"T_4430c_row3_col5\" class=\"data row3 col5\" >0.3837</td>\n",
       "      <td id=\"T_4430c_row3_col6\" class=\"data row3 col6\" >61.3232</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_4430c_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_4430c_row4_col0\" class=\"data row4 col0\" >3.8467</td>\n",
       "      <td id=\"T_4430c_row4_col1\" class=\"data row4 col1\" >39.5840</td>\n",
       "      <td id=\"T_4430c_row4_col2\" class=\"data row4 col2\" >6.2916</td>\n",
       "      <td id=\"T_4430c_row4_col3\" class=\"data row4 col3\" >0.9029</td>\n",
       "      <td id=\"T_4430c_row4_col4\" class=\"data row4 col4\" >0.6652</td>\n",
       "      <td id=\"T_4430c_row4_col5\" class=\"data row4 col5\" >0.5504</td>\n",
       "      <td id=\"T_4430c_row4_col6\" class=\"data row4 col6\" >52.0987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_4430c_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_4430c_row5_col0\" class=\"data row5 col0\" >3.4982</td>\n",
       "      <td id=\"T_4430c_row5_col1\" class=\"data row5 col1\" >17.2711</td>\n",
       "      <td id=\"T_4430c_row5_col2\" class=\"data row5 col2\" >4.1559</td>\n",
       "      <td id=\"T_4430c_row5_col3\" class=\"data row5 col3\" >0.9625</td>\n",
       "      <td id=\"T_4430c_row5_col4\" class=\"data row5 col4\" >0.5772</td>\n",
       "      <td id=\"T_4430c_row5_col5\" class=\"data row5 col5\" >0.5257</td>\n",
       "      <td id=\"T_4430c_row5_col6\" class=\"data row5 col6\" >44.4918</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_4430c_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_4430c_row6_col0\" class=\"data row6 col0\" >4.4456</td>\n",
       "      <td id=\"T_4430c_row6_col1\" class=\"data row6 col1\" >68.7649</td>\n",
       "      <td id=\"T_4430c_row6_col2\" class=\"data row6 col2\" >8.2925</td>\n",
       "      <td id=\"T_4430c_row6_col3\" class=\"data row6 col3\" >0.8674</td>\n",
       "      <td id=\"T_4430c_row6_col4\" class=\"data row6 col4\" >0.6298</td>\n",
       "      <td id=\"T_4430c_row6_col5\" class=\"data row6 col5\" >0.4706</td>\n",
       "      <td id=\"T_4430c_row6_col6\" class=\"data row6 col6\" >47.8732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_4430c_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_4430c_row7_col0\" class=\"data row7 col0\" >2.7378</td>\n",
       "      <td id=\"T_4430c_row7_col1\" class=\"data row7 col1\" >12.2460</td>\n",
       "      <td id=\"T_4430c_row7_col2\" class=\"data row7 col2\" >3.4994</td>\n",
       "      <td id=\"T_4430c_row7_col3\" class=\"data row7 col3\" >0.9688</td>\n",
       "      <td id=\"T_4430c_row7_col4\" class=\"data row7 col4\" >0.6377</td>\n",
       "      <td id=\"T_4430c_row7_col5\" class=\"data row7 col5\" >0.4409</td>\n",
       "      <td id=\"T_4430c_row7_col6\" class=\"data row7 col6\" >47.9188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_4430c_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_4430c_row8_col0\" class=\"data row8 col0\" >3.2332</td>\n",
       "      <td id=\"T_4430c_row8_col1\" class=\"data row8 col1\" >21.6720</td>\n",
       "      <td id=\"T_4430c_row8_col2\" class=\"data row8 col2\" >4.6553</td>\n",
       "      <td id=\"T_4430c_row8_col3\" class=\"data row8 col3\" >0.9446</td>\n",
       "      <td id=\"T_4430c_row8_col4\" class=\"data row8 col4\" >0.7376</td>\n",
       "      <td id=\"T_4430c_row8_col5\" class=\"data row8 col5\" >0.2916</td>\n",
       "      <td id=\"T_4430c_row8_col6\" class=\"data row8 col6\" >43.6864</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_4430c_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_4430c_row9_col0\" class=\"data row9 col0\" >3.4076</td>\n",
       "      <td id=\"T_4430c_row9_col1\" class=\"data row9 col1\" >26.5117</td>\n",
       "      <td id=\"T_4430c_row9_col2\" class=\"data row9 col2\" >5.1489</td>\n",
       "      <td id=\"T_4430c_row9_col3\" class=\"data row9 col3\" >0.9459</td>\n",
       "      <td id=\"T_4430c_row9_col4\" class=\"data row9 col4\" >0.6214</td>\n",
       "      <td id=\"T_4430c_row9_col5\" class=\"data row9 col5\" >0.3809</td>\n",
       "      <td id=\"T_4430c_row9_col6\" class=\"data row9 col6\" >41.3549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_4430c_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_4430c_row10_col0\" class=\"data row10 col0\" >3.3507</td>\n",
       "      <td id=\"T_4430c_row10_col1\" class=\"data row10 col1\" >24.8557</td>\n",
       "      <td id=\"T_4430c_row10_col2\" class=\"data row10 col2\" >4.7828</td>\n",
       "      <td id=\"T_4430c_row10_col3\" class=\"data row10 col3\" >0.9419</td>\n",
       "      <td id=\"T_4430c_row10_col4\" class=\"data row10 col4\" >0.6458</td>\n",
       "      <td id=\"T_4430c_row10_col5\" class=\"data row10 col5\" >0.4571</td>\n",
       "      <td id=\"T_4430c_row10_col6\" class=\"data row10 col6\" >46.9059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_4430c_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_4430c_row11_col0\" class=\"data row11 col0\" >0.4654</td>\n",
       "      <td id=\"T_4430c_row11_col1\" class=\"data row11 col1\" >16.5188</td>\n",
       "      <td id=\"T_4430c_row11_col2\" class=\"data row11 col2\" >1.4074</td>\n",
       "      <td id=\"T_4430c_row11_col3\" class=\"data row11 col3\" >0.0306</td>\n",
       "      <td id=\"T_4430c_row11_col4\" class=\"data row11 col4\" >0.1075</td>\n",
       "      <td id=\"T_4430c_row11_col5\" class=\"data row11 col5\" >0.0804</td>\n",
       "      <td id=\"T_4430c_row11_col6\" class=\"data row11 col6\" >6.9488</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2753946b820>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original model was better than the ensembled model, hence it will be returned. NOTE: The display metrics are for the ensembled model (not the original one).\n"
     ]
    }
   ],
   "source": [
    "dataset = pd.merge(clinical_data_train[['updrs_3',\n",
    "                                        'offset_updrs1_offset',\n",
    "                                        'offset_updrs2_offset',\n",
    "                                        'offset_updrs3_offset',\n",
    "                                        'offset_updrs4_offset',\n",
    "                                        'visit_month']],\n",
    "                   npx_features,\n",
    "                   how='inner',\n",
    "                   left_on=['visit_id'],\n",
    "                   right_on=['visit_id'])\n",
    "dataset.head()\n",
    "Y = dataset[dataset.columns[0]].values\n",
    "X = dataset[dataset.columns[6:]].values\n",
    "X_noscale = dataset[dataset.columns[1:6]].values\n",
    "X_scaled = StandardScaler().fit_transform(X)\n",
    "final_X = np.hstack((X_noscale, X_scaled))\n",
    "\n",
    "s = setup(data=final_X, target=Y)\n",
    "add_metric('SMAPE', 'SMAPE', smape, greater_is_better=False)\n",
    "b3 = create_model('et')\n",
    "t3 = tune_model(b3, choose_better=True, optimize='SMAPE')\n",
    "ensemble_u3 = ensemble_model(\n",
    "    t3,\n",
    "    choose_better=True,\n",
    "    n_estimators=20,\n",
    "    method='Boosting')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_18909\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_18909_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_18909_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_18909_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_18909_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_18909_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_18909_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_18909_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_18909_level0_col7\" class=\"col_heading level0 col7\" >SMAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_18909_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_18909_row0_col0\" class=\"data row0 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_18909_row0_col1\" class=\"data row0 col1\" >0.8558</td>\n",
       "      <td id=\"T_18909_row0_col2\" class=\"data row0 col2\" >19.0727</td>\n",
       "      <td id=\"T_18909_row0_col3\" class=\"data row0 col3\" >4.3672</td>\n",
       "      <td id=\"T_18909_row0_col4\" class=\"data row0 col4\" >0.9539</td>\n",
       "      <td id=\"T_18909_row0_col5\" class=\"data row0 col5\" >0.4343</td>\n",
       "      <td id=\"T_18909_row0_col6\" class=\"data row0 col6\" >0.0790</td>\n",
       "      <td id=\"T_18909_row0_col7\" class=\"data row0 col7\" >17.7302</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x275385c0070>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_37636\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_37636_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_37636_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_37636_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_37636_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_37636_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_37636_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_37636_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_37636_level0_col7\" class=\"col_heading level0 col7\" >SMAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_37636_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_37636_row0_col0\" class=\"data row0 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_37636_row0_col1\" class=\"data row0 col1\" >0.7279</td>\n",
       "      <td id=\"T_37636_row0_col2\" class=\"data row0 col2\" >18.9434</td>\n",
       "      <td id=\"T_37636_row0_col3\" class=\"data row0 col3\" >4.3524</td>\n",
       "      <td id=\"T_37636_row0_col4\" class=\"data row0 col4\" >0.9542</td>\n",
       "      <td id=\"T_37636_row0_col5\" class=\"data row0 col5\" >0.4494</td>\n",
       "      <td id=\"T_37636_row0_col6\" class=\"data row0 col6\" >0.0236</td>\n",
       "      <td id=\"T_37636_row0_col7\" class=\"data row0 col7\" >16.6275</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x275385c0cd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_cc0fd\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_cc0fd_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_cc0fd_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_cc0fd_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_cc0fd_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_cc0fd_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_cc0fd_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_cc0fd_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_cc0fd_level0_col7\" class=\"col_heading level0 col7\" >SMAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_cc0fd_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_cc0fd_row0_col0\" class=\"data row0 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_cc0fd_row0_col1\" class=\"data row0 col1\" >0.7279</td>\n",
       "      <td id=\"T_cc0fd_row0_col2\" class=\"data row0 col2\" >18.9434</td>\n",
       "      <td id=\"T_cc0fd_row0_col3\" class=\"data row0 col3\" >4.3524</td>\n",
       "      <td id=\"T_cc0fd_row0_col4\" class=\"data row0 col4\" >0.9542</td>\n",
       "      <td id=\"T_cc0fd_row0_col5\" class=\"data row0 col5\" >0.4494</td>\n",
       "      <td id=\"T_cc0fd_row0_col6\" class=\"data row0 col6\" >0.0236</td>\n",
