3_MLProject_Exploratory_ML_Models.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3) ML Project - Air Pollution - Exploratory ML Models (Auto-ML)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font color='blue'> Scripts by Kevin Jeswani, Junaid Zafar, & Laiba Shah \n",
    "<br>\n",
    "WeCloudData Bootcamp 2022 (Part-time Cohort)<br> </font>\n",
    "\n",
    "First we take a look at a dataset for air pollutions deaths due over time for different countries. We visualize data and develop the linear regression models on the regional and national level. The linear regression models are valuable at predicting the general trend in deaths due over time and be used to predict future deaths, where past data appears to be highly linear. It has severe limitations for many countries. Polynomial models could be explored. To explore the interaction with other features, we proceed to merge different datasets between death due to environmental factors and socio-demographic/economic factors. We then examine them in 3 tree-based machine learning models.\n",
    "\n",
    "The set of scripts in the folder 'AirPollution_Scripts' are segmented for the purpose of clarity & convenience <br>\n",
    "The following is the suggested order for running the scripts:\n",
    "- 'MLProject_AirPollution_Viz_LR.ipynb' - loading the initial Kaggle air pollution deaths dataset, processing, visualizing the base dataset, linear regression examination and future predictions.\n",
    "- 'MLProject_DataProcessing_OECD_WB_GBD.ipynb'  - loading raw data from world bank and GBD/OECD datasets, cleaning, and merging usable data. We test the use of fuzzy data set merging.\n",
    "- 'MLProject_Exploratory_ML_Models.ipynb' **This Script**- loading the cleaning & merged dataset, model selection with PyCaret, and discussion of models/insights in modelling data.\n",
    "- 'MLProject_ML_Models_refined.ipynb' - development of selected machine learning models (Extra Trees, Random Forest, XGBoost), cross-validation, & Ensemble methods; discussion of model results\n",
    "\n",
    "The aim is to explore the intial dataset (https://www.kaggle.com/datasets/akshat0giri/death-due-to-air-pollution-19902017?resource=download), supplement it, infer trends, and make predictions on future deaths due to environmental factors. \n",
    "The data is then supplement with:\n",
    "- Global Burden of Disease Study 2019 Results via OECD.stats: https://stats.oecd.org/index.aspx?lang=en# (where the Kaggle dataset seems to have extracted and cleaned data from)\n",
    "- World Bank Development Indicators: https://databank.worldbank.org/source/world-development-indicators#\n",
    "<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.0 Pre-processing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Steps to run PyCaret in Conda virtual environment: <br>\n",
    "- Open CMD.exe prompt in conda navigator <br>\n",
    "- create a conda environment: conda create --name pycaret3 python=3.9.16 <br>\n",
    "- activate conda environment: conda activate pycaret3 <br>\n",
    "- install pycaret: pip install pycaret <br>\n",
    "- conda install ipykernel\n",
    "- create notebook kernel: python -m ipykernel install --user --name pycaret3 --display-name \"display-name\" <br>\n",
    "<br>\n",
    "Install Fugue for parallel processing backend into Spark: <br>\n",
    "!pip install fugue <br>\n",
    "Install pyspark <br>\n",
    "!pip install pyspark\n",
    "Install Dask <br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!pip install fugue"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "!pip install pyspark"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.1 Standard Imports & Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Data Processing\n",
    "import numpy as np \n",
    "import numexpr\n",
    "\n",
    "import pandas as pd\n",
    "import csv\n",
    "\n",
    "#Plotting\n",
    "import plotly.express as px\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "sns.set()\n",
    "\n",
    "#Machine Learning - general tools\n",
    "from sklearn.model_selection import train_test_split #Test/Train and Holdout split\n",
    "from sklearn.preprocessing import OrdinalEncoder #PyCaret ordinal encoder causing issues "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# PyCaret\n",
    "from pycaret.regression import * #all regression submodules\n",
    "from pycaret.parallel import FugueBackend #for parallel processing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dask - parallel processing with pycaret\n",
    "# PyCaret fugue-dask does not always work\n",
    "# from dask import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create pyspark session for parallel processing\n",
    "# PyCaret fugue-spark does not always work\n",
    "# from pyspark.sql import SparkSession\n",
    "# spark = SparkSession.builder.getOrCreate()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.0 Data Examination & Processing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**NOTE** Steps in 2.0 are repeated in the 4th script for completeness"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "root = r'C:\\Users\\kevin\\Documents\\WCD_Bootcamp\\ML\\ML_Project\\Scripts'\n",
    "root_raw=r'C:\\Users\\kevin\\Documents\\WCD_Bootcamp\\ML\\ML_Project\\Scripts\\Raw Data'\n",
    "root_proc=r'C:\\Users\\kevin\\Documents\\WCD_Bootcamp\\ML\\ML_Project\\Scripts\\Processed'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.1 Load Data into DF & Examine Contents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>Unnamed: 0</th>\n",
       "      <th>COU</th>\n",
       "      <th>Country</th>\n",
       "      <th>Year</th>\n",
       "      <th>Deaths_OZone</th>\n",
       "      <th>Deaths_PM2.5</th>\n",
       "      <th>Deaths_Indoor</th>\n",
       "      <th>Deaths_Handwashing</th>\n",
       "      <th>Deaths_Sanitation</th>\n",
       "      <th>Deaths_Water</th>\n",
       "      <th>Deaths_HighTemp</th>\n",
       "      <th>Deaths_LowTemp</th>\n",
       "      <th>GDP_pCapita</th>\n",
       "      <th>Population</th>\n",
       "      <th>Income_Class</th>\n",
       "      <th>SoDI</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12</td>\n",
       "      <td>AFG</td>\n",
       "      <td>Afghanistan</td>\n",
       "      <td>2002</td>\n",
       "      <td>0.008418</td>\n",
       "      <td>0.069128</td>\n",
       "      <td>0.864084</td>\n",
       "      <td>0.154955</td>\n",
       "      <td>0.146771</td>\n",
       "      <td>0.195852</td>\n",
       "      <td>0.027761</td>\n",
       "      <td>7775.0</td>\n",
       "      <td>179.426579</td>\n",
       "      <td>22600774.0</td>\n",
       "      <td>L</td>\n",
       "      <td>0.194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>13</td>\n",
       "      <td>AFG</td>\n",
       "      <td>Afghanistan</td>\n",
       "      <td>2003</td>\n",
       "      <td>0.007989</td>\n",
       "      <td>0.065408</td>\n",
       "      <td>0.801610</td>\n",
       "      <td>0.151253</td>\n",
       "      <td>0.144548</td>\n",
       "      <td>0.193319</td>\n",
       "      <td>0.025447</td>\n",
       "      <td>8358.0</td>\n",
       "      <td>190.683814</td>\n",
       "      <td>23680871.0</td>\n",
       "      <td>L</td>\n",
       "      <td>0.202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>14</td>\n",
       "      <td>AFG</td>\n",
       "      <td>Afghanistan</td>\n",
       "      <td>2004</td>\n",
       "      <td>0.007311</td>\n",
       "      <td>0.060990</td>\n",
       "      <td>0.728383</td>\n",
       "      <td>0.138042</td>\n",
       "      <td>0.131303</td>\n",
       "      <td>0.176003</td>\n",
       "      <td>0.023667</td>\n",
       "      <td>7936.0</td>\n",
       "      <td>211.382074</td>\n",
       "      <td>24726689.0</td>\n",
       "      <td>L</td>\n",
       "      <td>0.209</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>15</td>\n",
       "      <td>AFG</td>\n",
       "      <td>Afghanistan</td>\n",
       "      <td>2005</td>\n",
       "      <td>0.006731</td>\n",
       "      <td>0.056720</td>\n",
       "      <td>0.657609</td>\n",
       "      <td>0.122056</td>\n",
       "      <td>0.114443</td>\n",
       "      <td>0.153736</td>\n",
       "      <td>0.020634</td>\n",
       "      <td>8244.0</td>\n",
       "      <td>242.031313</td>\n",
       "      <td>25654274.0</td>\n",
       "      <td>L</td>\n",
       "      <td>0.216</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>16</td>\n",
       "      <td>AFG</td>\n",
       "      <td>Afghanistan</td>\n",
       "      <td>2006</td>\n",
       "      <td>0.006097</td>\n",
       "      <td>0.054472</td>\n",
       "      <td>0.597933</td>\n",
       "      <td>0.107069</td>\n",
       "      <td>0.096464</td>\n",
       "      <td>0.129969</td>\n",
       "      <td>0.022527</td>\n",
       "      <td>7277.0</td>\n",
       "      <td>263.733602</td>\n",
       "      <td>26433058.0</td>\n",
       "      <td>L</td>\n",
       "      <td>0.224</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5707</th>\n",
       "      <td>6055</td>\n",
       "      <td>ZWE</td>\n",
       "      <td>Zimbabwe</td>\n",
       "      <td>2015</td>\n",
       "      <td>0.002620</td>\n",
       "      <td>0.145930</td>\n",
       "      <td>0.546781</td>\n",
       "      <td>0.226782</td>\n",
       "      <td>0.150856</td>\n",
       "      <td>0.227201</td>\n",
       "      <td>0.020383</td>\n",
       "      <td>742.0</td>\n",
       "      <td>1445.069702</td>\n",
       "      <td>13814642.0</td>\n",
       "      <td>L</td>\n",
       "      <td>0.452</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5708</th>\n",
       "      <td>6056</td>\n",
       "      <td>ZWE</td>\n",
       "      <td>Zimbabwe</td>\n",
       "      <td>2016</td>\n",
       "      <td>0.002946</td>\n",
       "      <td>0.138328</td>\n",
       "      <td>0.526542</td>\n",
       "      <td>0.218186</td>\n",
       "      <td>0.142138</td>\n",
       "      <td>0.215595</td>\n",
       "      <td>0.023571</td>\n",
       "      <td>762.0</td>\n",
       "      <td>1464.588957</td>\n",
       "      <td>14030338.0</td>\n",
       "      <td>L</td>\n",
       "      <td>0.459</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5709</th>\n",
       "      <td>6057</td>\n",
       "      <td>ZWE</td>\n",
       "      <td>Zimbabwe</td>\n",
       "      <td>2017</td>\n",
       "      <td>0.003454</td>\n",
       "      <td>0.129761</td>\n",
       "      <td>0.506067</td>\n",
       "      <td>0.209739</td>\n",
       "      <td>0.135385</td>\n",
       "      <td>0.206877</td>\n",
       "      <td>0.008881</td>\n",
       "      <td>891.0</td>\n",
       "      <td>1235.189032</td>\n",
       "      <td>14236599.0</td>\n",
       "      <td>L</td>\n",
       "      <td>0.465</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5710</th>\n",
       "      <td>6058</td>\n",
       "      <td>ZWE</td>\n",
       "      <td>Zimbabwe</td>\n",
       "      <td>2018</td>\n",
       "      <td>0.003310</td>\n",
       "      <td>0.124713</td>\n",
       "      <td>0.485084</td>\n",
       "      <td>0.199204</td>\n",
       "      <td>0.125096</td>\n",
       "      <td>0.192489</td>\n",
       "      <td>0.013527</td>\n",
       "      <td>767.0</td>\n",
       "      <td>1254.642265</td>\n",
       "      <td>14438812.0</td>\n",
       "      <td>LM</td>\n",
       "      <td>0.471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5711</th>\n",
       "      <td>6059</td>\n",
       "      <td>ZWE</td>\n",
       "      <td>Zimbabwe</td>\n",
       "      <td>2019</td>\n",
       "      <td>0.003403</td>\n",
       "      <td>0.121544</td>\n",
       "      <td>0.467108</td>\n",
       "      <td>0.191757</td>\n",
       "      <td>0.118001</td>\n",
       "      <td>0.182479</td>\n",
       "      <td>0.015199</td>\n",
       "      <td>744.0</td>\n",
       "      <td>1316.740657</td>\n",
       "      <td>14645473.0</td>\n",
       "      <td>LM</td>\n",
       "      <td>0.476</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5712 rows × 16 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Unnamed: 0  COU      Country  Year  Deaths_OZone  Deaths_PM2.5  \\\n",
       "0             12  AFG  Afghanistan  2002      0.008418      0.069128   \n",
       "1             13  AFG  Afghanistan  2003      0.007989      0.065408   \n",
       "2             14  AFG  Afghanistan  2004      0.007311      0.060990   \n",
       "3             15  AFG  Afghanistan  2005      0.006731      0.056720   \n",
       "4             16  AFG  Afghanistan  2006      0.006097      0.054472   \n",
       "...          ...  ...          ...   ...           ...           ...   \n",
       "5707        6055  ZWE     Zimbabwe  2015      0.002620      0.145930   \n",
       "5708        6056  ZWE     Zimbabwe  2016      0.002946      0.138328   \n",
       "5709        6057  ZWE     Zimbabwe  2017      0.003454      0.129761   \n",
       "5710        6058  ZWE     Zimbabwe  2018      0.003310      0.124713   \n",
       "5711        6059  ZWE     Zimbabwe  2019      0.003403      0.121544   \n",
       "\n",
       "      Deaths_Indoor  Deaths_Handwashing  Deaths_Sanitation  Deaths_Water  \\\n",
       "0          0.864084            0.154955           0.146771      0.195852   \n",
       "1          0.801610            0.151253           0.144548      0.193319   \n",
       "2          0.728383            0.138042           0.131303      0.176003   \n",
       "3          0.657609            0.122056           0.114443      0.153736   \n",
       "4          0.597933            0.107069           0.096464      0.129969   \n",
       "...             ...                 ...                ...           ...   \n",
       "5707       0.546781            0.226782           0.150856      0.227201   \n",
       "5708       0.526542            0.218186           0.142138      0.215595   \n",
       "5709       0.506067            0.209739           0.135385      0.206877   \n",
       "5710       0.485084            0.199204           0.125096      0.192489   \n",
       "5711       0.467108            0.191757           0.118001      0.182479   \n",
       "\n",
       "      Deaths_HighTemp  Deaths_LowTemp  GDP_pCapita  Population Income_Class  \\\n",
       "0            0.027761          7775.0   179.426579  22600774.0            L   \n",
       "1            0.025447          8358.0   190.683814  23680871.0            L   \n",
       "2            0.023667          7936.0   211.382074  24726689.0            L   \n",
       "3            0.020634          8244.0   242.031313  25654274.0            L   \n",
       "4            0.022527          7277.0   263.733602  26433058.0            L   \n",
       "...               ...             ...          ...         ...          ...   \n",
       "5707         0.020383           742.0  1445.069702  13814642.0            L   \n",
       "5708         0.023571           762.0  1464.588957  14030338.0            L   \n",
       "5709         0.008881           891.0  1235.189032  14236599.0            L   \n",
       "5710         0.013527           767.0  1254.642265  14438812.0           LM   \n",
       "5711         0.015199           744.0  1316.740657  14645473.0           LM   \n",
       "\n",
       "       SoDI  \n",
       "0     0.194  \n",
       "1     0.202  \n",
       "2     0.209  \n",
       "3     0.216  \n",
       "4     0.224  \n",
       "...     ...  \n",
       "5707  0.452  \n",
       "5708  0.459  \n",
       "5709  0.465  \n",
       "5710  0.471  \n",
       "5711  0.476  \n",
       "\n",
       "[5712 rows x 16 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fid = '\\MergedData_EnvDeaths_GDP_Pop_Income_SoDI_2.csv'\n",
    "dtypes = {'Population':'float','GDP_pCapita':'float'} #specify data types as required\n",
    "df_in = pd.read_csv(root_proc+fid,dtype=dtypes)\n",
    "df_in"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_in=df_in.drop(['Unnamed: 0'],axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.0 Use PyCaret for Model Selection - Regression (Boosted Trees)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**NOTE** EXERCISE CAUTION WHEN RUNNING, THE KERNEL MAY KEEP DYING, MODELS MAY NEED TO BE RUN ONE AT A TIME AND REPEATEDLY"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Feature Engineering & Encoding\n",
    "Ordinal Features: Income_class\n",
    "- Set ord_features=['Income_Class'] for ordinal_features in setup, if auto-encoding in PyCaret desired\n",
    "- To avoid dashboard issues, perform ordinal encoding with sklearn beforehand\n",
    "Feature Scaling:\n",
    "- By default PyCaret will use the StandardScaler, if nothing is specified in the setup()\n",
    "- This will take care of the difference in magnitude between the numeric features (i.e., population being much larger than the rest) <br>\n",
    "\n",
    "Regression with PyCaret Examples: \n",
    "- https://nbviewer.org/github/pycaret/pycaret/blob/master/tutorials/Tutorial%20-%20Regression.ipynb\n",
    "- https://pycaret.readthedocs.io/en/latest/api/regression.html#"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Encore Ordinal Features before PyCaret to avoid dashboard mismatch issues\n",
    "ordinal_encoder = OrdinalEncoder(categories=[['L', 'LM', 'UM', 'H']]) #Initialize encored on Income class col\n",
    "df_in['Income_Class_encoded'] = ordinal_encoder.fit_transform(df_in[['Income_Class']])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Independent Variables\n",
    "# Numeric features\n",
    "num_features=['GDP_pCapita','Population','SoDI','Income_Class_encoded']\n",
    "date_features=['Year'] #cast as datetime\n",
    "X=df_in.loc[:,num_features+date_features]\n",
    "X['Year'] = X['Year'].astype('datetime64[ns]') #cast the year as datetime64"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.1 Model Selection for Deaths due to Indoor Air Pollution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Dependent variable - classify the level of indoor pollution death\n",
    "y=df_in.loc[:,['Deaths_Indoor']]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "PyCaret conducts a 70-30 train-test split by default. The data saved for testing is not utilized during the cross-validation exercise. Thus, by default it saves the test set as a hold-out testing, which can be checked upon predicting target values on the full data set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = pd.concat([X,y],axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.1.1 Baseline Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Setup Parameters:\n",
    "- n_jobs=1 if parallel processing with fugue-spark/-dask;  required for spark/dark to avoid deadlocks; else =-1\n",
    "- 10-fold Cross Validation\n",
    "- session_id is equivalent to random seed to reproducibility\n",
    "- Specifying 80-20 train-test split\n",
    "- If using PyCaret ordinal features: ordinal_features={'Income_Class':['L','LM','UM','H']} = mismatch in dashboard output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_9ac44_row9_col1, #T_9ac44_row16_col1 {\n",
       "  background-color: lightgreen;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_9ac44\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_9ac44_level0_col0\" class=\"col_heading level0 col0\" >Description</th>\n",
       "      <th id=\"T_9ac44_level0_col1\" class=\"col_heading level0 col1\" >Value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_9ac44_row0_col0\" class=\"data row0 col0\" >Session id</td>\n",
       "      <td id=\"T_9ac44_row0_col1\" class=\"data row0 col1\" >123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_9ac44_row1_col0\" class=\"data row1 col0\" >Target</td>\n",
       "      <td id=\"T_9ac44_row1_col1\" class=\"data row1 col1\" >Deaths_Indoor</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_9ac44_row2_col0\" class=\"data row2 col0\" >Target type</td>\n",
       "      <td id=\"T_9ac44_row2_col1\" class=\"data row2 col1\" >Regression</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_9ac44_row3_col0\" class=\"data row3 col0\" >Original data shape</td>\n",
       "      <td id=\"T_9ac44_row3_col1\" class=\"data row3 col1\" >(5712, 6)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_9ac44_row4_col0\" class=\"data row4 col0\" >Transformed data shape</td>\n",
       "      <td id=\"T_9ac44_row4_col1\" class=\"data row4 col1\" >(5712, 8)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_9ac44_row5_col0\" class=\"data row5 col0\" >Transformed train set shape</td>\n",
       "      <td id=\"T_9ac44_row5_col1\" class=\"data row5 col1\" >(4569, 8)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_9ac44_row6_col0\" class=\"data row6 col0\" >Transformed test set shape</td>\n",
       "      <td id=\"T_9ac44_row6_col1\" class=\"data row6 col1\" >(1143, 8)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_9ac44_row7_col0\" class=\"data row7 col0\" >Numeric features</td>\n",
       "      <td id=\"T_9ac44_row7_col1\" class=\"data row7 col1\" >4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_9ac44_row8_col0\" class=\"data row8 col0\" >Date features</td>\n",
       "      <td id=\"T_9ac44_row8_col1\" class=\"data row8 col1\" >1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_9ac44_row9_col0\" class=\"data row9 col0\" >Preprocess</td>\n",
       "      <td id=\"T_9ac44_row9_col1\" class=\"data row9 col1\" >True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row10\" class=\"row_heading level0 row10\" >10</th>\n",
       "      <td id=\"T_9ac44_row10_col0\" class=\"data row10 col0\" >Imputation type</td>\n",
       "      <td id=\"T_9ac44_row10_col1\" class=\"data row10 col1\" >simple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row11\" class=\"row_heading level0 row11\" >11</th>\n",
       "      <td id=\"T_9ac44_row11_col0\" class=\"data row11 col0\" >Numeric imputation</td>\n",
       "      <td id=\"T_9ac44_row11_col1\" class=\"data row11 col1\" >mean</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row12\" class=\"row_heading level0 row12\" >12</th>\n",
       "      <td id=\"T_9ac44_row12_col0\" class=\"data row12 col0\" >Categorical imputation</td>\n",
       "      <td id=\"T_9ac44_row12_col1\" class=\"data row12 col1\" >mode</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row13\" class=\"row_heading level0 row13\" >13</th>\n",
       "      <td id=\"T_9ac44_row13_col0\" class=\"data row13 col0\" >Fold Generator</td>\n",
       "      <td id=\"T_9ac44_row13_col1\" class=\"data row13 col1\" >KFold</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row14\" class=\"row_heading level0 row14\" >14</th>\n",
       "      <td id=\"T_9ac44_row14_col0\" class=\"data row14 col0\" >Fold Number</td>\n",
       "      <td id=\"T_9ac44_row14_col1\" class=\"data row14 col1\" >10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row15\" class=\"row_heading level0 row15\" >15</th>\n",
       "      <td id=\"T_9ac44_row15_col0\" class=\"data row15 col0\" >CPU Jobs</td>\n",
       "      <td id=\"T_9ac44_row15_col1\" class=\"data row15 col1\" >-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row16\" class=\"row_heading level0 row16\" >16</th>\n",
       "      <td id=\"T_9ac44_row16_col0\" class=\"data row16 col0\" >Use GPU</td>\n",
       "      <td id=\"T_9ac44_row16_col1\" class=\"data row16 col1\" >True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row17\" class=\"row_heading level0 row17\" >17</th>\n",
       "      <td id=\"T_9ac44_row17_col0\" class=\"data row17 col0\" >Log Experiment</td>\n",
       "      <td id=\"T_9ac44_row17_col1\" class=\"data row17 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row18\" class=\"row_heading level0 row18\" >18</th>\n",
       "      <td id=\"T_9ac44_row18_col0\" class=\"data row18 col0\" >Experiment Name</td>\n",
       "      <td id=\"T_9ac44_row18_col1\" class=\"data row18 col1\" >reg-default-name</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_9ac44_level0_row19\" class=\"row_heading level0 row19\" >19</th>\n",
       "      <td id=\"T_9ac44_row19_col0\" class=\"data row19 col0\" >USI</td>\n",
       "      <td id=\"T_9ac44_row19_col1\" class=\"data row19 col1\" >f8c8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x248552f3d30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Initialize the PyCaret setup\n",
    "s = setup(data=dataset,target='Deaths_Indoor',\n",
    "          numeric_features=num_features,date_features=date_features,train_size=0.8,\n",
    "          use_gpu=True,fold=10,n_jobs=-1,session_id=123) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "K-Fold (with k=10) is also conducted: resampling is implemented k times without data replacement, to avoid overfitting. <br>\n",
    "The train set is split into sub train-test splits for the cross-validation. The test set will be saved for additional holdout testing after the model is developed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_246ba th {\n",
       "  text-align: left;\n",
       "}\n",
       "#T_246ba_row0_col0, #T_246ba_row0_col6, #T_246ba_row1_col0, #T_246ba_row1_col1, #T_246ba_row1_col2, #T_246ba_row1_col3, #T_246ba_row1_col4, #T_246ba_row1_col5, #T_246ba_row2_col0, #T_246ba_row2_col1, #T_246ba_row2_col2, #T_246ba_row2_col3, #T_246ba_row2_col4, #T_246ba_row2_col5, #T_246ba_row2_col6, #T_246ba_row3_col0, #T_246ba_row3_col1, #T_246ba_row3_col2, #T_246ba_row3_col3, #T_246ba_row3_col4, #T_246ba_row3_col5, #T_246ba_row3_col6, #T_246ba_row4_col0, #T_246ba_row4_col1, #T_246ba_row4_col2, #T_246ba_row4_col3, #T_246ba_row4_col4, #T_246ba_row4_col5, #T_246ba_row4_col6, #T_246ba_row5_col0, #T_246ba_row5_col1, #T_246ba_row5_col2, #T_246ba_row5_col3, #T_246ba_row5_col4, #T_246ba_row5_col5, #T_246ba_row5_col6, #T_246ba_row6_col0, #T_246ba_row6_col1, #T_246ba_row6_col2, #T_246ba_row6_col3, #T_246ba_row6_col4, #T_246ba_row6_col5, #T_246ba_row6_col6, #T_246ba_row7_col0, #T_246ba_row7_col1, #T_246ba_row7_col2, #T_246ba_row7_col3, #T_246ba_row7_col4, #T_246ba_row7_col5, #T_246ba_row7_col6, #T_246ba_row8_col0, #T_246ba_row8_col1, #T_246ba_row8_col2, #T_246ba_row8_col3, #T_246ba_row8_col4, #T_246ba_row8_col5, #T_246ba_row8_col6, #T_246ba_row9_col0, #T_246ba_row9_col1, #T_246ba_row9_col2, #T_246ba_row9_col3, #T_246ba_row9_col4, #T_246ba_row9_col5, #T_246ba_row9_col6, #T_246ba_row10_col0, #T_246ba_row10_col1, #T_246ba_row10_col2, #T_246ba_row10_col3, #T_246ba_row10_col4, #T_246ba_row10_col5, #T_246ba_row10_col6, #T_246ba_row11_col0, #T_246ba_row11_col1, #T_246ba_row11_col2, #T_246ba_row11_col3, #T_246ba_row11_col4, #T_246ba_row11_col5, #T_246ba_row11_col6, #T_246ba_row12_col0, #T_246ba_row12_col1, #T_246ba_row12_col2, #T_246ba_row12_col3, #T_246ba_row12_col4, #T_246ba_row12_col5, #T_246ba_row12_col6, #T_246ba_row13_col0, #T_246ba_row13_col1, #T_246ba_row13_col2, #T_246ba_row13_col3, #T_246ba_row13_col4, #T_246ba_row13_col5, #T_246ba_row13_col6, #T_246ba_row14_col0, #T_246ba_row14_col1, #T_246ba_row14_col2, #T_246ba_row14_col3, #T_246ba_row14_col4, #T_246ba_row14_col5, #T_246ba_row14_col6, #T_246ba_row15_col0, #T_246ba_row15_col1, #T_246ba_row15_col2, #T_246ba_row15_col3, #T_246ba_row15_col4, #T_246ba_row15_col5, #T_246ba_row15_col6, #T_246ba_row16_col0, #T_246ba_row16_col1, #T_246ba_row16_col2, #T_246ba_row16_col3, #T_246ba_row16_col4, #T_246ba_row16_col5, #T_246ba_row16_col6 {\n",
       "  text-align: left;\n",
       "}\n",
       "#T_246ba_row0_col1, #T_246ba_row0_col2, #T_246ba_row0_col3, #T_246ba_row0_col4, #T_246ba_row0_col5, #T_246ba_row1_col6 {\n",
       "  text-align: left;\n",
       "  background-color: yellow;\n",
       "}\n",
