FormulaAI_Hackathon_Notebook_ADS_AutoML.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "88567ea6",
   "metadata": {},
   "source": [
    "# Formula 1 Weather Forecasting using ADS and AutomL"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "da486634",
   "metadata": {},
   "source": [
    "# Overview:\n",
    "\n",
    "In the F1 sports E-game weather plays a vital role in the gameplay of the driver or a garage team. In this notebook we will be performing exploratory data analysis to understand the data, using the weather dataset. The weather dataset has 58 features in which a few of the features will contribute significantly in the prediction of weather. The target variables are continuous and dicrete values representing a quantitative and qualitative measure of weather and it's percentage.\n",
    "\n",
    "We will start from the data loading, then analysis of Data, after analysis we will transform the features and then apply the AutoML model. Finally, We will evaluate how well the model generalizes to the problem. \n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a8e3e035",
   "metadata": {},
   "source": [
    "\n",
    "# Contents:\n",
    "- <a>Data Loading</a>\n",
    "- <a>Exploratory Data Analysis</a>\n",
    "    - <a>Getting overall statistics</a>\n",
    "    - <a>Column based EDA</a>\n",
    "    - <a>ADS DatasetFactory</a>\n",
    "    - <a>Imputation Actions</a>\n",
    "    - <a>Normalizing Values</a>\n",
    "    - <a>Correlation evaluation with Target Variable</a>\n",
    "    - <a>Imputation Actions</a>\n",
    "- <a>AutoML Model</a>\n",
    "    - <a>Visualization and Evaluation</a> \n",
    "    - <a>Import model to pickle</a>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "375b508c",
   "metadata": {},
   "source": [
    "# Data Loading"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "254f86fc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load a number of libraries, including the Accelerated Data Science DatasetFactory and initialize it.\n",
    "# importig the libraries\n",
    "import pandas as pd\n",
    "import logging\n",
    "import numpy as np\n",
    "import ads\n",
    "import oci\n",
    "from oci.data_science import DataScienceClient\n",
    "from ads.dataset.factory import DatasetFactory\n",
    "\n",
    "# OCI ADS API\n",
    "ads.set_auth(auth='resource_principal')\n",
    "resource_principal = oci.auth.signers.get_resource_principals_signer()\n",
    "dsc = DataScienceClient(config={},signer=resource_principal)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f12ae4a4",
   "metadata": {},
   "outputs": [],
   "source": [
    "logging.basicConfig(format='%(levelname)s:%(message)s', level=logging.ERROR)\n",
    "weather_data = pd.read_csv('weather.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7b5a856",
   "metadata": {},
   "source": [
    "# Exploratory Data Analysis\n",
    "The intial investigation performed on data with the aim to get insights , patterns and to detect anamolies in a data is called EDA - Exploratory Data Analysis\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "968449b6",
   "metadata": {},
   "source": [
    "High level Exploratory Data Analysis (getting overall statistics of the data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "4bf2e4e2",
   "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>M_PACKET_FORMAT</th>\n",
       "      <th>M_GAME_MAJOR_VERSION</th>\n",
       "      <th>M_GAME_MINOR_VERSION</th>\n",
       "      <th>M_PACKET_VERSION</th>\n",
       "      <th>M_PACKET_ID</th>\n",
       "      <th>M_SESSION_UID</th>\n",
       "      <th>M_SESSION_TIME</th>\n",
       "      <th>M_FRAME_IDENTIFIER</th>\n",
       "      <th>M_PLAYER_CAR_INDEX</th>\n",
       "      <th>M_SECONDARY_PLAYER_CAR_INDEX</th>\n",
       "      <th>...</th>\n",
       "      <th>M_AI_DIFFICULTY</th>\n",
       "      <th>M_PIT_SPEED_LIMIT</th>\n",
       "      <th>M_NETWORK_GAME</th>\n",
       "      <th>M_TOTAL_LAPS</th>\n",
       "      <th>M_STEERING_ASSIST</th>\n",
       "      <th>M_IS_SPECTATING</th>\n",
       "      <th>M_DYNAMIC_RACING_LINE</th>\n",
       "      <th>M_DRSASSIST</th>\n",
       "      <th>M_NUM_MARSHAL_ZONES</th>\n",
       "      <th>Unnamed: 58</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2021</td>\n",
       "      <td>1</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.300210e+19</td>\n",
       "      <td>2803.836</td>\n",
       "      <td>82458</td>\n",
       "      <td>0</td>\n",
       "      <td>255</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>80</td>\n",
       "      <td>0.0</td>\n",
       "      <td>200.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2021</td>\n",
       "      <td>1</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.300210e+19</td>\n",
       "      <td>2803.836</td>\n",
       "      <td>82458</td>\n",
       "      <td>0</td>\n",
       "      <td>255</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>80</td>\n",
       "      <td>0.0</td>\n",
       "      <td>200.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2021</td>\n",
       "      <td>1</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.300210e+19</td>\n",
       "      <td>2803.836</td>\n",
       "      <td>82458</td>\n",
       "      <td>0</td>\n",
       "      <td>255</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>80</td>\n",
       "      <td>0.0</td>\n",
       "      <td>200.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2021</td>\n",
       "      <td>1</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.300210e+19</td>\n",
       "      <td>2803.836</td>\n",
       "      <td>82458</td>\n",
       "      <td>0</td>\n",
       "      <td>255</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>80</td>\n",
       "      <td>0.0</td>\n",
       "      <td>200.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2021</td>\n",
       "      <td>1</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.300210e+19</td>\n",
       "      <td>2803.836</td>\n",
       "      <td>82458</td>\n",
       "      <td>0</td>\n",
       "      <td>255</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>80</td>\n",
       "      <td>0.0</td>\n",
       "      <td>200.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 59 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   M_PACKET_FORMAT  M_GAME_MAJOR_VERSION  M_GAME_MINOR_VERSION  \\\n",
       "0             2021                     1                    14   \n",
       "1             2021                     1                    14   \n",
       "2             2021                     1                    14   \n",
       "3             2021                     1                    14   \n",
       "4             2021                     1                    14   \n",
       "\n",
       "   M_PACKET_VERSION  M_PACKET_ID  M_SESSION_UID  M_SESSION_TIME  \\\n",
       "0                 1            1   1.300210e+19        2803.836   \n",
       "1                 1            1   1.300210e+19        2803.836   \n",
       "2                 1            1   1.300210e+19        2803.836   \n",
       "3                 1            1   1.300210e+19        2803.836   \n",
       "4                 1            1   1.300210e+19        2803.836   \n",
       "\n",
       "   M_FRAME_IDENTIFIER  M_PLAYER_CAR_INDEX  M_SECONDARY_PLAYER_CAR_INDEX  ...  \\\n",
       "0               82458                   0                           255  ...   \n",
       "1               82458                   0                           255  ...   \n",
       "2               82458                   0                           255  ...   \n",
       "3               82458                   0                           255  ...   \n",
       "4               82458                   0                           255  ...   \n",
       "\n",
       "   M_AI_DIFFICULTY  M_PIT_SPEED_LIMIT  M_NETWORK_GAME  M_TOTAL_LAPS  \\\n",
       "0                0                 80             0.0         200.0   \n",
       "1                0                 80             0.0         200.0   \n",
       "2                0                 80             0.0         200.0   \n",
       "3                0                 80             0.0         200.0   \n",
       "4                0                 80             0.0         200.0   \n",
       "\n",
       "   M_STEERING_ASSIST  M_IS_SPECTATING  M_DYNAMIC_RACING_LINE  M_DRSASSIST  \\\n",
       "0                0.0              0.0                    0.0          0.0   \n",
       "1                0.0              0.0                    0.0          0.0   \n",
       "2                0.0              0.0                    0.0          0.0   \n",
       "3                0.0              0.0                    0.0          0.0   \n",
       "4                0.0              0.0                    0.0          0.0   \n",
       "\n",
       "   M_NUM_MARSHAL_ZONES  Unnamed: 58  \n",
       "0                 16.0          NaN  \n",
       "1                 16.0          NaN  \n",
       "2                 16.0          NaN  \n",
       "3                 16.0          NaN  \n",
       "4                 16.0          NaN  \n",
       "\n",
       "[5 rows x 59 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking the head(first five rows) of weather_data\n",
    "weather_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "7723c99a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3572328, 59)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's check the number of rows and columns in dataframe\n",
    "weather_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "20da13d1",
   "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>M_PACKET_FORMAT</th>\n",
       "      <th>M_GAME_MAJOR_VERSION</th>\n",
       "      <th>M_GAME_MINOR_VERSION</th>\n",
       "      <th>M_PACKET_VERSION</th>\n",
       "      <th>M_PACKET_ID</th>\n",
       "      <th>M_SESSION_UID</th>\n",
       "      <th>M_SESSION_TIME</th>\n",
       "      <th>M_FRAME_IDENTIFIER</th>\n",
       "      <th>M_PLAYER_CAR_INDEX</th>\n",
       "      <th>M_SECONDARY_PLAYER_CAR_INDEX</th>\n",
       "      <th>...</th>\n",
       "      <th>M_AI_DIFFICULTY</th>\n",
       "      <th>M_PIT_SPEED_LIMIT</th>\n",
       "      <th>M_NETWORK_GAME</th>\n",
       "      <th>M_TOTAL_LAPS</th>\n",
       "      <th>M_STEERING_ASSIST</th>\n",
       "      <th>M_IS_SPECTATING</th>\n",
       "      <th>M_DYNAMIC_RACING_LINE</th>\n",
       "      <th>M_DRSASSIST</th>\n",
       "      <th>M_NUM_MARSHAL_ZONES</th>\n",
       "      <th>Unnamed: 58</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>3572328.0</td>\n",
       "      <td>3572328.0</td>\n",
       "      <td>3.572328e+06</td>\n",
       "      <td>3572328.0</td>\n",
       "      <td>3572328.0</td>\n",
       "      <td>3.572328e+06</td>\n",
       "      <td>3.572328e+06</td>\n",
       "      <td>3.572328e+06</td>\n",
       "      <td>3.572328e+06</td>\n",
       "      <td>3572328.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.572328e+06</td>\n",
       "      <td>3.572328e+06</td>\n",
       "      <td>3.572327e+06</td>\n",
       "      <td>3.572327e+06</td>\n",
       "      <td>3.572327e+06</td>\n",
       "      <td>3.572327e+06</td>\n",
       "      <td>3.572327e+06</td>\n",
       "      <td>3.572327e+06</td>\n",
       "      <td>3.572327e+06</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2021.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.410704e+01</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.239394e+18</td>\n",
       "      <td>1.019926e+03</td>\n",
       "      <td>2.857427e+04</td>\n",
       "      <td>9.212292e+00</td>\n",
       "      <td>255.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.595344e+01</td>\n",
       "      <td>7.583221e+01</td>\n",
       "      <td>2.261691e-01</td>\n",
       "      <td>1.709807e+02</td>\n",
       "      <td>1.580382e-01</td>\n",
       "      <td>1.487266e-03</td>\n",
       "      <td>6.307732e-01</td>\n",
       "      <td>3.153866e-01</td>\n",
       "      <td>1.687091e+01</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.091641e-01</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.119260e+18</td>\n",
       "      <td>1.682487e+03</td>\n",
       "      <td>5.428773e+04</td>\n",
       "      <td>9.217495e+00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.811992e+01</td>\n",
       "      <td>8.123138e+00</td>\n",
       "      <td>4.183500e-01</td>\n",
       "      <td>6.422035e+01</td>\n",
       "      <td>3.647768e-01</td>\n",
       "      <td>3.853640e-02</td>\n",
       "      <td>9.293394e-01</td>\n",
       "      <td>4.646697e-01</td>\n",
       "      <td>1.092343e+00</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>2021.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.400000e+01</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.106082e+16</td>\n",
       "      <td>4.000000e-03</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>255.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>6.000000e+01</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>5.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.000000e+01</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2021.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.400000e+01</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.591802e+18</td>\n",
       "      <td>1.138000e+02</td>\n",
       "      <td>3.007000e+03</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>255.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.100000e+01</td>\n",
       "      <td>8.000000e+01</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.000000e+02</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.600000e+01</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2021.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.400000e+01</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7.802116e+18</td>\n",
       "      <td>4.319240e+02</td>\n",
       "      <td>1.174900e+04</td>\n",
       "      <td>2.000000e+00</td>\n",
       "      <td>255.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.100000e+01</td>\n",
       "      <td>8.000000e+01</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.000000e+02</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.700000e+01</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>2021.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.400000e+01</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.279207e+19</td>\n",
       "      <td>1.024212e+03</td>\n",
       "      <td>2.972400e+04</td>\n",
       "      <td>1.900000e+01</td>\n",
       "      <td>255.0</td>\n",
       "      <td>...</td>\n",
       "      <td>9.000000e+01</td>\n",
       "      <td>8.000000e+01</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.000000e+02</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>2.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.800000e+01</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2021.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.500000e+01</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.826297e+19</td>\n",
       "      <td>9.686959e+03</td>\n",
       "      <td>3.339170e+05</td>\n",