       "      <td id=\"T_cc0fd_row0_col7\" class=\"data row0 col7\" >16.6275</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x27539282280>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-3 {color: black;background-color: white;}#sk-container-id-3 pre{padding: 0;}#sk-container-id-3 div.sk-toggleable {background-color: white;}#sk-container-id-3 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-3 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-3 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-3 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-3 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-3 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-3 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-3 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-3 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-3 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-3 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-3 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-3 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-3 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-3 div.sk-item {position: relative;z-index: 1;}#sk-container-id-3 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-3 div.sk-item::before, #sk-container-id-3 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-3 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-3 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-3 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-3 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-3 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-3 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-3 div.sk-label-container {text-align: center;}#sk-container-id-3 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-3 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-3\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(memory=FastMemory(location=C:\\Users\\prith\\AppData\\Local\\Temp\\joblib),\n",
       "         steps=[(&#x27;numerical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[&#x27;feature_1&#x27;, &#x27;feature_2&#x27;,\n",
       "                                             &#x27;feature_3&#x27;, &#x27;feature_4&#x27;,\n",
       "                                             &#x27;feature_5&#x27;, &#x27;feature_6&#x27;,\n",
       "                                             &#x27;feature_7&#x27;, &#x27;feature_8&#x27;,\n",
       "                                             &#x27;feature_9&#x27;, &#x27;feature_10&#x27;,\n",
       "                                             &#x27;feature_11&#x27;, &#x27;feature_12&#x27;,\n",
       "                                             &#x27;feature_13&#x27;, &#x27;feature_14&#x27;,\n",
       "                                             &#x27;feature_15&#x27;, &#x27;feature_16&#x27;,\n",
       "                                             &#x27;feature_17&#x27;, &#x27;fea...\n",
       "                                             &#x27;feature_27&#x27;, &#x27;feature_28&#x27;,\n",
       "                                             &#x27;feature_29&#x27;, &#x27;feature_30&#x27;, ...],\n",
       "                                    transformer=SimpleImputer())),\n",
       "                (&#x27;categorical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[],\n",
       "                                    transformer=SimpleImputer(strategy=&#x27;most_frequent&#x27;))),\n",
       "                (&#x27;actual_estimator&#x27;,\n",
       "                 ExtraTreesRegressor(criterion=&#x27;absolute_error&#x27;, max_depth=9,\n",
       "                                     min_impurity_decrease=0.0001,\n",
       "                                     min_samples_leaf=3, n_estimators=260,\n",
       "                                     n_jobs=-1, random_state=7050))])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-17\" type=\"checkbox\" ><label for=\"sk-estimator-id-17\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(memory=FastMemory(location=C:\\Users\\prith\\AppData\\Local\\Temp\\joblib),\n",
       "         steps=[(&#x27;numerical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[&#x27;feature_1&#x27;, &#x27;feature_2&#x27;,\n",
       "                                             &#x27;feature_3&#x27;, &#x27;feature_4&#x27;,\n",
       "                                             &#x27;feature_5&#x27;, &#x27;feature_6&#x27;,\n",
       "                                             &#x27;feature_7&#x27;, &#x27;feature_8&#x27;,\n",
       "                                             &#x27;feature_9&#x27;, &#x27;feature_10&#x27;,\n",
       "                                             &#x27;feature_11&#x27;, &#x27;feature_12&#x27;,\n",
       "                                             &#x27;feature_13&#x27;, &#x27;feature_14&#x27;,\n",
       "                                             &#x27;feature_15&#x27;, &#x27;feature_16&#x27;,\n",
       "                                             &#x27;feature_17&#x27;, &#x27;fea...\n",
       "                                             &#x27;feature_27&#x27;, &#x27;feature_28&#x27;,\n",
       "                                             &#x27;feature_29&#x27;, &#x27;feature_30&#x27;, ...],\n",
       "                                    transformer=SimpleImputer())),\n",
       "                (&#x27;categorical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[],\n",
       "                                    transformer=SimpleImputer(strategy=&#x27;most_frequent&#x27;))),\n",
       "                (&#x27;actual_estimator&#x27;,\n",
       "                 ExtraTreesRegressor(criterion=&#x27;absolute_error&#x27;, max_depth=9,\n",
       "                                     min_impurity_decrease=0.0001,\n",
       "                                     min_samples_leaf=3, n_estimators=260,\n",
       "                                     n_jobs=-1, random_state=7050))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-18\" type=\"checkbox\" ><label for=\"sk-estimator-id-18\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">numerical_imputer: TransformerWrapper</label><div class=\"sk-toggleable__content\"><pre>TransformerWrapper(include=[&#x27;feature_1&#x27;, &#x27;feature_2&#x27;, &#x27;feature_3&#x27;, &#x27;feature_4&#x27;,\n",
       "                            &#x27;feature_5&#x27;, &#x27;feature_6&#x27;, &#x27;feature_7&#x27;, &#x27;feature_8&#x27;,\n",
       "                            &#x27;feature_9&#x27;, &#x27;feature_10&#x27;, &#x27;feature_11&#x27;,\n",
       "                            &#x27;feature_12&#x27;, &#x27;feature_13&#x27;, &#x27;feature_14&#x27;,\n",
       "                            &#x27;feature_15&#x27;, &#x27;feature_16&#x27;, &#x27;feature_17&#x27;,\n",
       "                            &#x27;feature_18&#x27;, &#x27;feature_19&#x27;, &#x27;feature_20&#x27;,\n",
       "                            &#x27;feature_21&#x27;, &#x27;feature_22&#x27;, &#x27;feature_23&#x27;,\n",
       "                            &#x27;feature_24&#x27;, &#x27;feature_25&#x27;, &#x27;feature_26&#x27;,\n",
       "                            &#x27;feature_27&#x27;, &#x27;feature_28&#x27;, &#x27;feature_29&#x27;,\n",
       "                            &#x27;feature_30&#x27;, ...],\n",
       "                   transformer=SimpleImputer())</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-19\" type=\"checkbox\" ><label for=\"sk-estimator-id-19\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">transformer: SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer()</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-20\" type=\"checkbox\" ><label for=\"sk-estimator-id-20\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-21\" type=\"checkbox\" ><label for=\"sk-estimator-id-21\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">categorical_imputer: TransformerWrapper</label><div class=\"sk-toggleable__content\"><pre>TransformerWrapper(include=[],\n",
       "                   transformer=SimpleImputer(strategy=&#x27;most_frequent&#x27;))</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-22\" type=\"checkbox\" ><label for=\"sk-estimator-id-22\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">transformer: SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(strategy=&#x27;most_frequent&#x27;)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-23\" type=\"checkbox\" ><label for=\"sk-estimator-id-23\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(strategy=&#x27;most_frequent&#x27;)</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-24\" type=\"checkbox\" ><label for=\"sk-estimator-id-24\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ExtraTreesRegressor</label><div class=\"sk-toggleable__content\"><pre>ExtraTreesRegressor(criterion=&#x27;absolute_error&#x27;, max_depth=9,\n",
       "                    min_impurity_decrease=0.0001, min_samples_leaf=3,\n",
       "                    n_estimators=260, n_jobs=-1, random_state=7050)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(memory=FastMemory(location=C:\\Users\\prith\\AppData\\Local\\Temp\\joblib),\n",
       "         steps=[('numerical_imputer',\n",
       "                 TransformerWrapper(include=['feature_1', 'feature_2',\n",
       "                                             'feature_3', 'feature_4',\n",
       "                                             'feature_5', 'feature_6',\n",
       "                                             'feature_7', 'feature_8',\n",
       "                                             'feature_9', 'feature_10',\n",
       "                                             'feature_11', 'feature_12',\n",
       "                                             'feature_13', 'feature_14',\n",
       "                                             'feature_15', 'feature_16',\n",
       "                                             'feature_17', 'fea...\n",
       "                                             'feature_27', 'feature_28',\n",
       "                                             'feature_29', 'feature_30', ...],\n",
       "                                    transformer=SimpleImputer())),\n",
       "                ('categorical_imputer',\n",
       "                 TransformerWrapper(include=[],\n",
       "                                    transformer=SimpleImputer(strategy='most_frequent'))),\n",
       "                ('actual_estimator',\n",
       "                 ExtraTreesRegressor(criterion='absolute_error', max_depth=9,\n",
       "                                     min_impurity_decrease=0.0001,\n",
       "                                     min_samples_leaf=3, n_estimators=260,\n",
       "                                     n_jobs=-1, random_state=7050))])"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "predict_model(b3)\n",
    "predict_model(t3)\n",
    "predict_model(ensemble_u3)\n",
    "finalize_model(ensemble_u3)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# UPDRS 4 Predictions with Proteins Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_b88dd_row8_col1 {\n",
       "  background-color: lightgreen;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_b88dd\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_b88dd_level0_col0\" class=\"col_heading level0 col0\" >Description</th>\n",
       "      <th id=\"T_b88dd_level0_col1\" class=\"col_heading level0 col1\" >Value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_b88dd_row0_col0\" class=\"data row0 col0\" >Session id</td>\n",
       "      <td id=\"T_b88dd_row0_col1\" class=\"data row0 col1\" >7109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_b88dd_row1_col0\" class=\"data row1 col0\" >Target</td>\n",
       "      <td id=\"T_b88dd_row1_col1\" class=\"data row1 col1\" >target</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_b88dd_row2_col0\" class=\"data row2 col0\" >Target type</td>\n",
       "      <td id=\"T_b88dd_row2_col1\" class=\"data row2 col1\" >Regression</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_b88dd_row3_col0\" class=\"data row3 col0\" >Original data shape</td>\n",
       "      <td id=\"T_b88dd_row3_col1\" class=\"data row3 col1\" >(1047, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_b88dd_row4_col0\" class=\"data row4 col0\" >Transformed data shape</td>\n",
       "      <td id=\"T_b88dd_row4_col1\" class=\"data row4 col1\" >(1047, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_b88dd_row5_col0\" class=\"data row5 col0\" >Transformed train set shape</td>\n",
       "      <td id=\"T_b88dd_row5_col1\" class=\"data row5 col1\" >(732, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_b88dd_row6_col0\" class=\"data row6 col0\" >Transformed test set shape</td>\n",
       "      <td id=\"T_b88dd_row6_col1\" class=\"data row6 col1\" >(315, 235)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_b88dd_row7_col0\" class=\"data row7 col0\" >Numeric features</td>\n",