       "#T_246ba_row0_col7, #T_246ba_row2_col7, #T_246ba_row3_col7, #T_246ba_row4_col7, #T_246ba_row5_col7, #T_246ba_row6_col7, #T_246ba_row7_col7, #T_246ba_row8_col7, #T_246ba_row9_col7, #T_246ba_row10_col7, #T_246ba_row11_col7, #T_246ba_row12_col7, #T_246ba_row13_col7, #T_246ba_row14_col7, #T_246ba_row15_col7, #T_246ba_row16_col7 {\n",
       "  text-align: left;\n",
       "  background-color: lightgrey;\n",
       "}\n",
       "#T_246ba_row1_col7 {\n",
       "  text-align: left;\n",
       "  background-color: yellow;\n",
       "  background-color: lightgrey;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_246ba\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_246ba_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_246ba_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_246ba_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_246ba_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_246ba_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_246ba_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_246ba_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_246ba_level0_col7\" class=\"col_heading level0 col7\" >TT (Sec)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row0\" class=\"row_heading level0 row0\" >rf</th>\n",
       "      <td id=\"T_246ba_row0_col0\" class=\"data row0 col0\" >Random Forest Regressor</td>\n",
       "      <td id=\"T_246ba_row0_col1\" class=\"data row0 col1\" >1.0062</td>\n",
       "      <td id=\"T_246ba_row0_col2\" class=\"data row0 col2\" >47.0806</td>\n",
       "      <td id=\"T_246ba_row0_col3\" class=\"data row0 col3\" >5.9146</td>\n",
       "      <td id=\"T_246ba_row0_col4\" class=\"data row0 col4\" >0.9815</td>\n",
       "      <td id=\"T_246ba_row0_col5\" class=\"data row0 col5\" >0.1739</td>\n",
       "      <td id=\"T_246ba_row0_col6\" class=\"data row0 col6\" >3.7086</td>\n",
       "      <td id=\"T_246ba_row0_col7\" class=\"data row0 col7\" >2.8530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row1\" class=\"row_heading level0 row1\" >dt</th>\n",
       "      <td id=\"T_246ba_row1_col0\" class=\"data row1 col0\" >Decision Tree Regressor</td>\n",
       "      <td id=\"T_246ba_row1_col1\" class=\"data row1 col1\" >1.0125</td>\n",
       "      <td id=\"T_246ba_row1_col2\" class=\"data row1 col2\" >75.1113</td>\n",
       "      <td id=\"T_246ba_row1_col3\" class=\"data row1 col3\" >7.0782</td>\n",
       "      <td id=\"T_246ba_row1_col4\" class=\"data row1 col4\" >0.9738</td>\n",
       "      <td id=\"T_246ba_row1_col5\" class=\"data row1 col5\" >0.1941</td>\n",
       "      <td id=\"T_246ba_row1_col6\" class=\"data row1 col6\" >2.5189</td>\n",
       "      <td id=\"T_246ba_row1_col7\" class=\"data row1 col7\" >2.3720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row2\" class=\"row_heading level0 row2\" >gbr</th>\n",
       "      <td id=\"T_246ba_row2_col0\" class=\"data row2 col0\" >Gradient Boosting Regressor</td>\n",
       "      <td id=\"T_246ba_row2_col1\" class=\"data row2 col1\" >1.9705</td>\n",
       "      <td id=\"T_246ba_row2_col2\" class=\"data row2 col2\" >74.4780</td>\n",
       "      <td id=\"T_246ba_row2_col3\" class=\"data row2 col3\" >7.5328</td>\n",
       "      <td id=\"T_246ba_row2_col4\" class=\"data row2 col4\" >0.9641</td>\n",
       "      <td id=\"T_246ba_row2_col5\" class=\"data row2 col5\" >0.4151</td>\n",
       "      <td id=\"T_246ba_row2_col6\" class=\"data row2 col6\" >173.6468</td>\n",
       "      <td id=\"T_246ba_row2_col7\" class=\"data row2 col7\" >2.5220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row3\" class=\"row_heading level0 row3\" >et</th>\n",
       "      <td id=\"T_246ba_row3_col0\" class=\"data row3 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_246ba_row3_col1\" class=\"data row3 col1\" >2.1216</td>\n",
       "      <td id=\"T_246ba_row3_col2\" class=\"data row3 col2\" >182.9834</td>\n",
       "      <td id=\"T_246ba_row3_col3\" class=\"data row3 col3\" >12.3444</td>\n",
       "      <td id=\"T_246ba_row3_col4\" class=\"data row3 col4\" >0.9263</td>\n",
       "      <td id=\"T_246ba_row3_col5\" class=\"data row3 col5\" >0.4611</td>\n",
       "      <td id=\"T_246ba_row3_col6\" class=\"data row3 col6\" >26.7672</td>\n",
       "      <td id=\"T_246ba_row3_col7\" class=\"data row3 col7\" >2.5570</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row4\" class=\"row_heading level0 row4\" >knn</th>\n",
       "      <td id=\"T_246ba_row4_col0\" class=\"data row4 col0\" >K Neighbors Regressor</td>\n",
       "      <td id=\"T_246ba_row4_col1\" class=\"data row4 col1\" >2.4124</td>\n",
       "      <td id=\"T_246ba_row4_col2\" class=\"data row4 col2\" >104.2100</td>\n",
       "      <td id=\"T_246ba_row4_col3\" class=\"data row4 col3\" >9.5871</td>\n",
       "      <td id=\"T_246ba_row4_col4\" class=\"data row4 col4\" >0.9386</td>\n",
       "      <td id=\"T_246ba_row4_col5\" class=\"data row4 col5\" >0.4968</td>\n",
       "      <td id=\"T_246ba_row4_col6\" class=\"data row4 col6\" >61.9074</td>\n",
       "      <td id=\"T_246ba_row4_col7\" class=\"data row4 col7\" >2.9640</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row5\" class=\"row_heading level0 row5\" >ada</th>\n",
       "      <td id=\"T_246ba_row5_col0\" class=\"data row5 col0\" >AdaBoost Regressor</td>\n",
       "      <td id=\"T_246ba_row5_col1\" class=\"data row5 col1\" >6.6035</td>\n",
       "      <td id=\"T_246ba_row5_col2\" class=\"data row5 col2\" >198.9587</td>\n",
       "      <td id=\"T_246ba_row5_col3\" class=\"data row5 col3\" >13.6576</td>\n",
       "      <td id=\"T_246ba_row5_col4\" class=\"data row5 col4\" >0.8306</td>\n",
       "      <td id=\"T_246ba_row5_col5\" class=\"data row5 col5\" >1.3401</td>\n",
       "      <td id=\"T_246ba_row5_col6\" class=\"data row5 col6\" >2183.5365</td>\n",
       "      <td id=\"T_246ba_row5_col7\" class=\"data row5 col7\" >2.4630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row6\" class=\"row_heading level0 row6\" >huber</th>\n",
       "      <td id=\"T_246ba_row6_col0\" class=\"data row6 col0\" >Huber Regressor</td>\n",
       "      <td id=\"T_246ba_row6_col1\" class=\"data row6 col1\" >7.9945</td>\n",
       "      <td id=\"T_246ba_row6_col2\" class=\"data row6 col2\" >3325.3974</td>\n",
       "      <td id=\"T_246ba_row6_col3\" class=\"data row6 col3\" >51.5802</td>\n",
       "      <td id=\"T_246ba_row6_col4\" class=\"data row6 col4\" >-0.0318</td>\n",
       "      <td id=\"T_246ba_row6_col5\" class=\"data row6 col5\" >1.2900</td>\n",
       "      <td id=\"T_246ba_row6_col6\" class=\"data row6 col6\" >248.8952</td>\n",
       "      <td id=\"T_246ba_row6_col7\" class=\"data row6 col7\" >3.2020</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row7\" class=\"row_heading level0 row7\" >dummy</th>\n",
       "      <td id=\"T_246ba_row7_col0\" class=\"data row7 col0\" >Dummy Regressor</td>\n",
       "      <td id=\"T_246ba_row7_col1\" class=\"data row7 col1\" >12.8199</td>\n",
       "      <td id=\"T_246ba_row7_col2\" class=\"data row7 col2\" >3264.8115</td>\n",
       "      <td id=\"T_246ba_row7_col3\" class=\"data row7 col3\" >51.1178</td>\n",
       "      <td id=\"T_246ba_row7_col4\" class=\"data row7 col4\" >-0.0181</td>\n",
       "      <td id=\"T_246ba_row7_col5\" class=\"data row7 col5\" >1.8368</td>\n",
       "      <td id=\"T_246ba_row7_col6\" class=\"data row7 col6\" >5762.4278</td>\n",
       "      <td id=\"T_246ba_row7_col7\" class=\"data row7 col7\" >3.1660</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row8\" class=\"row_heading level0 row8\" >omp</th>\n",
       "      <td id=\"T_246ba_row8_col0\" class=\"data row8 col0\" >Orthogonal Matching Pursuit</td>\n",
       "      <td id=\"T_246ba_row8_col1\" class=\"data row8 col1\" >12.9492</td>\n",
       "      <td id=\"T_246ba_row8_col2\" class=\"data row8 col2\" >3260.7586</td>\n",
       "      <td id=\"T_246ba_row8_col3\" class=\"data row8 col3\" >51.0847</td>\n",
       "      <td id=\"T_246ba_row8_col4\" class=\"data row8 col4\" >-0.0169</td>\n",
       "      <td id=\"T_246ba_row8_col5\" class=\"data row8 col5\" >1.8443</td>\n",
       "      <td id=\"T_246ba_row8_col6\" class=\"data row8 col6\" >4692.6749</td>\n",
       "      <td id=\"T_246ba_row8_col7\" class=\"data row8 col7\" >2.8980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row9\" class=\"row_heading level0 row9\" >ridge</th>\n",
       "      <td id=\"T_246ba_row9_col0\" class=\"data row9 col0\" >Ridge Regression</td>\n",
       "      <td id=\"T_246ba_row9_col1\" class=\"data row9 col1\" >13.1214</td>\n",
       "      <td id=\"T_246ba_row9_col2\" class=\"data row9 col2\" >3241.8840</td>\n",
       "      <td id=\"T_246ba_row9_col3\" class=\"data row9 col3\" >50.9577</td>\n",
       "      <td id=\"T_246ba_row9_col4\" class=\"data row9 col4\" >-0.0150</td>\n",
       "      <td id=\"T_246ba_row9_col5\" class=\"data row9 col5\" >1.7899</td>\n",
       "      <td id=\"T_246ba_row9_col6\" class=\"data row9 col6\" >2787.8545</td>\n",
       "      <td id=\"T_246ba_row9_col7\" class=\"data row9 col7\" >2.8250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row10\" class=\"row_heading level0 row10\" >lar</th>\n",
       "      <td id=\"T_246ba_row10_col0\" class=\"data row10 col0\" >Least Angle Regression</td>\n",
       "      <td id=\"T_246ba_row10_col1\" class=\"data row10 col1\" >13.1214</td>\n",
       "      <td id=\"T_246ba_row10_col2\" class=\"data row10 col2\" >3241.8978</td>\n",
       "      <td id=\"T_246ba_row10_col3\" class=\"data row10 col3\" >50.9589</td>\n",
       "      <td id=\"T_246ba_row10_col4\" class=\"data row10 col4\" >-0.0152</td>\n",
       "      <td id=\"T_246ba_row10_col5\" class=\"data row10 col5\" >1.7894</td>\n",
       "      <td id=\"T_246ba_row10_col6\" class=\"data row10 col6\" >2789.0198</td>\n",
       "      <td id=\"T_246ba_row10_col7\" class=\"data row10 col7\" >3.5220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row11\" class=\"row_heading level0 row11\" >lr</th>\n",
       "      <td id=\"T_246ba_row11_col0\" class=\"data row11 col0\" >Linear Regression</td>\n",
       "      <td id=\"T_246ba_row11_col1\" class=\"data row11 col1\" >13.1214</td>\n",
       "      <td id=\"T_246ba_row11_col2\" class=\"data row11 col2\" >3241.8978</td>\n",
       "      <td id=\"T_246ba_row11_col3\" class=\"data row11 col3\" >50.9589</td>\n",
       "      <td id=\"T_246ba_row11_col4\" class=\"data row11 col4\" >-0.0152</td>\n",
       "      <td id=\"T_246ba_row11_col5\" class=\"data row11 col5\" >1.7894</td>\n",
       "      <td id=\"T_246ba_row11_col6\" class=\"data row11 col6\" >2789.0198</td>\n",
       "      <td id=\"T_246ba_row11_col7\" class=\"data row11 col7\" >2.5150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row12\" class=\"row_heading level0 row12\" >br</th>\n",
       "      <td id=\"T_246ba_row12_col0\" class=\"data row12 col0\" >Bayesian Ridge</td>\n",
       "      <td id=\"T_246ba_row12_col1\" class=\"data row12 col1\" >13.1725</td>\n",
       "      <td id=\"T_246ba_row12_col2\" class=\"data row12 col2\" >3245.2187</td>\n",
       "      <td id=\"T_246ba_row12_col3\" class=\"data row12 col3\" >50.9692</td>\n",
       "      <td id=\"T_246ba_row12_col4\" class=\"data row12 col4\" >-0.0137</td>\n",
       "      <td id=\"T_246ba_row12_col5\" class=\"data row12 col5\" >1.8099</td>\n",
       "      <td id=\"T_246ba_row12_col6\" class=\"data row12 col6\" >2808.6163</td>\n",
       "      <td id=\"T_246ba_row12_col7\" class=\"data row12 col7\" >2.6930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row13\" class=\"row_heading level0 row13\" >lasso</th>\n",
       "      <td id=\"T_246ba_row13_col0\" class=\"data row13 col0\" >Lasso Regression</td>\n",
       "      <td id=\"T_246ba_row13_col1\" class=\"data row13 col1\" >13.1762</td>\n",
       "      <td id=\"T_246ba_row13_col2\" class=\"data row13 col2\" >3245.2239</td>\n",
       "      <td id=\"T_246ba_row13_col3\" class=\"data row13 col3\" >50.9696</td>\n",
       "      <td id=\"T_246ba_row13_col4\" class=\"data row13 col4\" >-0.0137</td>\n",
       "      <td id=\"T_246ba_row13_col5\" class=\"data row13 col5\" >1.8103</td>\n",
       "      <td id=\"T_246ba_row13_col6\" class=\"data row13 col6\" >2819.0522</td>\n",
       "      <td id=\"T_246ba_row13_col7\" class=\"data row13 col7\" >2.6550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row14\" class=\"row_heading level0 row14\" >en</th>\n",
       "      <td id=\"T_246ba_row14_col0\" class=\"data row14 col0\" >Elastic Net</td>\n",
       "      <td id=\"T_246ba_row14_col1\" class=\"data row14 col1\" >13.1762</td>\n",
       "      <td id=\"T_246ba_row14_col2\" class=\"data row14 col2\" >3245.2239</td>\n",
       "      <td id=\"T_246ba_row14_col3\" class=\"data row14 col3\" >50.9696</td>\n",
       "      <td id=\"T_246ba_row14_col4\" class=\"data row14 col4\" >-0.0137</td>\n",
       "      <td id=\"T_246ba_row14_col5\" class=\"data row14 col5\" >1.8103</td>\n",
       "      <td id=\"T_246ba_row14_col6\" class=\"data row14 col6\" >2819.0796</td>\n",
       "      <td id=\"T_246ba_row14_col7\" class=\"data row14 col7\" >3.0190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row15\" class=\"row_heading level0 row15\" >llar</th>\n",
       "      <td id=\"T_246ba_row15_col0\" class=\"data row15 col0\" >Lasso Least Angle Regression</td>\n",
       "      <td id=\"T_246ba_row15_col1\" class=\"data row15 col1\" >13.1762</td>\n",
       "      <td id=\"T_246ba_row15_col2\" class=\"data row15 col2\" >3245.2239</td>\n",
       "      <td id=\"T_246ba_row15_col3\" class=\"data row15 col3\" >50.9696</td>\n",
       "      <td id=\"T_246ba_row15_col4\" class=\"data row15 col4\" >-0.0137</td>\n",
       "      <td id=\"T_246ba_row15_col5\" class=\"data row15 col5\" >1.8103</td>\n",
       "      <td id=\"T_246ba_row15_col6\" class=\"data row15 col6\" >2819.0524</td>\n",
       "      <td id=\"T_246ba_row15_col7\" class=\"data row15 col7\" >2.8210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_246ba_level0_row16\" class=\"row_heading level0 row16\" >par</th>\n",
       "      <td id=\"T_246ba_row16_col0\" class=\"data row16 col0\" >Passive Aggressive Regressor</td>\n",
       "      <td id=\"T_246ba_row16_col1\" class=\"data row16 col1\" >4103.1268</td>\n",
       "      <td id=\"T_246ba_row16_col2\" class=\"data row16 col2\" >1962109334.2228</td>\n",
       "      <td id=\"T_246ba_row16_col3\" class=\"data row16 col3\" >15228.8593</td>\n",
       "      <td id=\"T_246ba_row16_col4\" class=\"data row16 col4\" >-236881.0080</td>\n",
       "      <td id=\"T_246ba_row16_col5\" class=\"data row16 col5\" >5.8849</td>\n",
       "      <td id=\"T_246ba_row16_col6\" class=\"data row16 col6\" >6369415.4462</td>\n",
       "      <td id=\"T_246ba_row16_col7\" class=\"data row16 col7\" >2.6200</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x24852f4f190>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "caa5a64953574bb0832ce64f2aa71439",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/49 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Multi-model training/test - Compare Mean Absolute Error\n",
    "best_indoor = compare_models(sort='MAE',budget_time=30,fold=10,exclude='lightgbm') #30 mins budget time\n",
    "# remove Light gradient boosting - causes error"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- parallel processing with fugue-spark; set n_jobs = 1 in setup()\n",
    "- set parallel = FugueBackend(\"dask\") or = FugueBackend; does not always work"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "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>RandomForestRegressor(n_jobs=-1, random_state=123)</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-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestRegressor</label><div class=\"sk-toggleable__content\"><pre>RandomForestRegressor(n_jobs=-1, random_state=123)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "RandomForestRegressor(n_jobs=-1, random_state=123)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_indoor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.1.2 Evaluate Baseline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Changing class type to RandomForestRegressionExplainer...\n",
      "Generating self.shap_explainer = shap.TreeExplainer(model)\n",
      "Building ExplainerDashboard..\n",
      "WARNING: the number of idxs (=1143) > max_idxs_in_dropdown(=1000). However with your installed version of dash(2.9.1) dropdown search may not work smoothly. You can downgrade to `pip install dash==2.6.2` which should work better for now...\n",
      "Warning: calculating shap interaction values can be slow! Pass shap_interaction=False to remove interactions tab.\n",
      "Generating layout...\n",
      "Calculating shap values...\n",
      "Calculating predictions...\n",
      "Calculating residuals...\n",
      "Calculating absolute residuals...\n",
      "Warning: mean-absolute-percentage-error is very large (54231862601686.51), you can hide it from the metrics by passing parameter show_metrics...\n",
      "Calculating shap interaction values...\n",
      "Reminder: TreeShap computational complexity is O(TLD^2), where T is the number of trees, L is the maximum number of leaves in any tree and D the maximal depth of any tree. So reducing these will speed up the calculation.\n",
      "Warning: mean-absolute-percentage-error is very large (54231862601686.51), you can hide it from the metrics by passing parameter show_metrics...\n",
      "Warning: mean-absolute-percentage-error is very large (54231862601686.51), you can hide it from the metrics by passing parameter show_metrics...\n",
      "Calculating dependencies...\n",
      "Calculating importances...\n",
      "Calculating ShadowDecTree for each individual decision tree...\n",
      "Reminder: you can store the explainer (including calculated dependencies) with explainer.dump('explainer.joblib') and reload with e.g. ClassifierExplainer.from_file('explainer.joblib')\n",
      "Registering callbacks...\n",
      "Starting ExplainerDashboard inline (terminate it with ExplainerDashboard.terminate(8050))\n",
      "Dash is running on http://127.0.0.1:8050/\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "        <iframe\n",
       "            width=\"1000\"\n",
       "            height=\"800\"\n",
       "            src=\"http://127.0.0.1:8050/\"\n",
       "            frameborder=\"0\"\n",
       "            allowfullscreen\n",
       "            \n",
       "        ></iframe>\n",
       "        "
      ],
      "text/plain": [
       "<IPython.lib.display.IFrame at 0x2487d429910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# View evaluation dashboard - show inline\n",
    "dashboard(best_indoor,display_format='inline')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It appears that Random forest/Decision Tree regressors are the best options. <br>\n",
    "Upon checking the SHAP results in the dashboard, the population and GDP per capita have the highest feature importance for the model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x550 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot learning Curve\n",
    "plot_model(best_indoor, plot = 'learning')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As shown in the learning curve, model training score doesn't increase greatly with with an increase in training instances between. In contrast, cross-validation model scores increase moderately from around 85% to 97%  (roughly 19% to 63% of the total test-train dataset). The model does not appear to be overfit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Deaths= 7.850 | Max Deaths= 1385.600\n"
     ]
    }
   ],
   "source": [
    "print('Mean Deaths= %.3f | Max Deaths= %.3f'%(round(np.mean(df_in['Deaths_Indoor']),2),round(np.max(df_in['Deaths_Indoor']),2)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x550 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot Residuals\n",
    "plot_model(best_indoor, plot = 'residuals')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The residuals plot shows the distribution of residual (delta between actual vs predicted indoor air pollution deaths), which are larely within +/-10 deaths per 100,000, which relatively poor on the individual country level considering the {mean,max} of the metric. However, the R2 is actually around 99%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x550 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Validation curve for tree depth\n",
    "plot_model(best_indoor, plot = 'vc')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The actual estimator (max depth) reprents the maximum number of decision tree levels were used in the cross-validation exercise. The cross-validation score has a significant increase from around 55% to 95% from a tree depth of 2 to 7. Typically a max depth around 5 to 7 is recommended in online resources."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot Feature Importance - Variable Importance (different from average importance shown in dashbaoard)\n",
    "plot_model(best_indoor, plot = 'feature')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Feature Importance plot shows that the Population drives the model at a variable importance of around 0.95, with GDP per capita coming in a distance second at around 0.1. The year does not seem to contribute much, which could be a problem when projecting the model forward into time."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.1.3 Tune Baseline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_e7ac0_row10_col0, #T_e7ac0_row10_col1, #T_e7ac0_row10_col2, #T_e7ac0_row10_col3, #T_e7ac0_row10_col4, #T_e7ac0_row10_col5 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_e7ac0\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_e7ac0_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_e7ac0_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_e7ac0_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_e7ac0_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_e7ac0_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_e7ac0_level0_col5\" class=\"col_heading level0 col5\" >MAPE</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_e7ac0_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_e7ac0_row0_col0\" class=\"data row0 col0\" >1.1214</td>\n",
       "      <td id=\"T_e7ac0_row0_col1\" class=\"data row0 col1\" >26.7613</td>\n",
       "      <td id=\"T_e7ac0_row0_col2\" class=\"data row0 col2\" >5.1731</td>\n",
       "      <td id=\"T_e7ac0_row0_col3\" class=\"data row0 col3\" >0.9890</td>\n",
       "      <td id=\"T_e7ac0_row0_col4\" class=\"data row0 col4\" >0.2590</td>\n",
       "      <td id=\"T_e7ac0_row0_col5\" class=\"data row0 col5\" >2.1161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e7ac0_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_e7ac0_row1_col0\" class=\"data row1 col0\" >1.1793</td>\n",
       "      <td id=\"T_e7ac0_row1_col1\" class=\"data row1 col1\" >43.6549</td>\n",
       "      <td id=\"T_e7ac0_row1_col2\" class=\"data row1 col2\" >6.6072</td>\n",
       "      <td id=\"T_e7ac0_row1_col3\" class=\"data row1 col3\" >0.9944</td>\n",
       "      <td id=\"T_e7ac0_row1_col4\" class=\"data row1 col4\" >0.1342</td>\n",
       "      <td id=\"T_e7ac0_row1_col5\" class=\"data row1 col5\" >3.1850</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e7ac0_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_e7ac0_row2_col0\" class=\"data row2 col0\" >0.4054</td>\n",
       "      <td id=\"T_e7ac0_row2_col1\" class=\"data row2 col1\" >4.8780</td>\n",
       "      <td id=\"T_e7ac0_row2_col2\" class=\"data row2 col2\" >2.2086</td>\n",
       "      <td id=\"T_e7ac0_row2_col3\" class=\"data row2 col3\" >0.9780</td>\n",
       "      <td id=\"T_e7ac0_row2_col4\" class=\"data row2 col4\" >0.1180</td>\n",
       "      <td id=\"T_e7ac0_row2_col5\" class=\"data row2 col5\" >5.5689</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e7ac0_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_e7ac0_row3_col0\" class=\"data row3 col0\" >0.6219</td>\n",
       "      <td id=\"T_e7ac0_row3_col1\" class=\"data row3 col1\" >10.1628</td>\n",
       "      <td id=\"T_e7ac0_row3_col2\" class=\"data row3 col2\" >3.1879</td>\n",
       "      <td id=\"T_e7ac0_row3_col3\" class=\"data row3 col3\" >0.9592</td>\n",
       "      <td id=\"T_e7ac0_row3_col4\" class=\"data row3 col4\" >0.2333</td>\n",
       "      <td id=\"T_e7ac0_row3_col5\" class=\"data row3 col5\" >0.8061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e7ac0_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_e7ac0_row4_col0\" class=\"data row4 col0\" >1.2171</td>\n",
       "      <td id=\"T_e7ac0_row4_col1\" class=\"data row4 col1\" >34.7624</td>\n",
       "      <td id=\"T_e7ac0_row4_col2\" class=\"data row4 col2\" >5.8960</td>\n",
       "      <td id=\"T_e7ac0_row4_col3\" class=\"data row4 col3\" >0.9857</td>\n",
       "      <td id=\"T_e7ac0_row4_col4\" class=\"data row4 col4\" >0.1761</td>\n",
       "      <td id=\"T_e7ac0_row4_col5\" class=\"data row4 col5\" >1.7779</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e7ac0_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_e7ac0_row5_col0\" class=\"data row5 col0\" >1.5510</td>\n",
       "      <td id=\"T_e7ac0_row5_col1\" class=\"data row5 col1\" >217.3038</td>\n",
       "      <td id=\"T_e7ac0_row5_col2\" class=\"data row5 col2\" >14.7412</td>\n",
       "      <td id=\"T_e7ac0_row5_col3\" class=\"data row5 col3\" >0.9497</td>\n",
       "      <td id=\"T_e7ac0_row5_col4\" class=\"data row5 col4\" >0.1773</td>\n",
       "      <td id=\"T_e7ac0_row5_col5\" class=\"data row5 col5\" >1.8205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e7ac0_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_e7ac0_row6_col0\" class=\"data row6 col0\" >1.2693</td>\n",
       "      <td id=\"T_e7ac0_row6_col1\" class=\"data row6 col1\" >35.3477</td>\n",
       "      <td id=\"T_e7ac0_row6_col2\" class=\"data row6 col2\" >5.9454</td>\n",
       "      <td id=\"T_e7ac0_row6_col3\" class=\"data row6 col3\" >0.9849</td>\n",
       "      <td id=\"T_e7ac0_row6_col4\" class=\"data row6 col4\" >0.1364</td>\n",
       "      <td id=\"T_e7ac0_row6_col5\" class=\"data row6 col5\" >1.5787</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e7ac0_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_e7ac0_row7_col0\" class=\"data row7 col0\" >0.6133</td>\n",
       "      <td id=\"T_e7ac0_row7_col1\" class=\"data row7 col1\" >7.7193</td>\n",
       "      <td id=\"T_e7ac0_row7_col2\" class=\"data row7 col2\" >2.7784</td>\n",
       "      <td id=\"T_e7ac0_row7_col3\" class=\"data row7 col3\" >0.9872</td>\n",
       "      <td id=\"T_e7ac0_row7_col4\" class=\"data row7 col4\" >0.1402</td>\n",
       "      <td id=\"T_e7ac0_row7_col5\" class=\"data row7 col5\" >1.9843</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e7ac0_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_e7ac0_row8_col0\" class=\"data row8 col0\" >0.7966</td>\n",
       "      <td id=\"T_e7ac0_row8_col1\" class=\"data row8 col1\" >15.9025</td>\n",
       "      <td id=\"T_e7ac0_row8_col2\" class=\"data row8 col2\" >3.9878</td>\n",
       "      <td id=\"T_e7ac0_row8_col3\" class=\"data row8 col3\" >0.9959</td>\n",
       "      <td id=\"T_e7ac0_row8_col4\" class=\"data row8 col4\" >0.1714</td>\n",
       "      <td id=\"T_e7ac0_row8_col5\" class=\"data row8 col5\" >3.6194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e7ac0_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_e7ac0_row9_col0\" class=\"data row9 col0\" >1.2869</td>\n",
       "      <td id=\"T_e7ac0_row9_col1\" class=\"data row9 col1\" >74.3128</td>\n",
       "      <td id=\"T_e7ac0_row9_col2\" class=\"data row9 col2\" >8.6205</td>\n",
       "      <td id=\"T_e7ac0_row9_col3\" class=\"data row9 col3\" >0.9911</td>\n",