       "      <td>1.900000e+01</td>\n",
       "      <td>255.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.100000e+02</td>\n",
       "      <td>8.000000e+01</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>2.000000e+02</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>2.000000e+00</td>\n",
       "      <td>1.000000e+00</td>\n",
       "      <td>1.900000e+01</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 58 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       M_PACKET_FORMAT  M_GAME_MAJOR_VERSION  M_GAME_MINOR_VERSION  \\\n",
       "count        3572328.0             3572328.0          3.572328e+06   \n",
       "mean            2021.0                   1.0          1.410704e+01   \n",
       "std                0.0                   0.0          3.091641e-01   \n",
       "min             2021.0                   1.0          1.400000e+01   \n",
       "25%             2021.0                   1.0          1.400000e+01   \n",
       "50%             2021.0                   1.0          1.400000e+01   \n",
       "75%             2021.0                   1.0          1.400000e+01   \n",
       "max             2021.0                   1.0          1.500000e+01   \n",
       "\n",
       "       M_PACKET_VERSION  M_PACKET_ID  M_SESSION_UID  M_SESSION_TIME  \\\n",
       "count         3572328.0    3572328.0   3.572328e+06    3.572328e+06   \n",
       "mean                1.0          1.0   8.239394e+18    1.019926e+03   \n",
       "std                 0.0          0.0   5.119260e+18    1.682487e+03   \n",
       "min                 1.0          1.0   2.106082e+16    4.000000e-03   \n",
       "25%                 1.0          1.0   3.591802e+18    1.138000e+02   \n",
       "50%                 1.0          1.0   7.802116e+18    4.319240e+02   \n",
       "75%                 1.0          1.0   1.279207e+19    1.024212e+03   \n",
       "max                 1.0          1.0   1.826297e+19    9.686959e+03   \n",
       "\n",
       "       M_FRAME_IDENTIFIER  M_PLAYER_CAR_INDEX  M_SECONDARY_PLAYER_CAR_INDEX  \\\n",
       "count        3.572328e+06        3.572328e+06                     3572328.0   \n",
       "mean         2.857427e+04        9.212292e+00                         255.0   \n",
       "std          5.428773e+04        9.217495e+00                           0.0   \n",
       "min          0.000000e+00        0.000000e+00                         255.0   \n",
       "25%          3.007000e+03        0.000000e+00                         255.0   \n",
       "50%          1.174900e+04        2.000000e+00                         255.0   \n",
       "75%          2.972400e+04        1.900000e+01                         255.0   \n",
       "max          3.339170e+05        1.900000e+01                         255.0   \n",
       "\n",
       "       ...  M_AI_DIFFICULTY  M_PIT_SPEED_LIMIT  M_NETWORK_GAME  M_TOTAL_LAPS  \\\n",
       "count  ...     3.572328e+06       3.572328e+06    3.572327e+06  3.572327e+06   \n",
       "mean   ...     4.595344e+01       7.583221e+01    2.261691e-01  1.709807e+02   \n",
       "std    ...     3.811992e+01       8.123138e+00    4.183500e-01  6.422035e+01   \n",
       "min    ...     0.000000e+00       6.000000e+01    0.000000e+00  5.000000e+00   \n",
       "25%    ...     3.100000e+01       8.000000e+01    0.000000e+00  2.000000e+02   \n",
       "50%    ...     3.100000e+01       8.000000e+01    0.000000e+00  2.000000e+02   \n",
       "75%    ...     9.000000e+01       8.000000e+01    0.000000e+00  2.000000e+02   \n",
       "max    ...     1.100000e+02       8.000000e+01    1.000000e+00  2.000000e+02   \n",
       "\n",
       "       M_STEERING_ASSIST  M_IS_SPECTATING  M_DYNAMIC_RACING_LINE  \\\n",
       "count       3.572327e+06     3.572327e+06           3.572327e+06   \n",
       "mean        1.580382e-01     1.487266e-03           6.307732e-01   \n",
       "std         3.647768e-01     3.853640e-02           9.293394e-01   \n",
       "min         0.000000e+00     0.000000e+00           0.000000e+00   \n",
       "25%         0.000000e+00     0.000000e+00           0.000000e+00   \n",
       "50%         0.000000e+00     0.000000e+00           0.000000e+00   \n",
       "75%         0.000000e+00     0.000000e+00           2.000000e+00   \n",
       "max         1.000000e+00     1.000000e+00           2.000000e+00   \n",
       "\n",
       "        M_DRSASSIST  M_NUM_MARSHAL_ZONES  Unnamed: 58  \n",
       "count  3.572327e+06         3.572327e+06          0.0  \n",
       "mean   3.153866e-01         1.687091e+01          NaN  \n",
       "std    4.646697e-01         1.092343e+00          NaN  \n",
       "min    0.000000e+00         1.000000e+01          NaN  \n",
       "25%    0.000000e+00         1.600000e+01          NaN  \n",
       "50%    0.000000e+00         1.700000e+01          NaN  \n",
       "75%    1.000000e+00         1.800000e+01          NaN  \n",
       "max    1.000000e+00         1.900000e+01          NaN  \n",
       "\n",
       "[8 rows x 58 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's see the stats of our dataframe\n",
    "weather_data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "a5ef56ed",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 3572328 entries, 0 to 3572327\n",
      "Data columns (total 59 columns):\n",
      " #   Column                                          Dtype  \n",
      "---  ------                                          -----  \n",
      " 0   M_PACKET_FORMAT                                 int64  \n",
      " 1   M_GAME_MAJOR_VERSION                            int64  \n",
      " 2   M_GAME_MINOR_VERSION                            int64  \n",
      " 3   M_PACKET_VERSION                                int64  \n",
      " 4   M_PACKET_ID                                     int64  \n",
      " 5   M_SESSION_UID                                   float64\n",
      " 6   M_SESSION_TIME                                  float64\n",
      " 7   M_FRAME_IDENTIFIER                              int64  \n",
      " 8   M_PLAYER_CAR_INDEX                              int64  \n",
      " 9   M_SECONDARY_PLAYER_CAR_INDEX                    int64  \n",
      " 10  M_BRAKING_ASSIST                                int64  \n",
      " 11  M_SESSION_LINK_IDENTIFIER                       int64  \n",
      " 12  M_PIT_RELEASE_ASSIST                            int64  \n",
      " 13  TIMESTAMP                                       float64\n",
      " 14  M_ZONE_START                                    float64\n",
      " 15  M_ZONE_FLAG                                     float64\n",
      " 16  M_PIT_STOP_WINDOW_IDEAL_LAP                     int64  \n",
      " 17  M_TRACK_TEMPERATURE                             int64  \n",
      " 18  M_TRACK_LENGTH                                  int64  \n",
      " 19  M_GAME_PAUSED                                   int64  \n",
      " 20  M_FORECAST_ACCURACY                             int64  \n",
      " 21  GAMEHOST                                        object \n",
      " 22  M_AIR_TEMPERATURE                               int64  \n",
      " 23  M_NUM_WEATHER_FORECAST_SAMPLES                  int64  \n",
      " 24  M_SLI_PRO_NATIVE_SUPPORT                        int64  \n",
      " 25  M_SAFETY_CAR_STATUS                             int64  \n",
      " 26  M_TRACK_ID                                      int64  \n",
      " 27  M_ERSASSIST                                     int64  \n",
      " 28  M_FORMULA                                       int64  \n",
      " 29  M_SEASON_LINK_IDENTIFIER                        int64  \n",
      " 30  M_PIT_ASSIST                                    int64  \n",
      " 31  M_GEARBOX_ASSIST                                int64  \n",
      " 32  M_SESSION_TYPE                                  int64  \n",
      " 33  M_SPECTATOR_CAR_INDEX                           int64  \n",
      " 34  M_PIT_STOP_WINDOW_LATEST_LAP                    int64  \n",
      " 35  M_WEEKEND_LINK_IDENTIFIER                       int64  \n",
      " 36  M_DYNAMIC_RACING_LINE_TYPE                      int64  \n",
      " 37  M_SESSION_TIME_LEFT                             int64  \n",
      " 38  M_SESSION_DURATION                              int64  \n",
      " 39  M_PIT_STOP_REJOIN_POSITION                      int64  \n",
      " 40  M_WEATHER_FORECAST_SAMPLES_M_SESSION_TYPE       float64\n",
      " 41  M_TIME_OFFSET                                   float64\n",
      " 42  M_WEATHER_FORECAST_SAMPLES_M_WEATHER            float64\n",
      " 43  M_WEATHER_FORECAST_SAMPLES_M_TRACK_TEMPERATURE  float64\n",
      " 44  M_TRACK_TEMPERATURE_CHANGE                      float64\n",
      " 45  M_WEATHER_FORECAST_SAMPLES_M_AIR_TEMPERATURE    float64\n",
      " 46  M_AIR_TEMPERATURE_CHANGE                        float64\n",
      " 47  M_RAIN_PERCENTAGE                               float64\n",
      " 48  M_WEATHER                                       int64  \n",
      " 49  M_AI_DIFFICULTY                                 int64  \n",
      " 50  M_PIT_SPEED_LIMIT                               int64  \n",
      " 51  M_NETWORK_GAME                                  float64\n",
      " 52  M_TOTAL_LAPS                                    float64\n",
      " 53  M_STEERING_ASSIST                               float64\n",
      " 54  M_IS_SPECTATING                                 float64\n",
      " 55  M_DYNAMIC_RACING_LINE                           float64\n",
      " 56  M_DRSASSIST                                     float64\n",
      " 57  M_NUM_MARSHAL_ZONES                             float64\n",
      " 58  Unnamed: 58                                     float64\n",
      "dtypes: float64(21), int64(37), object(1)\n",
      "memory usage: 1.6+ GB\n"
     ]
    }
   ],
   "source": [
    "# Information about the columns\n",
    "weather_data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "30ef7613",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "M_PACKET_FORMAT                                         0\n",
       "M_GAME_MAJOR_VERSION                                    0\n",
       "M_GAME_MINOR_VERSION                                    0\n",
       "M_PACKET_VERSION                                        0\n",
       "M_PACKET_ID                                             0\n",
       "M_SESSION_UID                                           0\n",
       "M_SESSION_TIME                                          0\n",
       "M_FRAME_IDENTIFIER                                      0\n",
       "M_PLAYER_CAR_INDEX                                      0\n",
       "M_SECONDARY_PLAYER_CAR_INDEX                            0\n",
       "M_BRAKING_ASSIST                                        0\n",
       "M_SESSION_LINK_IDENTIFIER                               0\n",
       "M_PIT_RELEASE_ASSIST                                    0\n",
       "TIMESTAMP                                               0\n",
       "M_ZONE_START                                      2598054\n",
       "M_ZONE_FLAG                                       2598054\n",
       "M_PIT_STOP_WINDOW_IDEAL_LAP                             0\n",
       "M_TRACK_TEMPERATURE                                     0\n",
       "M_TRACK_LENGTH                                          0\n",
       "M_GAME_PAUSED                                           0\n",
       "M_FORECAST_ACCURACY                                     0\n",
       "GAMEHOST                                           909216\n",
       "M_AIR_TEMPERATURE                                       0\n",
       "M_NUM_WEATHER_FORECAST_SAMPLES                          0\n",
       "M_SLI_PRO_NATIVE_SUPPORT                                0\n",
       "M_SAFETY_CAR_STATUS                                     0\n",
       "M_TRACK_ID                                              0\n",
       "M_ERSASSIST                                             0\n",
       "M_FORMULA                                               0\n",
       "M_SEASON_LINK_IDENTIFIER                                0\n",
       "M_PIT_ASSIST                                            0\n",
       "M_GEARBOX_ASSIST                                        0\n",
       "M_SESSION_TYPE                                          0\n",
       "M_SPECTATOR_CAR_INDEX                                   0\n",
       "M_PIT_STOP_WINDOW_LATEST_LAP                            0\n",
       "M_WEEKEND_LINK_IDENTIFIER                               0\n",
       "M_DYNAMIC_RACING_LINE_TYPE                              0\n",
       "M_SESSION_TIME_LEFT                                     0\n",
       "M_SESSION_DURATION                                      0\n",
       "M_PIT_STOP_REJOIN_POSITION                              0\n",
       "M_WEATHER_FORECAST_SAMPLES_M_SESSION_TYPE          974274\n",
       "M_TIME_OFFSET                                      974274\n",
       "M_WEATHER_FORECAST_SAMPLES_M_WEATHER               974274\n",
       "M_WEATHER_FORECAST_SAMPLES_M_TRACK_TEMPERATURE     974274\n",
       "M_TRACK_TEMPERATURE_CHANGE                         974274\n",
       "M_WEATHER_FORECAST_SAMPLES_M_AIR_TEMPERATURE       974274\n",
       "M_AIR_TEMPERATURE_CHANGE                           974274\n",
       "M_RAIN_PERCENTAGE                                  974274\n",
       "M_WEATHER                                               0\n",
       "M_AI_DIFFICULTY                                         0\n",
       "M_PIT_SPEED_LIMIT                                       0\n",
       "M_NETWORK_GAME                                          1\n",
       "M_TOTAL_LAPS                                            1\n",
       "M_STEERING_ASSIST                                       1\n",
       "M_IS_SPECTATING                                         1\n",
       "M_DYNAMIC_RACING_LINE                                   1\n",
       "M_DRSASSIST                                             1\n",
       "M_NUM_MARSHAL_ZONES                                     1\n",
       "Unnamed: 58                                       3572328\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# number of null values in every column\n",
    "weather_data.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f0b7190f",
   "metadata": {},
   "source": [
    "### Column based EDA \n",
    "(Viewing statistics for rest over columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "08a58cae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "M_PACKET_FORMAT                                       1\n",
       "M_GAME_MAJOR_VERSION                                  1\n",
       "M_GAME_MINOR_VERSION                                  2\n",
       "M_PACKET_VERSION                                      1\n",
       "M_PACKET_ID                                           1\n",
       "M_SESSION_UID                                       153\n",
       "M_SESSION_TIME                                    44500\n",
       "M_FRAME_IDENTIFIER                                28603\n",
       "M_PLAYER_CAR_INDEX                                    8\n",
       "M_SECONDARY_PLAYER_CAR_INDEX                          1\n",
       "M_BRAKING_ASSIST                                      3\n",
       "M_SESSION_LINK_IDENTIFIER                            50\n",