       "      <td id=\"T_b88dd_row7_col1\" class=\"data row7 col1\" >234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_b88dd_row8_col0\" class=\"data row8 col0\" >Preprocess</td>\n",
       "      <td id=\"T_b88dd_row8_col1\" class=\"data row8 col1\" >True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_b88dd_row9_col0\" class=\"data row9 col0\" >Imputation type</td>\n",
       "      <td id=\"T_b88dd_row9_col1\" class=\"data row9 col1\" >simple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row10\" class=\"row_heading level0 row10\" >10</th>\n",
       "      <td id=\"T_b88dd_row10_col0\" class=\"data row10 col0\" >Numeric imputation</td>\n",
       "      <td id=\"T_b88dd_row10_col1\" class=\"data row10 col1\" >mean</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row11\" class=\"row_heading level0 row11\" >11</th>\n",
       "      <td id=\"T_b88dd_row11_col0\" class=\"data row11 col0\" >Categorical imputation</td>\n",
       "      <td id=\"T_b88dd_row11_col1\" class=\"data row11 col1\" >mode</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row12\" class=\"row_heading level0 row12\" >12</th>\n",
       "      <td id=\"T_b88dd_row12_col0\" class=\"data row12 col0\" >Fold Generator</td>\n",
       "      <td id=\"T_b88dd_row12_col1\" class=\"data row12 col1\" >KFold</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row13\" class=\"row_heading level0 row13\" >13</th>\n",
       "      <td id=\"T_b88dd_row13_col0\" class=\"data row13 col0\" >Fold Number</td>\n",
       "      <td id=\"T_b88dd_row13_col1\" class=\"data row13 col1\" >10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row14\" class=\"row_heading level0 row14\" >14</th>\n",
       "      <td id=\"T_b88dd_row14_col0\" class=\"data row14 col0\" >CPU Jobs</td>\n",
       "      <td id=\"T_b88dd_row14_col1\" class=\"data row14 col1\" >-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row15\" class=\"row_heading level0 row15\" >15</th>\n",
       "      <td id=\"T_b88dd_row15_col0\" class=\"data row15 col0\" >Use GPU</td>\n",
       "      <td id=\"T_b88dd_row15_col1\" class=\"data row15 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row16\" class=\"row_heading level0 row16\" >16</th>\n",
       "      <td id=\"T_b88dd_row16_col0\" class=\"data row16 col0\" >Log Experiment</td>\n",
       "      <td id=\"T_b88dd_row16_col1\" class=\"data row16 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row17\" class=\"row_heading level0 row17\" >17</th>\n",
       "      <td id=\"T_b88dd_row17_col0\" class=\"data row17 col0\" >Experiment Name</td>\n",
       "      <td id=\"T_b88dd_row17_col1\" class=\"data row17 col1\" >reg-default-name</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b88dd_level0_row18\" class=\"row_heading level0 row18\" >18</th>\n",
       "      <td id=\"T_b88dd_row18_col0\" class=\"data row18 col0\" >USI</td>\n",
       "      <td id=\"T_b88dd_row18_col1\" class=\"data row18 col1\" >cc80</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x275396c33d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_bec77_row10_col0, #T_bec77_row10_col1, #T_bec77_row10_col2, #T_bec77_row10_col3, #T_bec77_row10_col4, #T_bec77_row10_col5, #T_bec77_row10_col6 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_bec77\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_bec77_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_bec77_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_bec77_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_bec77_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_bec77_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_bec77_level0_col5\" class=\"col_heading level0 col5\" >MAPE</th>\n",
       "      <th id=\"T_bec77_level0_col6\" class=\"col_heading level0 col6\" >SMAPE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >Fold</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "      <th class=\"blank col6\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_bec77_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_bec77_row0_col0\" class=\"data row0 col0\" >0.9595</td>\n",
       "      <td id=\"T_bec77_row0_col1\" class=\"data row0 col1\" >4.0526</td>\n",
       "      <td id=\"T_bec77_row0_col2\" class=\"data row0 col2\" >2.0131</td>\n",
       "      <td id=\"T_bec77_row0_col3\" class=\"data row0 col3\" >0.6132</td>\n",
       "      <td id=\"T_bec77_row0_col4\" class=\"data row0 col4\" >0.5156</td>\n",
       "      <td id=\"T_bec77_row0_col5\" class=\"data row0 col5\" >0.4870</td>\n",
       "      <td id=\"T_bec77_row0_col6\" class=\"data row0 col6\" >162.2561</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_bec77_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_bec77_row1_col0\" class=\"data row1 col0\" >0.8986</td>\n",
       "      <td id=\"T_bec77_row1_col1\" class=\"data row1 col1\" >1.8870</td>\n",
       "      <td id=\"T_bec77_row1_col2\" class=\"data row1 col2\" >1.3737</td>\n",
       "      <td id=\"T_bec77_row1_col3\" class=\"data row1 col3\" >0.6192</td>\n",
       "      <td id=\"T_bec77_row1_col4\" class=\"data row1 col4\" >0.5649</td>\n",
       "      <td id=\"T_bec77_row1_col5\" class=\"data row1 col5\" >0.4019</td>\n",
       "      <td id=\"T_bec77_row1_col6\" class=\"data row1 col6\" >158.1691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_bec77_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_bec77_row2_col0\" class=\"data row2 col0\" >0.9616</td>\n",
       "      <td id=\"T_bec77_row2_col1\" class=\"data row2 col1\" >2.7226</td>\n",
       "      <td id=\"T_bec77_row2_col2\" class=\"data row2 col2\" >1.6500</td>\n",
       "      <td id=\"T_bec77_row2_col3\" class=\"data row2 col3\" >0.6014</td>\n",
       "      <td id=\"T_bec77_row2_col4\" class=\"data row2 col4\" >0.6016</td>\n",
       "      <td id=\"T_bec77_row2_col5\" class=\"data row2 col5\" >0.4977</td>\n",
       "      <td id=\"T_bec77_row2_col6\" class=\"data row2 col6\" >162.4711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_bec77_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_bec77_row3_col0\" class=\"data row3 col0\" >1.0395</td>\n",
       "      <td id=\"T_bec77_row3_col1\" class=\"data row3 col1\" >2.9477</td>\n",
       "      <td id=\"T_bec77_row3_col2\" class=\"data row3 col2\" >1.7169</td>\n",
       "      <td id=\"T_bec77_row3_col3\" class=\"data row3 col3\" >-0.0257</td>\n",
       "      <td id=\"T_bec77_row3_col4\" class=\"data row3 col4\" >0.5919</td>\n",
       "      <td id=\"T_bec77_row3_col5\" class=\"data row3 col5\" >0.9023</td>\n",
       "      <td id=\"T_bec77_row3_col6\" class=\"data row3 col6\" >150.9868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_bec77_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_bec77_row4_col0\" class=\"data row4 col0\" >0.9153</td>\n",
       "      <td id=\"T_bec77_row4_col1\" class=\"data row4 col1\" >2.1300</td>\n",
       "      <td id=\"T_bec77_row4_col2\" class=\"data row4 col2\" >1.4595</td>\n",
       "      <td id=\"T_bec77_row4_col3\" class=\"data row4 col3\" >0.7141</td>\n",
       "      <td id=\"T_bec77_row4_col4\" class=\"data row4 col4\" >0.5335</td>\n",
       "      <td id=\"T_bec77_row4_col5\" class=\"data row4 col5\" >0.4429</td>\n",
       "      <td id=\"T_bec77_row4_col6\" class=\"data row4 col6\" >157.5491</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_bec77_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_bec77_row5_col0\" class=\"data row5 col0\" >0.6840</td>\n",
       "      <td id=\"T_bec77_row5_col1\" class=\"data row5 col1\" >1.4402</td>\n",
       "      <td id=\"T_bec77_row5_col2\" class=\"data row5 col2\" >1.2001</td>\n",
       "      <td id=\"T_bec77_row5_col3\" class=\"data row5 col3\" >0.6707</td>\n",
       "      <td id=\"T_bec77_row5_col4\" class=\"data row5 col4\" >0.4954</td>\n",
       "      <td id=\"T_bec77_row5_col5\" class=\"data row5 col5\" >0.5322</td>\n",
       "      <td id=\"T_bec77_row5_col6\" class=\"data row5 col6\" >171.1591</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_bec77_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_bec77_row6_col0\" class=\"data row6 col0\" >0.8616</td>\n",
       "      <td id=\"T_bec77_row6_col1\" class=\"data row6 col1\" >1.9488</td>\n",
       "      <td id=\"T_bec77_row6_col2\" class=\"data row6 col2\" >1.3960</td>\n",
       "      <td id=\"T_bec77_row6_col3\" class=\"data row6 col3\" >0.4243</td>\n",
       "      <td id=\"T_bec77_row6_col4\" class=\"data row6 col4\" >0.4892</td>\n",
       "      <td id=\"T_bec77_row6_col5\" class=\"data row6 col5\" >0.6628</td>\n",
       "      <td id=\"T_bec77_row6_col6\" class=\"data row6 col6\" >168.6644</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_bec77_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_bec77_row7_col0\" class=\"data row7 col0\" >1.0085</td>\n",
       "      <td id=\"T_bec77_row7_col1\" class=\"data row7 col1\" >2.7401</td>\n",
       "      <td id=\"T_bec77_row7_col2\" class=\"data row7 col2\" >1.6553</td>\n",
       "      <td id=\"T_bec77_row7_col3\" class=\"data row7 col3\" >0.5613</td>\n",
       "      <td id=\"T_bec77_row7_col4\" class=\"data row7 col4\" >0.6030</td>\n",
       "      <td id=\"T_bec77_row7_col5\" class=\"data row7 col5\" >0.6176</td>\n",
       "      <td id=\"T_bec77_row7_col6\" class=\"data row7 col6\" >170.6011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_bec77_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_bec77_row8_col0\" class=\"data row8 col0\" >1.0229</td>\n",
       "      <td id=\"T_bec77_row8_col1\" class=\"data row8 col1\" >3.7015</td>\n",
       "      <td id=\"T_bec77_row8_col2\" class=\"data row8 col2\" >1.9239</td>\n",
       "      <td id=\"T_bec77_row8_col3\" class=\"data row8 col3\" >0.4307</td>\n",
       "      <td id=\"T_bec77_row8_col4\" class=\"data row8 col4\" >0.5626</td>\n",
       "      <td id=\"T_bec77_row8_col5\" class=\"data row8 col5\" >0.7987</td>\n",
       "      <td id=\"T_bec77_row8_col6\" class=\"data row8 col6\" >167.5427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_bec77_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_bec77_row9_col0\" class=\"data row9 col0\" >1.3862</td>\n",
       "      <td id=\"T_bec77_row9_col1\" class=\"data row9 col1\" >7.7534</td>\n",
       "      <td id=\"T_bec77_row9_col2\" class=\"data row9 col2\" >2.7845</td>\n",
       "      <td id=\"T_bec77_row9_col3\" class=\"data row9 col3\" >0.3286</td>\n",
       "      <td id=\"T_bec77_row9_col4\" class=\"data row9 col4\" >0.6745</td>\n",
       "      <td id=\"T_bec77_row9_col5\" class=\"data row9 col5\" >0.6836</td>\n",
       "      <td id=\"T_bec77_row9_col6\" class=\"data row9 col6\" >166.0509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_bec77_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_bec77_row10_col0\" class=\"data row10 col0\" >0.9738</td>\n",
       "      <td id=\"T_bec77_row10_col1\" class=\"data row10 col1\" >3.1324</td>\n",
       "      <td id=\"T_bec77_row10_col2\" class=\"data row10 col2\" >1.7173</td>\n",
       "      <td id=\"T_bec77_row10_col3\" class=\"data row10 col3\" >0.4938</td>\n",
       "      <td id=\"T_bec77_row10_col4\" class=\"data row10 col4\" >0.5632</td>\n",
       "      <td id=\"T_bec77_row10_col5\" class=\"data row10 col5\" >0.6027</td>\n",
       "      <td id=\"T_bec77_row10_col6\" class=\"data row10 col6\" >163.5450</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_bec77_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_bec77_row11_col0\" class=\"data row11 col0\" >0.1685</td>\n",
       "      <td id=\"T_bec77_row11_col1\" class=\"data row11 col1\" >1.7234</td>\n",
       "      <td id=\"T_bec77_row11_col2\" class=\"data row11 col2\" >0.4281</td>\n",
       "      <td id=\"T_bec77_row11_col3\" class=\"data row11 col3\" >0.2078</td>\n",
       "      <td id=\"T_bec77_row11_col4\" class=\"data row11 col4\" >0.0543</td>\n",
       "      <td id=\"T_bec77_row11_col5\" class=\"data row11 col5\" >0.1529</td>\n",