       "      <td id=\"T_e7ac0_row9_col4\" class=\"data row9 col4\" >0.1927</td>\n",
       "      <td id=\"T_e7ac0_row9_col5\" class=\"data row9 col5\" >14.6291</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e7ac0_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_e7ac0_row10_col0\" class=\"data row10 col0\" >1.0062</td>\n",
       "      <td id=\"T_e7ac0_row10_col1\" class=\"data row10 col1\" >47.0806</td>\n",
       "      <td id=\"T_e7ac0_row10_col2\" class=\"data row10 col2\" >5.9146</td>\n",
       "      <td id=\"T_e7ac0_row10_col3\" class=\"data row10 col3\" >0.9815</td>\n",
       "      <td id=\"T_e7ac0_row10_col4\" class=\"data row10 col4\" >0.1739</td>\n",
       "      <td id=\"T_e7ac0_row10_col5\" class=\"data row10 col5\" >3.7086</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e7ac0_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_e7ac0_row11_col0\" class=\"data row11 col0\" >0.3520</td>\n",
       "      <td id=\"T_e7ac0_row11_col1\" class=\"data row11 col1\" >60.0807</td>\n",
       "      <td id=\"T_e7ac0_row11_col2\" class=\"data row11 col2\" >3.4782</td>\n",
       "      <td id=\"T_e7ac0_row11_col3\" class=\"data row11 col3\" >0.0145</td>\n",
       "      <td id=\"T_e7ac0_row11_col4\" class=\"data row11 col4\" >0.0429</td>\n",
       "      <td id=\"T_e7ac0_row11_col5\" class=\"data row11 col5\" >3.8570</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2487e5c09a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "82f4a8ed2c434719ada549c6124f1374",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/4 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create the Random Forest Regressor Model\n",
    "rf_indoor = create_model('rf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'bootstrap': True,\n",
       " 'ccp_alpha': 0.0,\n",
       " 'criterion': 'squared_error',\n",
       " 'max_depth': None,\n",
       " 'max_features': 1.0,\n",
       " 'max_leaf_nodes': None,\n",
       " 'max_samples': None,\n",
       " 'min_impurity_decrease': 0.0,\n",
       " 'min_samples_leaf': 1,\n",
       " 'min_samples_split': 2,\n",
       " 'min_weight_fraction_leaf': 0.0,\n",
       " 'n_estimators': 100,\n",
       " 'n_jobs': -1,\n",
       " 'oob_score': False,\n",
       " 'random_state': 123,\n",
       " 'verbose': 0,\n",
       " 'warm_start': False}"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check the default random forest parameters\n",
    "rf_indoor.get_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Comments on default parameters:\n",
    "- Note that max_depth is set to None, which means that the tree could be overfitting on the training data. <br>\n",
    "Commonly searched hyperparameters are:\n",
    "- n_estimators = number of trees in the forest\n",
    "- max_depth = tree depth\n",
    "- max_features = number of features considered when looking for best at every split; if None or 1.0 then max_features = n_features = bagged tree. Lower values = more randomnes. Can increase to sqrt(n total features)\n",
    "- min_samples_split = min sampled data points required to split a node (decision node); lower numbers tend to overfit\n",
    "- min_samples_leaf = min sampled data points required at each leaf (terminal) node; lower numbers tend to overfit\n",
    "- We will leave bootstrapping enabled, to permite for a more accurate estimation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define custom hyperparamter grid only on select parameters\n",
    "rf_grid = {'max_depth':[3,5,7], \n",
    "           'n_estimators':[25,50,100],\n",
    "           'max_features':[0.5,1,'sqrt']}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_d8533_row10_col0, #T_d8533_row10_col1, #T_d8533_row10_col2, #T_d8533_row10_col3, #T_d8533_row10_col4, #T_d8533_row10_col5 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_d8533\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_d8533_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_d8533_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_d8533_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_d8533_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_d8533_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_d8533_level0_col5\" class=\"col_heading level0 col5\" >MAPE</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_d8533_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_d8533_row0_col0\" class=\"data row0 col0\" >2.4282</td>\n",
       "      <td id=\"T_d8533_row0_col1\" class=\"data row0 col1\" >54.3249</td>\n",
       "      <td id=\"T_d8533_row0_col2\" class=\"data row0 col2\" >7.3705</td>\n",
       "      <td id=\"T_d8533_row0_col3\" class=\"data row0 col3\" >0.9776</td>\n",
       "      <td id=\"T_d8533_row0_col4\" class=\"data row0 col4\" >0.6369</td>\n",
       "      <td id=\"T_d8533_row0_col5\" class=\"data row0 col5\" >63.9489</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_d8533_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_d8533_row1_col0\" class=\"data row1 col0\" >3.7119</td>\n",
       "      <td id=\"T_d8533_row1_col1\" class=\"data row1 col1\" >332.0570</td>\n",
       "      <td id=\"T_d8533_row1_col2\" class=\"data row1 col2\" >18.2224</td>\n",
       "      <td id=\"T_d8533_row1_col3\" class=\"data row1 col3\" >0.9571</td>\n",
       "      <td id=\"T_d8533_row1_col4\" class=\"data row1 col4\" >0.6193</td>\n",
       "      <td id=\"T_d8533_row1_col5\" class=\"data row1 col5\" >79.3246</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_d8533_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_d8533_row2_col0\" class=\"data row2 col0\" >1.8073</td>\n",
       "      <td id=\"T_d8533_row2_col1\" class=\"data row2 col1\" >25.6951</td>\n",
       "      <td id=\"T_d8533_row2_col2\" class=\"data row2 col2\" >5.0690</td>\n",
       "      <td id=\"T_d8533_row2_col3\" class=\"data row2 col3\" >0.8844</td>\n",
       "      <td id=\"T_d8533_row2_col4\" class=\"data row2 col4\" >0.6282</td>\n",
       "      <td id=\"T_d8533_row2_col5\" class=\"data row2 col5\" >75.2500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_d8533_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_d8533_row3_col0\" class=\"data row3 col0\" >2.1386</td>\n",
       "      <td id=\"T_d8533_row3_col1\" class=\"data row3 col1\" >36.1435</td>\n",
       "      <td id=\"T_d8533_row3_col2\" class=\"data row3 col2\" >6.0119</td>\n",
       "      <td id=\"T_d8533_row3_col3\" class=\"data row3 col3\" >0.8547</td>\n",
       "      <td id=\"T_d8533_row3_col4\" class=\"data row3 col4\" >0.6054</td>\n",
       "      <td id=\"T_d8533_row3_col5\" class=\"data row3 col5\" >44.9476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_d8533_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_d8533_row4_col0\" class=\"data row4 col0\" >3.0918</td>\n",
       "      <td id=\"T_d8533_row4_col1\" class=\"data row4 col1\" >147.6153</td>\n",
       "      <td id=\"T_d8533_row4_col2\" class=\"data row4 col2\" >12.1497</td>\n",
       "      <td id=\"T_d8533_row4_col3\" class=\"data row4 col3\" >0.9395</td>\n",
       "      <td id=\"T_d8533_row4_col4\" class=\"data row4 col4\" >0.5440</td>\n",
       "      <td id=\"T_d8533_row4_col5\" class=\"data row4 col5\" >55.5251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_d8533_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_d8533_row5_col0\" class=\"data row5 col0\" >2.9600</td>\n",
       "      <td id=\"T_d8533_row5_col1\" class=\"data row5 col1\" >141.3657</td>\n",
       "      <td id=\"T_d8533_row5_col2\" class=\"data row5 col2\" >11.8897</td>\n",
       "      <td id=\"T_d8533_row5_col3\" class=\"data row5 col3\" >0.9673</td>\n",
       "      <td id=\"T_d8533_row5_col4\" class=\"data row5 col4\" >0.6700</td>\n",
       "      <td id=\"T_d8533_row5_col5\" class=\"data row5 col5\" >65.7365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_d8533_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_d8533_row6_col0\" class=\"data row6 col0\" >2.7337</td>\n",
       "      <td id=\"T_d8533_row6_col1\" class=\"data row6 col1\" >64.9904</td>\n",
       "      <td id=\"T_d8533_row6_col2\" class=\"data row6 col2\" >8.0617</td>\n",
       "      <td id=\"T_d8533_row6_col3\" class=\"data row6 col3\" >0.9722</td>\n",
       "      <td id=\"T_d8533_row6_col4\" class=\"data row6 col4\" >0.5586</td>\n",
       "      <td id=\"T_d8533_row6_col5\" class=\"data row6 col5\" >61.3956</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_d8533_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_d8533_row7_col0\" class=\"data row7 col0\" >2.5667</td>\n",
       "      <td id=\"T_d8533_row7_col1\" class=\"data row7 col1\" >63.0243</td>\n",
       "      <td id=\"T_d8533_row7_col2\" class=\"data row7 col2\" >7.9388</td>\n",
       "      <td id=\"T_d8533_row7_col3\" class=\"data row7 col3\" >0.8955</td>\n",
       "      <td id=\"T_d8533_row7_col4\" class=\"data row7 col4\" >0.6542</td>\n",
       "      <td id=\"T_d8533_row7_col5\" class=\"data row7 col5\" >86.3050</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_d8533_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_d8533_row8_col0\" class=\"data row8 col0\" >2.4097</td>\n",
       "      <td id=\"T_d8533_row8_col1\" class=\"data row8 col1\" >58.7306</td>\n",
       "      <td id=\"T_d8533_row8_col2\" class=\"data row8 col2\" >7.6636</td>\n",
       "      <td id=\"T_d8533_row8_col3\" class=\"data row8 col3\" >0.9849</td>\n",
       "      <td id=\"T_d8533_row8_col4\" class=\"data row8 col4\" >0.6169</td>\n",
       "      <td id=\"T_d8533_row8_col5\" class=\"data row8 col5\" >103.1619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_d8533_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_d8533_row9_col0\" class=\"data row9 col0\" >2.7507</td>\n",
       "      <td id=\"T_d8533_row9_col1\" class=\"data row9 col1\" >184.5570</td>\n",
       "      <td id=\"T_d8533_row9_col2\" class=\"data row9 col2\" >13.5852</td>\n",
       "      <td id=\"T_d8533_row9_col3\" class=\"data row9 col3\" >0.9779</td>\n",
       "      <td id=\"T_d8533_row9_col4\" class=\"data row9 col4\" >0.6049</td>\n",
       "      <td id=\"T_d8533_row9_col5\" class=\"data row9 col5\" >83.4195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_d8533_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_d8533_row10_col0\" class=\"data row10 col0\" >2.6599</td>\n",
       "      <td id=\"T_d8533_row10_col1\" class=\"data row10 col1\" >110.8504</td>\n",
       "      <td id=\"T_d8533_row10_col2\" class=\"data row10 col2\" >9.7963</td>\n",
       "      <td id=\"T_d8533_row10_col3\" class=\"data row10 col3\" >0.9411</td>\n",
       "      <td id=\"T_d8533_row10_col4\" class=\"data row10 col4\" >0.6138</td>\n",
       "      <td id=\"T_d8533_row10_col5\" class=\"data row10 col5\" >71.9015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_d8533_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_d8533_row11_col0\" class=\"data row11 col0\" >0.5020</td>\n",
       "      <td id=\"T_d8533_row11_col1\" class=\"data row11 col1\" >89.3210</td>\n",
       "      <td id=\"T_d8533_row11_col2\" class=\"data row11 col2\" >3.8579</td>\n",
       "      <td id=\"T_d8533_row11_col3\" class=\"data row11 col3\" >0.0439</td>\n",
       "      <td id=\"T_d8533_row11_col4\" class=\"data row11 col4\" >0.0369</td>\n",
       "      <td id=\"T_d8533_row11_col5\" class=\"data row11 col5\" >16.0934</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2487e6da160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2cfef6a972ff4fa8a96c5011c52b11ac",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/7 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 27 candidates, totalling 270 fits\n"
     ]
    }
   ],
   "source": [
    "# Optimize the model - optimize for MAE, pass custom hyperparameter grid; \n",
    "# force resulting model not to be overriden by baseline if performance is worse\n",
    "tuned_rf_indoor = tune_model(rf_indoor,optimize='MAE',custom_grid=rf_grid,search_algorithm='grid',choose_better=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'bootstrap': True,\n",
       " 'ccp_alpha': 0.0,\n",
       " 'criterion': 'squared_error',\n",
       " 'max_depth': 7,\n",
       " 'max_features': 0.5,\n",
       " 'max_leaf_nodes': None,\n",
       " 'max_samples': None,\n",
       " 'min_impurity_decrease': 0.0,\n",
       " 'min_samples_leaf': 1,\n",
       " 'min_samples_split': 2,\n",
       " 'min_weight_fraction_leaf': 0.0,\n",
       " 'n_estimators': 100,\n",
       " 'n_jobs': -1,\n",
       " 'oob_score': False,\n",
       " 'random_state': 123,\n",
       " 'verbose': 0,\n",
       " 'warm_start': False}"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tuned_rf_indoor.get_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The base random forest model with default parameters appears to still have less MAE than any of the models in the hyperparamter search space. Regardless of this, we will at least constrain the max_depth=7 to avoid overfitting."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.1.4 Predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_5e336\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_5e336_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_5e336_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_5e336_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_5e336_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_5e336_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_5e336_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_5e336_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_5e336_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_5e336_row0_col0\" class=\"data row0 col0\" >Random Forest Regressor</td>\n",
       "      <td id=\"T_5e336_row0_col1\" class=\"data row0 col1\" >2.7853</td>\n",
       "      <td id=\"T_5e336_row0_col2\" class=\"data row0 col2\" >71.9083</td>\n",
       "      <td id=\"T_5e336_row0_col3\" class=\"data row0 col3\" >8.4799</td>\n",
       "      <td id=\"T_5e336_row0_col4\" class=\"data row0 col4\" >0.9548</td>\n",
       "      <td id=\"T_5e336_row0_col5\" class=\"data row0 col5\" >0.6542</td>\n",
       "      <td id=\"T_5e336_row0_col6\" class=\"data row0 col6\" >80.7505</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2487d4b9d90>"
      ]
     },
     "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>GDP_pCapita</th>\n",
       "      <th>Population</th>\n",
       "      <th>SoDI</th>\n",
       "      <th>Income_Class_encoded</th>\n",
       "      <th>Year</th>\n",
       "      <th>Deaths_Indoor</th>\n",
       "      <th>prediction_label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1025</th>\n",
       "      <td>5386.467773</td>\n",
       "      <td>14587367.0</td>\n",
       "      <td>0.630</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001996</td>\n",
       "      <td>0.064852</td>\n",
       "      <td>0.425407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>935</th>\n",
       "      <td>21227.347656</td>\n",
       "      <td>29610218.0</td>\n",
       "      <td>0.809</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001996</td>\n",
       "      <td>0.001061</td>\n",
       "      <td>0.036093</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3024</th>\n",
       "      <td>117254.742188</td>\n",
       "      <td>607950.0</td>\n",
       "      <td>0.894</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002018</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.151802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3584</th>\n",
       "      <td>10171.480469</td>\n",
       "      <td>10289.0</td>\n",
       "      <td>0.573</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002014</td>\n",
       "      <td>94.461243</td>\n",
       "      <td>82.835495</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2131</th>\n",
       "      <td>477.762451</td>\n",
       "      <td>7081119.0</td>\n",
       "      <td>0.190</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001994</td>\n",
       "      <td>3.347876</td>\n",
       "      <td>2.131716</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>524</th>\n",
       "      <td>2993.033691</td>\n",
       "      <td>230248.0</td>\n",
       "      <td>0.494</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001998</td>\n",
       "      <td>8.676930</td>\n",
       "      <td>25.638869</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5227</th>\n",
       "      <td>590.393494</td>\n",
       "      <td>4413477.0</td>\n",
       "      <td>0.556</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001998</td>\n",
       "      <td>0.017455</td>\n",
       "      <td>1.261738</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2883</th>\n",
       "      <td>7948.685547</td>\n",
       "      <td>5537620.0</td>\n",
       "      <td>0.660</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002012</td>\n",
       "      <td>0.007826</td>\n",
       "      <td>0.309097</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3941</th>\n",
       "      <td>1664.245728</td>\n",
       "      <td>3494496.0</td>\n",
       "      <td>0.467</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002007</td>\n",
       "      <td>0.076977</td>\n",
       "      <td>4.326149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2083</th>\n",
       "      <td>26713.058594</td>\n",
       "      <td>158649.0</td>\n",
       "      <td>0.773</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002006</td>\n",
       "      <td>1.191918</td>\n",
       "      <td>0.970600</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1143 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        GDP_pCapita  Population   SoDI  Income_Class_encoded  \\\n",
       "1025    5386.467773  14587367.0  0.630                   2.0   \n",
       "935    21227.347656  29610218.0  0.809                   3.0   \n",
       "3024  117254.742188    607950.0  0.894                   3.0   \n",
       "3584   10171.480469     10289.0  0.573                   3.0   \n",
       "2131     477.762451   7081119.0  0.190                   0.0   \n",
       "...             ...         ...    ...                   ...   \n",
       "524     2993.033691    230248.0  0.494                   1.0   \n",
       "5227     590.393494   4413477.0  0.556                   0.0   \n",
       "2883    7948.685547   5537620.0  0.660                   2.0   \n",
       "3941    1664.245728   3494496.0  0.467                   1.0   \n",
       "2083   26713.058594    158649.0  0.773                   3.0   \n",
       "\n",
       "                              Year  Deaths_Indoor  prediction_label  \n",
       "1025 1970-01-01 00:00:00.000001996       0.064852          0.425407  \n",
       "935  1970-01-01 00:00:00.000001996       0.001061          0.036093  \n",
       "3024 1970-01-01 00:00:00.000002018       0.000000          0.151802  \n",
       "3584 1970-01-01 00:00:00.000002014      94.461243         82.835495  \n",
       "2131 1970-01-01 00:00:00.000001994       3.347876          2.131716  \n",
       "...                            ...            ...               ...  \n",
       "524  1970-01-01 00:00:00.000001998       8.676930         25.638869  \n",
       "5227 1970-01-01 00:00:00.000001998       0.017455          1.261738  \n",
       "2883 1970-01-01 00:00:00.000002012       0.007826          0.309097  \n",
       "3941 1970-01-01 00:00:00.000002007       0.076977          4.326149  \n",
       "2083 1970-01-01 00:00:00.000002006       1.191918          0.970600  \n",
       "\n",
       "[1143 rows x 7 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Use random forest model to predict holdout testing dataset indoor pollution deaths per 100,000\n",
    "predictions = predict_model(tuned_rf_indoor)\n",
    "predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Even with the hold-out testing dataset, the tuned model has RMSE,MAE,R2 not drastically different from the test-train dataset used to build the model, so it shows that the model was not overfitted to the input data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Transformation Pipeline and Model Successfully Saved\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(Pipeline(memory=FastMemory(location=C:\\Users\\kevin\\AppData\\Local\\Temp\\joblib),\n",
       "          steps=[('date_feature_extractor',\n",
       "                  TransformerWrapper(include=['Year'],\n",
       "                                     transformer=ExtractDateTimeFeatures())),\n",
       "                 ('numerical_imputer',\n",
       "                  TransformerWrapper(include=['GDP_pCapita', 'Population',\n",
       "                                              'SoDI', 'Income_Class_encoded'],\n",
       "                                     transformer=SimpleImputer())),\n",
       "                 ('categorical_imputer',\n",
       "                  TransformerWrapper(include=[],\n",
       "                                     transformer=SimpleImputer(strategy='most_frequent'))),\n",
       "                 ('trained_model',\n",
       "                  RandomForestRegressor(n_jobs=-1, random_state=123))]),\n",
       " 'Indoor Deaths Random Forest 2023-03-29.pkl')"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Save Model\n",
    "save_model(best_indoor,'Indoor Deaths Random Forest 2023-03-29')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**IMPORTANT NOTE** Since this is just a test for which model may offer the best solution, the fact that the data is time-series data was omitted. The first model actually just predicts deaths based mainly on socio-demographic and economic factors. The meaning of the 'Year' is not actually being interpreted as a data-time variable. <br>\n",
    "A potential method could be to set the fold_strategy = 'timeseries', which will use a rolling-window to develop the cross-validation folds. However, the issue with this is that all the timeseries of every nation is concatenated vertically. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.2 Model Selection for Outdoor Pollution - PM2.5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we will test models for predictions on outdoor PM2.5 pollution. First a baseline exercise will be conducted, where the max_depth will be capped at 7, as with the indoor deaths. Hyperparameter tuning will be omitted for this exercise. Next, a  blended model will be created with the top 3 models from the compare_models() process."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Dependent variable - classify the level of indoor pollution death\n",
    "y=df_in.loc[:,['Deaths_PM2.5']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = pd.concat([X,y],axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.2.1 Determine Top 3 Models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_6e392_row9_col1, #T_6e392_row16_col1 {\n",
       "  background-color: lightgreen;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_6e392\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_6e392_level0_col0\" class=\"col_heading level0 col0\" >Description</th>\n",
       "      <th id=\"T_6e392_level0_col1\" class=\"col_heading level0 col1\" >Value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_6e392_row0_col0\" class=\"data row0 col0\" >Session id</td>\n",
       "      <td id=\"T_6e392_row0_col1\" class=\"data row0 col1\" >123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_6e392_row1_col0\" class=\"data row1 col0\" >Target</td>\n",
       "      <td id=\"T_6e392_row1_col1\" class=\"data row1 col1\" >Deaths_PM2.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_6e392_row2_col0\" class=\"data row2 col0\" >Target type</td>\n",
       "      <td id=\"T_6e392_row2_col1\" class=\"data row2 col1\" >Regression</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_6e392_row3_col0\" class=\"data row3 col0\" >Original data shape</td>\n",
       "      <td id=\"T_6e392_row3_col1\" class=\"data row3 col1\" >(5712, 6)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_6e392_row4_col0\" class=\"data row4 col0\" >Transformed data shape</td>\n",
       "      <td id=\"T_6e392_row4_col1\" class=\"data row4 col1\" >(5712, 8)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_6e392_row5_col0\" class=\"data row5 col0\" >Transformed train set shape</td>\n",
       "      <td id=\"T_6e392_row5_col1\" class=\"data row5 col1\" >(4569, 8)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_6e392_row6_col0\" class=\"data row6 col0\" >Transformed test set shape</td>\n",
       "      <td id=\"T_6e392_row6_col1\" class=\"data row6 col1\" >(1143, 8)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_6e392_row7_col0\" class=\"data row7 col0\" >Numeric features</td>\n",
       "      <td id=\"T_6e392_row7_col1\" class=\"data row7 col1\" >4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_6e392_row8_col0\" class=\"data row8 col0\" >Date features</td>\n",
       "      <td id=\"T_6e392_row8_col1\" class=\"data row8 col1\" >1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_6e392_row9_col0\" class=\"data row9 col0\" >Preprocess</td>\n",
       "      <td id=\"T_6e392_row9_col1\" class=\"data row9 col1\" >True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row10\" class=\"row_heading level0 row10\" >10</th>\n",
       "      <td id=\"T_6e392_row10_col0\" class=\"data row10 col0\" >Imputation type</td>\n",
       "      <td id=\"T_6e392_row10_col1\" class=\"data row10 col1\" >simple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row11\" class=\"row_heading level0 row11\" >11</th>\n",
       "      <td id=\"T_6e392_row11_col0\" class=\"data row11 col0\" >Numeric imputation</td>\n",
       "      <td id=\"T_6e392_row11_col1\" class=\"data row11 col1\" >mean</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row12\" class=\"row_heading level0 row12\" >12</th>\n",
       "      <td id=\"T_6e392_row12_col0\" class=\"data row12 col0\" >Categorical imputation</td>\n",
       "      <td id=\"T_6e392_row12_col1\" class=\"data row12 col1\" >mode</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row13\" class=\"row_heading level0 row13\" >13</th>\n",
       "      <td id=\"T_6e392_row13_col0\" class=\"data row13 col0\" >Fold Generator</td>\n",
       "      <td id=\"T_6e392_row13_col1\" class=\"data row13 col1\" >KFold</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row14\" class=\"row_heading level0 row14\" >14</th>\n",
       "      <td id=\"T_6e392_row14_col0\" class=\"data row14 col0\" >Fold Number</td>\n",
       "      <td id=\"T_6e392_row14_col1\" class=\"data row14 col1\" >10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row15\" class=\"row_heading level0 row15\" >15</th>\n",
       "      <td id=\"T_6e392_row15_col0\" class=\"data row15 col0\" >CPU Jobs</td>\n",
       "      <td id=\"T_6e392_row15_col1\" class=\"data row15 col1\" >-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row16\" class=\"row_heading level0 row16\" >16</th>\n",
       "      <td id=\"T_6e392_row16_col0\" class=\"data row16 col0\" >Use GPU</td>\n",
       "      <td id=\"T_6e392_row16_col1\" class=\"data row16 col1\" >True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row17\" class=\"row_heading level0 row17\" >17</th>\n",
       "      <td id=\"T_6e392_row17_col0\" class=\"data row17 col0\" >Log Experiment</td>\n",
       "      <td id=\"T_6e392_row17_col1\" class=\"data row17 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row18\" class=\"row_heading level0 row18\" >18</th>\n",
       "      <td id=\"T_6e392_row18_col0\" class=\"data row18 col0\" >Experiment Name</td>\n",
       "      <td id=\"T_6e392_row18_col1\" class=\"data row18 col1\" >reg-default-name</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e392_level0_row19\" class=\"row_heading level0 row19\" >19</th>\n",