       "M_PIT_RELEASE_ASSIST                                  2\n",
       "TIMESTAMP                                         45718\n",
       "M_ZONE_START                                        192\n",
       "M_ZONE_FLAG                                           4\n",
       "M_PIT_STOP_WINDOW_IDEAL_LAP                          12\n",
       "M_TRACK_TEMPERATURE                                  12\n",
       "M_TRACK_LENGTH                                       12\n",
       "M_GAME_PAUSED                                         2\n",
       "M_FORECAST_ACCURACY                                   2\n",
       "GAMEHOST                                              2\n",
       "M_AIR_TEMPERATURE                                    13\n",
       "M_NUM_WEATHER_FORECAST_SAMPLES                        8\n",
       "M_SLI_PRO_NATIVE_SUPPORT                              1\n",
       "M_SAFETY_CAR_STATUS                                   1\n",
       "M_TRACK_ID                                           13\n",
       "M_ERSASSIST                                           2\n",
       "M_FORMULA                                             2\n",
       "M_SEASON_LINK_IDENTIFIER                             33\n",
       "M_PIT_ASSIST                                          2\n",
       "M_GEARBOX_ASSIST                                      2\n",
       "M_SESSION_TYPE                                        6\n",
       "M_SPECTATOR_CAR_INDEX                                 4\n",
       "M_PIT_STOP_WINDOW_LATEST_LAP                         11\n",
       "M_WEEKEND_LINK_IDENTIFIER                            38\n",
       "M_DYNAMIC_RACING_LINE_TYPE                            2\n",
       "M_SESSION_TIME_LEFT                                3497\n",
       "M_SESSION_DURATION                                    5\n",
       "M_PIT_STOP_REJOIN_POSITION                           21\n",
       "M_WEATHER_FORECAST_SAMPLES_M_SESSION_TYPE            10\n",
       "M_TIME_OFFSET                                         9\n",
       "M_WEATHER_FORECAST_SAMPLES_M_WEATHER                  6\n",
       "M_WEATHER_FORECAST_SAMPLES_M_TRACK_TEMPERATURE       14\n",
       "M_TRACK_TEMPERATURE_CHANGE                            3\n",
       "M_WEATHER_FORECAST_SAMPLES_M_AIR_TEMPERATURE         14\n",
       "M_AIR_TEMPERATURE_CHANGE                              3\n",
       "M_RAIN_PERCENTAGE                                    32\n",
       "M_WEATHER                                             4\n",
       "M_AI_DIFFICULTY                                       5\n",
       "M_PIT_SPEED_LIMIT                                     2\n",
       "M_NETWORK_GAME                                        2\n",
       "M_TOTAL_LAPS                                          8\n",
       "M_STEERING_ASSIST                                     2\n",
       "M_IS_SPECTATING                                       2\n",
       "M_DYNAMIC_RACING_LINE                                 2\n",
       "M_DRSASSIST                                           2\n",
       "M_NUM_MARSHAL_ZONES                                   7\n",
       "Unnamed: 58                                           0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# number of unique values in every column\n",
    "weather_data.nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "16b7809b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# drop all those columns, which have only one unique value\n",
    "weather_data_updated=weather_data\n",
    "for col in weather_data.columns:\n",
    "    if len(weather_data[col].unique()) == 1 and weather_data.shape[0]==weather_data[col].shape[0]:\n",
    "        weather_data_updated = weather_data_updated.drop(col,axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "8d36aa3e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Remaining Columns= 51\n"
     ]
    }
   ],
   "source": [
    "print(\"Remaining Columns=\",weather_data_updated.shape[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "3a4858da",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop those rows where num_weather_forecast_samples=0 and session type is 0\n",
    "weather_data_updated= weather_data_updated[(weather_data_updated['M_NUM_WEATHER_FORECAST_SAMPLES'] != 0)\n",
    "                                           & (weather_data_updated['M_SESSION_TYPE'] != 0)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "32279c70",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Remaining Rows= 2745117\n"
     ]
    }
   ],
   "source": [
    "print(\"Remaining Rows=\",weather_data_updated.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "7f54e2e9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    1880022\n",
       "1     763609\n",
       "2     101486\n",
       "Name: M_WEATHER, dtype: int64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# finding the count of every unique value in weather's column. its gonna be our target varriable\n",
    "weather_data_updated['M_WEATHER'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "a63760ba",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "M_GAME_MINOR_VERSION                                   NaN\n",
       "M_SESSION_UID                                     0.032278\n",
       "M_SESSION_TIME                                    0.213398\n",
       "M_FRAME_IDENTIFIER                                0.235064\n",
       "M_PLAYER_CAR_INDEX                               -0.528437\n",
       "M_BRAKING_ASSIST                                  0.129839\n",
       "M_SESSION_LINK_IDENTIFIER                         0.122606\n",
       "M_PIT_RELEASE_ASSIST                              0.150312\n",
       "TIMESTAMP                                        -0.306690\n",
       "M_ZONE_START                                      0.014273\n",
       "M_ZONE_FLAG                                       0.062567\n",
       "M_PIT_STOP_WINDOW_IDEAL_LAP                       0.228704\n",
       "M_TRACK_TEMPERATURE                              -0.703760\n",
       "M_TRACK_LENGTH                                    0.461109\n",
       "M_GAME_PAUSED                                     0.070961\n",
       "M_FORECAST_ACCURACY                              -0.105276\n",
       "M_AIR_TEMPERATURE                                -0.639371\n",
       "M_NUM_WEATHER_FORECAST_SAMPLES                   -0.074030\n",
       "M_TRACK_ID                                       -0.499124\n",
       "M_ERSASSIST                                       0.150312\n",
       "M_FORMULA                                        -0.038154\n",
       "M_SEASON_LINK_IDENTIFIER                          0.122606\n",
       "M_PIT_ASSIST                                      0.150312\n",
       "M_GEARBOX_ASSIST                                  0.150312\n",
       "M_SESSION_TYPE                                    0.119439\n",
       "M_SPECTATOR_CAR_INDEX                             0.028212\n",
       "M_PIT_STOP_WINDOW_LATEST_LAP                      0.231594\n",
       "M_WEEKEND_LINK_IDENTIFIER                         0.122606\n",
       "M_DYNAMIC_RACING_LINE_TYPE                        0.150312\n",
       "M_SESSION_TIME_LEFT                              -0.069390\n",
       "M_SESSION_DURATION                                0.136196\n",
       "M_PIT_STOP_REJOIN_POSITION                        0.068194\n",
       "M_WEATHER_FORECAST_SAMPLES_M_SESSION_TYPE        -0.015481\n",
       "M_TIME_OFFSET                                    -0.008562\n",
       "M_WEATHER_FORECAST_SAMPLES_M_WEATHER              0.126950\n",
       "M_WEATHER_FORECAST_SAMPLES_M_TRACK_TEMPERATURE   -0.062495\n",
       "M_TRACK_TEMPERATURE_CHANGE                       -0.051686\n",
       "M_WEATHER_FORECAST_SAMPLES_M_AIR_TEMPERATURE     -0.072029\n",
       "M_AIR_TEMPERATURE_CHANGE                         -0.055896\n",
       "M_RAIN_PERCENTAGE                                 0.119695\n",
       "M_WEATHER                                         1.000000\n",
       "M_AI_DIFFICULTY                                  -0.562013\n",
       "M_PIT_SPEED_LIMIT                                -0.323681\n",
       "M_NETWORK_GAME                                    0.294340\n",
       "M_TOTAL_LAPS                                     -0.119809\n",
       "M_STEERING_ASSIST                                -0.037586\n",
       "M_IS_SPECTATING                                  -0.028212\n",
       "M_DYNAMIC_RACING_LINE                             0.150312\n",
       "M_DRSASSIST                                       0.150312\n",
       "M_NUM_MARSHAL_ZONES                               0.302897\n",
       "dtype: float64"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Correlation of every column with target varriable M_WEATHER\n",
    "weather_data_updated.corrwith(weather_data_updated[\"M_WEATHER\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aa77b135",
   "metadata": {},
   "source": [
    "### ADS DatasetFactory\n",
    "Some automated operations that can be performed are\n",
    "\n",
    "- Dropping constant and ID columns, which has no predictive quality.\n",
    "\n",
    "- Imputation to fill in missing values in noisy data.\n",
    "\n",
    "- Dropping strongly co-correlated columns that tend to produce less generalizable models.\n",
    "\n",
    "- Balancing a dataset using up or down sampling.\n",
    "\n",
    "<a href='#rowops'>https://docs.oracle.com/en-us/iaas/tools/ads-sdk/latest/user_guide/data_transformation/data_transformation.html</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "b6e18c87",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "loop1:   0%|          | 0/4 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "loop1:   0%|          | 0/7 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/svg+xml": [
       "<svg height=\"690pt\" viewBox=\"0.00 0.00 586.00 690.00\" width=\"586pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g class=\"graph\" id=\"graph0\" transform=\"scale(1 1) rotate(0) translate(4 686)\">\n",
       "<title>%3</title>\n",
       "<polygon fill=\"white\" points=\"-4,5 -4,-686 583,-686 583,5 -4,5\" stroke=\"white\"/>\n",
       "<!-- partitions -->\n",
       "<g class=\"node\" id=\"node1\"><title>partitions</title>\n",
       "<polygon fill=\"dimgray\" points=\"241.25,-682 104.75,-682 104.75,-686 92.75,-686 92.75,-620 241.25,-620 241.25,-682\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"92.75,-682 104.75,-682 \" stroke=\"black\"/>\n",
       "<text fill=\"white\" font-family=\"Courier,monospace\" font-size=\"12.00\" text-anchor=\"start\" x=\"110.5\" y=\"-654.4\">- rows: 2745117</text>\n",
       "<text fill=\"white\" font-family=\"Courier,monospace\" font-size=\"12.00\" text-anchor=\"start\" x=\"110.5\" y=\"-641.4\">- columns: 34</text>\n",
       "</g>\n",
       "<!-- prepare -->\n",
       "<g class=\"node\" id=\"node2\"><title>prepare</title>\n",
       "<polygon fill=\"azure\" points=\"318.25,-669 259.75,-669 259.75,-665 255.75,-665 255.75,-661 259.75,-661 259.75,-641 255.75,-641 255.75,-637 259.75,-637 259.75,-633 318.25,-633 318.25,-669\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"259.75,-665 263.75,-665 263.75,-661 259.75,-661 \" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"259.75,-641 263.75,-641 263.75,-637 259.75,-637 \" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"middle\" x=\"289\" y=\"-648.5\">prepare</text>\n",
       "</g>\n",
       "<!-- type_discovery -->\n",
       "<g class=\"node\" id=\"node3\"><title>type_discovery</title>\n",
       "<polygon fill=\"azure\" points=\"339.25,-572 238.75,-572 238.75,-568 234.75,-568 234.75,-564 238.75,-564 238.75,-544 234.75,-544 234.75,-540 238.75,-540 238.75,-536 339.25,-536 339.25,-572\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"238.75,-568 242.75,-568 242.75,-564 238.75,-564 \" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"238.75,-544 242.75,-544 242.75,-540 238.75,-540 \" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"middle\" x=\"289\" y=\"-551.5\">type_discovery</text>\n",
       "</g>\n",
       "<!-- prepare&#45;&gt;type_discovery -->\n",
       "<g class=\"edge\" id=\"edge1\"><title>prepare-&gt;type_discovery</title>\n",
       "<path d=\"M289,-632.576C289,-618.648 289,-598.861 289,-582.757\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"292.5,-582.314 289,-572.314 285.5,-582.314 292.5,-582.314\" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"289.5\" y=\"-594.5\"> </text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" font-weight=\"bold\" text-anchor=\"start\" x=\"296.5\" y=\"-594.5\">fix_column_names</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"387.5\" y=\"-594.5\"> </text>\n",
       "</g>\n",
       "<!-- recommendations -->\n",
       "<g class=\"node\" id=\"node4\"><title>recommendations</title>\n",
       "<polygon fill=\"#eec900\" points=\"578,-487.5 0,-487.5 0,-483.5 -4,-483.5 -4,-479.5 0,-479.5 0,-188.5 -4,-188.5 -4,-184.5 0,-184.5 0,-180.5 578,-180.5 578,-487.5\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"0,-483.5 4,-483.5 4,-479.5 0,-479.5 \" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"0,-188.5 4,-188.5 4,-184.5 0,-184.5 \" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"219.5\" y=\"-462.5\"> Drop constant columns:</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"216.5\" y=\"-452.5\">['M_GAME_MINOR_VERSION',</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"255.5\" y=\"-442.5\">'GAMEHOST']</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"228.5\" y=\"-432.5\">Drop: &quot;M_ZONE_START&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"231.5\" y=\"-422.5\">Drop: &quot;M_ZONE_FLAG&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"33.5\" y=\"-412.5\">Fill missing values with frequent in M_WEATHER_FORECAST_SAMPLES_M_SESSION_TYPE: &quot;0.0&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"117.5\" y=\"-402.5\">Fill missing values with frequent in M_TIME_OFFSET: &quot;0.0&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"48.5\" y=\"-392.5\">Fill missing values with frequent in M_WEATHER_FORECAST_SAMPLES_M_WEATHER: &quot;0.0&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"18.5\" y=\"-382.5\">Fill missing values with frequent in M_WEATHER_FORECAST_SAMPLES_M_TRACK_TEMPERATURE: &quot;0.0&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"78.5\" y=\"-372.5\">Fill missing values with frequent in M_TRACK_TEMPERATURE_CHANGE: &quot;0.0&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"24.5\" y=\"-362.5\">Fill missing values with frequent in M_WEATHER_FORECAST_SAMPLES_M_AIR_TEMPERATURE: &quot;0.0&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"84.5\" y=\"-352.5\">Fill missing values with frequent in M_AIR_TEMPERATURE_CHANGE: &quot;0.0&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"105.5\" y=\"-342.5\">Fill missing values with frequent in M_RAIN_PERCENTAGE: &quot;0.0&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"222.5\" y=\"-332.5\">Drop: &quot;M_SESSION_TIME&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"189.5\" y=\"-322.5\">Drop: &quot;M_WEEKEND_LINK_IDENTIFIER&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"237.5\" y=\"-312.5\">Drop: &quot;TIMESTAMP&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"183.5\" y=\"-302.5\">Drop: &quot;M_PIT_STOP_WINDOW_IDEAL_LAP&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"213.5\" y=\"-292.5\">Drop: &quot;M_AIR_TEMPERATURE&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"234.5\" y=\"-282.5\">Drop: &quot;M_TRACK_ID&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"192.5\" y=\"-272.5\">Drop: &quot;M_SEASON_LINK_IDENTIFIER&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"207.5\" y=\"-262.5\">Drop: &quot;M_SESSION_TIME_LEFT&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"186.5\" y=\"-252.5\">Drop: &quot;M_PIT_STOP_REJOIN_POSITION&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"141.5\" y=\"-242.5\">Drop: &quot;M_WEATHER_FORECAST_SAMPLES_M_SESSION_TYPE&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"156.5\" y=\"-232.5\">Drop: &quot;M_WEATHER_FORECAST_SAMPLES_M_WEATHER&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"192.5\" y=\"-222.5\">Drop: &quot;M_AIR_TEMPERATURE_CHANGE&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"186.5\" y=\"-212.5\">Drop: &quot;M_TRACK_TEMPERATURE_CHANGE&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"189.5\" y=\"-202.5\">Drop: &quot;M_AIR_TEMPERATURE_CHANGE&quot; </text>\n",