       "      <td id=\"T_bec77_row11_col6\" class=\"data row11 col6\" >6.1774</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x27539909790>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_cd026_row10_col0, #T_cd026_row10_col1, #T_cd026_row10_col2, #T_cd026_row10_col3, #T_cd026_row10_col4, #T_cd026_row10_col5, #T_cd026_row10_col6 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_cd026\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_cd026_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_cd026_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_cd026_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_cd026_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_cd026_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_cd026_level0_col5\" class=\"col_heading level0 col5\" >MAPE</th>\n",
       "      <th id=\"T_cd026_level0_col6\" class=\"col_heading level0 col6\" >SMAPE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >Fold</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "      <th class=\"blank col6\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_cd026_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_cd026_row0_col0\" class=\"data row0 col0\" >1.1892</td>\n",
       "      <td id=\"T_cd026_row0_col1\" class=\"data row0 col1\" >11.8919</td>\n",
       "      <td id=\"T_cd026_row0_col2\" class=\"data row0 col2\" >3.4485</td>\n",
       "      <td id=\"T_cd026_row0_col3\" class=\"data row0 col3\" >-0.1350</td>\n",
       "      <td id=\"T_cd026_row0_col4\" class=\"data row0 col4\" >0.8220</td>\n",
       "      <td id=\"T_cd026_row0_col5\" class=\"data row0 col5\" >1.0000</td>\n",
       "      <td id=\"T_cd026_row0_col6\" class=\"data row0 col6\" >40.5405</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_cd026_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_cd026_row1_col0\" class=\"data row1 col0\" >1.1351</td>\n",
       "      <td id=\"T_cd026_row1_col1\" class=\"data row1 col1\" >6.2432</td>\n",
       "      <td id=\"T_cd026_row1_col2\" class=\"data row1 col2\" >2.4986</td>\n",
       "      <td id=\"T_cd026_row1_col3\" class=\"data row1 col3\" >-0.2601</td>\n",
       "      <td id=\"T_cd026_row1_col4\" class=\"data row1 col4\" >0.8435</td>\n",
       "      <td id=\"T_cd026_row1_col5\" class=\"data row1 col5\" >1.0000</td>\n",
       "      <td id=\"T_cd026_row1_col6\" class=\"data row1 col6\" >48.6486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_cd026_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_cd026_row2_col0\" class=\"data row2 col0\" >1.1370</td>\n",
       "      <td id=\"T_cd026_row2_col1\" class=\"data row2 col1\" >8.1233</td>\n",
       "      <td id=\"T_cd026_row2_col2\" class=\"data row2 col2\" >2.8501</td>\n",
       "      <td id=\"T_cd026_row2_col3\" class=\"data row2 col3\" >-0.1893</td>\n",
       "      <td id=\"T_cd026_row2_col4\" class=\"data row2 col4\" >0.8291</td>\n",
       "      <td id=\"T_cd026_row2_col5\" class=\"data row2 col5\" >1.0000</td>\n",
       "      <td id=\"T_cd026_row2_col6\" class=\"data row2 col6\" >43.8356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_cd026_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_cd026_row3_col0\" class=\"data row3 col0\" >0.9452</td>\n",
       "      <td id=\"T_cd026_row3_col1\" class=\"data row3 col1\" >3.7671</td>\n",
       "      <td id=\"T_cd026_row3_col2\" class=\"data row3 col2\" >1.9409</td>\n",
       "      <td id=\"T_cd026_row3_col3\" class=\"data row3 col3\" >-0.3109</td>\n",
       "      <td id=\"T_cd026_row3_col4\" class=\"data row3 col4\" >0.7712</td>\n",
       "      <td id=\"T_cd026_row3_col5\" class=\"data row3 col5\" >1.0000</td>\n",
       "      <td id=\"T_cd026_row3_col6\" class=\"data row3 col6\" >57.5342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_cd026_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_cd026_row4_col0\" class=\"data row4 col0\" >1.2055</td>\n",
       "      <td id=\"T_cd026_row4_col1\" class=\"data row4 col1\" >8.9041</td>\n",
       "      <td id=\"T_cd026_row4_col2\" class=\"data row4 col2\" >2.9840</td>\n",
       "      <td id=\"T_cd026_row4_col3\" class=\"data row4 col3\" >-0.1950</td>\n",
       "      <td id=\"T_cd026_row4_col4\" class=\"data row4 col4\" >0.8502</td>\n",
       "      <td id=\"T_cd026_row4_col5\" class=\"data row4 col5\" >1.0000</td>\n",
       "      <td id=\"T_cd026_row4_col6\" class=\"data row4 col6\" >46.5753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_cd026_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_cd026_row5_col0\" class=\"data row5 col0\" >0.6164</td>\n",
       "      <td id=\"T_cd026_row5_col1\" class=\"data row5 col1\" >4.7534</td>\n",
       "      <td id=\"T_cd026_row5_col2\" class=\"data row5 col2\" >2.1802</td>\n",
       "      <td id=\"T_cd026_row5_col3\" class=\"data row5 col3\" >-0.0869</td>\n",
       "      <td id=\"T_cd026_row5_col4\" class=\"data row5 col4\" >0.6012</td>\n",
       "      <td id=\"T_cd026_row5_col5\" class=\"data row5 col5\" >1.0000</td>\n",
       "      <td id=\"T_cd026_row5_col6\" class=\"data row5 col6\" >27.3973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_cd026_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_cd026_row6_col0\" class=\"data row6 col0\" >0.8904</td>\n",
       "      <td id=\"T_cd026_row6_col1\" class=\"data row6 col1\" >4.1781</td>\n",
       "      <td id=\"T_cd026_row6_col2\" class=\"data row6 col2\" >2.0440</td>\n",
       "      <td id=\"T_cd026_row6_col3\" class=\"data row6 col3\" >-0.2342</td>\n",
       "      <td id=\"T_cd026_row6_col4\" class=\"data row6 col4\" >0.7513</td>\n",
       "      <td id=\"T_cd026_row6_col5\" class=\"data row6 col5\" >1.0000</td>\n",
       "      <td id=\"T_cd026_row6_col6\" class=\"data row6 col6\" >43.8356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_cd026_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_cd026_row7_col0\" class=\"data row7 col0\" >0.9726</td>\n",
       "      <td id=\"T_cd026_row7_col1\" class=\"data row7 col1\" >7.1918</td>\n",
       "      <td id=\"T_cd026_row7_col2\" class=\"data row7 col2\" >2.6817</td>\n",
       "      <td id=\"T_cd026_row7_col3\" class=\"data row7 col3\" >-0.1515</td>\n",
       "      <td id=\"T_cd026_row7_col4\" class=\"data row7 col4\" >0.7652</td>\n",
       "      <td id=\"T_cd026_row7_col5\" class=\"data row7 col5\" >1.0000</td>\n",
       "      <td id=\"T_cd026_row7_col6\" class=\"data row7 col6\" >35.6164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_cd026_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_cd026_row8_col0\" class=\"data row8 col0\" >1.0685</td>\n",
       "      <td id=\"T_cd026_row8_col1\" class=\"data row8 col1\" >7.6438</td>\n",
       "      <td id=\"T_cd026_row8_col2\" class=\"data row8 col2\" >2.7647</td>\n",
       "      <td id=\"T_cd026_row8_col3\" class=\"data row8 col3\" >-0.1756</td>\n",
       "      <td id=\"T_cd026_row8_col4\" class=\"data row8 col4\" >0.7977</td>\n",
       "      <td id=\"T_cd026_row8_col5\" class=\"data row8 col5\" >1.0000</td>\n",
       "      <td id=\"T_cd026_row8_col6\" class=\"data row8 col6\" >46.5753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_cd026_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_cd026_row9_col0\" class=\"data row9 col0\" >1.6301</td>\n",
       "      <td id=\"T_cd026_row9_col1\" class=\"data row9 col1\" >14.2055</td>\n",
       "      <td id=\"T_cd026_row9_col2\" class=\"data row9 col2\" >3.7690</td>\n",
       "      <td id=\"T_cd026_row9_col3\" class=\"data row9 col3\" >-0.2301</td>\n",
       "      <td id=\"T_cd026_row9_col4\" class=\"data row9 col4\" >0.9792</td>\n",
       "      <td id=\"T_cd026_row9_col5\" class=\"data row9 col5\" >1.0000</td>\n",
       "      <td id=\"T_cd026_row9_col6\" class=\"data row9 col6\" >52.0548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_cd026_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_cd026_row10_col0\" class=\"data row10 col0\" >1.0790</td>\n",
       "      <td id=\"T_cd026_row10_col1\" class=\"data row10 col1\" >7.6902</td>\n",
       "      <td id=\"T_cd026_row10_col2\" class=\"data row10 col2\" >2.7162</td>\n",
       "      <td id=\"T_cd026_row10_col3\" class=\"data row10 col3\" >-0.1968</td>\n",
       "      <td id=\"T_cd026_row10_col4\" class=\"data row10 col4\" >0.8011</td>\n",
       "      <td id=\"T_cd026_row10_col5\" class=\"data row10 col5\" >1.0000</td>\n",
       "      <td id=\"T_cd026_row10_col6\" class=\"data row10 col6\" >44.2614</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_cd026_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_cd026_row11_col0\" class=\"data row11 col0\" >0.2484</td>\n",
       "      <td id=\"T_cd026_row11_col1\" class=\"data row11 col1\" >3.1695</td>\n",
       "      <td id=\"T_cd026_row11_col2\" class=\"data row11 col2\" >0.5590</td>\n",
       "      <td id=\"T_cd026_row11_col3\" class=\"data row11 col3\" >0.0616</td>\n",
       "      <td id=\"T_cd026_row11_col4\" class=\"data row11 col4\" >0.0904</td>\n",
       "      <td id=\"T_cd026_row11_col5\" class=\"data row11 col5\" >0.0000</td>\n",
       "      <td id=\"T_cd026_row11_col6\" class=\"data row11 col6\" >8.0013</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x27538dabfd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 10 candidates, totalling 100 fits\n"
     ]
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_2db42_row10_col0, #T_2db42_row10_col1, #T_2db42_row10_col2, #T_2db42_row10_col3, #T_2db42_row10_col4, #T_2db42_row10_col5, #T_2db42_row10_col6 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_2db42\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_2db42_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_2db42_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_2db42_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_2db42_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_2db42_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_2db42_level0_col5\" class=\"col_heading level0 col5\" >MAPE</th>\n",
       "      <th id=\"T_2db42_level0_col6\" class=\"col_heading level0 col6\" >SMAPE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >Fold</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "      <th class=\"blank col2\" >&nbsp;</th>\n",
       "      <th class=\"blank col3\" >&nbsp;</th>\n",
       "      <th class=\"blank col4\" >&nbsp;</th>\n",
       "      <th class=\"blank col5\" >&nbsp;</th>\n",
       "      <th class=\"blank col6\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_2db42_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_2db42_row0_col0\" class=\"data row0 col0\" >1.1892</td>\n",
       "      <td id=\"T_2db42_row0_col1\" class=\"data row0 col1\" >11.8919</td>\n",
       "      <td id=\"T_2db42_row0_col2\" class=\"data row0 col2\" >3.4485</td>\n",
       "      <td id=\"T_2db42_row0_col3\" class=\"data row0 col3\" >-0.1350</td>\n",
       "      <td id=\"T_2db42_row0_col4\" class=\"data row0 col4\" >0.8220</td>\n",
       "      <td id=\"T_2db42_row0_col5\" class=\"data row0 col5\" >1.0000</td>\n",
       "      <td id=\"T_2db42_row0_col6\" class=\"data row0 col6\" >40.5405</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_2db42_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_2db42_row1_col0\" class=\"data row1 col0\" >1.1351</td>\n",
       "      <td id=\"T_2db42_row1_col1\" class=\"data row1 col1\" >6.2432</td>\n",
       "      <td id=\"T_2db42_row1_col2\" class=\"data row1 col2\" >2.4986</td>\n",
       "      <td id=\"T_2db42_row1_col3\" class=\"data row1 col3\" >-0.2601</td>\n",
       "      <td id=\"T_2db42_row1_col4\" class=\"data row1 col4\" >0.8435</td>\n",
       "      <td id=\"T_2db42_row1_col5\" class=\"data row1 col5\" >1.0000</td>\n",
       "      <td id=\"T_2db42_row1_col6\" class=\"data row1 col6\" >48.6486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_2db42_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_2db42_row2_col0\" class=\"data row2 col0\" >1.1370</td>\n",
       "      <td id=\"T_2db42_row2_col1\" class=\"data row2 col1\" >8.1233</td>\n",
       "      <td id=\"T_2db42_row2_col2\" class=\"data row2 col2\" >2.8501</td>\n",
       "      <td id=\"T_2db42_row2_col3\" class=\"data row2 col3\" >-0.1893</td>\n",