       "      <td id=\"T_6e392_row19_col0\" class=\"data row19 col0\" >USI</td>\n",
       "      <td id=\"T_6e392_row19_col1\" class=\"data row19 col1\" >7bf4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2487e9eae20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Initialize the PyCaret setup\n",
    "s = setup(data=dataset,target='Deaths_PM2.5',\n",
    "          numeric_features=num_features,date_features=date_features,train_size=0.8,\n",
    "          use_gpu=True,fold=10,n_jobs=-1,session_id=123) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_286bc th {\n",
       "  text-align: left;\n",
       "}\n",
       "#T_286bc_row0_col0, #T_286bc_row0_col2, #T_286bc_row0_col3, #T_286bc_row0_col4, #T_286bc_row0_col5, #T_286bc_row1_col0, #T_286bc_row1_col1, #T_286bc_row1_col4, #T_286bc_row1_col6, #T_286bc_row2_col0, #T_286bc_row2_col1, #T_286bc_row2_col2, #T_286bc_row2_col3, #T_286bc_row2_col5, #T_286bc_row2_col6, #T_286bc_row3_col0, #T_286bc_row3_col1, #T_286bc_row3_col2, #T_286bc_row3_col3, #T_286bc_row3_col4, #T_286bc_row3_col5, #T_286bc_row3_col6, #T_286bc_row4_col0, #T_286bc_row4_col1, #T_286bc_row4_col2, #T_286bc_row4_col3, #T_286bc_row4_col4, #T_286bc_row4_col5, #T_286bc_row4_col6, #T_286bc_row5_col0, #T_286bc_row5_col1, #T_286bc_row5_col2, #T_286bc_row5_col3, #T_286bc_row5_col4, #T_286bc_row5_col5, #T_286bc_row5_col6, #T_286bc_row6_col0, #T_286bc_row6_col1, #T_286bc_row6_col2, #T_286bc_row6_col3, #T_286bc_row6_col4, #T_286bc_row6_col5, #T_286bc_row6_col6, #T_286bc_row7_col0, #T_286bc_row7_col1, #T_286bc_row7_col2, #T_286bc_row7_col3, #T_286bc_row7_col4, #T_286bc_row7_col5, #T_286bc_row7_col6, #T_286bc_row8_col0, #T_286bc_row8_col1, #T_286bc_row8_col2, #T_286bc_row8_col3, #T_286bc_row8_col4, #T_286bc_row8_col5, #T_286bc_row8_col6, #T_286bc_row9_col0, #T_286bc_row9_col1, #T_286bc_row9_col2, #T_286bc_row9_col3, #T_286bc_row9_col4, #T_286bc_row9_col5, #T_286bc_row9_col6, #T_286bc_row10_col0, #T_286bc_row10_col1, #T_286bc_row10_col2, #T_286bc_row10_col3, #T_286bc_row10_col4, #T_286bc_row10_col5, #T_286bc_row10_col6, #T_286bc_row11_col0, #T_286bc_row11_col1, #T_286bc_row11_col2, #T_286bc_row11_col3, #T_286bc_row11_col4, #T_286bc_row11_col5, #T_286bc_row11_col6, #T_286bc_row12_col0, #T_286bc_row12_col1, #T_286bc_row12_col2, #T_286bc_row12_col3, #T_286bc_row12_col4, #T_286bc_row12_col5, #T_286bc_row12_col6, #T_286bc_row13_col0, #T_286bc_row13_col1, #T_286bc_row13_col2, #T_286bc_row13_col3, #T_286bc_row13_col4, #T_286bc_row13_col5, #T_286bc_row13_col6, #T_286bc_row14_col0, #T_286bc_row14_col1, #T_286bc_row14_col2, #T_286bc_row14_col3, #T_286bc_row14_col4, #T_286bc_row14_col5, #T_286bc_row14_col6, #T_286bc_row15_col0, #T_286bc_row15_col1, #T_286bc_row15_col2, #T_286bc_row15_col3, #T_286bc_row15_col4, #T_286bc_row15_col5, #T_286bc_row15_col6, #T_286bc_row16_col0, #T_286bc_row16_col1, #T_286bc_row16_col2, #T_286bc_row16_col3, #T_286bc_row16_col4, #T_286bc_row16_col5, #T_286bc_row16_col6 {\n",
       "  text-align: left;\n",
       "}\n",
       "#T_286bc_row0_col1, #T_286bc_row0_col6, #T_286bc_row1_col2, #T_286bc_row1_col3, #T_286bc_row1_col5, #T_286bc_row2_col4 {\n",
       "  text-align: left;\n",
       "  background-color: yellow;\n",
       "}\n",
       "#T_286bc_row0_col7, #T_286bc_row1_col7, #T_286bc_row2_col7, #T_286bc_row3_col7, #T_286bc_row4_col7, #T_286bc_row5_col7, #T_286bc_row6_col7, #T_286bc_row7_col7, #T_286bc_row8_col7, #T_286bc_row9_col7, #T_286bc_row10_col7, #T_286bc_row12_col7, #T_286bc_row13_col7, #T_286bc_row14_col7, #T_286bc_row15_col7, #T_286bc_row16_col7 {\n",
       "  text-align: left;\n",
       "  background-color: lightgrey;\n",
       "}\n",
       "#T_286bc_row11_col7 {\n",
       "  text-align: left;\n",
       "  background-color: yellow;\n",
       "  background-color: lightgrey;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_286bc\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_286bc_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_286bc_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_286bc_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_286bc_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_286bc_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_286bc_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_286bc_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_286bc_level0_col7\" class=\"col_heading level0 col7\" >TT (Sec)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row0\" class=\"row_heading level0 row0\" >dt</th>\n",
       "      <td id=\"T_286bc_row0_col0\" class=\"data row0 col0\" >Decision Tree Regressor</td>\n",
       "      <td id=\"T_286bc_row0_col1\" class=\"data row0 col1\" >0.5401</td>\n",
       "      <td id=\"T_286bc_row0_col2\" class=\"data row0 col2\" >10.0574</td>\n",
       "      <td id=\"T_286bc_row0_col3\" class=\"data row0 col3\" >2.9965</td>\n",
       "      <td id=\"T_286bc_row0_col4\" class=\"data row0 col4\" >0.9626</td>\n",
       "      <td id=\"T_286bc_row0_col5\" class=\"data row0 col5\" >0.1710</td>\n",
       "      <td id=\"T_286bc_row0_col6\" class=\"data row0 col6\" >0.2056</td>\n",
       "      <td id=\"T_286bc_row0_col7\" class=\"data row0 col7\" >2.5480</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row1\" class=\"row_heading level0 row1\" >rf</th>\n",
       "      <td id=\"T_286bc_row1_col0\" class=\"data row1 col0\" >Random Forest Regressor</td>\n",
       "      <td id=\"T_286bc_row1_col1\" class=\"data row1 col1\" >0.5480</td>\n",
       "      <td id=\"T_286bc_row1_col2\" class=\"data row1 col2\" >7.0039</td>\n",
       "      <td id=\"T_286bc_row1_col3\" class=\"data row1 col3\" >2.4803</td>\n",
       "      <td id=\"T_286bc_row1_col4\" class=\"data row1 col4\" >0.9789</td>\n",
       "      <td id=\"T_286bc_row1_col5\" class=\"data row1 col5\" >0.1317</td>\n",
       "      <td id=\"T_286bc_row1_col6\" class=\"data row1 col6\" >0.2133</td>\n",
       "      <td id=\"T_286bc_row1_col7\" class=\"data row1 col7\" >2.9100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row2\" class=\"row_heading level0 row2\" >et</th>\n",
       "      <td id=\"T_286bc_row2_col0\" class=\"data row2 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_286bc_row2_col1\" class=\"data row2 col1\" >0.6299</td>\n",
       "      <td id=\"T_286bc_row2_col2\" class=\"data row2 col2\" >7.4491</td>\n",
       "      <td id=\"T_286bc_row2_col3\" class=\"data row2 col3\" >2.5118</td>\n",
       "      <td id=\"T_286bc_row2_col4\" class=\"data row2 col4\" >0.9790</td>\n",
       "      <td id=\"T_286bc_row2_col5\" class=\"data row2 col5\" >0.2004</td>\n",
       "      <td id=\"T_286bc_row2_col6\" class=\"data row2 col6\" >0.7481</td>\n",
       "      <td id=\"T_286bc_row2_col7\" class=\"data row2 col7\" >3.0600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row3\" class=\"row_heading level0 row3\" >gbr</th>\n",
       "      <td id=\"T_286bc_row3_col0\" class=\"data row3 col0\" >Gradient Boosting Regressor</td>\n",
       "      <td id=\"T_286bc_row3_col1\" class=\"data row3 col1\" >1.0531</td>\n",
       "      <td id=\"T_286bc_row3_col2\" class=\"data row3 col2\" >8.6165</td>\n",
       "      <td id=\"T_286bc_row3_col3\" class=\"data row3 col3\" >2.8567</td>\n",
       "      <td id=\"T_286bc_row3_col4\" class=\"data row3 col4\" >0.9696</td>\n",
       "      <td id=\"T_286bc_row3_col5\" class=\"data row3 col5\" >0.2462</td>\n",
       "      <td id=\"T_286bc_row3_col6\" class=\"data row3 col6\" >1.4319</td>\n",
       "      <td id=\"T_286bc_row3_col7\" class=\"data row3 col7\" >2.7630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row4\" class=\"row_heading level0 row4\" >knn</th>\n",
       "      <td id=\"T_286bc_row4_col0\" class=\"data row4 col0\" >K Neighbors Regressor</td>\n",
       "      <td id=\"T_286bc_row4_col1\" class=\"data row4 col1\" >1.0817</td>\n",
       "      <td id=\"T_286bc_row4_col2\" class=\"data row4 col2\" >13.6959</td>\n",
       "      <td id=\"T_286bc_row4_col3\" class=\"data row4 col3\" >3.5646</td>\n",
       "      <td id=\"T_286bc_row4_col4\" class=\"data row4 col4\" >0.9546</td>\n",
       "      <td id=\"T_286bc_row4_col5\" class=\"data row4 col5\" >0.2761</td>\n",
       "      <td id=\"T_286bc_row4_col6\" class=\"data row4 col6\" >0.7928</td>\n",
       "      <td id=\"T_286bc_row4_col7\" class=\"data row4 col7\" >4.6650</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row5\" class=\"row_heading level0 row5\" >ada</th>\n",
       "      <td id=\"T_286bc_row5_col0\" class=\"data row5 col0\" >AdaBoost Regressor</td>\n",
       "      <td id=\"T_286bc_row5_col1\" class=\"data row5 col1\" >2.2791</td>\n",
       "      <td id=\"T_286bc_row5_col2\" class=\"data row5 col2\" >26.5407</td>\n",
       "      <td id=\"T_286bc_row5_col3\" class=\"data row5 col3\" >5.0671</td>\n",
       "      <td id=\"T_286bc_row5_col4\" class=\"data row5 col4\" >0.9056</td>\n",
       "      <td id=\"T_286bc_row5_col5\" class=\"data row5 col5\" >0.5587</td>\n",
       "      <td id=\"T_286bc_row5_col6\" class=\"data row5 col6\" >11.2681</td>\n",
       "      <td id=\"T_286bc_row5_col7\" class=\"data row5 col7\" >2.7060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row6\" class=\"row_heading level0 row6\" >huber</th>\n",
       "      <td id=\"T_286bc_row6_col0\" class=\"data row6 col0\" >Huber Regressor</td>\n",
       "      <td id=\"T_286bc_row6_col1\" class=\"data row6 col1\" >5.3816</td>\n",
       "      <td id=\"T_286bc_row6_col2\" class=\"data row6 col2\" >348.8304</td>\n",
       "      <td id=\"T_286bc_row6_col3\" class=\"data row6 col3\" >18.1979</td>\n",
       "      <td id=\"T_286bc_row6_col4\" class=\"data row6 col4\" >-0.0879</td>\n",
       "      <td id=\"T_286bc_row6_col5\" class=\"data row6 col5\" >1.2715</td>\n",
       "      <td id=\"T_286bc_row6_col6\" class=\"data row6 col6\" >3.9441</td>\n",
       "      <td id=\"T_286bc_row6_col7\" class=\"data row6 col7\" >2.7920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row7\" class=\"row_heading level0 row7\" >br</th>\n",
       "      <td id=\"T_286bc_row7_col0\" class=\"data row7 col0\" >Bayesian Ridge</td>\n",
       "      <td id=\"T_286bc_row7_col1\" class=\"data row7 col1\" >7.6028</td>\n",
       "      <td id=\"T_286bc_row7_col2\" class=\"data row7 col2\" >314.2222</td>\n",
       "      <td id=\"T_286bc_row7_col3\" class=\"data row7 col3\" >17.2453</td>\n",
       "      <td id=\"T_286bc_row7_col4\" class=\"data row7 col4\" >0.0272</td>\n",
       "      <td id=\"T_286bc_row7_col5\" class=\"data row7 col5\" >1.4022</td>\n",
       "      <td id=\"T_286bc_row7_col6\" class=\"data row7 col6\" >50.8959</td>\n",
       "      <td id=\"T_286bc_row7_col7\" class=\"data row7 col7\" >2.4200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row8\" class=\"row_heading level0 row8\" >ridge</th>\n",
       "      <td id=\"T_286bc_row8_col0\" class=\"data row8 col0\" >Ridge Regression</td>\n",
       "      <td id=\"T_286bc_row8_col1\" class=\"data row8 col1\" >7.6064</td>\n",
       "      <td id=\"T_286bc_row8_col2\" class=\"data row8 col2\" >314.2063</td>\n",
       "      <td id=\"T_286bc_row8_col3\" class=\"data row8 col3\" >17.2454</td>\n",
       "      <td id=\"T_286bc_row8_col4\" class=\"data row8 col4\" >0.0271</td>\n",
       "      <td id=\"T_286bc_row8_col5\" class=\"data row8 col5\" >1.4017</td>\n",
       "      <td id=\"T_286bc_row8_col6\" class=\"data row8 col6\" >50.9735</td>\n",
       "      <td id=\"T_286bc_row8_col7\" class=\"data row8 col7\" >2.5750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row9\" class=\"row_heading level0 row9\" >lr</th>\n",
       "      <td id=\"T_286bc_row9_col0\" class=\"data row9 col0\" >Linear Regression</td>\n",
       "      <td id=\"T_286bc_row9_col1\" class=\"data row9 col1\" >7.6067</td>\n",
       "      <td id=\"T_286bc_row9_col2\" class=\"data row9 col2\" >314.2076</td>\n",
       "      <td id=\"T_286bc_row9_col3\" class=\"data row9 col3\" >17.2455</td>\n",
       "      <td id=\"T_286bc_row9_col4\" class=\"data row9 col4\" >0.0271</td>\n",
       "      <td id=\"T_286bc_row9_col5\" class=\"data row9 col5\" >1.4017</td>\n",
       "      <td id=\"T_286bc_row9_col6\" class=\"data row9 col6\" >50.9820</td>\n",
       "      <td id=\"T_286bc_row9_col7\" class=\"data row9 col7\" >2.3730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row10\" class=\"row_heading level0 row10\" >lar</th>\n",
       "      <td id=\"T_286bc_row10_col0\" class=\"data row10 col0\" >Least Angle Regression</td>\n",
       "      <td id=\"T_286bc_row10_col1\" class=\"data row10 col1\" >7.6078</td>\n",
       "      <td id=\"T_286bc_row10_col2\" class=\"data row10 col2\" >314.2209</td>\n",
       "      <td id=\"T_286bc_row10_col3\" class=\"data row10 col3\" >17.2460</td>\n",
       "      <td id=\"T_286bc_row10_col4\" class=\"data row10 col4\" >0.0270</td>\n",
       "      <td id=\"T_286bc_row10_col5\" class=\"data row10 col5\" >1.4014</td>\n",
       "      <td id=\"T_286bc_row10_col6\" class=\"data row10 col6\" >50.9752</td>\n",
       "      <td id=\"T_286bc_row10_col7\" class=\"data row10 col7\" >2.6830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row11\" class=\"row_heading level0 row11\" >lasso</th>\n",
       "      <td id=\"T_286bc_row11_col0\" class=\"data row11 col0\" >Lasso Regression</td>\n",
       "      <td id=\"T_286bc_row11_col1\" class=\"data row11 col1\" >7.6877</td>\n",
       "      <td id=\"T_286bc_row11_col2\" class=\"data row11 col2\" >315.8708</td>\n",
       "      <td id=\"T_286bc_row11_col3\" class=\"data row11 col3\" >17.2917</td>\n",
       "      <td id=\"T_286bc_row11_col4\" class=\"data row11 col4\" >0.0217</td>\n",
       "      <td id=\"T_286bc_row11_col5\" class=\"data row11 col5\" >1.4416</td>\n",
       "      <td id=\"T_286bc_row11_col6\" class=\"data row11 col6\" >53.8391</td>\n",
       "      <td id=\"T_286bc_row11_col7\" class=\"data row11 col7\" >2.2440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row12\" class=\"row_heading level0 row12\" >llar</th>\n",
       "      <td id=\"T_286bc_row12_col0\" class=\"data row12 col0\" >Lasso Least Angle Regression</td>\n",
       "      <td id=\"T_286bc_row12_col1\" class=\"data row12 col1\" >7.6877</td>\n",
       "      <td id=\"T_286bc_row12_col2\" class=\"data row12 col2\" >315.8708</td>\n",
       "      <td id=\"T_286bc_row12_col3\" class=\"data row12 col3\" >17.2917</td>\n",
       "      <td id=\"T_286bc_row12_col4\" class=\"data row12 col4\" >0.0217</td>\n",
       "      <td id=\"T_286bc_row12_col5\" class=\"data row12 col5\" >1.4416</td>\n",
       "      <td id=\"T_286bc_row12_col6\" class=\"data row12 col6\" >53.8391</td>\n",
       "      <td id=\"T_286bc_row12_col7\" class=\"data row12 col7\" >2.6220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row13\" class=\"row_heading level0 row13\" >en</th>\n",
       "      <td id=\"T_286bc_row13_col0\" class=\"data row13 col0\" >Elastic Net</td>\n",
       "      <td id=\"T_286bc_row13_col1\" class=\"data row13 col1\" >7.7125</td>\n",
       "      <td id=\"T_286bc_row13_col2\" class=\"data row13 col2\" >316.5579</td>\n",
       "      <td id=\"T_286bc_row13_col3\" class=\"data row13 col3\" >17.3124</td>\n",
       "      <td id=\"T_286bc_row13_col4\" class=\"data row13 col4\" >0.0191</td>\n",
       "      <td id=\"T_286bc_row13_col5\" class=\"data row13 col5\" >1.4490</td>\n",
       "      <td id=\"T_286bc_row13_col6\" class=\"data row13 col6\" >54.4763</td>\n",
       "      <td id=\"T_286bc_row13_col7\" class=\"data row13 col7\" >2.7290</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row14\" class=\"row_heading level0 row14\" >dummy</th>\n",
       "      <td id=\"T_286bc_row14_col0\" class=\"data row14 col0\" >Dummy Regressor</td>\n",
       "      <td id=\"T_286bc_row14_col1\" class=\"data row14 col1\" >7.8272</td>\n",
       "      <td id=\"T_286bc_row14_col2\" class=\"data row14 col2\" >323.0005</td>\n",
       "      <td id=\"T_286bc_row14_col3\" class=\"data row14 col3\" >17.5022</td>\n",
       "      <td id=\"T_286bc_row14_col4\" class=\"data row14 col4\" >-0.0048</td>\n",
       "      <td id=\"T_286bc_row14_col5\" class=\"data row14 col5\" >1.4837</td>\n",
       "      <td id=\"T_286bc_row14_col6\" class=\"data row14 col6\" >58.0146</td>\n",
       "      <td id=\"T_286bc_row14_col7\" class=\"data row14 col7\" >2.3860</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row15\" class=\"row_heading level0 row15\" >omp</th>\n",
       "      <td id=\"T_286bc_row15_col0\" class=\"data row15 col0\" >Orthogonal Matching Pursuit</td>\n",
       "      <td id=\"T_286bc_row15_col1\" class=\"data row15 col1\" >7.8679</td>\n",
       "      <td id=\"T_286bc_row15_col2\" class=\"data row15 col2\" >321.0936</td>\n",
       "      <td id=\"T_286bc_row15_col3\" class=\"data row15 col3\" >17.4489</td>\n",
       "      <td id=\"T_286bc_row15_col4\" class=\"data row15 col4\" >0.0016</td>\n",
       "      <td id=\"T_286bc_row15_col5\" class=\"data row15 col5\" >1.4829</td>\n",
       "      <td id=\"T_286bc_row15_col6\" class=\"data row15 col6\" >56.8566</td>\n",
       "      <td id=\"T_286bc_row15_col7\" class=\"data row15 col7\" >2.8110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_286bc_level0_row16\" class=\"row_heading level0 row16\" >par</th>\n",
       "      <td id=\"T_286bc_row16_col0\" class=\"data row16 col0\" >Passive Aggressive Regressor</td>\n",
       "      <td id=\"T_286bc_row16_col1\" class=\"data row16 col1\" >4246.5662</td>\n",
       "      <td id=\"T_286bc_row16_col2\" class=\"data row16 col2\" >809270356.9473</td>\n",
       "      <td id=\"T_286bc_row16_col3\" class=\"data row16 col3\" >15908.5749</td>\n",
       "      <td id=\"T_286bc_row16_col4\" class=\"data row16 col4\" >-5359446.3804</td>\n",
       "      <td id=\"T_286bc_row16_col5\" class=\"data row16 col5\" >5.2976</td>\n",
       "      <td id=\"T_286bc_row16_col6\" class=\"data row16 col6\" >358158.7250</td>\n",
       "      <td id=\"T_286bc_row16_col7\" class=\"data row16 col7\" >2.8250</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x248753f4dc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1cb098da1be3458e994ba36747485ce8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/51 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Multi-model training/test - Compare Mean Absolute Error\n",
    "# Select the best 3 models - Create a blended model this time\n",
    "best_3_outdoor = compare_models(sort='MAE',budget_time=30,fold=10,exclude='lightgbm',n_select=3) \n",
    "# 30 mins budget time\n",
    "# remove Light gradient boosting - causes error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'ccp_alpha': 0.0,\n",
       " 'criterion': 'squared_error',\n",
       " 'max_depth': None,\n",
       " 'max_features': None,\n",
       " 'max_leaf_nodes': None,\n",
       " 'min_impurity_decrease': 0.0,\n",
       " 'min_samples_leaf': 1,\n",
       " 'min_samples_split': 2,\n",
       " 'min_weight_fraction_leaf': 0.0,\n",
       " 'random_state': 123,\n",
       " 'splitter': 'best'}"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check the model parameters in top 3\n",
    "best_3_outdoor[0].get_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.2.2 Blended Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_f18be_row10_col0, #T_f18be_row10_col1, #T_f18be_row10_col2, #T_f18be_row10_col3, #T_f18be_row10_col4, #T_f18be_row10_col5 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_f18be\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_f18be_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_f18be_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_f18be_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_f18be_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_f18be_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_f18be_level0_col5\" class=\"col_heading level0 col5\" >MAPE</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_f18be_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_f18be_row0_col0\" class=\"data row0 col0\" >1.0453</td>\n",
       "      <td id=\"T_f18be_row0_col1\" class=\"data row0 col1\" >10.8500</td>\n",
       "      <td id=\"T_f18be_row0_col2\" class=\"data row0 col2\" >3.2939</td>\n",
       "      <td id=\"T_f18be_row0_col3\" class=\"data row0 col3\" >0.9685</td>\n",
       "      <td id=\"T_f18be_row0_col4\" class=\"data row0 col4\" >0.2301</td>\n",
       "      <td id=\"T_f18be_row0_col5\" class=\"data row0 col5\" >0.7956</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_f18be_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_f18be_row1_col0\" class=\"data row1 col0\" >0.9833</td>\n",
       "      <td id=\"T_f18be_row1_col1\" class=\"data row1 col1\" >7.3053</td>\n",
       "      <td id=\"T_f18be_row1_col2\" class=\"data row1 col2\" >2.7028</td>\n",
       "      <td id=\"T_f18be_row1_col3\" class=\"data row1 col3\" >0.9819</td>\n",
       "      <td id=\"T_f18be_row1_col4\" class=\"data row1 col4\" >0.2370</td>\n",
       "      <td id=\"T_f18be_row1_col5\" class=\"data row1 col5\" >0.8354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_f18be_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_f18be_row2_col0\" class=\"data row2 col0\" >0.7641</td>\n",
       "      <td id=\"T_f18be_row2_col1\" class=\"data row2 col1\" >4.0337</td>\n",
       "      <td id=\"T_f18be_row2_col2\" class=\"data row2 col2\" >2.0084</td>\n",
       "      <td id=\"T_f18be_row2_col3\" class=\"data row2 col3\" >0.9683</td>\n",
       "      <td id=\"T_f18be_row2_col4\" class=\"data row2 col4\" >0.2297</td>\n",
       "      <td id=\"T_f18be_row2_col5\" class=\"data row2 col5\" >0.7437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_f18be_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_f18be_row3_col0\" class=\"data row3 col0\" >0.9622</td>\n",
       "      <td id=\"T_f18be_row3_col1\" class=\"data row3 col1\" >11.4422</td>\n",
       "      <td id=\"T_f18be_row3_col2\" class=\"data row3 col2\" >3.3826</td>\n",
       "      <td id=\"T_f18be_row3_col3\" class=\"data row3 col3\" >0.9057</td>\n",
       "      <td id=\"T_f18be_row3_col4\" class=\"data row3 col4\" >0.2210</td>\n",
       "      <td id=\"T_f18be_row3_col5\" class=\"data row3 col5\" >0.8454</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_f18be_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_f18be_row4_col0\" class=\"data row4 col0\" >1.1177</td>\n",
       "      <td id=\"T_f18be_row4_col1\" class=\"data row4 col1\" >8.0929</td>\n",
       "      <td id=\"T_f18be_row4_col2\" class=\"data row4 col2\" >2.8448</td>\n",
       "      <td id=\"T_f18be_row4_col3\" class=\"data row4 col3\" >0.9875</td>\n",
       "      <td id=\"T_f18be_row4_col4\" class=\"data row4 col4\" >0.2377</td>\n",
       "      <td id=\"T_f18be_row4_col5\" class=\"data row4 col5\" >0.8318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_f18be_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_f18be_row5_col0\" class=\"data row5 col0\" >1.0524</td>\n",
       "      <td id=\"T_f18be_row5_col1\" class=\"data row5 col1\" >24.6800</td>\n",
       "      <td id=\"T_f18be_row5_col2\" class=\"data row5 col2\" >4.9679</td>\n",
       "      <td id=\"T_f18be_row5_col3\" class=\"data row5 col3\" >0.9017</td>\n",
       "      <td id=\"T_f18be_row5_col4\" class=\"data row5 col4\" >0.2196</td>\n",
       "      <td id=\"T_f18be_row5_col5\" class=\"data row5 col5\" >0.7730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_f18be_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_f18be_row6_col0\" class=\"data row6 col0\" >0.9995</td>\n",
       "      <td id=\"T_f18be_row6_col1\" class=\"data row6 col1\" >6.2995</td>\n",
       "      <td id=\"T_f18be_row6_col2\" class=\"data row6 col2\" >2.5099</td>\n",
       "      <td id=\"T_f18be_row6_col3\" class=\"data row6 col3\" >0.9838</td>\n",
       "      <td id=\"T_f18be_row6_col4\" class=\"data row6 col4\" >0.2298</td>\n",
       "      <td id=\"T_f18be_row6_col5\" class=\"data row6 col5\" >0.9554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_f18be_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_f18be_row7_col0\" class=\"data row7 col0\" >1.0068</td>\n",
       "      <td id=\"T_f18be_row7_col1\" class=\"data row7 col1\" >25.4069</td>\n",
       "      <td id=\"T_f18be_row7_col2\" class=\"data row7 col2\" >5.0405</td>\n",
       "      <td id=\"T_f18be_row7_col3\" class=\"data row7 col3\" >0.8950</td>\n",
       "      <td id=\"T_f18be_row7_col4\" class=\"data row7 col4\" >0.2189</td>\n",
       "      <td id=\"T_f18be_row7_col5\" class=\"data row7 col5\" >0.7519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_f18be_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_f18be_row8_col0\" class=\"data row8 col0\" >1.0575</td>\n",
       "      <td id=\"T_f18be_row8_col1\" class=\"data row8 col1\" >22.8713</td>\n",
       "      <td id=\"T_f18be_row8_col2\" class=\"data row8 col2\" >4.7824</td>\n",
       "      <td id=\"T_f18be_row8_col3\" class=\"data row8 col3\" >0.9392</td>\n",
       "      <td id=\"T_f18be_row8_col4\" class=\"data row8 col4\" >0.1988</td>\n",
       "      <td id=\"T_f18be_row8_col5\" class=\"data row8 col5\" >0.6752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_f18be_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_f18be_row9_col0\" class=\"data row9 col0\" >0.7316</td>\n",
       "      <td id=\"T_f18be_row9_col1\" class=\"data row9 col1\" >3.0321</td>\n",
       "      <td id=\"T_f18be_row9_col2\" class=\"data row9 col2\" >1.7413</td>\n",
       "      <td id=\"T_f18be_row9_col3\" class=\"data row9 col3\" >0.9903</td>\n",
       "      <td id=\"T_f18be_row9_col4\" class=\"data row9 col4\" >0.2243</td>\n",
       "      <td id=\"T_f18be_row9_col5\" class=\"data row9 col5\" >0.8025</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_f18be_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_f18be_row10_col0\" class=\"data row10 col0\" >0.9720</td>\n",
       "      <td id=\"T_f18be_row10_col1\" class=\"data row10 col1\" >12.4014</td>\n",
       "      <td id=\"T_f18be_row10_col2\" class=\"data row10 col2\" >3.3275</td>\n",
       "      <td id=\"T_f18be_row10_col3\" class=\"data row10 col3\" >0.9522</td>\n",
       "      <td id=\"T_f18be_row10_col4\" class=\"data row10 col4\" >0.2247</td>\n",
       "      <td id=\"T_f18be_row10_col5\" class=\"data row10 col5\" >0.8010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_f18be_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_f18be_row11_col0\" class=\"data row11 col0\" >0.1198</td>\n",
       "      <td id=\"T_f18be_row11_col1\" class=\"data row11 col1\" >8.2001</td>\n",
       "      <td id=\"T_f18be_row11_col2\" class=\"data row11 col2\" >1.1530</td>\n",
       "      <td id=\"T_f18be_row11_col3\" class=\"data row11 col3\" >0.0364</td>\n",
       "      <td id=\"T_f18be_row11_col4\" class=\"data row11 col4\" >0.0107</td>\n",
       "      <td id=\"T_f18be_row11_col5\" class=\"data row11 col5\" >0.0709</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2487e76ddc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d2f13b3e11724d92b47a576c1c2d5416",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/4 [00:00<?, ?it/s]"
      ]
     },
     "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_e3dc4_row10_col0, #T_e3dc4_row10_col1, #T_e3dc4_row10_col2, #T_e3dc4_row10_col3, #T_e3dc4_row10_col4, #T_e3dc4_row10_col5 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_e3dc4\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_e3dc4_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_e3dc4_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_e3dc4_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_e3dc4_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_e3dc4_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_e3dc4_level0_col5\" class=\"col_heading level0 col5\" >MAPE</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_e3dc4_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_e3dc4_row0_col0\" class=\"data row0 col0\" >1.0579</td>\n",