       "</g>\n",
       "<!-- type_discovery&#45;&gt;recommendations -->\n",
       "<g class=\"edge\" id=\"edge2\"><title>type_discovery-&gt;recommendations</title>\n",
       "<path d=\"M289,-535.894C289,-526.178 289,-512.955 289,-497.899\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"292.5,-497.759 289,-487.759 285.5,-497.759 292.5,-497.759\" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"289.5\" y=\"-510.5\"> </text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" font-weight=\"bold\" text-anchor=\"start\" x=\"296.5\" y=\"-510.5\">convert_columns</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"387.5\" y=\"-510.5\"> </text>\n",
       "</g>\n",
       "<!-- feature_engineering -->\n",
       "<g class=\"node\" id=\"node5\"><title>feature_engineering</title>\n",
       "<polygon fill=\"#eec900\" points=\"428,-131.5 150,-131.5 150,-127.5 146,-127.5 146,-123.5 150,-123.5 150,-92.5 146,-92.5 146,-88.5 150,-88.5 150,-84.5 428,-84.5 428,-131.5\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"150,-127.5 154,-127.5 154,-123.5 150,-123.5 \" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"150,-92.5 154,-92.5 154,-88.5 150,-88.5 \" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"168.5\" y=\"-106.5\"> No feature engineering transformations </text>\n",
       "</g>\n",
       "<!-- recommendations&#45;&gt;feature_engineering -->\n",
       "<g class=\"edge\" id=\"edge3\"><title>recommendations-&gt;feature_engineering</title>\n",
       "<path d=\"M289,-180.491C289,-166.315 289,-153.139 289,-142.06\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"292.5,-141.721 289,-131.721 285.5,-141.721 292.5,-141.721\" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"289.5\" y=\"-154.5\"> </text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" font-weight=\"bold\" text-anchor=\"start\" x=\"296.5\" y=\"-154.5\">RecommendationTransformer</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"447.5\" y=\"-154.5\"> </text>\n",
       "</g>\n",
       "<!-- terminus -->\n",
       "<g class=\"node\" id=\"node6\"><title>terminus</title>\n",
       "<polygon fill=\"white\" points=\"292,-18 292,-24 298,-24 298,-30 280,-30 280,-24 286,-24 286,-18 292,-18\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"262,-18 316,-18 \" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- feature_engineering&#45;&gt;terminus -->\n",
       "<g class=\"edge\" id=\"edge4\"><title>feature_engineering-&gt;terminus</title>\n",
       "<path d=\"M289,-84.3378C289,-72.6791 289,-58.3831 289,-46.1394\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"292.5,-46.0259 289,-36.026 285.5,-46.026 292.5,-46.0259\" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"289.5\" y=\"-58.5\"> </text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" font-weight=\"bold\" text-anchor=\"start\" x=\"296.5\" y=\"-58.5\">FeatureEngineeringTransformer</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"471.5\" y=\"-58.5\"> </text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Taking Suggestions from Dataset Facorty about our data cleaning\n",
    "dataSetFactorySuggestions = DatasetFactory.open(weather_data_updated, target=\"M_WEATHER\")\n",
    "suggestedData = dataSetFactorySuggestions.auto_transform()\n",
    "suggestedData.visualize_transforms()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "91a0dbb4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# dropping suggested columns\n",
    "weather_data_updated=weather_data_updated.drop(labels=['GAMEHOST','M_GAME_MINOR_VERSION','M_ZONE_START',\n",
    "                                        'M_ZONE_FLAG','M_SESSION_TIME',\n",
    "                                        'M_WEEKEND_LINK_IDENTIFIER','TIMESTAMP',\n",
    "                                        'M_PIT_STOP_WINDOW_IDEAL_LAP','M_AIR_TEMPERATURE',\n",
    "                                        'M_TRACK_ID','M_SEASON_LINK_IDENTIFIER','M_SESSION_TIME_LEFT',\n",
    "                                        'M_PIT_STOP_REJOIN_POSITION'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "09bc8fea",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Remaining Columns= 38\n"
     ]
    }
   ],
   "source": [
    "print(\"Remaining Columns=\",weather_data_updated.shape[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "218a4f74",
   "metadata": {},
   "source": [
    "### Imputation Actions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "46a8a32e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Fill the null values with suggested technique\n",
    "columns_list=['M_WEATHER_FORECAST_SAMPLES_M_SESSION_TYPE','M_TIME_OFFSET','M_WEATHER_FORECAST_SAMPLES_M_WEATHER',\n",
    "             'M_WEATHER_FORECAST_SAMPLES_M_TRACK_TEMPERATURE','M_TRACK_TEMPERATURE_CHANGE',\n",
    "              'M_WEATHER_FORECAST_SAMPLES_M_AIR_TEMPERATURE','M_AIR_TEMPERATURE_CHANGE','M_RAIN_PERCENTAGE']\n",
    "for col in columns_list:\n",
    "     weather_data_updated[col].fillna(0.0, inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "ca8c87b8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "M_SESSION_UID                                     0\n",
       "M_FRAME_IDENTIFIER                                0\n",
       "M_PLAYER_CAR_INDEX                                0\n",
       "M_BRAKING_ASSIST                                  0\n",
       "M_SESSION_LINK_IDENTIFIER                         0\n",
       "M_PIT_RELEASE_ASSIST                              0\n",
       "M_TRACK_TEMPERATURE                               0\n",
       "M_TRACK_LENGTH                                    0\n",
       "M_GAME_PAUSED                                     0\n",
       "M_FORECAST_ACCURACY                               0\n",
       "M_NUM_WEATHER_FORECAST_SAMPLES                    0\n",
       "M_ERSASSIST                                       0\n",
       "M_FORMULA                                         0\n",
       "M_PIT_ASSIST                                      0\n",
       "M_GEARBOX_ASSIST                                  0\n",
       "M_SESSION_TYPE                                    0\n",
       "M_SPECTATOR_CAR_INDEX                             0\n",
       "M_PIT_STOP_WINDOW_LATEST_LAP                      0\n",
       "M_DYNAMIC_RACING_LINE_TYPE                        0\n",
       "M_SESSION_DURATION                                0\n",
       "M_WEATHER_FORECAST_SAMPLES_M_SESSION_TYPE         0\n",
       "M_TIME_OFFSET                                     0\n",
       "M_WEATHER_FORECAST_SAMPLES_M_WEATHER              0\n",
       "M_WEATHER_FORECAST_SAMPLES_M_TRACK_TEMPERATURE    0\n",
       "M_TRACK_TEMPERATURE_CHANGE                        0\n",
       "M_WEATHER_FORECAST_SAMPLES_M_AIR_TEMPERATURE      0\n",
       "M_AIR_TEMPERATURE_CHANGE                          0\n",
       "M_RAIN_PERCENTAGE                                 0\n",
       "M_WEATHER                                         0\n",
       "M_AI_DIFFICULTY                                   0\n",
       "M_PIT_SPEED_LIMIT                                 0\n",
       "M_NETWORK_GAME                                    1\n",
       "M_TOTAL_LAPS                                      1\n",
       "M_STEERING_ASSIST                                 1\n",
       "M_IS_SPECTATING                                   1\n",
       "M_DYNAMIC_RACING_LINE                             1\n",
       "M_DRSASSIST                                       1\n",
       "M_NUM_MARSHAL_ZONES                               1\n",
       "dtype: int64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check the null values in every column\n",
    "weather_data_updated.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "4f1bc5f1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# imputing those columns where still nan values exist. Impute with most frquent value\n",
    "list_of_MoreColumns_to_impute=['M_NETWORK_GAME','M_TOTAL_LAPS','M_STEERING_ASSIST'\n",
    "                              ,'M_IS_SPECTATING','M_DYNAMIC_RACING_LINE','M_DRSASSIST',\n",
    "                              'M_NUM_MARSHAL_ZONES']\n",
    "for col in list_of_MoreColumns_to_impute:\n",
    "    weather_data_updated[col].fillna(weather_data_updated[col].value_counts().index[0], inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6378bdba",
   "metadata": {},
   "source": [
    "### Normalizing Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "6d16bb24",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import preprocessing\n",
    "\n",
    "dataFram_without_weather=resultantData.loc[:, resultantData.columns!='M_WEATHER']\n",
    "array_of_data = dataFram_without_weather.values #returns a numpy array\n",
    "min_max_scaler = preprocessing.MinMaxScaler()\n",
    "array_of_data_scaled = min_max_scaler.fit_transform(array_of_data)\n",
    "Final_data_frame = pd.DataFrame(array_of_data_scaled,columns=dataFram_without_weather.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "0df55809",
   "metadata": {},
   "outputs": [],
   "source": [
    "Final_data_frame=weather_data_updated"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5b9a2950",
   "metadata": {},
   "source": [
    "### Correlation evaluation with Target Variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "9e1ebe22",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "M_SESSION_UID                                     0.032278\n",
       "M_FRAME_IDENTIFIER                                0.235064\n",
       "M_PLAYER_CAR_INDEX                               -0.528437\n",
       "M_BRAKING_ASSIST                                  0.129839\n",
       "M_SESSION_LINK_IDENTIFIER                         0.122606\n",
       "M_PIT_RELEASE_ASSIST                              0.150312\n",
       "M_TRACK_TEMPERATURE                              -0.703760\n",
       "M_TRACK_LENGTH                                    0.461109\n",
       "M_GAME_PAUSED                                     0.070961\n",
       "M_FORECAST_ACCURACY                              -0.105276\n",
       "M_NUM_WEATHER_FORECAST_SAMPLES                   -0.074030\n",
       "M_ERSASSIST                                       0.150312\n",
       "M_FORMULA                                        -0.038154\n",
       "M_PIT_ASSIST                                      0.150312\n",
       "M_GEARBOX_ASSIST                                  0.150312\n",
       "M_SESSION_TYPE                                    0.119439\n",
       "M_SPECTATOR_CAR_INDEX                             0.028212\n",
       "M_PIT_STOP_WINDOW_LATEST_LAP                      0.231594\n",
       "M_DYNAMIC_RACING_LINE_TYPE                        0.150312\n",
       "M_SESSION_DURATION                                0.136196\n",
       "M_WEATHER_FORECAST_SAMPLES_M_SESSION_TYPE        -0.012612\n",
       "M_TIME_OFFSET                                    -0.007132\n",
       "M_WEATHER_FORECAST_SAMPLES_M_WEATHER              0.106547\n",
       "M_WEATHER_FORECAST_SAMPLES_M_TRACK_TEMPERATURE   -0.050163\n",
       "M_TRACK_TEMPERATURE_CHANGE                       -0.041778\n",
       "M_WEATHER_FORECAST_SAMPLES_M_AIR_TEMPERATURE     -0.057824\n",
       "M_AIR_TEMPERATURE_CHANGE                         -0.045154\n",
       "M_RAIN_PERCENTAGE                                 0.099948\n",
       "M_WEATHER                                         1.000000\n",
       "M_AI_DIFFICULTY                                  -0.562013\n",
       "M_PIT_SPEED_LIMIT                                -0.323681\n",
       "M_NETWORK_GAME                                    0.294340\n",
       "M_TOTAL_LAPS                                     -0.119809\n",
       "M_STEERING_ASSIST                                -0.037586\n",
       "M_IS_SPECTATING                                  -0.028212\n",
       "M_DYNAMIC_RACING_LINE                             0.150312\n",
       "M_DRSASSIST                                       0.150312\n",
       "M_NUM_MARSHAL_ZONES                               0.302897\n",
       "dtype: float64"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Correlation of every column with target varriable M_WEATHER\n",
    "Final_data_frame.corrwith(Final_data_frame[\"M_WEATHER\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "712e4ea6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "loop1:   0%|          | 0/4 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "loop1:   0%|          | 0/7 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/svg+xml": [
       "<svg height=\"470pt\" viewBox=\"0.00 0.00 394.00 470.00\" width=\"394pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g class=\"graph\" id=\"graph0\" transform=\"scale(1 1) rotate(0) translate(4 466)\">\n",
       "<title>%3</title>\n",
       "<polygon fill=\"white\" points=\"-4,5 -4,-466 391,-466 391,5 -4,5\" stroke=\"white\"/>\n",
       "<!-- partitions -->\n",
       "<g class=\"node\" id=\"node1\"><title>partitions</title>\n",
       "<polygon fill=\"dimgray\" points=\"148.25,-462 11.75,-462 11.75,-466 -0.25,-466 -0.25,-400 148.25,-400 148.25,-462\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"-0.25,-462 11.75,-462 \" stroke=\"black\"/>\n",
       "<text fill=\"white\" font-family=\"Courier,monospace\" font-size=\"12.00\" text-anchor=\"start\" x=\"17.5\" y=\"-434.4\">- rows: 2745117</text>\n",
       "<text fill=\"white\" font-family=\"Courier,monospace\" font-size=\"12.00\" text-anchor=\"start\" x=\"17.5\" y=\"-421.4\">- columns: 34</text>\n",
       "</g>\n",
       "<!-- prepare -->\n",
       "<g class=\"node\" id=\"node2\"><title>prepare</title>\n",
       "<polygon fill=\"azure\" points=\"225.25,-449 166.75,-449 166.75,-445 162.75,-445 162.75,-441 166.75,-441 166.75,-421 162.75,-421 162.75,-417 166.75,-417 166.75,-413 225.25,-413 225.25,-449\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"166.75,-445 170.75,-445 170.75,-441 166.75,-441 \" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"166.75,-421 170.75,-421 170.75,-417 166.75,-417 \" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"middle\" x=\"196\" y=\"-428.5\">prepare</text>\n",
       "</g>\n",
       "<!-- type_discovery -->\n",