       "      <td id=\"T_2db42_row2_col4\" class=\"data row2 col4\" >0.8291</td>\n",
       "      <td id=\"T_2db42_row2_col5\" class=\"data row2 col5\" >1.0000</td>\n",
       "      <td id=\"T_2db42_row2_col6\" class=\"data row2 col6\" >43.8356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_2db42_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_2db42_row3_col0\" class=\"data row3 col0\" >0.9452</td>\n",
       "      <td id=\"T_2db42_row3_col1\" class=\"data row3 col1\" >3.7671</td>\n",
       "      <td id=\"T_2db42_row3_col2\" class=\"data row3 col2\" >1.9409</td>\n",
       "      <td id=\"T_2db42_row3_col3\" class=\"data row3 col3\" >-0.3109</td>\n",
       "      <td id=\"T_2db42_row3_col4\" class=\"data row3 col4\" >0.7712</td>\n",
       "      <td id=\"T_2db42_row3_col5\" class=\"data row3 col5\" >1.0000</td>\n",
       "      <td id=\"T_2db42_row3_col6\" class=\"data row3 col6\" >57.5342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_2db42_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_2db42_row4_col0\" class=\"data row4 col0\" >1.2055</td>\n",
       "      <td id=\"T_2db42_row4_col1\" class=\"data row4 col1\" >8.9041</td>\n",
       "      <td id=\"T_2db42_row4_col2\" class=\"data row4 col2\" >2.9840</td>\n",
       "      <td id=\"T_2db42_row4_col3\" class=\"data row4 col3\" >-0.1950</td>\n",
       "      <td id=\"T_2db42_row4_col4\" class=\"data row4 col4\" >0.8502</td>\n",
       "      <td id=\"T_2db42_row4_col5\" class=\"data row4 col5\" >1.0000</td>\n",
       "      <td id=\"T_2db42_row4_col6\" class=\"data row4 col6\" >46.5753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_2db42_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_2db42_row5_col0\" class=\"data row5 col0\" >0.6164</td>\n",
       "      <td id=\"T_2db42_row5_col1\" class=\"data row5 col1\" >4.7534</td>\n",
       "      <td id=\"T_2db42_row5_col2\" class=\"data row5 col2\" >2.1802</td>\n",
       "      <td id=\"T_2db42_row5_col3\" class=\"data row5 col3\" >-0.0869</td>\n",
       "      <td id=\"T_2db42_row5_col4\" class=\"data row5 col4\" >0.6012</td>\n",
       "      <td id=\"T_2db42_row5_col5\" class=\"data row5 col5\" >1.0000</td>\n",
       "      <td id=\"T_2db42_row5_col6\" class=\"data row5 col6\" >27.3973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_2db42_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_2db42_row6_col0\" class=\"data row6 col0\" >0.8904</td>\n",
       "      <td id=\"T_2db42_row6_col1\" class=\"data row6 col1\" >4.1781</td>\n",
       "      <td id=\"T_2db42_row6_col2\" class=\"data row6 col2\" >2.0440</td>\n",
       "      <td id=\"T_2db42_row6_col3\" class=\"data row6 col3\" >-0.2342</td>\n",
       "      <td id=\"T_2db42_row6_col4\" class=\"data row6 col4\" >0.7513</td>\n",
       "      <td id=\"T_2db42_row6_col5\" class=\"data row6 col5\" >1.0000</td>\n",
       "      <td id=\"T_2db42_row6_col6\" class=\"data row6 col6\" >43.8356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_2db42_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_2db42_row7_col0\" class=\"data row7 col0\" >0.9726</td>\n",
       "      <td id=\"T_2db42_row7_col1\" class=\"data row7 col1\" >7.1918</td>\n",
       "      <td id=\"T_2db42_row7_col2\" class=\"data row7 col2\" >2.6817</td>\n",
       "      <td id=\"T_2db42_row7_col3\" class=\"data row7 col3\" >-0.1515</td>\n",
       "      <td id=\"T_2db42_row7_col4\" class=\"data row7 col4\" >0.7652</td>\n",
       "      <td id=\"T_2db42_row7_col5\" class=\"data row7 col5\" >1.0000</td>\n",
       "      <td id=\"T_2db42_row7_col6\" class=\"data row7 col6\" >35.6164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_2db42_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_2db42_row8_col0\" class=\"data row8 col0\" >1.0685</td>\n",
       "      <td id=\"T_2db42_row8_col1\" class=\"data row8 col1\" >7.6438</td>\n",
       "      <td id=\"T_2db42_row8_col2\" class=\"data row8 col2\" >2.7647</td>\n",
       "      <td id=\"T_2db42_row8_col3\" class=\"data row8 col3\" >-0.1756</td>\n",
       "      <td id=\"T_2db42_row8_col4\" class=\"data row8 col4\" >0.7977</td>\n",
       "      <td id=\"T_2db42_row8_col5\" class=\"data row8 col5\" >1.0000</td>\n",
       "      <td id=\"T_2db42_row8_col6\" class=\"data row8 col6\" >46.5753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_2db42_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_2db42_row9_col0\" class=\"data row9 col0\" >1.6301</td>\n",
       "      <td id=\"T_2db42_row9_col1\" class=\"data row9 col1\" >14.2055</td>\n",
       "      <td id=\"T_2db42_row9_col2\" class=\"data row9 col2\" >3.7690</td>\n",
       "      <td id=\"T_2db42_row9_col3\" class=\"data row9 col3\" >-0.2301</td>\n",
       "      <td id=\"T_2db42_row9_col4\" class=\"data row9 col4\" >0.9792</td>\n",
       "      <td id=\"T_2db42_row9_col5\" class=\"data row9 col5\" >1.0000</td>\n",
       "      <td id=\"T_2db42_row9_col6\" class=\"data row9 col6\" >52.0548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_2db42_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_2db42_row10_col0\" class=\"data row10 col0\" >1.0790</td>\n",
       "      <td id=\"T_2db42_row10_col1\" class=\"data row10 col1\" >7.6902</td>\n",
       "      <td id=\"T_2db42_row10_col2\" class=\"data row10 col2\" >2.7162</td>\n",
       "      <td id=\"T_2db42_row10_col3\" class=\"data row10 col3\" >-0.1968</td>\n",
       "      <td id=\"T_2db42_row10_col4\" class=\"data row10 col4\" >0.8011</td>\n",
       "      <td id=\"T_2db42_row10_col5\" class=\"data row10 col5\" >1.0000</td>\n",
       "      <td id=\"T_2db42_row10_col6\" class=\"data row10 col6\" >44.2614</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_2db42_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_2db42_row11_col0\" class=\"data row11 col0\" >0.2484</td>\n",
       "      <td id=\"T_2db42_row11_col1\" class=\"data row11 col1\" >3.1695</td>\n",
       "      <td id=\"T_2db42_row11_col2\" class=\"data row11 col2\" >0.5590</td>\n",
       "      <td id=\"T_2db42_row11_col3\" class=\"data row11 col3\" >0.0616</td>\n",
       "      <td id=\"T_2db42_row11_col4\" class=\"data row11 col4\" >0.0904</td>\n",
       "      <td id=\"T_2db42_row11_col5\" class=\"data row11 col5\" >0.0000</td>\n",
       "      <td id=\"T_2db42_row11_col6\" class=\"data row11 col6\" >8.0013</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x27539774fa0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original model was better than the ensembled model, hence it will be returned. NOTE: The display metrics are for the ensembled model (not the original one).\n"
     ]
    }
   ],
   "source": [
    "dataset = pd.merge(clinical_data_train[['updrs_4',\n",
    "                                        'offset_updrs1_offset',\n",
    "                                        'offset_updrs2_offset',\n",
    "                                        'offset_updrs3_offset',\n",
    "                                        'offset_updrs4_offset',\n",
    "                                        'visit_month']],\n",
    "                   npx_features,\n",
    "                   how='inner',\n",
    "                   left_on=['visit_id'],\n",
    "                   right_on=['visit_id'])\n",
    "dataset.head()\n",
    "Y = dataset[dataset.columns[0]].values\n",
    "X = dataset[dataset.columns[6:]].values\n",
    "X_noscale = dataset[dataset.columns[1:6]].values\n",
    "X_scaled = StandardScaler().fit_transform(X)\n",
    "final_X = np.hstack((X_noscale, X_scaled))\n",
    "\n",
    "s = setup(data=final_X, target=Y)\n",
    "add_metric('SMAPE', 'SMAPE', smape, greater_is_better=False)\n",
    "b4 = create_model('et')\n",
    "t4 = tune_model(b4, choose_better=True, optimize='SMAPE')\n",
    "ensemble_u4 = ensemble_model(\n",
    "    t4,\n",
    "    choose_better=True,\n",
    "    n_estimators=20,\n",
    "    method='Boosting')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_104d9\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_104d9_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_104d9_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_104d9_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_104d9_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_104d9_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_104d9_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_104d9_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_104d9_level0_col7\" class=\"col_heading level0 col7\" >SMAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_104d9_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_104d9_row0_col0\" class=\"data row0 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_104d9_row0_col1\" class=\"data row0 col1\" >0.9343</td>\n",
       "      <td id=\"T_104d9_row0_col2\" class=\"data row0 col2\" >3.1397</td>\n",
       "      <td id=\"T_104d9_row0_col3\" class=\"data row0 col3\" >1.7719</td>\n",
       "      <td id=\"T_104d9_row0_col4\" class=\"data row0 col4\" >0.3424</td>\n",
       "      <td id=\"T_104d9_row0_col5\" class=\"data row0 col5\" >0.5520</td>\n",
       "      <td id=\"T_104d9_row0_col6\" class=\"data row0 col6\" >0.6837</td>\n",
       "      <td id=\"T_104d9_row0_col7\" class=\"data row0 col7\" >165.3778</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2753855d250>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_58d9b\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_58d9b_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_58d9b_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_58d9b_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_58d9b_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_58d9b_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_58d9b_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_58d9b_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_58d9b_level0_col7\" class=\"col_heading level0 col7\" >SMAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_58d9b_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_58d9b_row0_col0\" class=\"data row0 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_58d9b_row0_col1\" class=\"data row0 col1\" >0.9016</td>\n",
       "      <td id=\"T_58d9b_row0_col2\" class=\"data row0 col2\" >5.5873</td>\n",
       "      <td id=\"T_58d9b_row0_col3\" class=\"data row0 col3\" >2.3637</td>\n",
       "      <td id=\"T_58d9b_row0_col4\" class=\"data row0 col4\" >-0.1703</td>\n",
       "      <td id=\"T_58d9b_row0_col5\" class=\"data row0 col5\" >0.7452</td>\n",
       "      <td id=\"T_58d9b_row0_col6\" class=\"data row0 col6\" >1.0000</td>\n",
       "      <td id=\"T_58d9b_row0_col7\" class=\"data row0 col7\" >37.4603</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x27538f467f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_d8cc1\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_d8cc1_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_d8cc1_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_d8cc1_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_d8cc1_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_d8cc1_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_d8cc1_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_d8cc1_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_d8cc1_level0_col7\" class=\"col_heading level0 col7\" >SMAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_d8cc1_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_d8cc1_row0_col0\" class=\"data row0 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_d8cc1_row0_col1\" class=\"data row0 col1\" >0.9016</td>\n",
       "      <td id=\"T_d8cc1_row0_col2\" class=\"data row0 col2\" >5.5873</td>\n",
       "      <td id=\"T_d8cc1_row0_col3\" class=\"data row0 col3\" >2.3637</td>\n",
       "      <td id=\"T_d8cc1_row0_col4\" class=\"data row0 col4\" >-0.1703</td>\n",
       "      <td id=\"T_d8cc1_row0_col5\" class=\"data row0 col5\" >0.7452</td>\n",
       "      <td id=\"T_d8cc1_row0_col6\" class=\"data row0 col6\" >1.0000</td>\n",