       "      <td id=\"T_e3dc4_row0_col1\" class=\"data row0 col1\" >14.0175</td>\n",
       "      <td id=\"T_e3dc4_row0_col2\" class=\"data row0 col2\" >3.7440</td>\n",
       "      <td id=\"T_e3dc4_row0_col3\" class=\"data row0 col3\" >0.9594</td>\n",
       "      <td id=\"T_e3dc4_row0_col4\" class=\"data row0 col4\" >0.2172</td>\n",
       "      <td id=\"T_e3dc4_row0_col5\" class=\"data row0 col5\" >0.7288</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3dc4_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_e3dc4_row1_col0\" class=\"data row1 col0\" >0.8583</td>\n",
       "      <td id=\"T_e3dc4_row1_col1\" class=\"data row1 col1\" >6.5010</td>\n",
       "      <td id=\"T_e3dc4_row1_col2\" class=\"data row1 col2\" >2.5497</td>\n",
       "      <td id=\"T_e3dc4_row1_col3\" class=\"data row1 col3\" >0.9839</td>\n",
       "      <td id=\"T_e3dc4_row1_col4\" class=\"data row1 col4\" >0.2137</td>\n",
       "      <td id=\"T_e3dc4_row1_col5\" class=\"data row1 col5\" >0.7977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3dc4_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_e3dc4_row2_col0\" class=\"data row2 col0\" >0.7770</td>\n",
       "      <td id=\"T_e3dc4_row2_col1\" class=\"data row2 col1\" >4.2761</td>\n",
       "      <td id=\"T_e3dc4_row2_col2\" class=\"data row2 col2\" >2.0679</td>\n",
       "      <td id=\"T_e3dc4_row2_col3\" class=\"data row2 col3\" >0.9664</td>\n",
       "      <td id=\"T_e3dc4_row2_col4\" class=\"data row2 col4\" >0.2078</td>\n",
       "      <td id=\"T_e3dc4_row2_col5\" class=\"data row2 col5\" >0.7152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3dc4_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_e3dc4_row3_col0\" class=\"data row3 col0\" >0.7732</td>\n",
       "      <td id=\"T_e3dc4_row3_col1\" class=\"data row3 col1\" >3.6403</td>\n",
       "      <td id=\"T_e3dc4_row3_col2\" class=\"data row3 col2\" >1.9080</td>\n",
       "      <td id=\"T_e3dc4_row3_col3\" class=\"data row3 col3\" >0.9700</td>\n",
       "      <td id=\"T_e3dc4_row3_col4\" class=\"data row3 col4\" >0.2008</td>\n",
       "      <td id=\"T_e3dc4_row3_col5\" class=\"data row3 col5\" >0.7737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3dc4_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_e3dc4_row4_col0\" class=\"data row4 col0\" >1.1226</td>\n",
       "      <td id=\"T_e3dc4_row4_col1\" class=\"data row4 col1\" >21.0379</td>\n",
       "      <td id=\"T_e3dc4_row4_col2\" class=\"data row4 col2\" >4.5867</td>\n",
       "      <td id=\"T_e3dc4_row4_col3\" class=\"data row4 col3\" >0.9676</td>\n",
       "      <td id=\"T_e3dc4_row4_col4\" class=\"data row4 col4\" >0.2193</td>\n",
       "      <td id=\"T_e3dc4_row4_col5\" class=\"data row4 col5\" >0.7677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3dc4_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_e3dc4_row5_col0\" class=\"data row5 col0\" >0.8107</td>\n",
       "      <td id=\"T_e3dc4_row5_col1\" class=\"data row5 col1\" >11.0711</td>\n",
       "      <td id=\"T_e3dc4_row5_col2\" class=\"data row5 col2\" >3.3273</td>\n",
       "      <td id=\"T_e3dc4_row5_col3\" class=\"data row5 col3\" >0.9559</td>\n",
       "      <td id=\"T_e3dc4_row5_col4\" class=\"data row5 col4\" >0.2071</td>\n",
       "      <td id=\"T_e3dc4_row5_col5\" class=\"data row5 col5\" >0.7263</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3dc4_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_e3dc4_row6_col0\" class=\"data row6 col0\" >0.9201</td>\n",
       "      <td id=\"T_e3dc4_row6_col1\" class=\"data row6 col1\" >4.7424</td>\n",
       "      <td id=\"T_e3dc4_row6_col2\" class=\"data row6 col2\" >2.1777</td>\n",
       "      <td id=\"T_e3dc4_row6_col3\" class=\"data row6 col3\" >0.9878</td>\n",
       "      <td id=\"T_e3dc4_row6_col4\" class=\"data row6 col4\" >0.2203</td>\n",
       "      <td id=\"T_e3dc4_row6_col5\" class=\"data row6 col5\" >0.8882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3dc4_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_e3dc4_row7_col0\" class=\"data row7 col0\" >0.8583</td>\n",
       "      <td id=\"T_e3dc4_row7_col1\" class=\"data row7 col1\" >7.6696</td>\n",
       "      <td id=\"T_e3dc4_row7_col2\" class=\"data row7 col2\" >2.7694</td>\n",
       "      <td id=\"T_e3dc4_row7_col3\" class=\"data row7 col3\" >0.9683</td>\n",
       "      <td id=\"T_e3dc4_row7_col4\" class=\"data row7 col4\" >0.1997</td>\n",
       "      <td id=\"T_e3dc4_row7_col5\" class=\"data row7 col5\" >0.6948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3dc4_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_e3dc4_row8_col0\" class=\"data row8 col0\" >0.9592</td>\n",
       "      <td id=\"T_e3dc4_row8_col1\" class=\"data row8 col1\" >10.3014</td>\n",
       "      <td id=\"T_e3dc4_row8_col2\" class=\"data row8 col2\" >3.2096</td>\n",
       "      <td id=\"T_e3dc4_row8_col3\" class=\"data row8 col3\" >0.9726</td>\n",
       "      <td id=\"T_e3dc4_row8_col4\" class=\"data row8 col4\" >0.1972</td>\n",
       "      <td id=\"T_e3dc4_row8_col5\" class=\"data row8 col5\" >0.6586</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3dc4_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_e3dc4_row9_col0\" class=\"data row9 col0\" >0.8478</td>\n",
       "      <td id=\"T_e3dc4_row9_col1\" class=\"data row9 col1\" >6.8215</td>\n",
       "      <td id=\"T_e3dc4_row9_col2\" class=\"data row9 col2\" >2.6118</td>\n",
       "      <td id=\"T_e3dc4_row9_col3\" class=\"data row9 col3\" >0.9783</td>\n",
       "      <td id=\"T_e3dc4_row9_col4\" class=\"data row9 col4\" >0.2179</td>\n",
       "      <td id=\"T_e3dc4_row9_col5\" class=\"data row9 col5\" >0.7550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3dc4_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_e3dc4_row10_col0\" class=\"data row10 col0\" >0.8985</td>\n",
       "      <td id=\"T_e3dc4_row10_col1\" class=\"data row10 col1\" >9.0079</td>\n",
       "      <td id=\"T_e3dc4_row10_col2\" class=\"data row10 col2\" >2.8952</td>\n",
       "      <td id=\"T_e3dc4_row10_col3\" class=\"data row10 col3\" >0.9710</td>\n",
       "      <td id=\"T_e3dc4_row10_col4\" class=\"data row10 col4\" >0.2101</td>\n",
       "      <td id=\"T_e3dc4_row10_col5\" class=\"data row10 col5\" >0.7506</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_e3dc4_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_e3dc4_row11_col0\" class=\"data row11 col0\" >0.1112</td>\n",
       "      <td id=\"T_e3dc4_row11_col1\" class=\"data row11 col1\" >5.0794</td>\n",
       "      <td id=\"T_e3dc4_row11_col2\" class=\"data row11 col2\" >0.7910</td>\n",
       "      <td id=\"T_e3dc4_row11_col3\" class=\"data row11 col3\" >0.0095</td>\n",
       "      <td id=\"T_e3dc4_row11_col4\" class=\"data row11 col4\" >0.0083</td>\n",
       "      <td id=\"T_e3dc4_row11_col5\" class=\"data row11 col5\" >0.0599</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2487e5d2d90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c4d63e34d32e4ed89bb2e488fccddafd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/4 [00:00<?, ?it/s]"
      ]
     },
     "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_ea7da_row10_col0, #T_ea7da_row10_col1, #T_ea7da_row10_col2, #T_ea7da_row10_col3, #T_ea7da_row10_col4, #T_ea7da_row10_col5 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_ea7da\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_ea7da_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_ea7da_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_ea7da_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_ea7da_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_ea7da_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_ea7da_level0_col5\" class=\"col_heading level0 col5\" >MAPE</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_ea7da_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_ea7da_row0_col0\" class=\"data row0 col0\" >5.6420</td>\n",
       "      <td id=\"T_ea7da_row0_col1\" class=\"data row0 col1\" >230.4316</td>\n",
       "      <td id=\"T_ea7da_row0_col2\" class=\"data row0 col2\" >15.1800</td>\n",
       "      <td id=\"T_ea7da_row0_col3\" class=\"data row0 col3\" >0.3320</td>\n",
       "      <td id=\"T_ea7da_row0_col4\" class=\"data row0 col4\" >0.9792</td>\n",
       "      <td id=\"T_ea7da_row0_col5\" class=\"data row0 col5\" >14.1895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ea7da_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_ea7da_row1_col0\" class=\"data row1 col0\" >5.6213</td>\n",
       "      <td id=\"T_ea7da_row1_col1\" class=\"data row1 col1\" >298.0491</td>\n",
       "      <td id=\"T_ea7da_row1_col2\" class=\"data row1 col2\" >17.2641</td>\n",
       "      <td id=\"T_ea7da_row1_col3\" class=\"data row1 col3\" >0.2622</td>\n",
       "      <td id=\"T_ea7da_row1_col4\" class=\"data row1 col4\" >0.9602</td>\n",
       "      <td id=\"T_ea7da_row1_col5\" class=\"data row1 col5\" >12.4515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ea7da_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_ea7da_row2_col0\" class=\"data row2 col0\" >4.2524</td>\n",
       "      <td id=\"T_ea7da_row2_col1\" class=\"data row2 col1\" >60.7441</td>\n",
       "      <td id=\"T_ea7da_row2_col2\" class=\"data row2 col2\" >7.7939</td>\n",
       "      <td id=\"T_ea7da_row2_col3\" class=\"data row2 col3\" >0.5232</td>\n",
       "      <td id=\"T_ea7da_row2_col4\" class=\"data row2 col4\" >1.0383</td>\n",
       "      <td id=\"T_ea7da_row2_col5\" class=\"data row2 col5\" >14.9541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ea7da_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_ea7da_row3_col0\" class=\"data row3 col0\" >4.2823</td>\n",
       "      <td id=\"T_ea7da_row3_col1\" class=\"data row3 col1\" >59.4442</td>\n",
       "      <td id=\"T_ea7da_row3_col2\" class=\"data row3 col2\" >7.7100</td>\n",
       "      <td id=\"T_ea7da_row3_col3\" class=\"data row3 col3\" >0.5103</td>\n",
       "      <td id=\"T_ea7da_row3_col4\" class=\"data row3 col4\" >0.9419</td>\n",
       "      <td id=\"T_ea7da_row3_col5\" class=\"data row3 col5\" >13.3897</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ea7da_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_ea7da_row4_col0\" class=\"data row4 col0\" >6.7079</td>\n",
       "      <td id=\"T_ea7da_row4_col1\" class=\"data row4 col1\" >496.7091</td>\n",
       "      <td id=\"T_ea7da_row4_col2\" class=\"data row4 col2\" >22.2870</td>\n",
       "      <td id=\"T_ea7da_row4_col3\" class=\"data row4 col3\" >0.2341</td>\n",
       "      <td id=\"T_ea7da_row4_col4\" class=\"data row4 col4\" >0.9332</td>\n",
       "      <td id=\"T_ea7da_row4_col5\" class=\"data row4 col5\" >10.2214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ea7da_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_ea7da_row5_col0\" class=\"data row5 col0\" >4.7124</td>\n",
       "      <td id=\"T_ea7da_row5_col1\" class=\"data row5 col1\" >153.9898</td>\n",
       "      <td id=\"T_ea7da_row5_col2\" class=\"data row5 col2\" >12.4093</td>\n",
       "      <td id=\"T_ea7da_row5_col3\" class=\"data row5 col3\" >0.3867</td>\n",
       "      <td id=\"T_ea7da_row5_col4\" class=\"data row5 col4\" >0.9878</td>\n",
       "      <td id=\"T_ea7da_row5_col5\" class=\"data row5 col5\" >14.6409</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ea7da_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_ea7da_row6_col0\" class=\"data row6 col0\" >5.2275</td>\n",
       "      <td id=\"T_ea7da_row6_col1\" class=\"data row6 col1\" >295.1385</td>\n",
       "      <td id=\"T_ea7da_row6_col2\" class=\"data row6 col2\" >17.1796</td>\n",
       "      <td id=\"T_ea7da_row6_col3\" class=\"data row6 col3\" >0.2397</td>\n",
       "      <td id=\"T_ea7da_row6_col4\" class=\"data row6 col4\" >0.9404</td>\n",
       "      <td id=\"T_ea7da_row6_col5\" class=\"data row6 col5\" >12.5175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ea7da_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_ea7da_row7_col0\" class=\"data row7 col0\" >4.3438</td>\n",
       "      <td id=\"T_ea7da_row7_col1\" class=\"data row7 col1\" >163.1672</td>\n",
       "      <td id=\"T_ea7da_row7_col2\" class=\"data row7 col2\" >12.7737</td>\n",
       "      <td id=\"T_ea7da_row7_col3\" class=\"data row7 col3\" >0.3259</td>\n",
       "      <td id=\"T_ea7da_row7_col4\" class=\"data row7 col4\" >0.9319</td>\n",
       "      <td id=\"T_ea7da_row7_col5\" class=\"data row7 col5\" >13.0943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ea7da_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_ea7da_row8_col0\" class=\"data row8 col0\" >5.4413</td>\n",
       "      <td id=\"T_ea7da_row8_col1\" class=\"data row8 col1\" >285.9416</td>\n",
       "      <td id=\"T_ea7da_row8_col2\" class=\"data row8 col2\" >16.9098</td>\n",
       "      <td id=\"T_ea7da_row8_col3\" class=\"data row8 col3\" >0.2403</td>\n",
       "      <td id=\"T_ea7da_row8_col4\" class=\"data row8 col4\" >0.9596</td>\n",
       "      <td id=\"T_ea7da_row8_col5\" class=\"data row8 col5\" >13.7237</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ea7da_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_ea7da_row9_col0\" class=\"data row9 col0\" >5.0547</td>\n",
       "      <td id=\"T_ea7da_row9_col1\" class=\"data row9 col1\" >200.6964</td>\n",
       "      <td id=\"T_ea7da_row9_col2\" class=\"data row9 col2\" >14.1667</td>\n",
       "      <td id=\"T_ea7da_row9_col3\" class=\"data row9 col3\" >0.3606</td>\n",
       "      <td id=\"T_ea7da_row9_col4\" class=\"data row9 col4\" >0.9526</td>\n",
       "      <td id=\"T_ea7da_row9_col5\" class=\"data row9 col5\" >12.0229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ea7da_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_ea7da_row10_col0\" class=\"data row10 col0\" >5.1285</td>\n",
       "      <td id=\"T_ea7da_row10_col1\" class=\"data row10 col1\" >224.4312</td>\n",
       "      <td id=\"T_ea7da_row10_col2\" class=\"data row10 col2\" >14.3674</td>\n",
       "      <td id=\"T_ea7da_row10_col3\" class=\"data row10 col3\" >0.3415</td>\n",
       "      <td id=\"T_ea7da_row10_col4\" class=\"data row10 col4\" >0.9625</td>\n",
       "      <td id=\"T_ea7da_row10_col5\" class=\"data row10 col5\" >13.1205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ea7da_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_ea7da_row11_col0\" class=\"data row11 col0\" >0.7340</td>\n",
       "      <td id=\"T_ea7da_row11_col1\" class=\"data row11 col1\" >123.1530</td>\n",
       "      <td id=\"T_ea7da_row11_col2\" class=\"data row11 col2\" >4.2437</td>\n",
       "      <td id=\"T_ea7da_row11_col3\" class=\"data row11 col3\" >0.1013</td>\n",
       "      <td id=\"T_ea7da_row11_col4\" class=\"data row11 col4\" >0.0308</td>\n",
       "      <td id=\"T_ea7da_row11_col5\" class=\"data row11 col5\" >1.3294</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2487e865490>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "47971aa8fd66473697abaedf2938a269",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/4 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Initialize 3 best models, Capping max_depth at 7\n",
    "dt_outdoor = create_model('dt',max_depth=7)\n",
    "rf_outdoor = create_model('rf',max_depth=7)\n",
    "et_outdoor = create_model('et',max_depth=7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_0f73b_row10_col0, #T_0f73b_row10_col1, #T_0f73b_row10_col2, #T_0f73b_row10_col3, #T_0f73b_row10_col4, #T_0f73b_row10_col5 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_0f73b\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_0f73b_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_0f73b_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_0f73b_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_0f73b_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_0f73b_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_0f73b_level0_col5\" class=\"col_heading level0 col5\" >MAPE</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_0f73b_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_0f73b_row0_col0\" class=\"data row0 col0\" >2.0879</td>\n",
       "      <td id=\"T_0f73b_row0_col1\" class=\"data row0 col1\" >27.6458</td>\n",
       "      <td id=\"T_0f73b_row0_col2\" class=\"data row0 col2\" >5.2579</td>\n",
       "      <td id=\"T_0f73b_row0_col3\" class=\"data row0 col3\" >0.9199</td>\n",
       "      <td id=\"T_0f73b_row0_col4\" class=\"data row0 col4\" >0.5040</td>\n",
       "      <td id=\"T_0f73b_row0_col5\" class=\"data row0 col5\" >5.1160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0f73b_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_0f73b_row1_col0\" class=\"data row1 col0\" >2.2131</td>\n",
       "      <td id=\"T_0f73b_row1_col1\" class=\"data row1 col1\" >38.6490</td>\n",
       "      <td id=\"T_0f73b_row1_col2\" class=\"data row1 col2\" >6.2168</td>\n",
       "      <td id=\"T_0f73b_row1_col3\" class=\"data row1 col3\" >0.9043</td>\n",
       "      <td id=\"T_0f73b_row1_col4\" class=\"data row1 col4\" >0.4962</td>\n",
       "      <td id=\"T_0f73b_row1_col5\" class=\"data row1 col5\" >4.5813</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0f73b_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_0f73b_row2_col0\" class=\"data row2 col0\" >1.7012</td>\n",
       "      <td id=\"T_0f73b_row2_col1\" class=\"data row2 col1\" >10.9323</td>\n",
       "      <td id=\"T_0f73b_row2_col2\" class=\"data row2 col2\" >3.3064</td>\n",
       "      <td id=\"T_0f73b_row2_col3\" class=\"data row2 col3\" >0.9142</td>\n",
       "      <td id=\"T_0f73b_row2_col4\" class=\"data row2 col4\" >0.5508</td>\n",
       "      <td id=\"T_0f73b_row2_col5\" class=\"data row2 col5\" >5.3433</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0f73b_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_0f73b_row3_col0\" class=\"data row3 col0\" >1.6091</td>\n",
       "      <td id=\"T_0f73b_row3_col1\" class=\"data row3 col1\" >8.5854</td>\n",
       "      <td id=\"T_0f73b_row3_col2\" class=\"data row3 col2\" >2.9301</td>\n",
       "      <td id=\"T_0f73b_row3_col3\" class=\"data row3 col3\" >0.9293</td>\n",
       "      <td id=\"T_0f73b_row3_col4\" class=\"data row3 col4\" >0.4889</td>\n",
       "      <td id=\"T_0f73b_row3_col5\" class=\"data row3 col5\" >4.8893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0f73b_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_0f73b_row4_col0\" class=\"data row4 col0\" >2.6406</td>\n",
       "      <td id=\"T_0f73b_row4_col1\" class=\"data row4 col1\" >79.1854</td>\n",
       "      <td id=\"T_0f73b_row4_col2\" class=\"data row4 col2\" >8.8986</td>\n",
       "      <td id=\"T_0f73b_row4_col3\" class=\"data row4 col3\" >0.8779</td>\n",
       "      <td id=\"T_0f73b_row4_col4\" class=\"data row4 col4\" >0.4693</td>\n",
       "      <td id=\"T_0f73b_row4_col5\" class=\"data row4 col5\" >3.8109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0f73b_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_0f73b_row5_col0\" class=\"data row5 col0\" >1.8067</td>\n",
       "      <td id=\"T_0f73b_row5_col1\" class=\"data row5 col1\" >23.0395</td>\n",
       "      <td id=\"T_0f73b_row5_col2\" class=\"data row5 col2\" >4.7999</td>\n",
       "      <td id=\"T_0f73b_row5_col3\" class=\"data row5 col3\" >0.9082</td>\n",
       "      <td id=\"T_0f73b_row5_col4\" class=\"data row5 col4\" >0.5133</td>\n",
       "      <td id=\"T_0f73b_row5_col5\" class=\"data row5 col5\" >5.2553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0f73b_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_0f73b_row6_col0\" class=\"data row6 col0\" >2.0767</td>\n",
       "      <td id=\"T_0f73b_row6_col1\" class=\"data row6 col1\" >38.2177</td>\n",
       "      <td id=\"T_0f73b_row6_col2\" class=\"data row6 col2\" >6.1820</td>\n",
       "      <td id=\"T_0f73b_row6_col3\" class=\"data row6 col3\" >0.9016</td>\n",
       "      <td id=\"T_0f73b_row6_col4\" class=\"data row6 col4\" >0.4952</td>\n",
       "      <td id=\"T_0f73b_row6_col5\" class=\"data row6 col5\" >4.6822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0f73b_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_0f73b_row7_col0\" class=\"data row7 col0\" >1.6258</td>\n",
       "      <td id=\"T_0f73b_row7_col1\" class=\"data row7 col1\" >18.9040</td>\n",
       "      <td id=\"T_0f73b_row7_col2\" class=\"data row7 col2\" >4.3479</td>\n",
       "      <td id=\"T_0f73b_row7_col3\" class=\"data row7 col3\" >0.9219</td>\n",
       "      <td id=\"T_0f73b_row7_col4\" class=\"data row7 col4\" >0.4822</td>\n",
       "      <td id=\"T_0f73b_row7_col5\" class=\"data row7 col5\" >4.6957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0f73b_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_0f73b_row8_col0\" class=\"data row8 col0\" >2.2113</td>\n",
       "      <td id=\"T_0f73b_row8_col1\" class=\"data row8 col1\" >55.3550</td>\n",
       "      <td id=\"T_0f73b_row8_col2\" class=\"data row8 col2\" >7.4401</td>\n",
       "      <td id=\"T_0f73b_row8_col3\" class=\"data row8 col3\" >0.8529</td>\n",
       "      <td id=\"T_0f73b_row8_col4\" class=\"data row8 col4\" >0.4881</td>\n",
       "      <td id=\"T_0f73b_row8_col5\" class=\"data row8 col5\" >4.8990</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0f73b_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_0f73b_row9_col0\" class=\"data row9 col0\" >1.8833</td>\n",
       "      <td id=\"T_0f73b_row9_col1\" class=\"data row9 col1\" >23.3095</td>\n",
       "      <td id=\"T_0f73b_row9_col2\" class=\"data row9 col2\" >4.8280</td>\n",
       "      <td id=\"T_0f73b_row9_col3\" class=\"data row9 col3\" >0.9257</td>\n",
       "      <td id=\"T_0f73b_row9_col4\" class=\"data row9 col4\" >0.4896</td>\n",
       "      <td id=\"T_0f73b_row9_col5\" class=\"data row9 col5\" >4.3912</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0f73b_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_0f73b_row10_col0\" class=\"data row10 col0\" >1.9856</td>\n",
       "      <td id=\"T_0f73b_row10_col1\" class=\"data row10 col1\" >32.3824</td>\n",
       "      <td id=\"T_0f73b_row10_col2\" class=\"data row10 col2\" >5.4208</td>\n",
       "      <td id=\"T_0f73b_row10_col3\" class=\"data row10 col3\" >0.9056</td>\n",
       "      <td id=\"T_0f73b_row10_col4\" class=\"data row10 col4\" >0.4978</td>\n",
       "      <td id=\"T_0f73b_row10_col5\" class=\"data row10 col5\" >4.7664</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0f73b_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_0f73b_row11_col0\" class=\"data row11 col0\" >0.3074</td>\n",
       "      <td id=\"T_0f73b_row11_col1\" class=\"data row11 col1\" >20.4737</td>\n",
       "      <td id=\"T_0f73b_row11_col2\" class=\"data row11 col2\" >1.7313</td>\n",
       "      <td id=\"T_0f73b_row11_col3\" class=\"data row11 col3\" >0.0225</td>\n",
       "      <td id=\"T_0f73b_row11_col4\" class=\"data row11 col4\" >0.0210</td>\n",
       "      <td id=\"T_0f73b_row11_col5\" class=\"data row11 col5\" >0.4274</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2480958a0d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6b406c7ae4344f559261ccea666d9b9b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/6 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Develop a blended model with voting regressor\n",
    "blended_outdoor = blend_models(estimator_list=[dt_outdoor,rf_outdoor,et_outdoor],optimize='MAE')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this particular case the 2 original individual models decision tree and random forest models have better MAE and R2 compared to the blended model, when extra-tree is limited to a max_depth of 7."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.2.3 Evaluate Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Assume that random forest model is the most appropriate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x550 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot learning Curve\n",
    "plot_model(rf_outdoor, plot = 'learning')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As shown in the learning curve, model training score doesn't increase greatly with with an increase in training instances between. In contrast, cross-validation model scores increase moderately from around 93% to 97%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x550 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot Residuals\n",
    "plot_model(rf_outdoor, plot = 'residuals')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Deaths= 5.360 | Max Deaths= 229.070\n"
     ]
    }
   ],
   "source": [
    "print('Mean Deaths= %.3f | Max Deaths= %.3f'%(round(np.mean(df_in['Deaths_PM2.5']),2),round(np.max(df_in['Deaths_PM2.5']),2)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The residuals plot shows the distribution of residual (delta between actual vs predicted indoor air pollution deaths), which are largely within +/-5 deaths per 100,000, which is quite poor on the individual country level considering the {mean,max} of the metric. Despite this the R2 is around 97%. The highly right-skewed target values may be causing issues."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot Feature Importance - Variable Importance (different from average importance shown in dashbaoard)\n",
    "plot_model(rf_outdoor, plot = 'feature')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Feature Importance plot shows that the Population drives the model at a variable importance of around 0.95, with GDP per capita coming in a distance second at around 0.1. The year does not seem to contribute much, which could be a problem when projecting the model forward into time."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.1.4 Predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_85108\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_85108_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_85108_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_85108_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_85108_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_85108_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_85108_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_85108_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_85108_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_85108_row0_col0\" class=\"data row0 col0\" >Random Forest Regressor</td>\n",
       "      <td id=\"T_85108_row0_col1\" class=\"data row0 col1\" >0.8053</td>\n",
       "      <td id=\"T_85108_row0_col2\" class=\"data row0 col2\" >5.0340</td>\n",
       "      <td id=\"T_85108_row0_col3\" class=\"data row0 col3\" >2.2436</td>\n",
       "      <td id=\"T_85108_row0_col4\" class=\"data row0 col4\" >0.9767</td>\n",
       "      <td id=\"T_85108_row0_col5\" class=\"data row0 col5\" >0.2001</td>\n",
       "      <td id=\"T_85108_row0_col6\" class=\"data row0 col6\" >0.7372</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2480978f460>"
      ]
     },
     "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>GDP_pCapita</th>\n",
       "      <th>Population</th>\n",
       "      <th>SoDI</th>\n",
       "      <th>Income_Class_encoded</th>\n",
       "      <th>Year</th>\n",
       "      <th>Deaths_PM2.5</th>\n",
       "      <th>prediction_label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1025</th>\n",
       "      <td>5386.467773</td>\n",
       "      <td>14587367.0</td>\n",
       "      <td>0.630</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001996</td>\n",
       "      <td>0.182761</td>\n",
       "      <td>0.161715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>935</th>\n",