       "<g class=\"node\" id=\"node3\"><title>type_discovery</title>\n",
       "<polygon fill=\"azure\" points=\"246.25,-352 145.75,-352 145.75,-348 141.75,-348 141.75,-344 145.75,-344 145.75,-324 141.75,-324 141.75,-320 145.75,-320 145.75,-316 246.25,-316 246.25,-352\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"145.75,-348 149.75,-348 149.75,-344 145.75,-344 \" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"145.75,-324 149.75,-324 149.75,-320 145.75,-320 \" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"middle\" x=\"196\" y=\"-331.5\">type_discovery</text>\n",
       "</g>\n",
       "<!-- prepare&#45;&gt;type_discovery -->\n",
       "<g class=\"edge\" id=\"edge1\"><title>prepare-&gt;type_discovery</title>\n",
       "<path d=\"M196,-412.576C196,-398.648 196,-378.861 196,-362.757\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"199.5,-362.314 196,-352.314 192.5,-362.314 199.5,-362.314\" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"196.5\" y=\"-374.5\"> </text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" font-weight=\"bold\" text-anchor=\"start\" x=\"203.5\" y=\"-374.5\">fix_column_names</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"294.5\" y=\"-374.5\"> </text>\n",
       "</g>\n",
       "<!-- recommendations -->\n",
       "<g class=\"node\" id=\"node4\"><title>recommendations</title>\n",
       "<polygon fill=\"#eec900\" points=\"365,-267.5 27,-267.5 27,-263.5 23,-263.5 23,-259.5 27,-259.5 27,-188.5 23,-188.5 23,-184.5 27,-184.5 27,-180.5 365,-180.5 365,-267.5\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"27,-263.5 31,-263.5 31,-259.5 27,-259.5 \" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"27,-188.5 31,-188.5 31,-184.5 27,-184.5 \" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"45.5\" y=\"-242.5\"> Drop: &quot;M_WEATHER_FORECAST_SAMPLES_M_SESSION_TYPE&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"63.5\" y=\"-232.5\">Drop: &quot;M_WEATHER_FORECAST_SAMPLES_M_WEATHER&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"99.5\" y=\"-222.5\">Drop: &quot;M_AIR_TEMPERATURE_CHANGE&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"93.5\" y=\"-212.5\">Drop: &quot;M_TRACK_TEMPERATURE_CHANGE&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"96.5\" y=\"-202.5\">Drop: &quot;M_AIR_TEMPERATURE_CHANGE&quot; </text>\n",
       "</g>\n",
       "<!-- type_discovery&#45;&gt;recommendations -->\n",
       "<g class=\"edge\" id=\"edge2\"><title>type_discovery-&gt;recommendations</title>\n",
       "<path d=\"M196,-315.648C196,-305.385 196,-291.777 196,-278.234\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"199.5,-277.946 196,-267.946 192.5,-277.946 199.5,-277.946\" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"196.5\" y=\"-290.5\"> </text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" font-weight=\"bold\" text-anchor=\"start\" x=\"203.5\" y=\"-290.5\">convert_columns</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"294.5\" y=\"-290.5\"> </text>\n",
       "</g>\n",
       "<!-- feature_engineering -->\n",
       "<g class=\"node\" id=\"node5\"><title>feature_engineering</title>\n",
       "<polygon fill=\"#eec900\" points=\"335,-131.5 57,-131.5 57,-127.5 53,-127.5 53,-123.5 57,-123.5 57,-92.5 53,-92.5 53,-88.5 57,-88.5 57,-84.5 335,-84.5 335,-131.5\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"57,-127.5 61,-127.5 61,-123.5 57,-123.5 \" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"57,-92.5 61,-92.5 61,-88.5 57,-88.5 \" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"75.5\" y=\"-106.5\"> No feature engineering transformations </text>\n",
       "</g>\n",
       "<!-- recommendations&#45;&gt;feature_engineering -->\n",
       "<g class=\"edge\" id=\"edge3\"><title>recommendations-&gt;feature_engineering</title>\n",
       "<path d=\"M196,-180.226C196,-167.424 196,-153.598 196,-141.564\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"199.5,-141.522 196,-131.522 192.5,-141.522 199.5,-141.522\" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"196.5\" y=\"-154.5\"> </text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" font-weight=\"bold\" text-anchor=\"start\" x=\"203.5\" y=\"-154.5\">RecommendationTransformer</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"354.5\" y=\"-154.5\"> </text>\n",
       "</g>\n",
       "<!-- terminus -->\n",
       "<g class=\"node\" id=\"node6\"><title>terminus</title>\n",
       "<polygon fill=\"white\" points=\"199,-18 199,-24 205,-24 205,-30 187,-30 187,-24 193,-24 193,-18 199,-18\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"169,-18 223,-18 \" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- feature_engineering&#45;&gt;terminus -->\n",
       "<g class=\"edge\" id=\"edge4\"><title>feature_engineering-&gt;terminus</title>\n",
       "<path d=\"M196,-84.3378C196,-72.6791 196,-58.3831 196,-46.1394\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"199.5,-46.0259 196,-36.026 192.5,-46.026 199.5,-46.0259\" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"196.5\" y=\"-58.5\"> </text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" font-weight=\"bold\" text-anchor=\"start\" x=\"203.5\" y=\"-58.5\">FeatureEngineeringTransformer</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"378.5\" y=\"-58.5\"> </text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dataSetFactoryFinal = DatasetFactory.open(Final_data_frame, target=\"M_WEATHER\")\n",
    "dataset_transformed_final = dataSetFactoryFinal.auto_transform()\n",
    "dataset_transformed_final.visualize_transforms()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "202b9cd4",
   "metadata": {},
   "outputs": [],
   "source": [
    "train, test = dataSetFactoryFinal.train_test_split(test_size=0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6c88b1e",
   "metadata": {},
   "source": [
    "# ADS and AutoML"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "8ca81b9d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:Local ProcessPool execution (n_jobs=2)\n",
      "WARNING:Random state (42) is used for model builds\n",
      "WARNING:Dataset shape: (1756874, 37)\n",
      "WARNING:Running Auto-Preprocessing\n",
      "WARNING:Preprocessing completed. Updated Dataset shape: (1756874, 37), cv: None\n",
      "WARNING:SVC, KNeighborsClassifier are disabled for datasets with > 10K samples or > 1K features\n",
      "WARNING:Running Model Selection\n",
      "WARNING:Time budget exceeded by 4.16s, resetting XEngine\n",
      "WARNING:Timebudget Exceeded or Timedout completed 7/8, 0 tasks timedout\n",
      "WARNING:Model Selection completed. Selected model: ['RandomForestClassifier']\n",
      "WARNING:Running Adaptive Sampling. Dataset Shape: (1756874, 37), Valid Shape: (439219, 37), CV: None, Class counts: [1203252  488765   64857]\n",
      "WARNING:Timebudget exhausted. Skipping Adaptive Sampling\n",
      "WARNING:Adaptive Sampling Completed. Updated Dataset Shape: (1756874, 37), Valid Shape: (439219, 37), CV: None, Class counts: [1203252  488765   64857]\n",
      "WARNING:Starting Feature Selection 0. Dataset Shape: (1756874, 37)\n",
      "WARNING:Timebudget exhausted. Skipping Feature Selection\n",
      "WARNING:Using all features: Index(['M_SESSION_UID', 'M_FRAME_IDENTIFIER', 'M_PLAYER_CAR_INDEX',\n",
      "       'M_BRAKING_ASSIST', 'M_SESSION_LINK_IDENTIFIER', 'M_PIT_RELEASE_ASSIST',\n",
      "       'M_TRACK_TEMPERATURE', 'M_TRACK_LENGTH', 'M_GAME_PAUSED',\n",
      "       'M_FORECAST_ACCURACY', 'M_NUM_WEATHER_FORECAST_SAMPLES', 'M_ERSASSIST',\n",
      "       'M_FORMULA', 'M_PIT_ASSIST', 'M_GEARBOX_ASSIST', 'M_SESSION_TYPE',\n",
      "       'M_SPECTATOR_CAR_INDEX', 'M_PIT_STOP_WINDOW_LATEST_LAP',\n",
      "       'M_DYNAMIC_RACING_LINE_TYPE', 'M_SESSION_DURATION',\n",
      "       'M_WEATHER_FORECAST_SAMPLES_M_SESSION_TYPE', 'M_TIME_OFFSET',\n",
      "       'M_WEATHER_FORECAST_SAMPLES_M_WEATHER',\n",
      "       'M_WEATHER_FORECAST_SAMPLES_M_TRACK_TEMPERATURE',\n",
      "       'M_TRACK_TEMPERATURE_CHANGE',\n",
      "       'M_WEATHER_FORECAST_SAMPLES_M_AIR_TEMPERATURE',\n",
      "       'M_AIR_TEMPERATURE_CHANGE', 'M_RAIN_PERCENTAGE', 'M_AI_DIFFICULTY',\n",
      "       'M_PIT_SPEED_LIMIT', 'M_NETWORK_GAME', 'M_TOTAL_LAPS',\n",
      "       'M_STEERING_ASSIST', 'M_IS_SPECTATING', 'M_DYNAMIC_RACING_LINE',\n",
      "       'M_DRSASSIST', 'M_NUM_MARSHAL_ZONES'],\n",
      "      dtype='object')\n",
      "WARNING:Feature Selection 0 completed. Updated Dataset shape: (1756874, 37)\n",
      "WARNING:Timebudget exhausted. Skipping Hyperparameter Optimization for RandomForestClassifier\n",
      "WARNING:(Re)fitting Pipeline\n",
      "WARNING:Running Auto-Preprocessing\n",
      "WARNING:Preprocessing completed. Updated Dataset shape: (2196093, 37), cv: auto\n",
      "WARNING:Local ProcessPool execution (n_jobs=2)\n",
      "WARNING:AutoML completed. Time taken - 637.908 sec\n",
      "ERROR:Unfortunately, there were no trials found, so we cannot visualize it.\n"
     ]
    }
   ],
   "source": [
    "# Now, let's see how well AutoML can perform the same task.\n",
    "# AutoML will evaluate several algorithms in combination with different values for their \n",
    "# hyperparameters.\n",
    "# Note the time parameter of 160 seconds, this means we allow AutoML this amount of time to find the best solution.\n",
    "\n",
    "from ads.automl.provider import OracleAutoMLProvider\n",
    "from ads.automl.driver import AutoML\n",
    "ml_engine = OracleAutoMLProvider(n_jobs = -1) # n_jobs = -1 means that we are not going to limit the resources and let the AutoML to do this for us and take as much as possible\n",
    "automl = AutoML(train, provider=ml_engine)\n",
    "automl_model1, baseline = automl.train(time_budget=160, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "01895f18",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>Algorithm</th>\n",
       "      <th>#Samples</th>\n",
       "      <th>#Features</th>\n",
       "      <th>Mean Validation Score</th>\n",
       "      <th>Hyperparameters</th>\n",
       "      <th>CPU Time</th>\n",
       "      <th>Memory Usage (GB)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>RandomForestClassifier_AS</td>\n",
       "      <td>1756874</td>\n",
       "      <td>37</td>\n",
       "      <td>-0.0018</td>\n",
       "      <td>{'n_estimators': 100, 'class_weight': 'balanced', 'max_features': 0.777777778, 'min_samples_leaf': 0.000625, 'min_samples_split': 0.00125}</td>\n",
       "      <td>33.7094</td>\n",
       "      <td>2.9697e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>LGBMClassifier_AS</td>\n",
       "      <td>1756874</td>\n",
       "      <td>37</td>\n",
       "      <td>-0.0018</td>\n",
       "      <td>{'boosting_type': 'gbdt', 'learning_rate': 0.1, 'max_depth': -1, 'min_child_weight': 0.001, 'n_estimators': 100, 'num_leaves': 31, 'reg_alpha': 0, 'reg_lambda': 1, 'class_weight': 'balanced'}</td>\n",
       "      <td>10.5318</td>\n",
       "      <td>1.4830e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>DecisionTreeClassifier_AS</td>\n",
       "      <td>1756874</td>\n",
       "      <td>37</td>\n",
       "      <td>-0.0021</td>\n",
       "      <td>{'class_weight': None, 'max_features': 1.0, 'min_samples_leaf': 0.000625, 'min_samples_split': 0.00125}</td>\n",
       "      <td>0.9453</td>\n",
       "      <td>6.1148e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>ExtraTreesClassifier_AS</td>\n",
       "      <td>1756874</td>\n",
       "      <td>37</td>\n",
       "      <td>-0.0049</td>\n",
       "      <td>{'n_estimators': 100, 'class_weight': 'balanced', 'criterion': 'gini', 'max_features': 0.777777778, 'min_samples_leaf': 0.000625, 'min_samples_split': 0.00125}</td>\n",
       "      <td>29.3446</td>\n",
       "      <td>5.7560e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>XGBClassifier_AS</td>\n",
       "      <td>1756874</td>\n",
       "      <td>37</td>\n",
       "      <td>-0.0054</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>44.3426</td>\n",
       "      <td>1.0722e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>LogisticRegression_AS</td>\n",
       "      <td>1756874</td>\n",
       "      <td>37</td>\n",
       "      <td>-0.0626</td>\n",
       "      <td>{'C': 1.0, 'class_weight': 'balanced', 'solver': 'liblinear'}</td>\n",
       "      <td>26.6323</td>\n",
       "      <td>1.2848e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>GaussianNB_AS</td>\n",
       "      <td>1756874</td>\n",
       "      <td>37</td>\n",
       "      <td>-6.3571</td>\n",
       "      <td>{}</td>\n",
       "      <td>0.3911</td>\n",
       "      <td>7.6294e-06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "automl.print_trials(max_rows=10, sort_column='Mean Validation Score')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "29cd555b",
   "metadata": {},
   "source": [
    "### Visualization and Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "09f32043",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig1 =automl.visualize_algorithm_selection_trials()\n",
    "import matplotlib.pyplot as plt\n",
    "plt.savefig('fig1')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "c230eb32",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig2 = automl.visualize_feature_selection_trials()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "273d364e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.savefig('fig2')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "01e276ba",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:Random state (7) is used for model builds\n",
      "WARNING:Dataset shape: (1756874, 37)\n",
      "WARNING:Running Auto-Preprocessing\n",
      "WARNING:Preprocessing completed. Updated Dataset shape: (1756874, 37), cv: None\n",
      "WARNING:SVC, KNeighborsClassifier are disabled for datasets with > 10K samples or > 1K features\n",
      "WARNING:Running Model Selection\n",
      "WARNING:Algorithm Selection Disabled\n",
      "WARNING:User provided model RandomForestClassifier is the only desired model\n",
      "WARNING:Model Selection completed. Selected model: ['RandomForestClassifier']\n",
      "WARNING:Running Adaptive Sampling. Dataset Shape: (1756874, 37), Valid Shape: (439219, 37), CV: None, Class counts: [1203252  488765   64857]\n",
      "WARNING:Adaptive Sampling Completed. Updated Dataset Shape: (647383, 37), Valid Shape: (77593, 37), CV: None, Class counts: [291263 291263  64857]\n",
      "WARNING:Starting Feature Selection 0. Dataset Shape: (647383, 37)\n",
      "WARNING:Feature Selection 0 completed. Updated Dataset shape: (647383, 9)\n",
      "WARNING:Tuning RandomForestClassifier\n",
      "WARNING:Tuning completed. Best params: {'class_weight': 'balanced_subsample', 'max_features': 0.1111111111111111, 'min_samples_leaf': 1.544680660443663e-06, 'min_samples_split': 3.089361320887326e-06, 'n_estimators': 5}\n",
      "WARNING:(Re)fitting Pipeline\n",
      "WARNING:Running Auto-Preprocessing\n",
      "WARNING:Preprocessing completed. Updated Dataset shape: (2196093, 37), cv: auto\n",
      "WARNING:Local ProcessPool execution (n_jobs=2)\n",
      "WARNING:AutoML completed. Time taken - 6586.258 sec\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>Training Dataset size</td>\n",