       "      <td id=\"T_d8cc1_row0_col7\" class=\"data row0 col7\" >37.4603</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x275385d1970>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-4 {color: black;background-color: white;}#sk-container-id-4 pre{padding: 0;}#sk-container-id-4 div.sk-toggleable {background-color: white;}#sk-container-id-4 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-4 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-4 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-4 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-4 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-4 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-4 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-4 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-4 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-4 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-4 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-4 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-4 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-4 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-4 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-4 div.sk-item {position: relative;z-index: 1;}#sk-container-id-4 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-4 div.sk-item::before, #sk-container-id-4 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-4 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-4 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-4 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-4 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-4 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-4 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-4 div.sk-label-container {text-align: center;}#sk-container-id-4 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-4 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-4\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(memory=FastMemory(location=C:\\Users\\prith\\AppData\\Local\\Temp\\joblib),\n",
       "         steps=[(&#x27;numerical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[&#x27;feature_1&#x27;, &#x27;feature_2&#x27;,\n",
       "                                             &#x27;feature_3&#x27;, &#x27;feature_4&#x27;,\n",
       "                                             &#x27;feature_5&#x27;, &#x27;feature_6&#x27;,\n",
       "                                             &#x27;feature_7&#x27;, &#x27;feature_8&#x27;,\n",
       "                                             &#x27;feature_9&#x27;, &#x27;feature_10&#x27;,\n",
       "                                             &#x27;feature_11&#x27;, &#x27;feature_12&#x27;,\n",
       "                                             &#x27;feature_13&#x27;, &#x27;feature_14&#x27;,\n",
       "                                             &#x27;feature_15&#x27;, &#x27;feature_16&#x27;,\n",
       "                                             &#x27;feature_17&#x27;, &#x27;fea...\n",
       "                                             &#x27;feature_29&#x27;, &#x27;feature_30&#x27;, ...],\n",
       "                                    transformer=SimpleImputer())),\n",
       "                (&#x27;categorical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[],\n",
       "                                    transformer=SimpleImputer(strategy=&#x27;most_frequent&#x27;))),\n",
       "                (&#x27;actual_estimator&#x27;,\n",
       "                 ExtraTreesRegressor(criterion=&#x27;absolute_error&#x27;, max_depth=5,\n",
       "                                     max_features=&#x27;sqrt&#x27;,\n",
       "                                     min_impurity_decrease=0.5,\n",
       "                                     min_samples_leaf=3, n_estimators=20,\n",
       "                                     n_jobs=-1, random_state=7109))])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-25\" type=\"checkbox\" ><label for=\"sk-estimator-id-25\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(memory=FastMemory(location=C:\\Users\\prith\\AppData\\Local\\Temp\\joblib),\n",
       "         steps=[(&#x27;numerical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[&#x27;feature_1&#x27;, &#x27;feature_2&#x27;,\n",
       "                                             &#x27;feature_3&#x27;, &#x27;feature_4&#x27;,\n",
       "                                             &#x27;feature_5&#x27;, &#x27;feature_6&#x27;,\n",
       "                                             &#x27;feature_7&#x27;, &#x27;feature_8&#x27;,\n",
       "                                             &#x27;feature_9&#x27;, &#x27;feature_10&#x27;,\n",
       "                                             &#x27;feature_11&#x27;, &#x27;feature_12&#x27;,\n",
       "                                             &#x27;feature_13&#x27;, &#x27;feature_14&#x27;,\n",
       "                                             &#x27;feature_15&#x27;, &#x27;feature_16&#x27;,\n",
       "                                             &#x27;feature_17&#x27;, &#x27;fea...\n",
       "                                             &#x27;feature_29&#x27;, &#x27;feature_30&#x27;, ...],\n",
       "                                    transformer=SimpleImputer())),\n",
       "                (&#x27;categorical_imputer&#x27;,\n",
       "                 TransformerWrapper(include=[],\n",
       "                                    transformer=SimpleImputer(strategy=&#x27;most_frequent&#x27;))),\n",
       "                (&#x27;actual_estimator&#x27;,\n",
       "                 ExtraTreesRegressor(criterion=&#x27;absolute_error&#x27;, max_depth=5,\n",
       "                                     max_features=&#x27;sqrt&#x27;,\n",
       "                                     min_impurity_decrease=0.5,\n",
       "                                     min_samples_leaf=3, n_estimators=20,\n",
       "                                     n_jobs=-1, random_state=7109))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-26\" type=\"checkbox\" ><label for=\"sk-estimator-id-26\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">numerical_imputer: TransformerWrapper</label><div class=\"sk-toggleable__content\"><pre>TransformerWrapper(include=[&#x27;feature_1&#x27;, &#x27;feature_2&#x27;, &#x27;feature_3&#x27;, &#x27;feature_4&#x27;,\n",
       "                            &#x27;feature_5&#x27;, &#x27;feature_6&#x27;, &#x27;feature_7&#x27;, &#x27;feature_8&#x27;,\n",
       "                            &#x27;feature_9&#x27;, &#x27;feature_10&#x27;, &#x27;feature_11&#x27;,\n",
       "                            &#x27;feature_12&#x27;, &#x27;feature_13&#x27;, &#x27;feature_14&#x27;,\n",
       "                            &#x27;feature_15&#x27;, &#x27;feature_16&#x27;, &#x27;feature_17&#x27;,\n",
       "                            &#x27;feature_18&#x27;, &#x27;feature_19&#x27;, &#x27;feature_20&#x27;,\n",
       "                            &#x27;feature_21&#x27;, &#x27;feature_22&#x27;, &#x27;feature_23&#x27;,\n",
       "                            &#x27;feature_24&#x27;, &#x27;feature_25&#x27;, &#x27;feature_26&#x27;,\n",
       "                            &#x27;feature_27&#x27;, &#x27;feature_28&#x27;, &#x27;feature_29&#x27;,\n",
       "                            &#x27;feature_30&#x27;, ...],\n",
       "                   transformer=SimpleImputer())</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-27\" type=\"checkbox\" ><label for=\"sk-estimator-id-27\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">transformer: SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer()</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-28\" type=\"checkbox\" ><label for=\"sk-estimator-id-28\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-29\" type=\"checkbox\" ><label for=\"sk-estimator-id-29\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">categorical_imputer: TransformerWrapper</label><div class=\"sk-toggleable__content\"><pre>TransformerWrapper(include=[],\n",
       "                   transformer=SimpleImputer(strategy=&#x27;most_frequent&#x27;))</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-30\" type=\"checkbox\" ><label for=\"sk-estimator-id-30\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">transformer: SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(strategy=&#x27;most_frequent&#x27;)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-31\" type=\"checkbox\" ><label for=\"sk-estimator-id-31\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer(strategy=&#x27;most_frequent&#x27;)</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-32\" type=\"checkbox\" ><label for=\"sk-estimator-id-32\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ExtraTreesRegressor</label><div class=\"sk-toggleable__content\"><pre>ExtraTreesRegressor(criterion=&#x27;absolute_error&#x27;, max_depth=5,\n",
       "                    max_features=&#x27;sqrt&#x27;, min_impurity_decrease=0.5,\n",
       "                    min_samples_leaf=3, n_estimators=20, n_jobs=-1,\n",
       "                    random_state=7109)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(memory=FastMemory(location=C:\\Users\\prith\\AppData\\Local\\Temp\\joblib),\n",
       "         steps=[('numerical_imputer',\n",
       "                 TransformerWrapper(include=['feature_1', 'feature_2',\n",
       "                                             'feature_3', 'feature_4',\n",
       "                                             'feature_5', 'feature_6',\n",
       "                                             'feature_7', 'feature_8',\n",
       "                                             'feature_9', 'feature_10',\n",
       "                                             'feature_11', 'feature_12',\n",
       "                                             'feature_13', 'feature_14',\n",
       "                                             'feature_15', 'feature_16',\n",
       "                                             'feature_17', 'fea...\n",
       "                                             'feature_29', 'feature_30', ...],\n",
       "                                    transformer=SimpleImputer())),\n",
       "                ('categorical_imputer',\n",
       "                 TransformerWrapper(include=[],\n",
       "                                    transformer=SimpleImputer(strategy='most_frequent'))),\n",
       "                ('actual_estimator',\n",
       "                 ExtraTreesRegressor(criterion='absolute_error', max_depth=5,\n",
       "                                     max_features='sqrt',\n",
       "                                     min_impurity_decrease=0.5,\n",
       "                                     min_samples_leaf=3, n_estimators=20,\n",
       "                                     n_jobs=-1, random_state=7109))])"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "predict_model(b4)\n",
    "predict_model(t4)\n",
    "predict_model(ensemble_u4)\n",