       "      <td>21227.347656</td>\n",
       "      <td>29610218.0</td>\n",
       "      <td>0.809</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001996</td>\n",
       "      <td>0.067681</td>\n",
       "      <td>0.145434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3024</th>\n",
       "      <td>117254.742188</td>\n",
       "      <td>607950.0</td>\n",
       "      <td>0.894</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002018</td>\n",
       "      <td>2.326820</td>\n",
       "      <td>6.191070</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3584</th>\n",
       "      <td>10171.480469</td>\n",
       "      <td>10289.0</td>\n",
       "      <td>0.573</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002014</td>\n",
       "      <td>94.461243</td>\n",
       "      <td>94.875011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2131</th>\n",
       "      <td>477.762451</td>\n",
       "      <td>7081119.0</td>\n",
       "      <td>0.190</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001994</td>\n",
       "      <td>0.266641</td>\n",
       "      <td>0.265382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>524</th>\n",
       "      <td>2993.033691</td>\n",
       "      <td>230248.0</td>\n",
       "      <td>0.494</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001998</td>\n",
       "      <td>9.620074</td>\n",
       "      <td>7.299584</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5227</th>\n",
       "      <td>590.393494</td>\n",
       "      <td>4413477.0</td>\n",
       "      <td>0.556</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001998</td>\n",
       "      <td>1.620225</td>\n",
       "      <td>1.279303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2883</th>\n",
       "      <td>7948.685547</td>\n",
       "      <td>5537620.0</td>\n",
       "      <td>0.660</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002012</td>\n",
       "      <td>1.067985</td>\n",
       "      <td>1.064601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3941</th>\n",
       "      <td>1664.245728</td>\n",
       "      <td>3494496.0</td>\n",
       "      <td>0.467</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002007</td>\n",
       "      <td>1.166933</td>\n",
       "      <td>0.832874</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2083</th>\n",
       "      <td>26713.058594</td>\n",
       "      <td>158649.0</td>\n",
       "      <td>0.773</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002006</td>\n",
       "      <td>7.151511</td>\n",
       "      <td>7.601745</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1143 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        GDP_pCapita  Population   SoDI  Income_Class_encoded  \\\n",
       "1025    5386.467773  14587367.0  0.630                   2.0   \n",
       "935    21227.347656  29610218.0  0.809                   3.0   \n",
       "3024  117254.742188    607950.0  0.894                   3.0   \n",
       "3584   10171.480469     10289.0  0.573                   3.0   \n",
       "2131     477.762451   7081119.0  0.190                   0.0   \n",
       "...             ...         ...    ...                   ...   \n",
       "524     2993.033691    230248.0  0.494                   1.0   \n",
       "5227     590.393494   4413477.0  0.556                   0.0   \n",
       "2883    7948.685547   5537620.0  0.660                   2.0   \n",
       "3941    1664.245728   3494496.0  0.467                   1.0   \n",
       "2083   26713.058594    158649.0  0.773                   3.0   \n",
       "\n",
       "                              Year  Deaths_PM2.5  prediction_label  \n",
       "1025 1970-01-01 00:00:00.000001996      0.182761          0.161715  \n",
       "935  1970-01-01 00:00:00.000001996      0.067681          0.145434  \n",
       "3024 1970-01-01 00:00:00.000002018      2.326820          6.191070  \n",
       "3584 1970-01-01 00:00:00.000002014     94.461243         94.875011  \n",
       "2131 1970-01-01 00:00:00.000001994      0.266641          0.265382  \n",
       "...                            ...           ...               ...  \n",
       "524  1970-01-01 00:00:00.000001998      9.620074          7.299584  \n",
       "5227 1970-01-01 00:00:00.000001998      1.620225          1.279303  \n",
       "2883 1970-01-01 00:00:00.000002012      1.067985          1.064601  \n",
       "3941 1970-01-01 00:00:00.000002007      1.166933          0.832874  \n",
       "2083 1970-01-01 00:00:00.000002006      7.151511          7.601745  \n",
       "\n",
       "[1143 rows x 7 columns]"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Use random forest model to predict holdout testing dataset indoor pollution deaths per 100,000\n",
    "predictions = predict_model(rf_outdoor)\n",
    "predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Even with the hold-out testing dataset, the tuned model has RMSE,MAE,R2 not drastically different from the test-train dataset used to build the model, so it shows that the model was not overfitted to the input data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Transformation Pipeline and Model Successfully Saved\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(Pipeline(memory=FastMemory(location=C:\\Users\\kevin\\AppData\\Local\\Temp\\joblib),\n",
       "          steps=[('date_feature_extractor',\n",
       "                  TransformerWrapper(include=['Year'],\n",
       "                                     transformer=ExtractDateTimeFeatures())),\n",
       "                 ('numerical_imputer',\n",
       "                  TransformerWrapper(include=['GDP_pCapita', 'Population',\n",
       "                                              'SoDI', 'Income_Class_encoded'],\n",
       "                                     transformer=SimpleImputer())),\n",
       "                 ('categorical_imputer',\n",
       "                  TransformerWrapper(include=[],\n",
       "                                     transformer=SimpleImputer(strategy='most_frequent'))),\n",
       "                 ('trained_model',\n",
       "                  RandomForestRegressor(max_depth=7, n_jobs=-1,\n",
       "                                        random_state=123))]),\n",
       " 'OutdoorPM2.5 Deaths Random Forest 2023-03-29.pkl')"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Save Model\n",
    "save_model(rf_outdoor,'OutdoorPM2.5 Deaths Random Forest 2023-03-29')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.3 Model Selection for Outdoor Pollution - OZone"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we will test models for predictions on outdoor OZone pollution deaths."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Without target transformation:** <br>\n",
    "Predicting deaths due to OZone pollution appears to be more difficult. All MAE/MSE/RMSE/RMSLE values are all close to 0 because the amplitudes of the target is very small. If we scale the target as shown above, the model performance is even worse. Extra Trees Regressor appears to have an acceptable R2 of 0.7-0.8. <br>\n",
    "**With target transformation**: <br>\n",
    "Extra Trees Regressor R2 = 0.65, which is worse."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Dependent variable - classify the level of indoor pollution death\n",
    "y=df_in.loc[:,['Deaths_OZone']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = pd.concat([X,y],axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3.1 Determine Top 3 Models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_7defe_row9_col1, #T_7defe_row16_col1 {\n",
       "  background-color: lightgreen;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_7defe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_7defe_level0_col0\" class=\"col_heading level0 col0\" >Description</th>\n",
       "      <th id=\"T_7defe_level0_col1\" class=\"col_heading level0 col1\" >Value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_7defe_row0_col0\" class=\"data row0 col0\" >Session id</td>\n",
       "      <td id=\"T_7defe_row0_col1\" class=\"data row0 col1\" >123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_7defe_row1_col0\" class=\"data row1 col0\" >Target</td>\n",
       "      <td id=\"T_7defe_row1_col1\" class=\"data row1 col1\" >Deaths_OZone</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_7defe_row2_col0\" class=\"data row2 col0\" >Target type</td>\n",
       "      <td id=\"T_7defe_row2_col1\" class=\"data row2 col1\" >Regression</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_7defe_row3_col0\" class=\"data row3 col0\" >Original data shape</td>\n",
       "      <td id=\"T_7defe_row3_col1\" class=\"data row3 col1\" >(5712, 6)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_7defe_row4_col0\" class=\"data row4 col0\" >Transformed data shape</td>\n",
       "      <td id=\"T_7defe_row4_col1\" class=\"data row4 col1\" >(5712, 8)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_7defe_row5_col0\" class=\"data row5 col0\" >Transformed train set shape</td>\n",
       "      <td id=\"T_7defe_row5_col1\" class=\"data row5 col1\" >(4569, 8)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_7defe_row6_col0\" class=\"data row6 col0\" >Transformed test set shape</td>\n",
       "      <td id=\"T_7defe_row6_col1\" class=\"data row6 col1\" >(1143, 8)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_7defe_row7_col0\" class=\"data row7 col0\" >Numeric features</td>\n",
       "      <td id=\"T_7defe_row7_col1\" class=\"data row7 col1\" >4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_7defe_row8_col0\" class=\"data row8 col0\" >Date features</td>\n",
       "      <td id=\"T_7defe_row8_col1\" class=\"data row8 col1\" >1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_7defe_row9_col0\" class=\"data row9 col0\" >Preprocess</td>\n",
       "      <td id=\"T_7defe_row9_col1\" class=\"data row9 col1\" >True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row10\" class=\"row_heading level0 row10\" >10</th>\n",
       "      <td id=\"T_7defe_row10_col0\" class=\"data row10 col0\" >Imputation type</td>\n",
       "      <td id=\"T_7defe_row10_col1\" class=\"data row10 col1\" >simple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row11\" class=\"row_heading level0 row11\" >11</th>\n",
       "      <td id=\"T_7defe_row11_col0\" class=\"data row11 col0\" >Numeric imputation</td>\n",
       "      <td id=\"T_7defe_row11_col1\" class=\"data row11 col1\" >mean</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row12\" class=\"row_heading level0 row12\" >12</th>\n",
       "      <td id=\"T_7defe_row12_col0\" class=\"data row12 col0\" >Categorical imputation</td>\n",
       "      <td id=\"T_7defe_row12_col1\" class=\"data row12 col1\" >mode</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row13\" class=\"row_heading level0 row13\" >13</th>\n",
       "      <td id=\"T_7defe_row13_col0\" class=\"data row13 col0\" >Fold Generator</td>\n",
       "      <td id=\"T_7defe_row13_col1\" class=\"data row13 col1\" >KFold</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row14\" class=\"row_heading level0 row14\" >14</th>\n",
       "      <td id=\"T_7defe_row14_col0\" class=\"data row14 col0\" >Fold Number</td>\n",
       "      <td id=\"T_7defe_row14_col1\" class=\"data row14 col1\" >10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row15\" class=\"row_heading level0 row15\" >15</th>\n",
       "      <td id=\"T_7defe_row15_col0\" class=\"data row15 col0\" >CPU Jobs</td>\n",
       "      <td id=\"T_7defe_row15_col1\" class=\"data row15 col1\" >-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row16\" class=\"row_heading level0 row16\" >16</th>\n",
       "      <td id=\"T_7defe_row16_col0\" class=\"data row16 col0\" >Use GPU</td>\n",
       "      <td id=\"T_7defe_row16_col1\" class=\"data row16 col1\" >True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row17\" class=\"row_heading level0 row17\" >17</th>\n",
       "      <td id=\"T_7defe_row17_col0\" class=\"data row17 col0\" >Log Experiment</td>\n",
       "      <td id=\"T_7defe_row17_col1\" class=\"data row17 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row18\" class=\"row_heading level0 row18\" >18</th>\n",
       "      <td id=\"T_7defe_row18_col0\" class=\"data row18 col0\" >Experiment Name</td>\n",
       "      <td id=\"T_7defe_row18_col1\" class=\"data row18 col1\" >reg-default-name</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_7defe_level0_row19\" class=\"row_heading level0 row19\" >19</th>\n",
       "      <td id=\"T_7defe_row19_col0\" class=\"data row19 col0\" >USI</td>\n",
       "      <td id=\"T_7defe_row19_col1\" class=\"data row19 col1\" >6632</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x248096d8a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Initialize the PyCaret setup\n",
    "s = setup(data=dataset,target='Deaths_OZone',\n",
    "          numeric_features=num_features,date_features=date_features,train_size=0.8,\n",
    "          use_gpu=True,fold=10,n_jobs=-1,session_id=123) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_6e4f2 th {\n",
       "  text-align: left;\n",
       "}\n",
       "#T_6e4f2_row0_col0, #T_6e4f2_row0_col2, #T_6e4f2_row0_col3, #T_6e4f2_row0_col4, #T_6e4f2_row0_col6, #T_6e4f2_row1_col0, #T_6e4f2_row1_col1, #T_6e4f2_row1_col2, #T_6e4f2_row1_col3, #T_6e4f2_row1_col4, #T_6e4f2_row1_col5, #T_6e4f2_row2_col0, #T_6e4f2_row2_col1, #T_6e4f2_row2_col2, #T_6e4f2_row2_col3, #T_6e4f2_row2_col4, #T_6e4f2_row2_col5, #T_6e4f2_row2_col6, #T_6e4f2_row3_col0, #T_6e4f2_row3_col1, #T_6e4f2_row3_col5, #T_6e4f2_row3_col6, #T_6e4f2_row4_col0, #T_6e4f2_row4_col1, #T_6e4f2_row4_col2, #T_6e4f2_row4_col3, #T_6e4f2_row4_col4, #T_6e4f2_row4_col5, #T_6e4f2_row4_col6, #T_6e4f2_row5_col0, #T_6e4f2_row5_col1, #T_6e4f2_row5_col2, #T_6e4f2_row5_col3, #T_6e4f2_row5_col4, #T_6e4f2_row5_col5, #T_6e4f2_row5_col6, #T_6e4f2_row6_col0, #T_6e4f2_row6_col1, #T_6e4f2_row6_col2, #T_6e4f2_row6_col3, #T_6e4f2_row6_col4, #T_6e4f2_row6_col5, #T_6e4f2_row6_col6, #T_6e4f2_row7_col0, #T_6e4f2_row7_col1, #T_6e4f2_row7_col2, #T_6e4f2_row7_col3, #T_6e4f2_row7_col4, #T_6e4f2_row7_col5, #T_6e4f2_row7_col6, #T_6e4f2_row8_col0, #T_6e4f2_row8_col1, #T_6e4f2_row8_col2, #T_6e4f2_row8_col3, #T_6e4f2_row8_col4, #T_6e4f2_row8_col5, #T_6e4f2_row8_col6, #T_6e4f2_row9_col0, #T_6e4f2_row9_col1, #T_6e4f2_row9_col2, #T_6e4f2_row9_col3, #T_6e4f2_row9_col4, #T_6e4f2_row9_col5, #T_6e4f2_row9_col6, #T_6e4f2_row10_col0, #T_6e4f2_row10_col1, #T_6e4f2_row10_col2, #T_6e4f2_row10_col3, #T_6e4f2_row10_col4, #T_6e4f2_row10_col5, #T_6e4f2_row10_col6, #T_6e4f2_row11_col0, #T_6e4f2_row11_col1, #T_6e4f2_row11_col2, #T_6e4f2_row11_col3, #T_6e4f2_row11_col4, #T_6e4f2_row11_col5, #T_6e4f2_row11_col6, #T_6e4f2_row12_col0, #T_6e4f2_row12_col1, #T_6e4f2_row12_col2, #T_6e4f2_row12_col3, #T_6e4f2_row12_col4, #T_6e4f2_row12_col5, #T_6e4f2_row12_col6, #T_6e4f2_row13_col0, #T_6e4f2_row13_col1, #T_6e4f2_row13_col2, #T_6e4f2_row13_col3, #T_6e4f2_row13_col4, #T_6e4f2_row13_col5, #T_6e4f2_row13_col6, #T_6e4f2_row14_col0, #T_6e4f2_row14_col1, #T_6e4f2_row14_col2, #T_6e4f2_row14_col3, #T_6e4f2_row14_col4, #T_6e4f2_row14_col5, #T_6e4f2_row14_col6, #T_6e4f2_row15_col0, #T_6e4f2_row15_col1, #T_6e4f2_row15_col2, #T_6e4f2_row15_col3, #T_6e4f2_row15_col4, #T_6e4f2_row15_col5, #T_6e4f2_row15_col6, #T_6e4f2_row16_col0, #T_6e4f2_row16_col1, #T_6e4f2_row16_col2, #T_6e4f2_row16_col3, #T_6e4f2_row16_col4, #T_6e4f2_row16_col5, #T_6e4f2_row16_col6 {\n",
       "  text-align: left;\n",
       "}\n",
       "#T_6e4f2_row0_col1, #T_6e4f2_row0_col5, #T_6e4f2_row1_col6, #T_6e4f2_row3_col2, #T_6e4f2_row3_col3, #T_6e4f2_row3_col4 {\n",
       "  text-align: left;\n",
       "  background-color: yellow;\n",
       "}\n",
       "#T_6e4f2_row0_col7, #T_6e4f2_row1_col7, #T_6e4f2_row2_col7, #T_6e4f2_row3_col7, #T_6e4f2_row4_col7, #T_6e4f2_row5_col7, #T_6e4f2_row6_col7, #T_6e4f2_row7_col7, #T_6e4f2_row8_col7, #T_6e4f2_row9_col7, #T_6e4f2_row10_col7, #T_6e4f2_row11_col7, #T_6e4f2_row12_col7, #T_6e4f2_row14_col7, #T_6e4f2_row15_col7, #T_6e4f2_row16_col7 {\n",
       "  text-align: left;\n",
       "  background-color: lightgrey;\n",
       "}\n",
       "#T_6e4f2_row13_col7 {\n",
       "  text-align: left;\n",
       "  background-color: yellow;\n",
       "  background-color: lightgrey;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_6e4f2\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_6e4f2_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_6e4f2_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_6e4f2_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_6e4f2_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_6e4f2_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_6e4f2_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_6e4f2_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "      <th id=\"T_6e4f2_level0_col7\" class=\"col_heading level0 col7\" >TT (Sec)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row0\" class=\"row_heading level0 row0\" >rf</th>\n",
       "      <td id=\"T_6e4f2_row0_col0\" class=\"data row0 col0\" >Random Forest Regressor</td>\n",
       "      <td id=\"T_6e4f2_row0_col1\" class=\"data row0 col1\" >0.0305</td>\n",
       "      <td id=\"T_6e4f2_row0_col2\" class=\"data row0 col2\" >0.0682</td>\n",
       "      <td id=\"T_6e4f2_row0_col3\" class=\"data row0 col3\" >0.2051</td>\n",
       "      <td id=\"T_6e4f2_row0_col4\" class=\"data row0 col4\" >0.8531</td>\n",
       "      <td id=\"T_6e4f2_row0_col5\" class=\"data row0 col5\" >0.0685</td>\n",
       "      <td id=\"T_6e4f2_row0_col6\" class=\"data row0 col6\" >0.7843</td>\n",
       "      <td id=\"T_6e4f2_row0_col7\" class=\"data row0 col7\" >3.8210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row1\" class=\"row_heading level0 row1\" >dt</th>\n",
       "      <td id=\"T_6e4f2_row1_col0\" class=\"data row1 col0\" >Decision Tree Regressor</td>\n",
       "      <td id=\"T_6e4f2_row1_col1\" class=\"data row1 col1\" >0.0308</td>\n",
       "      <td id=\"T_6e4f2_row1_col2\" class=\"data row1 col2\" >0.1158</td>\n",
       "      <td id=\"T_6e4f2_row1_col3\" class=\"data row1 col3\" >0.2738</td>\n",
       "      <td id=\"T_6e4f2_row1_col4\" class=\"data row1 col4\" >0.7371</td>\n",
       "      <td id=\"T_6e4f2_row1_col5\" class=\"data row1 col5\" >0.0974</td>\n",
       "      <td id=\"T_6e4f2_row1_col6\" class=\"data row1 col6\" >0.5801</td>\n",
       "      <td id=\"T_6e4f2_row1_col7\" class=\"data row1 col7\" >2.3920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row2\" class=\"row_heading level0 row2\" >et</th>\n",
       "      <td id=\"T_6e4f2_row2_col0\" class=\"data row2 col0\" >Extra Trees Regressor</td>\n",
       "      <td id=\"T_6e4f2_row2_col1\" class=\"data row2 col1\" >0.0357</td>\n",
       "      <td id=\"T_6e4f2_row2_col2\" class=\"data row2 col2\" >0.0805</td>\n",
       "      <td id=\"T_6e4f2_row2_col3\" class=\"data row2 col3\" >0.2226</td>\n",
       "      <td id=\"T_6e4f2_row2_col4\" class=\"data row2 col4\" >0.8393</td>\n",
       "      <td id=\"T_6e4f2_row2_col5\" class=\"data row2 col5\" >0.0738</td>\n",
       "      <td id=\"T_6e4f2_row2_col6\" class=\"data row2 col6\" >1.2412</td>\n",
       "      <td id=\"T_6e4f2_row2_col7\" class=\"data row2 col7\" >3.0330</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row3\" class=\"row_heading level0 row3\" >knn</th>\n",
       "      <td id=\"T_6e4f2_row3_col0\" class=\"data row3 col0\" >K Neighbors Regressor</td>\n",
       "      <td id=\"T_6e4f2_row3_col1\" class=\"data row3 col1\" >0.0462</td>\n",
       "      <td id=\"T_6e4f2_row3_col2\" class=\"data row3 col2\" >0.0500</td>\n",
       "      <td id=\"T_6e4f2_row3_col3\" class=\"data row3 col3\" >0.1942</td>\n",
       "      <td id=\"T_6e4f2_row3_col4\" class=\"data row3 col4\" >0.8823</td>\n",
       "      <td id=\"T_6e4f2_row3_col5\" class=\"data row3 col5\" >0.0921</td>\n",
       "      <td id=\"T_6e4f2_row3_col6\" class=\"data row3 col6\" >1.9850</td>\n",
       "      <td id=\"T_6e4f2_row3_col7\" class=\"data row3 col7\" >2.9620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row4\" class=\"row_heading level0 row4\" >gbr</th>\n",
       "      <td id=\"T_6e4f2_row4_col0\" class=\"data row4 col0\" >Gradient Boosting Regressor</td>\n",
       "      <td id=\"T_6e4f2_row4_col1\" class=\"data row4 col1\" >0.0598</td>\n",
       "      <td id=\"T_6e4f2_row4_col2\" class=\"data row4 col2\" >0.0996</td>\n",
       "      <td id=\"T_6e4f2_row4_col3\" class=\"data row4 col3\" >0.2551</td>\n",
       "      <td id=\"T_6e4f2_row4_col4\" class=\"data row4 col4\" >0.7820</td>\n",
       "      <td id=\"T_6e4f2_row4_col5\" class=\"data row4 col5\" >0.1001</td>\n",
       "      <td id=\"T_6e4f2_row4_col6\" class=\"data row4 col6\" >2.5866</td>\n",
       "      <td id=\"T_6e4f2_row4_col7\" class=\"data row4 col7\" >3.2160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row5\" class=\"row_heading level0 row5\" >ada</th>\n",
       "      <td id=\"T_6e4f2_row5_col0\" class=\"data row5 col0\" >AdaBoost Regressor</td>\n",
       "      <td id=\"T_6e4f2_row5_col1\" class=\"data row5 col1\" >0.0950</td>\n",
       "      <td id=\"T_6e4f2_row5_col2\" class=\"data row5 col2\" >0.1196</td>\n",
       "      <td id=\"T_6e4f2_row5_col3\" class=\"data row5 col3\" >0.2983</td>\n",
       "      <td id=\"T_6e4f2_row5_col4\" class=\"data row5 col4\" >0.7475</td>\n",
       "      <td id=\"T_6e4f2_row5_col5\" class=\"data row5 col5\" >0.1313</td>\n",
       "      <td id=\"T_6e4f2_row5_col6\" class=\"data row5 col6\" >17.0098</td>\n",
       "      <td id=\"T_6e4f2_row5_col7\" class=\"data row5 col7\" >2.6670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row6\" class=\"row_heading level0 row6\" >huber</th>\n",
       "      <td id=\"T_6e4f2_row6_col0\" class=\"data row6 col0\" >Huber Regressor</td>\n",
       "      <td id=\"T_6e4f2_row6_col1\" class=\"data row6 col1\" >0.1797</td>\n",
       "      <td id=\"T_6e4f2_row6_col2\" class=\"data row6 col2\" >0.5137</td>\n",
       "      <td id=\"T_6e4f2_row6_col3\" class=\"data row6 col3\" >0.6994</td>\n",
       "      <td id=\"T_6e4f2_row6_col4\" class=\"data row6 col4\" >0.0305</td>\n",
       "      <td id=\"T_6e4f2_row6_col5\" class=\"data row6 col5\" >0.2387</td>\n",
       "      <td id=\"T_6e4f2_row6_col6\" class=\"data row6 col6\" >25.7144</td>\n",
       "      <td id=\"T_6e4f2_row6_col7\" class=\"data row6 col7\" >2.4580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row7\" class=\"row_heading level0 row7\" >lar</th>\n",
       "      <td id=\"T_6e4f2_row7_col0\" class=\"data row7 col0\" >Least Angle Regression</td>\n",
       "      <td id=\"T_6e4f2_row7_col1\" class=\"data row7 col1\" >0.2084</td>\n",
       "      <td id=\"T_6e4f2_row7_col2\" class=\"data row7 col2\" >0.4939</td>\n",
       "      <td id=\"T_6e4f2_row7_col3\" class=\"data row7 col3\" >0.6870</td>\n",
       "      <td id=\"T_6e4f2_row7_col4\" class=\"data row7 col4\" >0.0607</td>\n",
       "      <td id=\"T_6e4f2_row7_col5\" class=\"data row7 col5\" >0.2530</td>\n",
       "      <td id=\"T_6e4f2_row7_col6\" class=\"data row7 col6\" >30.6889</td>\n",
       "      <td id=\"T_6e4f2_row7_col7\" class=\"data row7 col7\" >2.4930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row8\" class=\"row_heading level0 row8\" >ridge</th>\n",
       "      <td id=\"T_6e4f2_row8_col0\" class=\"data row8 col0\" >Ridge Regression</td>\n",
       "      <td id=\"T_6e4f2_row8_col1\" class=\"data row8 col1\" >0.2084</td>\n",
       "      <td id=\"T_6e4f2_row8_col2\" class=\"data row8 col2\" >0.4939</td>\n",
       "      <td id=\"T_6e4f2_row8_col3\" class=\"data row8 col3\" >0.6870</td>\n",
       "      <td id=\"T_6e4f2_row8_col4\" class=\"data row8 col4\" >0.0608</td>\n",
       "      <td id=\"T_6e4f2_row8_col5\" class=\"data row8 col5\" >0.2530</td>\n",
       "      <td id=\"T_6e4f2_row8_col6\" class=\"data row8 col6\" >30.6878</td>\n",
       "      <td id=\"T_6e4f2_row8_col7\" class=\"data row8 col7\" >2.9150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row9\" class=\"row_heading level0 row9\" >lr</th>\n",
       "      <td id=\"T_6e4f2_row9_col0\" class=\"data row9 col0\" >Linear Regression</td>\n",
       "      <td id=\"T_6e4f2_row9_col1\" class=\"data row9 col1\" >0.2084</td>\n",
       "      <td id=\"T_6e4f2_row9_col2\" class=\"data row9 col2\" >0.4939</td>\n",
       "      <td id=\"T_6e4f2_row9_col3\" class=\"data row9 col3\" >0.6870</td>\n",
       "      <td id=\"T_6e4f2_row9_col4\" class=\"data row9 col4\" >0.0607</td>\n",
       "      <td id=\"T_6e4f2_row9_col5\" class=\"data row9 col5\" >0.2530</td>\n",
       "      <td id=\"T_6e4f2_row9_col6\" class=\"data row9 col6\" >30.6889</td>\n",
       "      <td id=\"T_6e4f2_row9_col7\" class=\"data row9 col7\" >2.4520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row10\" class=\"row_heading level0 row10\" >lasso</th>\n",
       "      <td id=\"T_6e4f2_row10_col0\" class=\"data row10 col0\" >Lasso Regression</td>\n",
       "      <td id=\"T_6e4f2_row10_col1\" class=\"data row10 col1\" >0.2089</td>\n",
       "      <td id=\"T_6e4f2_row10_col2\" class=\"data row10 col2\" >0.4936</td>\n",
       "      <td id=\"T_6e4f2_row10_col3\" class=\"data row10 col3\" >0.6867</td>\n",
       "      <td id=\"T_6e4f2_row10_col4\" class=\"data row10 col4\" >0.0620</td>\n",
       "      <td id=\"T_6e4f2_row10_col5\" class=\"data row10 col5\" >0.2533</td>\n",
       "      <td id=\"T_6e4f2_row10_col6\" class=\"data row10 col6\" >28.8511</td>\n",
       "      <td id=\"T_6e4f2_row10_col7\" class=\"data row10 col7\" >2.6360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row11\" class=\"row_heading level0 row11\" >en</th>\n",
       "      <td id=\"T_6e4f2_row11_col0\" class=\"data row11 col0\" >Elastic Net</td>\n",
       "      <td id=\"T_6e4f2_row11_col1\" class=\"data row11 col1\" >0.2089</td>\n",
       "      <td id=\"T_6e4f2_row11_col2\" class=\"data row11 col2\" >0.4936</td>\n",
       "      <td id=\"T_6e4f2_row11_col3\" class=\"data row11 col3\" >0.6867</td>\n",
       "      <td id=\"T_6e4f2_row11_col4\" class=\"data row11 col4\" >0.0620</td>\n",
       "      <td id=\"T_6e4f2_row11_col5\" class=\"data row11 col5\" >0.2533</td>\n",
       "      <td id=\"T_6e4f2_row11_col6\" class=\"data row11 col6\" >28.8482</td>\n",
       "      <td id=\"T_6e4f2_row11_col7\" class=\"data row11 col7\" >2.6080</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row12\" class=\"row_heading level0 row12\" >llar</th>\n",
       "      <td id=\"T_6e4f2_row12_col0\" class=\"data row12 col0\" >Lasso Least Angle Regression</td>\n",
       "      <td id=\"T_6e4f2_row12_col1\" class=\"data row12 col1\" >0.2089</td>\n",
       "      <td id=\"T_6e4f2_row12_col2\" class=\"data row12 col2\" >0.4936</td>\n",
       "      <td id=\"T_6e4f2_row12_col3\" class=\"data row12 col3\" >0.6867</td>\n",
       "      <td id=\"T_6e4f2_row12_col4\" class=\"data row12 col4\" >0.0620</td>\n",
       "      <td id=\"T_6e4f2_row12_col5\" class=\"data row12 col5\" >0.2533</td>\n",
       "      <td id=\"T_6e4f2_row12_col6\" class=\"data row12 col6\" >28.8511</td>\n",
       "      <td id=\"T_6e4f2_row12_col7\" class=\"data row12 col7\" >2.7230</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row13\" class=\"row_heading level0 row13\" >br</th>\n",
       "      <td id=\"T_6e4f2_row13_col0\" class=\"data row13 col0\" >Bayesian Ridge</td>\n",
       "      <td id=\"T_6e4f2_row13_col1\" class=\"data row13 col1\" >0.2089</td>\n",
       "      <td id=\"T_6e4f2_row13_col2\" class=\"data row13 col2\" >0.4936</td>\n",
       "      <td id=\"T_6e4f2_row13_col3\" class=\"data row13 col3\" >0.6867</td>\n",