       "      <td>(2196093, 37)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Validation Dataset size</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>CV</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Target variable</td>\n",
       "      <td>M_WEATHER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Optimization Metric</td>\n",
       "      <td>neg_log_loss</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Initial number of Features</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected number of Features</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected Features</td>\n",
       "      <td>Index(['M_SESSION_UID', 'M_FRAME_IDENTIFIER', 'M_SESSION_LINK_IDENTIFIER',\n",
       "       'M_TRACK_TEMPERATURE', 'M_TRACK_LENGTH',\n",
       "       'M_NUM_WEATHER_FORECAST_SAMPLES', 'M_SESSION_TYPE',\n",
       "       'M_PIT_STOP_WINDOW_LATEST_LAP', 'M_SESSION_DURATION'],\n",
       "      dtype='object')</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected Algorithm</td>\n",
       "      <td>RandomForestClassifier</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>End-to-end Elapsed Time (seconds)</td>\n",
       "      <td>6595.3337</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected Hyperparameters</td>\n",
       "      <td>{'bootstrap': True, 'ccp_alpha': 0.0, 'class_weight': 'balanced_subsample', 'criterion': 'gini', 'max_depth': None, 'max_features': 0.1111111111111111, 'max_leaf_nodes': None, 'max_samples': None, 'min_impurity_decrease': 0.0, 'min_samples_leaf': 1.544680660443663e-06, 'min_samples_split': 3.089361320887326e-06, 'min_weight_fraction_leaf': 0.0, 'n_estimators': 5, 'n_jobs': 4, 'oob_score': False, 'random_state': 7, 'verbose': 0, 'warm_start': False}</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Mean Validation Score</td>\n",
       "      <td>-0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>AutoML n_jobs</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>AutoML version</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Python version</td>\n",
       "      <td>3.7.12 | packaged by conda-forge | (default, Oct 26 2021, 06:08:53) \\n[GCC 9.4.0]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>Rank based on Performance</th>\n",
       "      <th>Algorithm</th>\n",
       "      <th>#Samples</th>\n",
       "      <th>#Features</th>\n",
       "      <th>Mean Validation Score</th>\n",
       "      <th>Hyperparameters</th>\n",
       "      <th>CPU Time</th>\n",
       "      <th>Memory Usage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>RandomForestClassifier_HT</td>\n",
       "      <td>647383</td>\n",
       "      <td>9</td>\n",
       "      <td>-2.1094e-15</td>\n",
       "      <td>{'class_weight': None, 'max_features': 0.7777866668888889, 'min_samples_leaf': 1.544680660443663e-06, 'min_samples_split': 3.089361320887326e-06, 'n_estimators': 5}</td>\n",
       "      <td>4.4085</td>\n",
       "      <td>0.0434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>RandomForestClassifier_HT</td>\n",
       "      <td>647383</td>\n",
       "      <td>9</td>\n",
       "      <td>-2.1094e-15</td>\n",
       "      <td>{'class_weight': 'balanced_subsample', 'max_features': 0.1111111111111111, 'min_samples_leaf': 1.544680660443663e-06, 'min_samples_split': 3.089361320887326e-06, 'n_estimators': 6}</td>\n",
       "      <td>4.2161</td>\n",
       "      <td>0.0481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>RandomForestClassifier_HT</td>\n",
       "      <td>647383</td>\n",
       "      <td>9</td>\n",
       "      <td>-2.1094e-15</td>\n",
       "      <td>{'class_weight': None, 'max_features': 0.5773618158486156, 'min_samples_leaf': 1.544680660443663e-06, 'min_samples_split': 3.089361320887326e-06, 'n_estimators': 5}</td>\n",
       "      <td>3.4905</td>\n",
       "      <td>0.0434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>RandomForestClassifier_HT</td>\n",
       "      <td>647383</td>\n",
       "      <td>9</td>\n",
       "      <td>-2.1094e-15</td>\n",
       "      <td>{'class_weight': None, 'max_features': 0.5773707047375044, 'min_samples_leaf': 1.544680660443663e-06, 'min_samples_split': 3.089361320887326e-06, 'n_estimators': 5}</td>\n",
       "      <td>3.5289</td>\n",
       "      <td>0.0434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>RandomForestClassifier_HT</td>\n",
       "      <td>647383</td>\n",
       "      <td>9</td>\n",
       "      <td>-2.1094e-15</td>\n",
       "      <td>{'class_weight': None, 'max_features': 0.777777778, 'min_samples_leaf': 1.544680660443663e-06, 'min_samples_split': 3.089361320887326e-06, 'n_estimators': 5}</td>\n",
       "      <td>4.3340</td>\n",
       "      <td>0.0434</td>\n",
       "    </tr>\n",
       "    <tr>\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",
       "      <td>142</td>\n",
       "      <td>RandomForestClassifier_AdaBoostClassifier_FS</td>\n",
       "      <td>647383</td>\n",
       "      <td>3</td>\n",
       "      <td>-1.5874e-01</td>\n",
       "      <td>{'n_estimators': 100, 'class_weight': 'balanced', 'max_features': 0.777777778, 'min_samples_leaf': 0.000625, 'min_samples_split': 0.00125}</td>\n",
       "      <td>36.3696</td>\n",
       "      <td>0.0658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>143</td>\n",
       "      <td>RandomForestClassifier_HT</td>\n",
       "      <td>647383</td>\n",
       "      <td>9</td>\n",
       "      <td>-1.7126e-01</td>\n",
       "      <td>{'class_weight': None, 'max_features': 0.1111111111111111, 'min_samples_leaf': 1.544680660443663e-06, 'min_samples_split': 0.06014913107074945, 'n_estimators': 5}</td>\n",
       "      <td>2.0049</td>\n",
       "      <td>0.0434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>144</td>\n",
       "      <td>RandomForestClassifier_HT</td>\n",
       "      <td>647383</td>\n",
       "      <td>9</td>\n",
       "      <td>-1.7126e-01</td>\n",
       "      <td>{'class_weight': None, 'max_features': 0.1111111111111111, 'min_samples_leaf': 1.544680660443663e-06, 'min_samples_split': 0.06015913103985584, 'n_estimators': 5}</td>\n",
       "      <td>2.1113</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>145</td>\n",
       "      <td>RandomForestClassifier_AdaBoostClassifier_FS</td>\n",
       "      <td>647383</td>\n",
       "      <td>2</td>\n",
       "      <td>-1.8167e-01</td>\n",
       "      <td>{'n_estimators': 100, 'class_weight': 'balanced', 'max_features': 0.777777778, 'min_samples_leaf': 0.000625, 'min_samples_split': 0.00125}</td>\n",
       "      <td>23.6297</td>\n",
       "      <td>0.0534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>146</td>\n",
       "      <td>RandomForestClassifier_ANOVAF_FS</td>\n",
       "      <td>647383</td>\n",
       "      <td>1</td>\n",
       "      <td>-2.7231e-01</td>\n",
       "      <td>{'n_estimators': 100, 'class_weight': 'balanced', 'max_features': 0.777777778, 'min_samples_leaf': 0.000625, 'min_samples_split': 0.00125}</td>\n",
       "      <td>23.5087</td>\n",
       "      <td>0.0660</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "automl_model2, _ = automl.train(model_list=['RandomForestClassifier'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "a49a53ba",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:Random state (7) is used for model builds\n",
      "WARNING:Dataset shape: (1756874, 37)\n",
      "WARNING:Running Auto-Preprocessing\n",
      "WARNING:Preprocessing completed. Updated Dataset shape: (1756874, 37), cv: None\n",
      "WARNING:SVC, KNeighborsClassifier are disabled for datasets with > 10K samples or > 1K features\n",
      "WARNING:Running Model Selection\n",
      "WARNING:Model Selection completed. Selected model: ['XGBClassifier']\n",
      "WARNING:Running Adaptive Sampling. Dataset Shape: (1756874, 37), Valid Shape: (439219, 37), CV: None, Class counts: [1203252  488765   64857]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[18:39:53] WARNING: ../src/learner.cc:1115: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:Adaptive Sampling Completed. Updated Dataset Shape: (51198, 37), Valid Shape: (29406, 37), CV: None, Class counts: [17066 17066 17066]\n",
      "WARNING:Starting Feature Selection 0. Dataset Shape: (51198, 37)\n",
      "WARNING:Feature Selection 0 completed. Updated Dataset shape: (51198, 7)\n",
      "WARNING:Tuning XGBClassifier\n",
      "WARNING:Tuning completed. Best params: {'booster': 'dart', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}\n",
      "WARNING:(Re)fitting Pipeline\n",
      "WARNING:Running Auto-Preprocessing\n",
      "WARNING:Preprocessing completed. Updated Dataset shape: (2196093, 37), cv: auto\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[18:55:34] WARNING: ../src/learner.cc:1115: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:Local ProcessPool execution (n_jobs=2)\n",
      "WARNING:AutoML completed. Time taken - 2736.745 sec\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>Training Dataset size</td>\n",
       "      <td>(2196093, 37)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Validation Dataset size</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>CV</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Target variable</td>\n",
       "      <td>M_WEATHER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Optimization Metric</td>\n",
       "      <td>f1_macro</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Initial number of Features</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected number of Features</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected Features</td>\n",
       "      <td>Index(['M_SESSION_UID', 'M_FRAME_IDENTIFIER', 'M_PLAYER_CAR_INDEX',\n",
       "       'M_SESSION_LINK_IDENTIFIER', 'M_TRACK_TEMPERATURE', 'M_TRACK_LENGTH',\n",
       "       'M_NUM_MARSHAL_ZONES'],\n",
       "      dtype='object')</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected Algorithm</td>\n",
       "      <td>XGBClassifier</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>End-to-end Elapsed Time (seconds)</td>\n",
       "      <td>2745.2891</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected Hyperparameters</td>\n",
       "      <td>{'objective': 'multi:softprob', 'use_label_encoder': True, 'base_score': 0.5, 'booster': 'dart', 'colsample_bylevel': 1, 'colsample_bynode': 1, 'colsample_bytree': 1, 'enable_categorical': False, 'gamma': 0, 'gpu_id': -1, 'importance_type': None, 'interaction_constraints': '', 'learning_rate': 0.1, 'max_delta_step': 0, 'max_depth': 3, 'min_child_weight': 1, 'missing': nan, 'monotone_constraints': '()', 'n_estimators': 100, 'n_jobs': 4, 'num_parallel_tree': 1, 'predictor': 'auto', 'random_state': 7, 'reg_alpha': 0, 'reg_lambda': 1, 'scale_pos_weight': None, 'subsample': 1, 'tree_method': 'exact', 'validate_parameters': 1, 'verbosity': None}</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Mean Validation Score</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>AutoML n_jobs</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>AutoML version</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Python version</td>\n",
       "      <td>3.7.12 | packaged by conda-forge | (default, Oct 26 2021, 06:08:53) \\n[GCC 9.4.0]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>Rank based on Performance</th>\n",
       "      <th>Algorithm</th>\n",
       "      <th>#Samples</th>\n",
       "      <th>#Features</th>\n",
       "      <th>Mean Validation Score</th>\n",
       "      <th>Hyperparameters</th>\n",
       "      <th>CPU Time</th>\n",
       "      <th>Memory Usage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>XGBClassifier_ADS</td>\n",
       "      <td>3000</td>\n",
       "      <td>37</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'objective': 'multi:softprob', 'use_label_encoder': True, 'base_score': 0.5, 'booster': 'gbtree', 'colsample_bylevel': 1, 'colsample_bynode': 1, 'colsample_bytree': 1, 'enable_categorical': False, 'gamma': 0, 'gpu_id': -1, 'importance_type': None, 'interaction_constraints': '', 'learning_rate': 0.1, 'max_delta_step': 0, 'max_depth': 3, 'min_child_weight': 1, 'missing': nan, 'monotone_constraints': '()', 'n_estimators': 100, 'n_jobs': 4, 'num_parallel_tree': 1, 'predictor': 'auto', 'random_state': 7, 'reg_alpha': 0, 'reg_lambda': 1, 'scale_pos_weight': None, 'subsample': 1, 'tree_method': 'exact', 'validate_parameters': 1, 'verbosity': None}</td>\n",
       "      <td>0.0244</td>\n",
       "      <td>1.1902e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>XGBClassifier_ADS</td>\n",
       "      <td>9393</td>\n",
       "      <td>37</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'objective': 'multi:softprob', 'use_label_encoder': True, 'base_score': 0.5, 'booster': 'gbtree', 'colsample_bylevel': 1, 'colsample_bynode': 1, 'colsample_bytree': 1, 'enable_categorical': False, 'gamma': 0, 'gpu_id': -1, 'importance_type': None, 'interaction_constraints': '', 'learning_rate': 0.1, 'max_delta_step': 0, 'max_depth': 3, 'min_child_weight': 1, 'missing': nan, 'monotone_constraints': '()', 'n_estimators': 100, 'n_jobs': 4, 'num_parallel_tree': 1, 'predictor': 'auto', 'random_state': 7, 'reg_alpha': 0, 'reg_lambda': 1, 'scale_pos_weight': None, 'subsample': 1, 'tree_method': 'exact', 'validate_parameters': 1, 'verbosity': None}</td>\n",
       "      <td>0.0379</td>\n",
       "      <td>1.1902e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>XGBClassifier_AS</td>\n",
       "      <td>1756874</td>\n",
       "      <td>37</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>44.8743</td>\n",
       "      <td>8.6636e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>XGBClassifier_HT</td>\n",
       "      <td>51198</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>16.0609</td>\n",
       "      <td>0.0000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>XGBClassifier_HT</td>\n",
       "      <td>51198</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 275, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>22.3556</td>\n",
       "      <td>0.0000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\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",
       "      <td>144</td>\n",
       "      <td>XGBClassifier_HT</td>\n",
       "      <td>51198</td>\n",
       "      <td>7</td>\n",
       "      <td>0.9392</td>\n",
       "      <td>{'booster': 'dart', 'learning_rate': 0.0001, 'max_depth': 2, 'min_child_weight': 0, 'n_estimators': 50, 'reg_alpha': 0.0, 'reg_lambda': 0.0}</td>\n",
       "      <td>16.8380</td>\n",
       "      <td>0.0000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>145</td>\n",
       "      <td>XGBClassifier_MIClassification_FS</td>\n",
       "      <td>51198</td>\n",
       "      <td>2</td>\n",
       "      <td>0.9337</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>8.8340</td>\n",
       "      <td>0.0000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>146</td>\n",
       "      <td>XGBClassifier_MIClassification_FS</td>\n",
       "      <td>51198</td>\n",
       "      <td>1</td>\n",
       "      <td>0.9273</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>6.0464</td>\n",
       "      <td>0.0000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>147</td>\n",
       "      <td>XGBClassifier_ANOVAF_FS</td>\n",