    "finalize_model(ensemble_u4)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-5 {color: black;background-color: white;}#sk-container-id-5 pre{padding: 0;}#sk-container-id-5 div.sk-toggleable {background-color: white;}#sk-container-id-5 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-5 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-5 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-5 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-5 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-5 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-5 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-5 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-5 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-5 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-5 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-5 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-5 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-5 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-5 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-5 div.sk-item {position: relative;z-index: 1;}#sk-container-id-5 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-5 div.sk-item::before, #sk-container-id-5 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-5 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-5 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-5 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-5 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-5 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-5 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-5 div.sk-label-container {text-align: center;}#sk-container-id-5 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-5 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-5\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>ExtraTreesRegressor(n_jobs=-1, random_state=5597)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-33\" type=\"checkbox\" checked><label for=\"sk-estimator-id-33\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ExtraTreesRegressor</label><div class=\"sk-toggleable__content\"><pre>ExtraTreesRegressor(n_jobs=-1, random_state=5597)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "ExtraTreesRegressor(n_jobs=-1, random_state=5597)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ensemble_u1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>visit_id</th>\n",
       "      <th>updrs_1</th>\n",
       "      <th>prediction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>23391_0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>23391_6</td>\n",
       "      <td>2.0</td>\n",
       "      <td>8.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>23391_12</td>\n",
       "      <td>4.0</td>\n",
       "      <td>10.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>23391_24</td>\n",
       "      <td>4.0</td>\n",
       "      <td>17.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>23391_36</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>23391_48</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>23391_60</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>23391_72</td>\n",
       "      <td>8.0</td>\n",
       "      <td>2.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>23391_84</td>\n",
       "      <td>11.0</td>\n",
       "      <td>11.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>27468_0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>18.82</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>27468_6</td>\n",
       "      <td>11.0</td>\n",
       "      <td>20.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>27468_12</td>\n",
       "      <td>12.0</td>\n",
       "      <td>19.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>27468_24</td>\n",
       "      <td>10.0</td>\n",
       "      <td>2.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>27468_36</td>\n",
       "      <td>16.0</td>\n",
       "      <td>2.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>27468_48</td>\n",
       "      <td>15.0</td>\n",
       "      <td>3.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>27468_60</td>\n",
       "      <td>21.0</td>\n",
       "      <td>2.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>27468_72</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>40200_0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>40200_12</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>40200_24</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>40200_36</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    visit_id  updrs_1  prediction\n",
       "0    23391_0      0.0        5.23\n",
       "1    23391_6      2.0        8.00\n",
       "2   23391_12      4.0       10.00\n",
       "3   23391_24      4.0       17.00\n",
       "4   23391_36      6.0        8.00\n",
       "5   23391_48      5.0        5.00\n",
       "6   23391_60      4.0        2.00\n",
       "7   23391_72      8.0        2.00\n",
       "8   23391_84     11.0       11.00\n",
       "9    27468_0      9.0       18.82\n",
       "10   27468_6     11.0       20.00\n",
       "11  27468_12     12.0       19.68\n",
       "12  27468_24     10.0        2.00\n",
       "13  27468_36     16.0        2.00\n",
       "14  27468_48     15.0        3.00\n",
       "15  27468_60     21.0        2.00\n",
       "16  27468_72     14.0        0.00\n",
       "17   40200_0      0.0        2.00\n",
       "18  40200_12      1.0        6.00\n",
       "19  40200_24      2.0        6.39\n",
       "20  40200_36      0.0        9.00"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "updrs1_test = pd.merge(clinical_data_holdout[['updrs_1',\n",
    "                                              'offset_updrs1_offset',\n",
    "                                              'offset_updrs2_offset',\n",
    "                                              'offset_updrs3_offset',\n",
    "                                              'offset_updrs4_offset',\n",
    "                                              'visit_month']],\n",
    "                       npx_features,\n",
    "                       how='inner',\n",
    "                       left_on=['visit_id'],\n",
    "                       right_on=['visit_id'])\n",
    "Y_test = updrs1_test[updrs1_test.columns[0]].values\n",
    "X_test = updrs1_test[updrs1_test.columns[6:]].values\n",
    "X_noscale_test = updrs1_test[updrs1_test.columns[1:6]].values\n",
    "X_scaled_test = StandardScaler().fit(X).transform(X_test)\n",
    "final_X_test = np.hstack((X_noscale, X_scaled))\n",
    "\n",
    "y_predict = predict_model(ensemble_u1, data=pd.DataFrame(final_X_test, columns=[\n",
    "                          f'feature_{i+1}' for i in range(final_X_test.shape[1])]))['prediction_label']\n",
    "updrs1_test.reset_index(inplace=True)\n",
    "updrs1_test['prediction'] = y_predict\n",
    "plot1 = updrs1_test[['visit_id', 'updrs_1', 'prediction']]\n",
    "plot1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>visit_id</th>\n",
       "      <th>updrs_2</th>\n",
       "      <th>prediction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>23391_0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.929650</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>23391_6</td>\n",
       "      <td>4.0</td>\n",
       "      <td>7.703273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>23391_12</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8.585667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>23391_24</td>\n",
       "      <td>7.0</td>\n",
       "      <td>13.017826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>23391_36</td>\n",
       "      <td>7.0</td>\n",
       "      <td>2.921820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>23391_48</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.720842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>23391_60</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.569220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>23391_72</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.784492</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>23391_84</td>\n",
       "      <td>15.0</td>\n",
       "      <td>2.915067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>27468_0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>12.941766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>27468_6</td>\n",
       "      <td>12.0</td>\n",
       "      <td>14.668456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>27468_12</td>\n",
       "      <td>7.0</td>\n",
       "      <td>17.686631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>27468_24</td>\n",
       "      <td>12.0</td>\n",
       "      <td>2.876168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>27468_36</td>\n",
       "      <td>13.0</td>\n",
       "      <td>2.241683</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>27468_48</td>\n",
       "      <td>14.0</td>\n",
       "      <td>1.361661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>27468_60</td>\n",
       "      <td>15.0</td>\n",
       "      <td>3.069718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>27468_72</td>\n",
       "      <td>14.0</td>\n",
       "      <td>1.600991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>40200_0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.459800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>40200_12</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1.474078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>40200_24</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.268124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>40200_36</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.080378</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    visit_id  updrs_2  prediction\n",
       "0    23391_0      3.0    2.929650\n",
       "1    23391_6      4.0    7.703273\n",
       "2   23391_12      5.0    8.585667\n",
       "3   23391_24      7.0   13.017826\n",
       "4   23391_36      7.0    2.921820\n",
       "5   23391_48      6.0    2.720842\n",
       "6   23391_60      5.0    5.569220\n",
       "7   23391_72      6.0    3.784492\n",
       "8   23391_84     15.0    2.915067\n",
       "9    27468_0      8.0   12.941766\n",
       "10   27468_6     12.0   14.668456\n",
       "11  27468_12      7.0   17.686631\n",
       "12  27468_24     12.0    2.876168\n",
       "13  27468_36     13.0    2.241683\n",
       "14  27468_48     14.0    1.361661\n",
       "15  27468_60     15.0    3.069718\n",
       "16  27468_72     14.0    1.600991\n",
       "17   40200_0      4.0    1.459800\n",
       "18  40200_12     10.0    1.474078\n",
       "19  40200_24      3.0    4.268124\n",
       "20  40200_36      7.0    8.080378"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "updrs2_test = pd.merge(clinical_data_holdout[['updrs_2',\n",
    "                                              'offset_updrs1_offset',\n",
    "                                              'offset_updrs2_offset',\n",
    "                                              'offset_updrs3_offset',\n",
    "                                              'offset_updrs4_offset',\n",
    "                                              'visit_month']],\n",
    "                       npx_features,\n",
    "                       how='inner',\n",
    "                       left_on=['visit_id'],\n",
    "                       right_on=['visit_id'])\n",
    "Y_test = updrs2_test[updrs2_test.columns[0]].values\n",
    "X_test = updrs2_test[updrs2_test.columns[6:]].values\n",
    "X_noscale_test = updrs2_test[updrs2_test.columns[1:6]].values\n",
    "X_scaled_test = StandardScaler().fit(X).transform(X_test)\n",
    "final_X_test = np.hstack((X_noscale, X_scaled))\n",
    "\n",
    "y_predict = predict_model(ensemble_u2, data=pd.DataFrame(final_X_test, columns=[\n",
    "                          f'feature_{i+1}' for i in range(final_X_test.shape[1])]))['prediction_label']\n",
    "updrs2_test.reset_index(inplace=True)\n",
    "updrs2_test['prediction'] = y_predict\n",
    "plot2 = updrs2_test[['visit_id', 'updrs_2', 'prediction']]\n",
    "plot2\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>visit_id</th>\n",
       "      <th>updrs_3</th>\n",
       "      <th>prediction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>23391_0</td>\n",
       "      <td>6</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>23391_6</td>\n",
       "      <td>58</td>\n",
       "      <td>58.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>23391_12</td>\n",
       "      <td>2</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>23391_24</td>\n",
       "      <td>16</td>\n",
       "      <td>9.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>23391_36</td>\n",
       "      <td>9</td>\n",
       "      <td>58.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>23391_48</td>\n",
       "      <td>51</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>23391_60</td>\n",
       "      <td>0</td>\n",
       "      <td>16.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>23391_72</td>\n",
       "      <td>58</td>\n",