       "      <td id=\"T_6e4f2_row13_col4\" class=\"data row13 col4\" >0.0620</td>\n",
       "      <td id=\"T_6e4f2_row13_col5\" class=\"data row13 col5\" >0.2533</td>\n",
       "      <td id=\"T_6e4f2_row13_col6\" class=\"data row13 col6\" >28.8536</td>\n",
       "      <td id=\"T_6e4f2_row13_col7\" class=\"data row13 col7\" >2.2400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row14\" class=\"row_heading level0 row14\" >dummy</th>\n",
       "      <td id=\"T_6e4f2_row14_col0\" class=\"data row14 col0\" >Dummy Regressor</td>\n",
       "      <td id=\"T_6e4f2_row14_col1\" class=\"data row14 col1\" >0.2154</td>\n",
       "      <td id=\"T_6e4f2_row14_col2\" class=\"data row14 col2\" >0.5289</td>\n",
       "      <td id=\"T_6e4f2_row14_col3\" class=\"data row14 col3\" >0.7103</td>\n",
       "      <td id=\"T_6e4f2_row14_col4\" class=\"data row14 col4\" >-0.0022</td>\n",
       "      <td id=\"T_6e4f2_row14_col5\" class=\"data row14 col5\" >0.2533</td>\n",
       "      <td id=\"T_6e4f2_row14_col6\" class=\"data row14 col6\" >53.2468</td>\n",
       "      <td id=\"T_6e4f2_row14_col7\" class=\"data row14 col7\" >2.2790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row15\" class=\"row_heading level0 row15\" >omp</th>\n",
       "      <td id=\"T_6e4f2_row15_col0\" class=\"data row15 col0\" >Orthogonal Matching Pursuit</td>\n",
       "      <td id=\"T_6e4f2_row15_col1\" class=\"data row15 col1\" >0.2177</td>\n",
       "      <td id=\"T_6e4f2_row15_col2\" class=\"data row15 col2\" >0.5278</td>\n",
       "      <td id=\"T_6e4f2_row15_col3\" class=\"data row15 col3\" >0.7096</td>\n",
       "      <td id=\"T_6e4f2_row15_col4\" class=\"data row15 col4\" >0.0001</td>\n",
       "      <td id=\"T_6e4f2_row15_col5\" class=\"data row15 col5\" >0.2531</td>\n",
       "      <td id=\"T_6e4f2_row15_col6\" class=\"data row15 col6\" >51.1778</td>\n",
       "      <td id=\"T_6e4f2_row15_col7\" class=\"data row15 col7\" >3.1160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_6e4f2_level0_row16\" class=\"row_heading level0 row16\" >par</th>\n",
       "      <td id=\"T_6e4f2_row16_col0\" class=\"data row16 col0\" >Passive Aggressive Regressor</td>\n",
       "      <td id=\"T_6e4f2_row16_col1\" class=\"data row16 col1\" >14.2480</td>\n",
       "      <td id=\"T_6e4f2_row16_col2\" class=\"data row16 col2\" >13256.5484</td>\n",
       "      <td id=\"T_6e4f2_row16_col3\" class=\"data row16 col3\" >55.6178</td>\n",
       "      <td id=\"T_6e4f2_row16_col4\" class=\"data row16 col4\" >-37865.3985</td>\n",
       "      <td id=\"T_6e4f2_row16_col5\" class=\"data row16 col5\" >1.4325</td>\n",
       "      <td id=\"T_6e4f2_row16_col6\" class=\"data row16 col6\" >18658.7069</td>\n",
       "      <td id=\"T_6e4f2_row16_col7\" class=\"data row16 col7\" >2.3790</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2482a9640d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dd7c7040f09d4ffb9c99294bbec1eeab",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/51 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Multi-model training/test - Compare Mean Absolute Error\n",
    "# Select the best 3 models - Create a blended model this time\n",
    "best_3_ozone = compare_models(sort='MAE',budget_time=30,fold=10,exclude='lightgbm',n_select=3) \n",
    "# 30 mins budget time\n",
    "# remove Light gradient boosting - causes error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[RandomForestRegressor(n_jobs=-1, random_state=123),\n",
       " DecisionTreeRegressor(random_state=123),\n",
       " ExtraTreesRegressor(n_jobs=-1, random_state=123)]"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_3_ozone"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'ccp_alpha': 0.0,\n",
       " 'criterion': 'squared_error',\n",
       " 'max_depth': None,\n",
       " 'max_features': None,\n",
       " 'max_leaf_nodes': None,\n",
       " 'min_impurity_decrease': 0.0,\n",
       " 'min_samples_leaf': 1,\n",
       " 'min_samples_split': 2,\n",
       " 'min_weight_fraction_leaf': 0.0,\n",
       " 'random_state': 123,\n",
       " 'splitter': 'best'}"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check the model parameters in top 3\n",
    "best_3_ozone[1].get_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3.2 Blended Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_5772f_row10_col0, #T_5772f_row10_col1, #T_5772f_row10_col2, #T_5772f_row10_col3, #T_5772f_row10_col4, #T_5772f_row10_col5 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_5772f\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_5772f_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_5772f_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_5772f_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_5772f_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_5772f_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_5772f_level0_col5\" class=\"col_heading level0 col5\" >MAPE</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_5772f_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_5772f_row0_col0\" class=\"data row0 col0\" >0.0353</td>\n",
       "      <td id=\"T_5772f_row0_col1\" class=\"data row0 col1\" >0.0110</td>\n",
       "      <td id=\"T_5772f_row0_col2\" class=\"data row0 col2\" >0.1049</td>\n",
       "      <td id=\"T_5772f_row0_col3\" class=\"data row0 col3\" >0.9837</td>\n",
       "      <td id=\"T_5772f_row0_col4\" class=\"data row0 col4\" >0.0732</td>\n",
       "      <td id=\"T_5772f_row0_col5\" class=\"data row0 col5\" >1.5804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_5772f_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_5772f_row1_col0\" class=\"data row1 col0\" >0.0505</td>\n",
       "      <td id=\"T_5772f_row1_col1\" class=\"data row1 col1\" >0.0446</td>\n",
       "      <td id=\"T_5772f_row1_col2\" class=\"data row1 col2\" >0.2111</td>\n",
       "      <td id=\"T_5772f_row1_col3\" class=\"data row1 col3\" >0.8435</td>\n",
       "      <td id=\"T_5772f_row1_col4\" class=\"data row1 col4\" >0.1181</td>\n",
       "      <td id=\"T_5772f_row1_col5\" class=\"data row1 col5\" >1.6742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_5772f_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_5772f_row2_col0\" class=\"data row2 col0\" >0.0269</td>\n",
       "      <td id=\"T_5772f_row2_col1\" class=\"data row2 col1\" >0.0093</td>\n",
       "      <td id=\"T_5772f_row2_col2\" class=\"data row2 col2\" >0.0967</td>\n",
       "      <td id=\"T_5772f_row2_col3\" class=\"data row2 col3\" >0.9741</td>\n",
       "      <td id=\"T_5772f_row2_col4\" class=\"data row2 col4\" >0.0654</td>\n",
       "      <td id=\"T_5772f_row2_col5\" class=\"data row2 col5\" >1.5017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_5772f_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_5772f_row3_col0\" class=\"data row3 col0\" >0.0415</td>\n",
       "      <td id=\"T_5772f_row3_col1\" class=\"data row3 col1\" >0.0282</td>\n",
       "      <td id=\"T_5772f_row3_col2\" class=\"data row3 col2\" >0.1680</td>\n",
       "      <td id=\"T_5772f_row3_col3\" class=\"data row3 col3\" >0.9666</td>\n",
       "      <td id=\"T_5772f_row3_col4\" class=\"data row3 col4\" >0.0872</td>\n",
       "      <td id=\"T_5772f_row3_col5\" class=\"data row3 col5\" >1.1859</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_5772f_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_5772f_row4_col0\" class=\"data row4 col0\" >0.0603</td>\n",
       "      <td id=\"T_5772f_row4_col1\" class=\"data row4 col1\" >0.0937</td>\n",
       "      <td id=\"T_5772f_row4_col2\" class=\"data row4 col2\" >0.3060</td>\n",
       "      <td id=\"T_5772f_row4_col3\" class=\"data row4 col3\" >0.8362</td>\n",
       "      <td id=\"T_5772f_row4_col4\" class=\"data row4 col4\" >0.1398</td>\n",
       "      <td id=\"T_5772f_row4_col5\" class=\"data row4 col5\" >1.8607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_5772f_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_5772f_row5_col0\" class=\"data row5 col0\" >0.0295</td>\n",
       "      <td id=\"T_5772f_row5_col1\" class=\"data row5 col1\" >0.0148</td>\n",
       "      <td id=\"T_5772f_row5_col2\" class=\"data row5 col2\" >0.1217</td>\n",
       "      <td id=\"T_5772f_row5_col3\" class=\"data row5 col3\" >0.9655</td>\n",
       "      <td id=\"T_5772f_row5_col4\" class=\"data row5 col4\" >0.0708</td>\n",
       "      <td id=\"T_5772f_row5_col5\" class=\"data row5 col5\" >1.4705</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_5772f_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_5772f_row6_col0\" class=\"data row6 col0\" >0.0488</td>\n",
       "      <td id=\"T_5772f_row6_col1\" class=\"data row6 col1\" >0.0246</td>\n",
       "      <td id=\"T_5772f_row6_col2\" class=\"data row6 col2\" >0.1569</td>\n",
       "      <td id=\"T_5772f_row6_col3\" class=\"data row6 col3\" >0.9695</td>\n",
       "      <td id=\"T_5772f_row6_col4\" class=\"data row6 col4\" >0.1069</td>\n",
       "      <td id=\"T_5772f_row6_col5\" class=\"data row6 col5\" >1.8135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_5772f_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_5772f_row7_col0\" class=\"data row7 col0\" >0.0487</td>\n",
       "      <td id=\"T_5772f_row7_col1\" class=\"data row7 col1\" >0.0899</td>\n",
       "      <td id=\"T_5772f_row7_col2\" class=\"data row7 col2\" >0.2999</td>\n",
       "      <td id=\"T_5772f_row7_col3\" class=\"data row7 col3\" >0.5672</td>\n",
       "      <td id=\"T_5772f_row7_col4\" class=\"data row7 col4\" >0.1305</td>\n",
       "      <td id=\"T_5772f_row7_col5\" class=\"data row7 col5\" >1.6167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_5772f_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_5772f_row8_col0\" class=\"data row8 col0\" >0.0692</td>\n",
       "      <td id=\"T_5772f_row8_col1\" class=\"data row8 col1\" >0.4483</td>\n",
       "      <td id=\"T_5772f_row8_col2\" class=\"data row8 col2\" >0.6695</td>\n",
       "      <td id=\"T_5772f_row8_col3\" class=\"data row8 col3\" >0.4019</td>\n",
       "      <td id=\"T_5772f_row8_col4\" class=\"data row8 col4\" >0.1471</td>\n",
       "      <td id=\"T_5772f_row8_col5\" class=\"data row8 col5\" >1.0859</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_5772f_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_5772f_row9_col0\" class=\"data row9 col0\" >0.0675</td>\n",
       "      <td id=\"T_5772f_row9_col1\" class=\"data row9 col1\" >0.4190</td>\n",
       "      <td id=\"T_5772f_row9_col2\" class=\"data row9 col2\" >0.6473</td>\n",
       "      <td id=\"T_5772f_row9_col3\" class=\"data row9 col3\" >-0.2101</td>\n",
       "      <td id=\"T_5772f_row9_col4\" class=\"data row9 col4\" >0.1451</td>\n",
       "      <td id=\"T_5772f_row9_col5\" class=\"data row9 col5\" >1.6548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_5772f_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_5772f_row10_col0\" class=\"data row10 col0\" >0.0478</td>\n",
       "      <td id=\"T_5772f_row10_col1\" class=\"data row10 col1\" >0.1183</td>\n",
       "      <td id=\"T_5772f_row10_col2\" class=\"data row10 col2\" >0.2782</td>\n",
       "      <td id=\"T_5772f_row10_col3\" class=\"data row10 col3\" >0.7298</td>\n",
       "      <td id=\"T_5772f_row10_col4\" class=\"data row10 col4\" >0.1084</td>\n",
       "      <td id=\"T_5772f_row10_col5\" class=\"data row10 col5\" >1.5444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_5772f_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_5772f_row11_col0\" class=\"data row11 col0\" >0.0141</td>\n",
       "      <td id=\"T_5772f_row11_col1\" class=\"data row11 col1\" >0.1604</td>\n",
       "      <td id=\"T_5772f_row11_col2\" class=\"data row11 col2\" >0.2023</td>\n",
       "      <td id=\"T_5772f_row11_col3\" class=\"data row11 col3\" >0.3654</td>\n",
       "      <td id=\"T_5772f_row11_col4\" class=\"data row11 col4\" >0.0306</td>\n",
       "      <td id=\"T_5772f_row11_col5\" class=\"data row11 col5\" >0.2354</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2487e58ae80>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9d76339f283a4565a3f87ac5a12c38fe",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/4 [00:00<?, ?it/s]"
      ]
     },
     "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_873f6_row10_col0, #T_873f6_row10_col1, #T_873f6_row10_col2, #T_873f6_row10_col3, #T_873f6_row10_col4, #T_873f6_row10_col5 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_873f6\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_873f6_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_873f6_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_873f6_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_873f6_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_873f6_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_873f6_level0_col5\" class=\"col_heading level0 col5\" >MAPE</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_873f6_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_873f6_row0_col0\" class=\"data row0 col0\" >0.0382</td>\n",
       "      <td id=\"T_873f6_row0_col1\" class=\"data row0 col1\" >0.0120</td>\n",
       "      <td id=\"T_873f6_row0_col2\" class=\"data row0 col2\" >0.1097</td>\n",
       "      <td id=\"T_873f6_row0_col3\" class=\"data row0 col3\" >0.9822</td>\n",
       "      <td id=\"T_873f6_row0_col4\" class=\"data row0 col4\" >0.0699</td>\n",
       "      <td id=\"T_873f6_row0_col5\" class=\"data row0 col5\" >1.5989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_873f6_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_873f6_row1_col0\" class=\"data row1 col0\" >0.0494</td>\n",
       "      <td id=\"T_873f6_row1_col1\" class=\"data row1 col1\" >0.0259</td>\n",
       "      <td id=\"T_873f6_row1_col2\" class=\"data row1 col2\" >0.1610</td>\n",
       "      <td id=\"T_873f6_row1_col3\" class=\"data row1 col3\" >0.9090</td>\n",
       "      <td id=\"T_873f6_row1_col4\" class=\"data row1 col4\" >0.0917</td>\n",
       "      <td id=\"T_873f6_row1_col5\" class=\"data row1 col5\" >1.7343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_873f6_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_873f6_row2_col0\" class=\"data row2 col0\" >0.0318</td>\n",
       "      <td id=\"T_873f6_row2_col1\" class=\"data row2 col1\" >0.0082</td>\n",
       "      <td id=\"T_873f6_row2_col2\" class=\"data row2 col2\" >0.0905</td>\n",
       "      <td id=\"T_873f6_row2_col3\" class=\"data row2 col3\" >0.9773</td>\n",
       "      <td id=\"T_873f6_row2_col4\" class=\"data row2 col4\" >0.0641</td>\n",
       "      <td id=\"T_873f6_row2_col5\" class=\"data row2 col5\" >1.6632</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_873f6_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_873f6_row3_col0\" class=\"data row3 col0\" >0.0419</td>\n",
       "      <td id=\"T_873f6_row3_col1\" class=\"data row3 col1\" >0.0185</td>\n",
       "      <td id=\"T_873f6_row3_col2\" class=\"data row3 col2\" >0.1358</td>\n",
       "      <td id=\"T_873f6_row3_col3\" class=\"data row3 col3\" >0.9782</td>\n",
       "      <td id=\"T_873f6_row3_col4\" class=\"data row3 col4\" >0.0705</td>\n",
       "      <td id=\"T_873f6_row3_col5\" class=\"data row3 col5\" >1.3048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_873f6_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_873f6_row4_col0\" class=\"data row4 col0\" >0.0485</td>\n",
       "      <td id=\"T_873f6_row4_col1\" class=\"data row4 col1\" >0.0329</td>\n",
       "      <td id=\"T_873f6_row4_col2\" class=\"data row4 col2\" >0.1813</td>\n",
       "      <td id=\"T_873f6_row4_col3\" class=\"data row4 col3\" >0.9425</td>\n",
       "      <td id=\"T_873f6_row4_col4\" class=\"data row4 col4\" >0.0796</td>\n",
       "      <td id=\"T_873f6_row4_col5\" class=\"data row4 col5\" >1.9051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_873f6_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_873f6_row5_col0\" class=\"data row5 col0\" >0.0442</td>\n",
       "      <td id=\"T_873f6_row5_col1\" class=\"data row5 col1\" >0.0553</td>\n",
       "      <td id=\"T_873f6_row5_col2\" class=\"data row5 col2\" >0.2352</td>\n",
       "      <td id=\"T_873f6_row5_col3\" class=\"data row5 col3\" >0.8712</td>\n",
       "      <td id=\"T_873f6_row5_col4\" class=\"data row5 col4\" >0.0693</td>\n",
       "      <td id=\"T_873f6_row5_col5\" class=\"data row5 col5\" >1.6845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_873f6_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_873f6_row6_col0\" class=\"data row6 col0\" >0.0461</td>\n",
       "      <td id=\"T_873f6_row6_col1\" class=\"data row6 col1\" >0.0157</td>\n",
       "      <td id=\"T_873f6_row6_col2\" class=\"data row6 col2\" >0.1255</td>\n",
       "      <td id=\"T_873f6_row6_col3\" class=\"data row6 col3\" >0.9805</td>\n",
       "      <td id=\"T_873f6_row6_col4\" class=\"data row6 col4\" >0.0907</td>\n",
       "      <td id=\"T_873f6_row6_col5\" class=\"data row6 col5\" >1.6854</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_873f6_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_873f6_row7_col0\" class=\"data row7 col0\" >0.0332</td>\n",
       "      <td id=\"T_873f6_row7_col1\" class=\"data row7 col1\" >0.0117</td>\n",
       "      <td id=\"T_873f6_row7_col2\" class=\"data row7 col2\" >0.1082</td>\n",
       "      <td id=\"T_873f6_row7_col3\" class=\"data row7 col3\" >0.9436</td>\n",
       "      <td id=\"T_873f6_row7_col4\" class=\"data row7 col4\" >0.0662</td>\n",
       "      <td id=\"T_873f6_row7_col5\" class=\"data row7 col5\" >1.4661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_873f6_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_873f6_row8_col0\" class=\"data row8 col0\" >0.0590</td>\n",
       "      <td id=\"T_873f6_row8_col1\" class=\"data row8 col1\" >0.2573</td>\n",
       "      <td id=\"T_873f6_row8_col2\" class=\"data row8 col2\" >0.5072</td>\n",
       "      <td id=\"T_873f6_row8_col3\" class=\"data row8 col3\" >0.6567</td>\n",
       "      <td id=\"T_873f6_row8_col4\" class=\"data row8 col4\" >0.0917</td>\n",
       "      <td id=\"T_873f6_row8_col5\" class=\"data row8 col5\" >1.0293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_873f6_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_873f6_row9_col0\" class=\"data row9 col0\" >0.0635</td>\n",
       "      <td id=\"T_873f6_row9_col1\" class=\"data row9 col1\" >0.2643</td>\n",
       "      <td id=\"T_873f6_row9_col2\" class=\"data row9 col2\" >0.5141</td>\n",
       "      <td id=\"T_873f6_row9_col3\" class=\"data row9 col3\" >0.2368</td>\n",
       "      <td id=\"T_873f6_row9_col4\" class=\"data row9 col4\" >0.1339</td>\n",
       "      <td id=\"T_873f6_row9_col5\" class=\"data row9 col5\" >1.5704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_873f6_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_873f6_row10_col0\" class=\"data row10 col0\" >0.0456</td>\n",
       "      <td id=\"T_873f6_row10_col1\" class=\"data row10 col1\" >0.0702</td>\n",
       "      <td id=\"T_873f6_row10_col2\" class=\"data row10 col2\" >0.2168</td>\n",
       "      <td id=\"T_873f6_row10_col3\" class=\"data row10 col3\" >0.8478</td>\n",
       "      <td id=\"T_873f6_row10_col4\" class=\"data row10 col4\" >0.0828</td>\n",
       "      <td id=\"T_873f6_row10_col5\" class=\"data row10 col5\" >1.5642</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_873f6_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_873f6_row11_col0\" class=\"data row11 col0\" >0.0097</td>\n",
       "      <td id=\"T_873f6_row11_col1\" class=\"data row11 col1\" >0.0962</td>\n",
       "      <td id=\"T_873f6_row11_col2\" class=\"data row11 col2\" >0.1522</td>\n",
       "      <td id=\"T_873f6_row11_col3\" class=\"data row11 col3\" >0.2240</td>\n",
       "      <td id=\"T_873f6_row11_col4\" class=\"data row11 col4\" >0.0199</td>\n",
       "      <td id=\"T_873f6_row11_col5\" class=\"data row11 col5\" >0.2342</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2487e881cd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "79318362961a46248da163f37ec33ee9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/4 [00:00<?, ?it/s]"
      ]
     },
     "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_ac2c6_row10_col0, #T_ac2c6_row10_col1, #T_ac2c6_row10_col2, #T_ac2c6_row10_col3, #T_ac2c6_row10_col4, #T_ac2c6_row10_col5 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_ac2c6\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_ac2c6_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_ac2c6_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_ac2c6_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_ac2c6_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_ac2c6_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_ac2c6_level0_col5\" class=\"col_heading level0 col5\" >MAPE</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_ac2c6_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_ac2c6_row0_col0\" class=\"data row0 col0\" >0.1447</td>\n",
       "      <td id=\"T_ac2c6_row0_col1\" class=\"data row0 col1\" >0.2753</td>\n",
       "      <td id=\"T_ac2c6_row0_col2\" class=\"data row0 col2\" >0.5247</td>\n",
       "      <td id=\"T_ac2c6_row0_col3\" class=\"data row0 col3\" >0.5929</td>\n",
       "      <td id=\"T_ac2c6_row0_col4\" class=\"data row0 col4\" >0.1617</td>\n",
       "      <td id=\"T_ac2c6_row0_col5\" class=\"data row0 col5\" >13.1139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ac2c6_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_ac2c6_row1_col0\" class=\"data row1 col0\" >0.1197</td>\n",
       "      <td id=\"T_ac2c6_row1_col1\" class=\"data row1 col1\" >0.0903</td>\n",
       "      <td id=\"T_ac2c6_row1_col2\" class=\"data row1 col2\" >0.3005</td>\n",
       "      <td id=\"T_ac2c6_row1_col3\" class=\"data row1 col3\" >0.6828</td>\n",
       "      <td id=\"T_ac2c6_row1_col4\" class=\"data row1 col4\" >0.1550</td>\n",
       "      <td id=\"T_ac2c6_row1_col5\" class=\"data row1 col5\" >12.3724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ac2c6_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_ac2c6_row2_col0\" class=\"data row2 col0\" >0.1307</td>\n",
       "      <td id=\"T_ac2c6_row2_col1\" class=\"data row2 col1\" >0.1455</td>\n",
       "      <td id=\"T_ac2c6_row2_col2\" class=\"data row2 col2\" >0.3814</td>\n",
       "      <td id=\"T_ac2c6_row2_col3\" class=\"data row2 col3\" >0.5976</td>\n",
       "      <td id=\"T_ac2c6_row2_col4\" class=\"data row2 col4\" >0.1686</td>\n",
       "      <td id=\"T_ac2c6_row2_col5\" class=\"data row2 col5\" >14.2388</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ac2c6_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_ac2c6_row3_col0\" class=\"data row3 col0\" >0.1583</td>\n",
       "      <td id=\"T_ac2c6_row3_col1\" class=\"data row3 col1\" >0.3035</td>\n",
       "      <td id=\"T_ac2c6_row3_col2\" class=\"data row3 col2\" >0.5509</td>\n",
       "      <td id=\"T_ac2c6_row3_col3\" class=\"data row3 col3\" >0.6410</td>\n",
       "      <td id=\"T_ac2c6_row3_col4\" class=\"data row3 col4\" >0.1735</td>\n",
       "      <td id=\"T_ac2c6_row3_col5\" class=\"data row3 col5\" >11.7671</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ac2c6_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_ac2c6_row4_col0\" class=\"data row4 col0\" >0.1365</td>\n",
       "      <td id=\"T_ac2c6_row4_col1\" class=\"data row4 col1\" >0.1574</td>\n",
       "      <td id=\"T_ac2c6_row4_col2\" class=\"data row4 col2\" >0.3967</td>\n",
       "      <td id=\"T_ac2c6_row4_col3\" class=\"data row4 col3\" >0.7247</td>\n",
       "      <td id=\"T_ac2c6_row4_col4\" class=\"data row4 col4\" >0.1563</td>\n",
       "      <td id=\"T_ac2c6_row4_col5\" class=\"data row4 col5\" >14.4973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ac2c6_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_ac2c6_row5_col0\" class=\"data row5 col0\" >0.1207</td>\n",
       "      <td id=\"T_ac2c6_row5_col1\" class=\"data row5 col1\" >0.2396</td>\n",
       "      <td id=\"T_ac2c6_row5_col2\" class=\"data row5 col2\" >0.4895</td>\n",
       "      <td id=\"T_ac2c6_row5_col3\" class=\"data row5 col3\" >0.4421</td>\n",
       "      <td id=\"T_ac2c6_row5_col4\" class=\"data row5 col4\" >0.1534</td>\n",
       "      <td id=\"T_ac2c6_row5_col5\" class=\"data row5 col5\" >12.5807</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ac2c6_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_ac2c6_row6_col0\" class=\"data row6 col0\" >0.1560</td>\n",
       "      <td id=\"T_ac2c6_row6_col1\" class=\"data row6 col1\" >0.3620</td>\n",
       "      <td id=\"T_ac2c6_row6_col2\" class=\"data row6 col2\" >0.6017</td>\n",
       "      <td id=\"T_ac2c6_row6_col3\" class=\"data row6 col3\" >0.5513</td>\n",
       "      <td id=\"T_ac2c6_row6_col4\" class=\"data row6 col4\" >0.1737</td>\n",
       "      <td id=\"T_ac2c6_row6_col5\" class=\"data row6 col5\" >11.6353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ac2c6_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_ac2c6_row7_col0\" class=\"data row7 col0\" >0.1075</td>\n",
       "      <td id=\"T_ac2c6_row7_col1\" class=\"data row7 col1\" >0.0948</td>\n",
       "      <td id=\"T_ac2c6_row7_col2\" class=\"data row7 col2\" >0.3078</td>\n",
       "      <td id=\"T_ac2c6_row7_col3\" class=\"data row7 col3\" >0.5440</td>\n",
       "      <td id=\"T_ac2c6_row7_col4\" class=\"data row7 col4\" >0.1412</td>\n",
       "      <td id=\"T_ac2c6_row7_col5\" class=\"data row7 col5\" >14.4467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ac2c6_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_ac2c6_row8_col0\" class=\"data row8 col0\" >0.1436</td>\n",
       "      <td id=\"T_ac2c6_row8_col1\" class=\"data row8 col1\" >0.4006</td>\n",
       "      <td id=\"T_ac2c6_row8_col2\" class=\"data row8 col2\" >0.6330</td>\n",
       "      <td id=\"T_ac2c6_row8_col3\" class=\"data row8 col3\" >0.4654</td>\n",
       "      <td id=\"T_ac2c6_row8_col4\" class=\"data row8 col4\" >0.1639</td>\n",
       "      <td id=\"T_ac2c6_row8_col5\" class=\"data row8 col5\" >11.9561</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ac2c6_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_ac2c6_row9_col0\" class=\"data row9 col0\" >0.1109</td>\n",
       "      <td id=\"T_ac2c6_row9_col1\" class=\"data row9 col1\" >0.1020</td>\n",
       "      <td id=\"T_ac2c6_row9_col2\" class=\"data row9 col2\" >0.3194</td>\n",
       "      <td id=\"T_ac2c6_row9_col3\" class=\"data row9 col3\" >0.7055</td>\n",
       "      <td id=\"T_ac2c6_row9_col4\" class=\"data row9 col4\" >0.1451</td>\n",
       "      <td id=\"T_ac2c6_row9_col5\" class=\"data row9 col5\" >11.6260</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ac2c6_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_ac2c6_row10_col0\" class=\"data row10 col0\" >0.1329</td>\n",
       "      <td id=\"T_ac2c6_row10_col1\" class=\"data row10 col1\" >0.2171</td>\n",
       "      <td id=\"T_ac2c6_row10_col2\" class=\"data row10 col2\" >0.4506</td>\n",
       "      <td id=\"T_ac2c6_row10_col3\" class=\"data row10 col3\" >0.5947</td>\n",
       "      <td id=\"T_ac2c6_row10_col4\" class=\"data row10 col4\" >0.1593</td>\n",
       "      <td id=\"T_ac2c6_row10_col5\" class=\"data row10 col5\" >12.8234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_ac2c6_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_ac2c6_row11_col0\" class=\"data row11 col0\" >0.0170</td>\n",
       "      <td id=\"T_ac2c6_row11_col1\" class=\"data row11 col1\" >0.1091</td>\n",
       "      <td id=\"T_ac2c6_row11_col2\" class=\"data row11 col2\" >0.1187</td>\n",
       "      <td id=\"T_ac2c6_row11_col3\" class=\"data row11 col3\" >0.0914</td>\n",
       "      <td id=\"T_ac2c6_row11_col4\" class=\"data row11 col4\" >0.0106</td>\n",