       "      <td>51198</td>\n",
       "      <td>1</td>\n",
       "      <td>0.9169</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>6.0031</td>\n",
       "      <td>0.0000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>148</td>\n",
       "      <td>GaussianNB_AS</td>\n",
       "      <td>1756874</td>\n",
       "      <td>37</td>\n",
       "      <td>0.7164</td>\n",
       "      <td>{}</td>\n",
       "      <td>0.3621</td>\n",
       "      <td>3.8147e-06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "automl_model3, _ = automl.train(score_metric='f1_macro')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "b09a2844",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:Random state (7) is used for model builds\n",
      "WARNING:Dataset shape: (1756874, 37)\n",
      "WARNING:Running Auto-Preprocessing\n",
      "WARNING:Preprocessing completed. Updated Dataset shape: (1756874, 37), cv: None\n",
      "WARNING:SVC, KNeighborsClassifier are disabled for datasets with > 10K samples or > 1K features\n",
      "WARNING:Running Model Selection\n",
      "WARNING:Model Selection completed. Selected model: ['XGBClassifier']\n",
      "WARNING:Running Adaptive Sampling. Dataset Shape: (1756874, 37), Valid Shape: (439219, 37), CV: None, Class counts: [1203252  488765   64857]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[19:50:32] WARNING: ../src/learner.cc:1115: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:Adaptive Sampling Completed. Updated Dataset Shape: (51198, 37), Valid Shape: (29406, 37), CV: None, Class counts: [17066 17066 17066]\n",
      "WARNING:Starting Feature Selection 0. Dataset Shape: (51198, 37)\n",
      "WARNING:Feature Selection 0 completed. Updated Dataset shape: (51198, 7)\n",
      "WARNING:Tuning XGBClassifier\n",
      "WARNING:Tuning completed. Best params: {'booster': 'dart', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}\n",
      "WARNING:(Re)fitting Pipeline\n",
      "WARNING:Running Auto-Preprocessing\n",
      "WARNING:Preprocessing completed. Updated Dataset shape: (2196093, 37), cv: auto\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[20:06:08] WARNING: ../src/learner.cc:1115: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:Local ProcessPool execution (n_jobs=2)\n",
      "WARNING:AutoML completed. Time taken - 2734.287 sec\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>Training Dataset size</td>\n",
       "      <td>(2196093, 37)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Validation Dataset size</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>CV</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Target variable</td>\n",
       "      <td>M_WEATHER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Optimization Metric</td>\n",
       "      <td>accuracy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Initial number of Features</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected number of Features</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected Features</td>\n",
       "      <td>Index(['M_SESSION_UID', 'M_FRAME_IDENTIFIER', 'M_PLAYER_CAR_INDEX',\n",
       "       'M_SESSION_LINK_IDENTIFIER', 'M_TRACK_TEMPERATURE', 'M_TRACK_LENGTH',\n",
       "       'M_NUM_MARSHAL_ZONES'],\n",
       "      dtype='object')</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected Algorithm</td>\n",
       "      <td>XGBClassifier</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>End-to-end Elapsed Time (seconds)</td>\n",
       "      <td>2742.7593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected Hyperparameters</td>\n",
       "      <td>{'objective': 'multi:softprob', 'use_label_encoder': True, 'base_score': 0.5, 'booster': 'dart', 'colsample_bylevel': 1, 'colsample_bynode': 1, 'colsample_bytree': 1, 'enable_categorical': False, 'gamma': 0, 'gpu_id': -1, 'importance_type': None, 'interaction_constraints': '', 'learning_rate': 0.1, 'max_delta_step': 0, 'max_depth': 3, 'min_child_weight': 1, 'missing': nan, 'monotone_constraints': '()', 'n_estimators': 100, 'n_jobs': 4, 'num_parallel_tree': 1, 'predictor': 'auto', 'random_state': 7, 'reg_alpha': 0, 'reg_lambda': 1, 'scale_pos_weight': None, 'subsample': 1, 'tree_method': 'exact', 'validate_parameters': 1, 'verbosity': None}</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Mean Validation Score</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>AutoML n_jobs</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>AutoML version</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Python version</td>\n",
       "      <td>3.7.12 | packaged by conda-forge | (default, Oct 26 2021, 06:08:53) \\n[GCC 9.4.0]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>Rank based on Performance</th>\n",
       "      <th>Algorithm</th>\n",
       "      <th>#Samples</th>\n",
       "      <th>#Features</th>\n",
       "      <th>Mean Validation Score</th>\n",
       "      <th>Hyperparameters</th>\n",
       "      <th>CPU Time</th>\n",
       "      <th>Memory Usage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>XGBClassifier_ADS</td>\n",
       "      <td>3000</td>\n",
       "      <td>37</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'objective': 'multi:softprob', 'use_label_encoder': True, 'base_score': 0.5, 'booster': 'gbtree', 'colsample_bylevel': 1, 'colsample_bynode': 1, 'colsample_bytree': 1, 'enable_categorical': False, 'gamma': 0, 'gpu_id': -1, 'importance_type': None, 'interaction_constraints': '', 'learning_rate': 0.1, 'max_delta_step': 0, 'max_depth': 3, 'min_child_weight': 1, 'missing': nan, 'monotone_constraints': '()', 'n_estimators': 100, 'n_jobs': 4, 'num_parallel_tree': 1, 'predictor': 'auto', 'random_state': 7, 'reg_alpha': 0, 'reg_lambda': 1, 'scale_pos_weight': None, 'subsample': 1, 'tree_method': 'exact', 'validate_parameters': 1, 'verbosity': None}</td>\n",
       "      <td>0.0226</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>XGBClassifier_ADS</td>\n",
       "      <td>9393</td>\n",
       "      <td>37</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'objective': 'multi:softprob', 'use_label_encoder': True, 'base_score': 0.5, 'booster': 'gbtree', 'colsample_bylevel': 1, 'colsample_bynode': 1, 'colsample_bytree': 1, 'enable_categorical': False, 'gamma': 0, 'gpu_id': -1, 'importance_type': None, 'interaction_constraints': '', 'learning_rate': 0.1, 'max_delta_step': 0, 'max_depth': 3, 'min_child_weight': 1, 'missing': nan, 'monotone_constraints': '()', 'n_estimators': 100, 'n_jobs': 4, 'num_parallel_tree': 1, 'predictor': 'auto', 'random_state': 7, 'reg_alpha': 0, 'reg_lambda': 1, 'scale_pos_weight': None, 'subsample': 1, 'tree_method': 'exact', 'validate_parameters': 1, 'verbosity': None}</td>\n",
       "      <td>0.0392</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>XGBClassifier_HT</td>\n",
       "      <td>51198</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>15.4911</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>XGBClassifier_HT</td>\n",
       "      <td>51198</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 275, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>23.3805</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>XGBClassifier_HT</td>\n",
       "      <td>51198</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 5.623413251903491}</td>\n",
       "      <td>7.7344</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\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",
       "      <td>144</td>\n",
       "      <td>XGBClassifier_HT</td>\n",
       "      <td>51198</td>\n",
       "      <td>7</td>\n",
       "      <td>0.9400</td>\n",
       "      <td>{'booster': 'dart', 'learning_rate': 0.0001, 'max_depth': 2, 'min_child_weight': 0, 'n_estimators': 50, 'reg_alpha': 0.0, 'reg_lambda': 0.0}</td>\n",
       "      <td>14.6204</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>145</td>\n",
       "      <td>XGBClassifier_MIClassification_FS</td>\n",
       "      <td>51198</td>\n",
       "      <td>2</td>\n",
       "      <td>0.9344</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>6.7040</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>146</td>\n",
       "      <td>XGBClassifier_MIClassification_FS</td>\n",
       "      <td>51198</td>\n",
       "      <td>1</td>\n",
       "      <td>0.9279</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>5.5445</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>147</td>\n",
       "      <td>XGBClassifier_ANOVAF_FS</td>\n",
       "      <td>51198</td>\n",
       "      <td>1</td>\n",
       "      <td>0.9177</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>6.7135</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>148</td>\n",
       "      <td>GaussianNB_AS</td>\n",
       "      <td>1756874</td>\n",
       "      <td>37</td>\n",
       "      <td>0.7239</td>\n",
       "      <td>{}</td>\n",
       "      <td>0.3964</td>\n",
       "      <td>0.0172</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "automl_model4, _ = automl.train(score_metric='accuracy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "87671c8a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig3 =automl.visualize_algorithm_selection_trials()\n",
    "import matplotlib.pyplot as plt\n",
    "plt.savefig('fig3.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "2ddc9ef9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "loop1:   0%|          | 0/4 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "loop1:   0%|          | 0/7 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/svg+xml": [
       "<svg height=\"470pt\" viewBox=\"0.00 0.00 394.00 470.00\" width=\"394pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n",
       "<g class=\"graph\" id=\"graph0\" transform=\"scale(1 1) rotate(0) translate(4 466)\">\n",
       "<title>%3</title>\n",
       "<polygon fill=\"white\" points=\"-4,5 -4,-466 391,-466 391,5 -4,5\" stroke=\"white\"/>\n",
       "<!-- partitions -->\n",
       "<g class=\"node\" id=\"node1\"><title>partitions</title>\n",
       "<polygon fill=\"dimgray\" points=\"148.25,-462 11.75,-462 11.75,-466 -0.25,-466 -0.25,-400 148.25,-400 148.25,-462\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"-0.25,-462 11.75,-462 \" stroke=\"black\"/>\n",
       "<text fill=\"white\" font-family=\"Courier,monospace\" font-size=\"12.00\" text-anchor=\"start\" x=\"17.5\" y=\"-434.4\">- rows: 2745117</text>\n",
       "<text fill=\"white\" font-family=\"Courier,monospace\" font-size=\"12.00\" text-anchor=\"start\" x=\"17.5\" y=\"-421.4\">- columns: 34</text>\n",
       "</g>\n",
       "<!-- prepare -->\n",
       "<g class=\"node\" id=\"node2\"><title>prepare</title>\n",
       "<polygon fill=\"azure\" points=\"225.25,-449 166.75,-449 166.75,-445 162.75,-445 162.75,-441 166.75,-441 166.75,-421 162.75,-421 162.75,-417 166.75,-417 166.75,-413 225.25,-413 225.25,-449\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"166.75,-445 170.75,-445 170.75,-441 166.75,-441 \" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"166.75,-421 170.75,-421 170.75,-417 166.75,-417 \" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"middle\" x=\"196\" y=\"-428.5\">prepare</text>\n",
       "</g>\n",
       "<!-- type_discovery -->\n",
       "<g class=\"node\" id=\"node3\"><title>type_discovery</title>\n",
       "<polygon fill=\"azure\" points=\"246.25,-352 145.75,-352 145.75,-348 141.75,-348 141.75,-344 145.75,-344 145.75,-324 141.75,-324 141.75,-320 145.75,-320 145.75,-316 246.25,-316 246.25,-352\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"145.75,-348 149.75,-348 149.75,-344 145.75,-344 \" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"145.75,-324 149.75,-324 149.75,-320 145.75,-320 \" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"middle\" x=\"196\" y=\"-331.5\">type_discovery</text>\n",
       "</g>\n",
       "<!-- prepare&#45;&gt;type_discovery -->\n",
       "<g class=\"edge\" id=\"edge1\"><title>prepare-&gt;type_discovery</title>\n",
       "<path d=\"M196,-412.576C196,-398.648 196,-378.861 196,-362.757\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"199.5,-362.314 196,-352.314 192.5,-362.314 199.5,-362.314\" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"196.5\" y=\"-374.5\"> </text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" font-weight=\"bold\" text-anchor=\"start\" x=\"203.5\" y=\"-374.5\">fix_column_names</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"294.5\" y=\"-374.5\"> </text>\n",
       "</g>\n",
       "<!-- recommendations -->\n",
       "<g class=\"node\" id=\"node4\"><title>recommendations</title>\n",
       "<polygon fill=\"#eec900\" points=\"365,-267.5 27,-267.5 27,-263.5 23,-263.5 23,-259.5 27,-259.5 27,-188.5 23,-188.5 23,-184.5 27,-184.5 27,-180.5 365,-180.5 365,-267.5\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"27,-263.5 31,-263.5 31,-259.5 27,-259.5 \" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"27,-188.5 31,-188.5 31,-184.5 27,-184.5 \" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"45.5\" y=\"-242.5\"> Drop: &quot;M_WEATHER_FORECAST_SAMPLES_M_SESSION_TYPE&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"63.5\" y=\"-232.5\">Drop: &quot;M_WEATHER_FORECAST_SAMPLES_M_WEATHER&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"99.5\" y=\"-222.5\">Drop: &quot;M_AIR_TEMPERATURE_CHANGE&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"93.5\" y=\"-212.5\">Drop: &quot;M_TRACK_TEMPERATURE_CHANGE&quot;</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"96.5\" y=\"-202.5\">Drop: &quot;M_AIR_TEMPERATURE_CHANGE&quot; </text>\n",
       "</g>\n",
       "<!-- type_discovery&#45;&gt;recommendations -->\n",
       "<g class=\"edge\" id=\"edge2\"><title>type_discovery-&gt;recommendations</title>\n",
       "<path d=\"M196,-315.648C196,-305.385 196,-291.777 196,-278.234\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"199.5,-277.946 196,-267.946 192.5,-277.946 199.5,-277.946\" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"196.5\" y=\"-290.5\"> </text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" font-weight=\"bold\" text-anchor=\"start\" x=\"203.5\" y=\"-290.5\">convert_columns</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"294.5\" y=\"-290.5\"> </text>\n",
       "</g>\n",
       "<!-- feature_engineering -->\n",
       "<g class=\"node\" id=\"node5\"><title>feature_engineering</title>\n",
       "<polygon fill=\"#eec900\" points=\"335,-131.5 57,-131.5 57,-127.5 53,-127.5 53,-123.5 57,-123.5 57,-92.5 53,-92.5 53,-88.5 57,-88.5 57,-84.5 335,-84.5 335,-131.5\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"57,-127.5 61,-127.5 61,-123.5 57,-123.5 \" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"57,-92.5 61,-92.5 61,-88.5 57,-88.5 \" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"75.5\" y=\"-106.5\"> No feature engineering transformations </text>\n",
       "</g>\n",
       "<!-- recommendations&#45;&gt;feature_engineering -->\n",
       "<g class=\"edge\" id=\"edge3\"><title>recommendations-&gt;feature_engineering</title>\n",