       "      <td>51.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>23391_84</td>\n",
       "      <td>0</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>27468_0</td>\n",
       "      <td>6</td>\n",
       "      <td>16.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>27468_6</td>\n",
       "      <td>58</td>\n",
       "      <td>9.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>27468_12</td>\n",
       "      <td>2</td>\n",
       "      <td>1.692308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>27468_24</td>\n",
       "      <td>16</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>27468_36</td>\n",
       "      <td>9</td>\n",
       "      <td>16.053846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>27468_48</td>\n",
       "      <td>51</td>\n",
       "      <td>9.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>27468_60</td>\n",
       "      <td>0</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>27468_72</td>\n",
       "      <td>58</td>\n",
       "      <td>2.161538</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>40200_0</td>\n",
       "      <td>6</td>\n",
       "      <td>15.965385</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>40200_12</td>\n",
       "      <td>2</td>\n",
       "      <td>51.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>40200_24</td>\n",
       "      <td>16</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>40200_36</td>\n",
       "      <td>9</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    visit_id  updrs_3  prediction\n",
       "0    23391_0        6    6.000000\n",
       "1    23391_6       58   58.000000\n",
       "2   23391_12        2    2.000000\n",
       "3   23391_24       16    9.000000\n",
       "4   23391_36        9   58.000000\n",
       "5   23391_48       51    2.000000\n",
       "6   23391_60        0   16.000000\n",
       "7   23391_72       58   51.000000\n",
       "8   23391_84        0    6.000000\n",
       "9    27468_0        6   16.000000\n",
       "10   27468_6       58    9.000000\n",
       "11  27468_12        2    1.692308\n",
       "12  27468_24       16    6.000000\n",
       "13  27468_36        9   16.053846\n",
       "14  27468_48       51    9.000000\n",
       "15  27468_60        0    6.000000\n",
       "16  27468_72       58    2.161538\n",
       "17   40200_0        6   15.965385\n",
       "18  40200_12        2   51.000000\n",
       "19  40200_24       16    6.000000\n",
       "20  40200_36        9    2.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "updrs3_test = pd.merge(clinical_data_holdout[['updrs_3',\n",
    "                                              'offset_updrs1_offset',\n",
    "                                              'offset_updrs2_offset',\n",
    "                                              'offset_updrs3_offset',\n",
    "                                              'offset_updrs4_offset',\n",
    "                                              'visit_month']],\n",
    "                       npx_features,\n",
    "                       how='inner',\n",
    "                       left_on=['visit_id'],\n",
    "                       right_on=['visit_id'])\n",
    "Y_test = updrs3_test[updrs3_test.columns[0]].values\n",
    "X_test = updrs3_test[updrs3_test.columns[6:]].values\n",
    "X_noscale_test = updrs3_test[updrs3_test.columns[1:6]].values\n",
    "X_scaled_test = StandardScaler().fit(X).transform(X_test)\n",
    "final_X_test = np.hstack((X_noscale, X_scaled))\n",
    "\n",
    "y_predict = predict_model(ensemble_u3, data=pd.DataFrame(final_X_test, columns=[\n",
    "                          f'feature_{i+1}' for i in range(final_X_test.shape[1])]))['prediction_label']\n",
    "updrs3_test.reset_index(inplace=True)\n",
    "updrs3_test['prediction'] = y_predict\n",
    "plot3 = updrs3_test[['visit_id', 'updrs_3', 'prediction']]\n",
    "plot3\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>visit_id</th>\n",
       "      <th>updrs_4</th>\n",
       "      <th>prediction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>23391_0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>23391_6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>23391_12</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>23391_24</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>23391_36</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>23391_48</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>23391_60</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>23391_72</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>23391_84</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>27468_0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>27468_6</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>27468_12</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>27468_24</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>27468_36</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>27468_48</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>27468_60</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>27468_72</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>40200_0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>40200_12</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>40200_24</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>40200_36</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    visit_id  updrs_4  prediction\n",
       "0    23391_0      0.0         0.0\n",
       "1    23391_6      0.0         0.0\n",
       "2   23391_12      0.0         0.0\n",
       "3   23391_24      0.0         0.0\n",
       "4   23391_36      0.0         0.0\n",
       "5   23391_48      0.0         0.0\n",
       "6   23391_60      0.0         0.0\n",
       "7   23391_72      1.0         0.0\n",
       "8   23391_84      0.0         0.0\n",
       "9    27468_0      0.0         0.0\n",
       "10   27468_6      0.0         0.0\n",
       "11  27468_12      0.0         0.0\n",
       "12  27468_24      0.0         0.0\n",
       "13  27468_36      0.0         0.0\n",
       "14  27468_48      3.0         0.0\n",
       "15  27468_60      7.0         0.0\n",
       "16  27468_72      1.0         0.0\n",
       "17   40200_0      0.0         0.0\n",
       "18  40200_12      0.0         0.0\n",
       "19  40200_24      0.0         0.0\n",
       "20  40200_36      0.0         0.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "updrs4_test = pd.merge(clinical_data_holdout[['updrs_4',\n",
    "                                              'offset_updrs1_offset',\n",
    "                                              'offset_updrs2_offset',\n",
    "                                              'offset_updrs3_offset',\n",
    "                                              'offset_updrs4_offset',\n",
    "                                              'visit_month']],\n",
    "                       npx_features,\n",
    "                       how='inner',\n",
    "                       left_on=['visit_id'],\n",
    "                       right_on=['visit_id'])\n",
    "Y_test = updrs4_test[updrs4_test.columns[0]].values\n",
    "X_test = updrs4_test[updrs4_test.columns[6:]].values\n",
    "X_noscale_test = updrs4_test[updrs4_test.columns[1:6]].values\n",
    "X_scaled_test = StandardScaler().fit(X).transform(X_test)\n",
    "final_X_test = np.hstack((X_noscale, X_scaled))\n",
    "\n",
    "y_predict = predict_model(ensemble_u4, data=pd.DataFrame(final_X_test, columns=[\n",
    "                          f'feature_{i+1}' for i in range(final_X_test.shape[1])]))['prediction_label']\n",
    "updrs4_test.reset_index(inplace=True)\n",
    "updrs4_test['prediction'] = y_predict\n",
    "plot4 = updrs4_test[['visit_id', 'updrs_4', 'prediction']]\n",
    "plot4\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot1['patient'] = plot1['visit_id'].apply(lambda x: x.split('_')[0])\n",
    "plot2['patient'] = plot2['visit_id'].apply(lambda x: x.split('_')[0])\n",
    "plot3['patient'] = plot3['visit_id'].apply(lambda x: x.split('_')[0])\n",
    "plot4['patient'] = plot4['visit_id'].apply(lambda x: x.split('_')[0])\n",
    "plot1['month'] = plot1['visit_id'].apply(lambda x: int(x.split('_')[1]))\n",
    "plot2['month'] = plot2['visit_id'].apply(lambda x: int(x.split('_')[1]))\n",
    "plot3['month'] = plot3['visit_id'].apply(lambda x: int(x.split('_')[1]))\n",
    "plot4['month'] = plot4['visit_id'].apply(lambda x: int(x.split('_')[1]))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "pid = '23391'\n",
    "fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, figsize=(10, 5))\n",
    "plot1[plot1['patient'] == pid].plot.line(x='month', y='prediction', ax=ax1)\n",
    "plot1[plot1['patient'] == pid].plot.line(\n",
    "    x='month', y='updrs_1', ax=ax1, ls=\":\")\n",
    "plot2[plot2['patient'] == pid].plot.line(x='month', y='prediction', ax=ax2)\n",
    "plot2[plot2['patient'] == pid].plot.line(\n",
    "    x='month', y='updrs_2', ax=ax2, ls=':')\n",
    "plot3[plot3['patient'] == pid].plot.line(x='month', y='prediction', ax=ax3)\n",
    "plot3[plot3['patient'] == pid].plot.line(\n",
    "    x='month', y='updrs_3', ax=ax3, ls=':')\n",
    "plot4[plot4['patient'] == pid].plot.line(x='month', y='prediction', ax=ax4)\n",
    "plot4[plot4['patient'] == pid].plot.line(\n",
    "    x='month', y='updrs_4', ax=ax4, ls=':')\n",
    "plt.suptitle(f'Predictions for Patient {pid}')\n",
    "ax1.set_title('UPDRS 1')\n",
    "ax2.set_title('UPDRS 2')\n",
    "ax3.set_title('UPDRS 3')\n",
    "ax4.set_title('UPDRS 4')\n",
    "fig.tight_layout()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x550 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "pid = '40200'\n",
    "fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2)\n",
    "plot1[plot1['patient'] == pid].plot.line(x='month', y='prediction', ax=ax1)\n",
    "plot1[plot1['patient'] == pid].plot.line(\n",
    "    x='month', y='updrs_1', ax=ax1, ls=\":\")\n",
    "plot2[plot2['patient'] == pid].plot.line(x='month', y='prediction', ax=ax2)\n",
    "plot2[plot2['patient'] == pid].plot.line(\n",
    "    x='month', y='updrs_2', ax=ax2, ls=':')\n",
    "plot3[plot3['patient'] == pid].plot.line(x='month', y='prediction', ax=ax3)\n",
    "plot3[plot3['patient'] == pid].plot.line(\n",
    "    x='month', y='updrs_3', ax=ax3, ls=':')\n",
    "plot4[plot4['patient'] == pid].plot.line(x='month', y='prediction', ax=ax4)\n",
    "plot4[plot4['patient'] == pid].plot.line(\n",
    "    x='month', y='updrs_4', ax=ax4, ls=':')\n",
    "plt.suptitle(f'Predictions for Patient {pid}')\n",
    "ax1.set_title('UPDRS 1')\n",
    "ax2.set_title('UPDRS 2')\n",
    "ax3.set_title('UPDRS 3')\n",
    "ax4.set_title('UPDRS 4')\n",
    "fig.tight_layout()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x550 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "pid = '27468'\n",
    "fig, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2)\n",
    "plot1[plot1['patient'] == pid].plot.line(x='month', y='prediction', ax=ax1)\n",
    "plot1[plot1['patient'] == pid].plot.line(\n",
    "    x='month', y='updrs_1', ax=ax1, ls=\":\")\n",
    "plot2[plot2['patient'] == pid].plot.line(x='month', y='prediction', ax=ax2)\n",
    "plot2[plot2['patient'] == pid].plot.line(\n",
    "    x='month', y='updrs_2', ax=ax2, ls=':')\n",
    "plot3[plot3['patient'] == pid].plot.line(x='month', y='prediction', ax=ax3)\n",
    "plot3[plot3['patient'] == pid].plot.line(\n",
    "    x='month', y='updrs_3', ax=ax3, ls=':')\n",
    "plot4[plot4['patient'] == pid].plot.line(x='month', y='prediction', ax=ax4)\n",
    "plot4[plot4['patient'] == pid].plot.line(\n",
    "    x='month', y='updrs_4', ax=ax4, ls=':')\n",
    "plt.suptitle(f'Predictions for Patient {pid}')\n",
    "ax1.set_title('UPDRS 1')\n",
    "ax2.set_title('UPDRS 2')\n",
    "ax3.set_title('UPDRS 3')\n",
    "ax4.set_title('UPDRS 4')\n",
    "fig.tight_layout()\n"
   ]
  }
 ],
 "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": "3.10.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}