       "      <td id=\"T_ac2c6_row11_col5\" class=\"data row11 col5\" >1.1181</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2487e78efa0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a06e2cb92ebf48f68fd3fabfc9e15372",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/4 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Initialize 3 best models, Capping max_depth at 7\n",
    "dt_ozone = create_model('dt',max_depth=7)\n",
    "rf_ozone = create_model('rf',max_depth=7)\n",
    "et_ozone = create_model('et',max_depth=7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_34d5d_row10_col0, #T_34d5d_row10_col1, #T_34d5d_row10_col2, #T_34d5d_row10_col3, #T_34d5d_row10_col4, #T_34d5d_row10_col5 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_34d5d\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_34d5d_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_34d5d_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_34d5d_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_34d5d_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_34d5d_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_34d5d_level0_col5\" class=\"col_heading level0 col5\" >MAPE</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_34d5d_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_34d5d_row0_col0\" class=\"data row0 col0\" >0.0684</td>\n",
       "      <td id=\"T_34d5d_row0_col1\" class=\"data row0 col1\" >0.0441</td>\n",
       "      <td id=\"T_34d5d_row0_col2\" class=\"data row0 col2\" >0.2101</td>\n",
       "      <td id=\"T_34d5d_row0_col3\" class=\"data row0 col3\" >0.9347</td>\n",
       "      <td id=\"T_34d5d_row0_col4\" class=\"data row0 col4\" >0.0849</td>\n",
       "      <td id=\"T_34d5d_row0_col5\" class=\"data row0 col5\" >5.3115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_34d5d_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_34d5d_row1_col0\" class=\"data row1 col0\" >0.0681</td>\n",
       "      <td id=\"T_34d5d_row1_col1\" class=\"data row1 col1\" >0.0377</td>\n",
       "      <td id=\"T_34d5d_row1_col2\" class=\"data row1 col2\" >0.1942</td>\n",
       "      <td id=\"T_34d5d_row1_col3\" class=\"data row1 col3\" >0.8676</td>\n",
       "      <td id=\"T_34d5d_row1_col4\" class=\"data row1 col4\" >0.1046</td>\n",
       "      <td id=\"T_34d5d_row1_col5\" class=\"data row1 col5\" >5.1505</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_34d5d_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_34d5d_row2_col0\" class=\"data row2 col0\" >0.0582</td>\n",
       "      <td id=\"T_34d5d_row2_col1\" class=\"data row2 col1\" >0.0222</td>\n",
       "      <td id=\"T_34d5d_row2_col2\" class=\"data row2 col2\" >0.1490</td>\n",
       "      <td id=\"T_34d5d_row2_col3\" class=\"data row2 col3\" >0.9386</td>\n",
       "      <td id=\"T_34d5d_row2_col4\" class=\"data row2 col4\" >0.0815</td>\n",
       "      <td id=\"T_34d5d_row2_col5\" class=\"data row2 col5\" >5.6781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_34d5d_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_34d5d_row3_col0\" class=\"data row3 col0\" >0.0752</td>\n",
       "      <td id=\"T_34d5d_row3_col1\" class=\"data row3 col1\" >0.0568</td>\n",
       "      <td id=\"T_34d5d_row3_col2\" class=\"data row3 col2\" >0.2383</td>\n",
       "      <td id=\"T_34d5d_row3_col3\" class=\"data row3 col3\" >0.9328</td>\n",
       "      <td id=\"T_34d5d_row3_col4\" class=\"data row3 col4\" >0.0937</td>\n",
       "      <td id=\"T_34d5d_row3_col5\" class=\"data row3 col5\" >4.6550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_34d5d_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_34d5d_row4_col0\" class=\"data row4 col0\" >0.0772</td>\n",
       "      <td id=\"T_34d5d_row4_col1\" class=\"data row4 col1\" >0.0640</td>\n",
       "      <td id=\"T_34d5d_row4_col2\" class=\"data row4 col2\" >0.2529</td>\n",
       "      <td id=\"T_34d5d_row4_col3\" class=\"data row4 col3\" >0.8881</td>\n",
       "      <td id=\"T_34d5d_row4_col4\" class=\"data row4 col4\" >0.1008</td>\n",
       "      <td id=\"T_34d5d_row4_col5\" class=\"data row4 col5\" >5.9912</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_34d5d_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_34d5d_row5_col0\" class=\"data row5 col0\" >0.0602</td>\n",
       "      <td id=\"T_34d5d_row5_col1\" class=\"data row5 col1\" >0.0700</td>\n",
       "      <td id=\"T_34d5d_row5_col2\" class=\"data row5 col2\" >0.2646</td>\n",
       "      <td id=\"T_34d5d_row5_col3\" class=\"data row5 col3\" >0.8369</td>\n",
       "      <td id=\"T_34d5d_row5_col4\" class=\"data row5 col4\" >0.0813</td>\n",
       "      <td id=\"T_34d5d_row5_col5\" class=\"data row5 col5\" >5.1246</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_34d5d_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_34d5d_row6_col0\" class=\"data row6 col0\" >0.0778</td>\n",
       "      <td id=\"T_34d5d_row6_col1\" class=\"data row6 col1\" >0.0496</td>\n",
       "      <td id=\"T_34d5d_row6_col2\" class=\"data row6 col2\" >0.2226</td>\n",
       "      <td id=\"T_34d5d_row6_col3\" class=\"data row6 col3\" >0.9386</td>\n",
       "      <td id=\"T_34d5d_row6_col4\" class=\"data row6 col4\" >0.1023</td>\n",
       "      <td id=\"T_34d5d_row6_col5\" class=\"data row6 col5\" >4.9438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_34d5d_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_34d5d_row7_col0\" class=\"data row7 col0\" >0.0592</td>\n",
       "      <td id=\"T_34d5d_row7_col1\" class=\"data row7 col1\" >0.0380</td>\n",
       "      <td id=\"T_34d5d_row7_col2\" class=\"data row7 col2\" >0.1949</td>\n",
       "      <td id=\"T_34d5d_row7_col3\" class=\"data row7 col3\" >0.8171</td>\n",
       "      <td id=\"T_34d5d_row7_col4\" class=\"data row7 col4\" >0.0974</td>\n",
       "      <td id=\"T_34d5d_row7_col5\" class=\"data row7 col5\" >5.7316</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_34d5d_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_34d5d_row8_col0\" class=\"data row8 col0\" >0.0847</td>\n",
       "      <td id=\"T_34d5d_row8_col1\" class=\"data row8 col1\" >0.3421</td>\n",
       "      <td id=\"T_34d5d_row8_col2\" class=\"data row8 col2\" >0.5849</td>\n",
       "      <td id=\"T_34d5d_row8_col3\" class=\"data row8 col3\" >0.5436</td>\n",
       "      <td id=\"T_34d5d_row8_col4\" class=\"data row8 col4\" >0.1122</td>\n",
       "      <td id=\"T_34d5d_row8_col5\" class=\"data row8 col5\" >4.5609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_34d5d_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_34d5d_row9_col0\" class=\"data row9 col0\" >0.0742</td>\n",
       "      <td id=\"T_34d5d_row9_col1\" class=\"data row9 col1\" >0.1972</td>\n",
       "      <td id=\"T_34d5d_row9_col2\" class=\"data row9 col2\" >0.4440</td>\n",
       "      <td id=\"T_34d5d_row9_col3\" class=\"data row9 col3\" >0.4307</td>\n",
       "      <td id=\"T_34d5d_row9_col4\" class=\"data row9 col4\" >0.1323</td>\n",
       "      <td id=\"T_34d5d_row9_col5\" class=\"data row9 col5\" >4.8318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_34d5d_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_34d5d_row10_col0\" class=\"data row10 col0\" >0.0703</td>\n",
       "      <td id=\"T_34d5d_row10_col1\" class=\"data row10 col1\" >0.0922</td>\n",
       "      <td id=\"T_34d5d_row10_col2\" class=\"data row10 col2\" >0.2756</td>\n",
       "      <td id=\"T_34d5d_row10_col3\" class=\"data row10 col3\" >0.8129</td>\n",
       "      <td id=\"T_34d5d_row10_col4\" class=\"data row10 col4\" >0.0991</td>\n",
       "      <td id=\"T_34d5d_row10_col5\" class=\"data row10 col5\" >5.1979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_34d5d_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_34d5d_row11_col0\" class=\"data row11 col0\" >0.0085</td>\n",
       "      <td id=\"T_34d5d_row11_col1\" class=\"data row11 col1\" >0.0954</td>\n",
       "      <td id=\"T_34d5d_row11_col2\" class=\"data row11 col2\" >0.1274</td>\n",
       "      <td id=\"T_34d5d_row11_col3\" class=\"data row11 col3\" >0.1699</td>\n",
       "      <td id=\"T_34d5d_row11_col4\" class=\"data row11 col4\" >0.0148</td>\n",
       "      <td id=\"T_34d5d_row11_col5\" class=\"data row11 col5\" >0.4544</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2487e7ee6d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2eba04831edf42feb9bbf2043cfec79c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/6 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Develop a blended model with voting regressor\n",
    "blended_ozone = blend_models(estimator_list=[dt_ozone,rf_outdoor,et_ozone],optimize='MAE')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this case the random forest individual models has a better MAE and R2 compared to the blended model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3.3 Stacked Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A meta model, linear regression by default, is trained over the 3 selected regressors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_b14ce_row10_col0, #T_b14ce_row10_col1, #T_b14ce_row10_col2, #T_b14ce_row10_col3, #T_b14ce_row10_col4, #T_b14ce_row10_col5 {\n",
       "  background: yellow;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_b14ce\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_b14ce_level0_col0\" class=\"col_heading level0 col0\" >MAE</th>\n",
       "      <th id=\"T_b14ce_level0_col1\" class=\"col_heading level0 col1\" >MSE</th>\n",
       "      <th id=\"T_b14ce_level0_col2\" class=\"col_heading level0 col2\" >RMSE</th>\n",
       "      <th id=\"T_b14ce_level0_col3\" class=\"col_heading level0 col3\" >R2</th>\n",
       "      <th id=\"T_b14ce_level0_col4\" class=\"col_heading level0 col4\" >RMSLE</th>\n",
       "      <th id=\"T_b14ce_level0_col5\" class=\"col_heading level0 col5\" >MAPE</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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_b14ce_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_b14ce_row0_col0\" class=\"data row0 col0\" >0.0384</td>\n",
       "      <td id=\"T_b14ce_row0_col1\" class=\"data row0 col1\" >0.0103</td>\n",
       "      <td id=\"T_b14ce_row0_col2\" class=\"data row0 col2\" >0.1016</td>\n",
       "      <td id=\"T_b14ce_row0_col3\" class=\"data row0 col3\" >0.9847</td>\n",
       "      <td id=\"T_b14ce_row0_col4\" class=\"data row0 col4\" >0.0686</td>\n",
       "      <td id=\"T_b14ce_row0_col5\" class=\"data row0 col5\" >1.9161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b14ce_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_b14ce_row1_col0\" class=\"data row1 col0\" >0.0520</td>\n",
       "      <td id=\"T_b14ce_row1_col1\" class=\"data row1 col1\" >0.0306</td>\n",
       "      <td id=\"T_b14ce_row1_col2\" class=\"data row1 col2\" >0.1750</td>\n",
       "      <td id=\"T_b14ce_row1_col3\" class=\"data row1 col3\" >0.8924</td>\n",
       "      <td id=\"T_b14ce_row1_col4\" class=\"data row1 col4\" >0.0968</td>\n",
       "      <td id=\"T_b14ce_row1_col5\" class=\"data row1 col5\" >2.7537</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b14ce_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_b14ce_row2_col0\" class=\"data row2 col0\" >0.0314</td>\n",
       "      <td id=\"T_b14ce_row2_col1\" class=\"data row2 col1\" >0.0078</td>\n",
       "      <td id=\"T_b14ce_row2_col2\" class=\"data row2 col2\" >0.0883</td>\n",
       "      <td id=\"T_b14ce_row2_col3\" class=\"data row2 col3\" >0.9785</td>\n",
       "      <td id=\"T_b14ce_row2_col4\" class=\"data row2 col4\" >0.0610</td>\n",
       "      <td id=\"T_b14ce_row2_col5\" class=\"data row2 col5\" >2.1320</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b14ce_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_b14ce_row3_col0\" class=\"data row3 col0\" >0.0431</td>\n",
       "      <td id=\"T_b14ce_row3_col1\" class=\"data row3 col1\" >0.0189</td>\n",
       "      <td id=\"T_b14ce_row3_col2\" class=\"data row3 col2\" >0.1375</td>\n",
       "      <td id=\"T_b14ce_row3_col3\" class=\"data row3 col3\" >0.9776</td>\n",
       "      <td id=\"T_b14ce_row3_col4\" class=\"data row3 col4\" >0.0714</td>\n",
       "      <td id=\"T_b14ce_row3_col5\" class=\"data row3 col5\" >2.4320</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b14ce_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_b14ce_row4_col0\" class=\"data row4 col0\" >0.0499</td>\n",
       "      <td id=\"T_b14ce_row4_col1\" class=\"data row4 col1\" >0.0316</td>\n",
       "      <td id=\"T_b14ce_row4_col2\" class=\"data row4 col2\" >0.1778</td>\n",
       "      <td id=\"T_b14ce_row4_col3\" class=\"data row4 col3\" >0.9447</td>\n",
       "      <td id=\"T_b14ce_row4_col4\" class=\"data row4 col4\" >0.0799</td>\n",
       "      <td id=\"T_b14ce_row4_col5\" class=\"data row4 col5\" >3.0693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b14ce_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_b14ce_row5_col0\" class=\"data row5 col0\" >0.0469</td>\n",
       "      <td id=\"T_b14ce_row5_col1\" class=\"data row5 col1\" >0.0600</td>\n",
       "      <td id=\"T_b14ce_row5_col2\" class=\"data row5 col2\" >0.2449</td>\n",
       "      <td id=\"T_b14ce_row5_col3\" class=\"data row5 col3\" >0.8603</td>\n",
       "      <td id=\"T_b14ce_row5_col4\" class=\"data row5 col4\" >0.0712</td>\n",
       "      <td id=\"T_b14ce_row5_col5\" class=\"data row5 col5\" >2.2730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b14ce_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_b14ce_row6_col0\" class=\"data row6 col0\" >0.0512</td>\n",
       "      <td id=\"T_b14ce_row6_col1\" class=\"data row6 col1\" >0.0191</td>\n",
       "      <td id=\"T_b14ce_row6_col2\" class=\"data row6 col2\" >0.1383</td>\n",
       "      <td id=\"T_b14ce_row6_col3\" class=\"data row6 col3\" >0.9763</td>\n",
       "      <td id=\"T_b14ce_row6_col4\" class=\"data row6 col4\" >0.0922</td>\n",
       "      <td id=\"T_b14ce_row6_col5\" class=\"data row6 col5\" >2.6935</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b14ce_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_b14ce_row7_col0\" class=\"data row7 col0\" >0.0376</td>\n",
       "      <td id=\"T_b14ce_row7_col1\" class=\"data row7 col1\" >0.0191</td>\n",
       "      <td id=\"T_b14ce_row7_col2\" class=\"data row7 col2\" >0.1382</td>\n",
       "      <td id=\"T_b14ce_row7_col3\" class=\"data row7 col3\" >0.9080</td>\n",
       "      <td id=\"T_b14ce_row7_col4\" class=\"data row7 col4\" >0.0785</td>\n",
       "      <td id=\"T_b14ce_row7_col5\" class=\"data row7 col5\" >2.6018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b14ce_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_b14ce_row8_col0\" class=\"data row8 col0\" >0.0642</td>\n",
       "      <td id=\"T_b14ce_row8_col1\" class=\"data row8 col1\" >0.2677</td>\n",
       "      <td id=\"T_b14ce_row8_col2\" class=\"data row8 col2\" >0.5174</td>\n",
       "      <td id=\"T_b14ce_row8_col3\" class=\"data row8 col3\" >0.6428</td>\n",
       "      <td id=\"T_b14ce_row8_col4\" class=\"data row8 col4\" >0.0939</td>\n",
       "      <td id=\"T_b14ce_row8_col5\" class=\"data row8 col5\" >3.2557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b14ce_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_b14ce_row9_col0\" class=\"data row9 col0\" >0.0699</td>\n",
       "      <td id=\"T_b14ce_row9_col1\" class=\"data row9 col1\" >0.2506</td>\n",
       "      <td id=\"T_b14ce_row9_col2\" class=\"data row9 col2\" >0.5006</td>\n",
       "      <td id=\"T_b14ce_row9_col3\" class=\"data row9 col3\" >0.2764</td>\n",
       "      <td id=\"T_b14ce_row9_col4\" class=\"data row9 col4\" >0.1358</td>\n",
       "      <td id=\"T_b14ce_row9_col5\" class=\"data row9 col5\" >3.4226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b14ce_level0_row10\" class=\"row_heading level0 row10\" >Mean</th>\n",
       "      <td id=\"T_b14ce_row10_col0\" class=\"data row10 col0\" >0.0485</td>\n",
       "      <td id=\"T_b14ce_row10_col1\" class=\"data row10 col1\" >0.0716</td>\n",
       "      <td id=\"T_b14ce_row10_col2\" class=\"data row10 col2\" >0.2220</td>\n",
       "      <td id=\"T_b14ce_row10_col3\" class=\"data row10 col3\" >0.8442</td>\n",
       "      <td id=\"T_b14ce_row10_col4\" class=\"data row10 col4\" >0.0849</td>\n",
       "      <td id=\"T_b14ce_row10_col5\" class=\"data row10 col5\" >2.6550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_b14ce_level0_row11\" class=\"row_heading level0 row11\" >Std</th>\n",
       "      <td id=\"T_b14ce_row11_col0\" class=\"data row11 col0\" >0.0113</td>\n",
       "      <td id=\"T_b14ce_row11_col1\" class=\"data row11 col1\" >0.0949</td>\n",
       "      <td id=\"T_b14ce_row11_col2\" class=\"data row11 col2\" >0.1494</td>\n",
       "      <td id=\"T_b14ce_row11_col3\" class=\"data row11 col3\" >0.2128</td>\n",
       "      <td id=\"T_b14ce_row11_col4\" class=\"data row11 col4\" >0.0204</td>\n",
       "      <td id=\"T_b14ce_row11_col5\" class=\"data row11 col5\" >0.4638</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2487e86f9a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4269af184aaf4a36acf8024ccb4c46fe",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing:   0%|          | 0/6 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Develop a blended model with voting regressor\n",
    "stacked_ozone = stack_models(estimator_list=[dt_ozone,rf_outdoor,et_ozone],optimize='MAE')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This model has roughly equivalent results to the random forest model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3.3 Evaluate Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Assume that random forest model is the most appropriate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x550 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plot learning Curve\n",
    "plot_model(rf_ozone, plot = 'learning')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Cross-validation model scores hardly increase from 85% to 90% with an increase in training instances."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x550 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot Residuals\n",
    "plot_model(rf_ozone, plot = 'residuals')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The residuals plot shows the distribution of residual (delta between actual vs predicted indoor air pollution deaths), which are largely within +/-10 deaths per 100,000, which is reasonable on the individual country level considering the {mean,max} of the metric = {20,2830}, considering the R2 is actually around 97%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot Feature Importance - Variable Importance (different from average importance shown in dashbaoard)\n",
    "plot_model(rf_ozone, plot = 'feature')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Feature Importance plot shows that the SoDI drives the model at a variable importance of around 0.45, with Population in second at around 0.35, and third with GDP per capita at around 0.2."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3.4 Predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_c58df\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_c58df_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_c58df_level0_col1\" class=\"col_heading level0 col1\" >MAE</th>\n",
       "      <th id=\"T_c58df_level0_col2\" class=\"col_heading level0 col2\" >MSE</th>\n",
       "      <th id=\"T_c58df_level0_col3\" class=\"col_heading level0 col3\" >RMSE</th>\n",
       "      <th id=\"T_c58df_level0_col4\" class=\"col_heading level0 col4\" >R2</th>\n",
       "      <th id=\"T_c58df_level0_col5\" class=\"col_heading level0 col5\" >RMSLE</th>\n",
       "      <th id=\"T_c58df_level0_col6\" class=\"col_heading level0 col6\" >MAPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_c58df_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_c58df_row0_col0\" class=\"data row0 col0\" >Random Forest Regressor</td>\n",
       "      <td id=\"T_c58df_row0_col1\" class=\"data row0 col1\" >0.0571</td>\n",
       "      <td id=\"T_c58df_row0_col2\" class=\"data row0 col2\" >0.0935</td>\n",
       "      <td id=\"T_c58df_row0_col3\" class=\"data row0 col3\" >0.3058</td>\n",
       "      <td id=\"T_c58df_row0_col4\" class=\"data row0 col4\" >0.9068</td>\n",
       "      <td id=\"T_c58df_row0_col5\" class=\"data row0 col5\" >0.0776</td>\n",
       "      <td id=\"T_c58df_row0_col6\" class=\"data row0 col6\" >1.5183</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2480978f8b0>"
      ]
     },
     "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>GDP_pCapita</th>\n",
       "      <th>Population</th>\n",
       "      <th>SoDI</th>\n",
       "      <th>Income_Class_encoded</th>\n",
       "      <th>Year</th>\n",
       "      <th>Deaths_OZone</th>\n",
       "      <th>prediction_label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1025</th>\n",
       "      <td>5386.467773</td>\n",
       "      <td>14587367.0</td>\n",
       "      <td>0.630</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001996</td>\n",
       "      <td>0.004042</td>\n",
       "      <td>0.005366</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>935</th>\n",
       "      <td>21227.347656</td>\n",
       "      <td>29610218.0</td>\n",
       "      <td>0.809</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001996</td>\n",
       "      <td>0.004505</td>\n",
       "      <td>0.004721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3024</th>\n",
       "      <td>117254.742188</td>\n",
       "      <td>607950.0</td>\n",
       "      <td>0.894</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002018</td>\n",
       "      <td>0.324673</td>\n",
       "      <td>0.461738</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3584</th>\n",
       "      <td>10171.480469</td>\n",
       "      <td>10289.0</td>\n",
       "      <td>0.573</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002014</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2131</th>\n",
       "      <td>477.762451</td>\n",
       "      <td>7081119.0</td>\n",
       "      <td>0.190</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001994</td>\n",
       "      <td>0.005983</td>\n",
       "      <td>0.009433</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>524</th>\n",
       "      <td>2993.033691</td>\n",
       "      <td>230248.0</td>\n",
       "      <td>0.494</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001998</td>\n",
       "      <td>0.188629</td>\n",
       "      <td>0.162806</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5227</th>\n",
       "      <td>590.393494</td>\n",
       "      <td>4413477.0</td>\n",
       "      <td>0.556</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1970-01-01 00:00:00.000001998</td>\n",
       "      <td>0.024129</td>\n",
       "      <td>0.054696</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2883</th>\n",
       "      <td>7948.685547</td>\n",
       "      <td>5537620.0</td>\n",
       "      <td>0.660</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002012</td>\n",
       "      <td>0.029349</td>\n",
       "      <td>0.022826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3941</th>\n",
       "      <td>1664.245728</td>\n",
       "      <td>3494496.0</td>\n",
       "      <td>0.467</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002007</td>\n",
       "      <td>0.022929</td>\n",
       "      <td>0.029863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2083</th>\n",
       "      <td>26713.058594</td>\n",
       "      <td>158649.0</td>\n",
       "      <td>0.773</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1970-01-01 00:00:00.000002006</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.032571</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1143 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        GDP_pCapita  Population   SoDI  Income_Class_encoded  \\\n",
       "1025    5386.467773  14587367.0  0.630                   2.0   \n",
       "935    21227.347656  29610218.0  0.809                   3.0   \n",
       "3024  117254.742188    607950.0  0.894                   3.0   \n",
       "3584   10171.480469     10289.0  0.573                   3.0   \n",
       "2131     477.762451   7081119.0  0.190                   0.0   \n",
       "...             ...         ...    ...                   ...   \n",
       "524     2993.033691    230248.0  0.494                   1.0   \n",
       "5227     590.393494   4413477.0  0.556                   0.0   \n",
       "2883    7948.685547   5537620.0  0.660                   2.0   \n",
       "3941    1664.245728   3494496.0  0.467                   1.0   \n",
       "2083   26713.058594    158649.0  0.773                   3.0   \n",
       "\n",
       "                              Year  Deaths_OZone  prediction_label  \n",
       "1025 1970-01-01 00:00:00.000001996      0.004042          0.005366  \n",
       "935  1970-01-01 00:00:00.000001996      0.004505          0.004721  \n",
       "3024 1970-01-01 00:00:00.000002018      0.324673          0.461738  \n",
       "3584 1970-01-01 00:00:00.000002014      0.000000          0.000000  \n",
       "2131 1970-01-01 00:00:00.000001994      0.005983          0.009433  \n",
       "...                            ...           ...               ...  \n",
       "524  1970-01-01 00:00:00.000001998      0.188629          0.162806  \n",
       "5227 1970-01-01 00:00:00.000001998      0.024129          0.054696  \n",
       "2883 1970-01-01 00:00:00.000002012      0.029349          0.022826  \n",
       "3941 1970-01-01 00:00:00.000002007      0.022929          0.029863  \n",
       "2083 1970-01-01 00:00:00.000002006      0.000000          0.032571  \n",
       "\n",
       "[1143 rows x 7 columns]"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Use random forest model to predict holdout testing dataset indoor pollution deaths per 100,000\n",
    "predictions = predict_model(rf_ozone)\n",
    "predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Even with the hold-out testing dataset, the tuned model has RMSE,MAE,R2 not drastically different from the test-train dataset used to build the model, so it shows that the model was not overfitted to the input data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Transformation Pipeline and Model Successfully Saved\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(Pipeline(memory=FastMemory(location=C:\\Users\\kevin\\AppData\\Local\\Temp\\joblib),\n",
       "          steps=[('date_feature_extractor',\n",
       "                  TransformerWrapper(include=['Year'],\n",
       "                                     transformer=ExtractDateTimeFeatures())),\n",
       "                 ('numerical_imputer',\n",
       "                  TransformerWrapper(include=['GDP_pCapita', 'Population',\n",
       "                                              'SoDI', 'Income_Class_encoded'],\n",
       "                                     transformer=SimpleImputer())),\n",
       "                 ('categorical_imputer',\n",
       "                  TransformerWrapper(include=[],\n",
       "                                     transformer=SimpleImputer(strategy='most_frequent'))),\n",
       "                 ('trained_model',\n",
       "                  RandomForestRegressor(max_depth=7, n_jobs=-1,\n",
       "                                        random_state=123))]),\n",
       " 'Ozone Deaths Random Forest 2023-03-29.pkl')"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Save Model\n",
    "save_model(rf_ozone,'Ozone Deaths Random Forest 2023-03-29')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}