       "<path d=\"M196,-180.226C196,-167.424 196,-153.598 196,-141.564\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"199.5,-141.522 196,-131.522 192.5,-141.522 199.5,-141.522\" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"196.5\" y=\"-154.5\"> </text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" font-weight=\"bold\" text-anchor=\"start\" x=\"203.5\" y=\"-154.5\">RecommendationTransformer</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"354.5\" y=\"-154.5\"> </text>\n",
       "</g>\n",
       "<!-- terminus -->\n",
       "<g class=\"node\" id=\"node6\"><title>terminus</title>\n",
       "<polygon fill=\"white\" points=\"199,-18 199,-24 205,-24 205,-30 187,-30 187,-24 193,-24 193,-18 199,-18\" stroke=\"black\"/>\n",
       "<polyline fill=\"none\" points=\"169,-18 223,-18 \" stroke=\"black\"/>\n",
       "</g>\n",
       "<!-- feature_engineering&#45;&gt;terminus -->\n",
       "<g class=\"edge\" id=\"edge4\"><title>feature_engineering-&gt;terminus</title>\n",
       "<path d=\"M196,-84.3378C196,-72.6791 196,-58.3831 196,-46.1394\" fill=\"none\" stroke=\"black\"/>\n",
       "<polygon fill=\"black\" points=\"199.5,-46.0259 196,-36.026 192.5,-46.026 199.5,-46.0259\" stroke=\"black\"/>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"196.5\" y=\"-58.5\"> </text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" font-weight=\"bold\" text-anchor=\"start\" x=\"203.5\" y=\"-58.5\">FeatureEngineeringTransformer</text>\n",
       "<text font-family=\"Courier,monospace\" font-size=\"10.00\" text-anchor=\"start\" x=\"378.5\" y=\"-58.5\"> </text>\n",
       "</g>\n",
       "</g>\n",
       "</svg>"
      ],
      "text/plain": [
       "<IPython.core.display.SVG object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dataSetFactory_Rain = DatasetFactory.open(Final_data_frame, target=\"M_RAIN_PERCENTAGE\")\n",
    "dataset_transformed_Rain = dataSetFactoryFinal.auto_transform()\n",
    "dataset_transformed_Rain.visualize_transforms()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "3a09314e",
   "metadata": {},
   "outputs": [],
   "source": [
    "train, test = dataset_transformed_Rain.train_test_split(test_size=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "55012797",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:Execution engine (local) has already been initialized. Reinitializing!\n",
      "WARNING:Local ProcessPool execution (n_jobs=2)\n",
      "WARNING:Random state (42) is used for model builds\n",
      "WARNING:Dataset shape: (1756874, 33)\n",
      "WARNING:Running Auto-Preprocessing\n",
      "WARNING:Preprocessing completed. Updated Dataset shape: (1756874, 33), cv: None\n",
      "WARNING:SVC, KNeighborsClassifier are disabled for datasets with > 10K samples or > 1K features\n",
      "WARNING:Running Model Selection\n",
      "WARNING:Time budget exceeded by 3.37s, resetting XEngine\n",
      "WARNING:Timebudget Exceeded or Timedout completed 7/8, 0 tasks timedout\n",
      "WARNING:Model Selection completed. Selected model: ['LGBMClassifier']\n",
      "WARNING:Running Adaptive Sampling. Dataset Shape: (1756874, 33), Valid Shape: (439219, 33), CV: None, Class counts: [1203252  488765   64857]\n",
      "WARNING:Timebudget exhausted. Skipping Adaptive Sampling\n",
      "WARNING:Adaptive Sampling Completed. Updated Dataset Shape: (1756874, 33), Valid Shape: (439219, 33), CV: None, Class counts: [1203252  488765   64857]\n",
      "WARNING:Starting Feature Selection 0. Dataset Shape: (1756874, 33)\n",
      "WARNING:Timebudget exhausted. Skipping Feature Selection\n",
      "WARNING:Using all features: Index(['M_SESSION_UID', 'M_FRAME_IDENTIFIER', 'M_PLAYER_CAR_INDEX',\n",
      "       'M_BRAKING_ASSIST', 'M_SESSION_LINK_IDENTIFIER', 'M_PIT_RELEASE_ASSIST',\n",
      "       'M_TRACK_TEMPERATURE', 'M_TRACK_LENGTH', 'M_GAME_PAUSED',\n",
      "       'M_FORECAST_ACCURACY', 'M_NUM_WEATHER_FORECAST_SAMPLES', 'M_ERSASSIST',\n",
      "       'M_FORMULA', 'M_PIT_ASSIST', 'M_GEARBOX_ASSIST', 'M_SESSION_TYPE',\n",
      "       'M_SPECTATOR_CAR_INDEX', 'M_PIT_STOP_WINDOW_LATEST_LAP',\n",
      "       'M_DYNAMIC_RACING_LINE_TYPE', 'M_SESSION_DURATION', 'M_TIME_OFFSET',\n",
      "       'M_WEATHER_FORECAST_SAMPLES_M_TRACK_TEMPERATURE',\n",
      "       'M_WEATHER_FORECAST_SAMPLES_M_AIR_TEMPERATURE', 'M_RAIN_PERCENTAGE',\n",
      "       'M_AI_DIFFICULTY', 'M_PIT_SPEED_LIMIT', 'M_NETWORK_GAME',\n",
      "       'M_TOTAL_LAPS', 'M_STEERING_ASSIST', 'M_IS_SPECTATING',\n",
      "       'M_DYNAMIC_RACING_LINE', 'M_DRSASSIST', 'M_NUM_MARSHAL_ZONES'],\n",
      "      dtype='object')\n",
      "WARNING:Feature Selection 0 completed. Updated Dataset shape: (1756874, 33)\n",
      "WARNING:Timebudget exhausted. Skipping Hyperparameter Optimization for LGBMClassifier\n",
      "WARNING:(Re)fitting Pipeline\n",
      "WARNING:Running Auto-Preprocessing\n",
      "WARNING:Preprocessing completed. Updated Dataset shape: (2196093, 33), cv: auto\n",
      "WARNING:Local ProcessPool execution (n_jobs=2)\n",
      "WARNING:AutoML completed. Time taken - 452.043 sec\n",
      "ERROR:Unfortunately, there were no trials found, so we cannot visualize it.\n"
     ]
    }
   ],
   "source": [
    "# Now, let's see how well AutoML can perform the same task.\n",
    "# AutoML will evaluate several algorithms in combination with different values for their \n",
    "# hyperparameters.\n",
    "# Note the time parameter of 160 seconds, this means we allow AutoML this amount of time to find the best solution.\n",
    "\n",
    "from ads.automl.provider import OracleAutoMLProvider\n",
    "from ads.automl.driver import AutoML\n",
    "ml_engine = OracleAutoMLProvider(n_jobs = -1) # n_jobs = -1 means that we are not going to limit the resources and let the AutoML to do this for us and take as much as possible\n",
    "automl = AutoML(train, provider=ml_engine)\n",
    "automl_model1, baseline = automl.train(time_budget=160, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "ec7060ff",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:Random state (7) is used for model builds\n",
      "WARNING:Dataset shape: (1756874, 33)\n",
      "WARNING:Running Auto-Preprocessing\n",
      "WARNING:Preprocessing completed. Updated Dataset shape: (1756874, 33), cv: None\n",
      "WARNING:SVC, KNeighborsClassifier are disabled for datasets with > 10K samples or > 1K features\n",
      "WARNING:Running Model Selection\n",
      "WARNING:Model Selection completed. Selected model: ['XGBClassifier']\n",
      "WARNING:Running Adaptive Sampling. Dataset Shape: (1756874, 33), Valid Shape: (439219, 33), CV: None, Class counts: [1203252  488765   64857]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[21:31:29] WARNING: ../src/learner.cc:1115: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:Adaptive Sampling Completed. Updated Dataset Shape: (51198, 33), Valid Shape: (29406, 33), CV: None, Class counts: [17066 17066 17066]\n",
      "WARNING:Starting Feature Selection 0. Dataset Shape: (51198, 33)\n",
      "WARNING:Feature Selection 0 completed. Updated Dataset shape: (51198, 7)\n",
      "WARNING:Tuning XGBClassifier\n",
      "WARNING:Tuning completed. Best params: {'booster': 'dart', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}\n",
      "WARNING:(Re)fitting Pipeline\n",
      "WARNING:Running Auto-Preprocessing\n",
      "WARNING:Preprocessing completed. Updated Dataset shape: (2196093, 33), cv: auto\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[21:46:14] WARNING: ../src/learner.cc:1115: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'multi:softprob' was changed from 'merror' to 'mlogloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:Local ProcessPool execution (n_jobs=2)\n",
      "WARNING:AutoML completed. Time taken - 2683.089 sec\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>Training Dataset size</td>\n",
       "      <td>(2196093, 33)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Validation Dataset size</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>CV</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Target variable</td>\n",
       "      <td>M_WEATHER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Optimization Metric</td>\n",
       "      <td>accuracy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Initial number of Features</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected number of Features</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected Features</td>\n",
       "      <td>Index(['M_SESSION_UID', 'M_FRAME_IDENTIFIER', 'M_PLAYER_CAR_INDEX',\n",
       "       'M_SESSION_LINK_IDENTIFIER', 'M_TRACK_TEMPERATURE', 'M_TRACK_LENGTH',\n",
       "       'M_NUM_MARSHAL_ZONES'],\n",
       "      dtype='object')</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected Algorithm</td>\n",
       "      <td>XGBClassifier</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>End-to-end Elapsed Time (seconds)</td>\n",
       "      <td>2691.0106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Selected Hyperparameters</td>\n",
       "      <td>{'objective': 'multi:softprob', 'use_label_encoder': True, 'base_score': 0.5, 'booster': 'dart', 'colsample_bylevel': 1, 'colsample_bynode': 1, 'colsample_bytree': 1, 'enable_categorical': False, 'gamma': 0, 'gpu_id': -1, 'importance_type': None, 'interaction_constraints': '', 'learning_rate': 0.1, 'max_delta_step': 0, 'max_depth': 3, 'min_child_weight': 1, 'missing': nan, 'monotone_constraints': '()', 'n_estimators': 100, 'n_jobs': 4, 'num_parallel_tree': 1, 'predictor': 'auto', 'random_state': 7, 'reg_alpha': 0, 'reg_lambda': 1, 'scale_pos_weight': None, 'subsample': 1, 'tree_method': 'exact', 'validate_parameters': 1, 'verbosity': None}</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Mean Validation Score</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>AutoML n_jobs</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>AutoML version</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Python version</td>\n",
       "      <td>3.7.12 | packaged by conda-forge | (default, Oct 26 2021, 06:08:53) \\n[GCC 9.4.0]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>Rank based on Performance</th>\n",
       "      <th>Algorithm</th>\n",
       "      <th>#Samples</th>\n",
       "      <th>#Features</th>\n",
       "      <th>Mean Validation Score</th>\n",
       "      <th>Hyperparameters</th>\n",
       "      <th>CPU Time</th>\n",
       "      <th>Memory Usage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>XGBClassifier_ADS</td>\n",
       "      <td>3000</td>\n",
       "      <td>33</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'objective': 'multi:softprob', 'use_label_encoder': True, 'base_score': 0.5, 'booster': 'gbtree', 'colsample_bylevel': 1, 'colsample_bynode': 1, 'colsample_bytree': 1, 'enable_categorical': False, 'gamma': 0, 'gpu_id': -1, 'importance_type': None, 'interaction_constraints': '', 'learning_rate': 0.1, 'max_delta_step': 0, 'max_depth': 3, 'min_child_weight': 1, 'missing': nan, 'monotone_constraints': '()', 'n_estimators': 100, 'n_jobs': 4, 'num_parallel_tree': 1, 'predictor': 'auto', 'random_state': 7, 'reg_alpha': 0, 'reg_lambda': 1, 'scale_pos_weight': None, 'subsample': 1, 'tree_method': 'exact', 'validate_parameters': 1, 'verbosity': None}</td>\n",
       "      <td>0.0226</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>XGBClassifier_ADS</td>\n",
       "      <td>9393</td>\n",
       "      <td>33</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'objective': 'multi:softprob', 'use_label_encoder': True, 'base_score': 0.5, 'booster': 'gbtree', 'colsample_bylevel': 1, 'colsample_bynode': 1, 'colsample_bytree': 1, 'enable_categorical': False, 'gamma': 0, 'gpu_id': -1, 'importance_type': None, 'interaction_constraints': '', 'learning_rate': 0.1, 'max_delta_step': 0, 'max_depth': 3, 'min_child_weight': 1, 'missing': nan, 'monotone_constraints': '()', 'n_estimators': 100, 'n_jobs': 4, 'num_parallel_tree': 1, 'predictor': 'auto', 'random_state': 7, 'reg_alpha': 0, 'reg_lambda': 1, 'scale_pos_weight': None, 'subsample': 1, 'tree_method': 'exact', 'validate_parameters': 1, 'verbosity': None}</td>\n",
       "      <td>0.0370</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>XGBClassifier_HT</td>\n",
       "      <td>51198</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 4, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>10.6554</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>XGBClassifier_HT</td>\n",
       "      <td>51198</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 276, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>22.9144</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>XGBClassifier_HT</td>\n",
       "      <td>51198</td>\n",
       "      <td>7</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>13.5170</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\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",
       "      <td>137</td>\n",
       "      <td>XGBClassifier_HT</td>\n",
       "      <td>51198</td>\n",
       "      <td>7</td>\n",
       "      <td>0.9400</td>\n",
       "      <td>{'booster': 'dart', 'learning_rate': 0.0001, 'max_depth': 2, 'min_child_weight': 0, 'n_estimators': 50, 'reg_alpha': 0.0, 'reg_lambda': 0.0}</td>\n",
       "      <td>14.1886</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>138</td>\n",
       "      <td>XGBClassifier_MIClassification_FS</td>\n",
       "      <td>51198</td>\n",
       "      <td>2</td>\n",
       "      <td>0.9344</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>6.5754</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>139</td>\n",
       "      <td>XGBClassifier_MIClassification_FS</td>\n",
       "      <td>51198</td>\n",
       "      <td>1</td>\n",
       "      <td>0.9279</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>6.8171</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>140</td>\n",
       "      <td>XGBClassifier_ANOVAF_FS</td>\n",
       "      <td>51198</td>\n",
       "      <td>1</td>\n",
       "      <td>0.9177</td>\n",
       "      <td>{'booster': 'gbtree', 'learning_rate': 0.1, 'max_depth': 3, 'min_child_weight': 1, 'n_estimators': 100, 'reg_alpha': 0, 'reg_lambda': 1}</td>\n",
       "      <td>5.3285</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>141</td>\n",
       "      <td>GaussianNB_AS</td>\n",
       "      <td>1756874</td>\n",
       "      <td>33</td>\n",
       "      <td>0.7215</td>\n",
       "      <td>{}</td>\n",
       "      <td>0.3815</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "automl_model4, _ = automl.train(score_metric='accuracy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "d8b1ffbd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig4 =automl.visualize_algorithm_selection_trials()\n",
    "import matplotlib.pyplot as plt\n",
    "plt.savefig('fig4.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fcf54a52",
   "metadata": {},
   "source": [
    "### Import Model to pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "20b04e50",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    " \n",
    "# Save the trained model as a pickle string.\n",
    "saved_model = pickle.dumps(automl_model4)\n",
    "pickle.dump(automl_model4, open('model.pkl', 'wb'))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:generalmachinelearningforcpusonpython3_7v1_000]",
   "language": "python",
   "name": "conda-env-generalmachinelearningforcpusonpython3_7v1_000-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}