Multiple Linear Regression using Scikit-Learn

Multiple Linear Regression using Scikit-Learn.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Multiple Linear Regression using Scikit-Learn\n",
    "\n",
    "\n",
    "This project is about Multiple Linear Regression which is a machine learning algorithm. I build a multiple linear regression model to estimate the relative cpu performance of computer hardware dataset.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Table of contents\n",
    "\n",
    "The contents of this project are divided into various categories which are given as follows:-\n",
    "\n",
    "\n",
    "1.\tIntroduction\n",
    "\n",
    "2.\tLinear regression intuition\n",
    "\n",
    "3.\tIndependent and dependent variables\n",
    "\n",
    "4.\tAssumptions of  linear regression\n",
    "\n",
    "5.\tThe dataset description\n",
    "\n",
    "6.\tThe problem statement\n",
    "\n",
    "7.\tImport the Python libraries\n",
    "\n",
    "8.\tImport the dataset\n",
    "\n",
    "9.\tExploratory Data Analysis\n",
    "    - Explore types of variables\n",
    "    \n",
    "    - Estimate correlation coefficients\n",
    "    \n",
    "     - Correlation heat map\n",
    "     \n",
    "10.\tDetect problems within variables\n",
    "     - Detect missing values\n",
    "     \n",
    "     - Outliers in discrete variables\n",
    "     \n",
    "     - Number of labels – cardinality\n",
    "     \n",
    "11.\tLinear Regression modeling\n",
    "    - Divide the dataset into categorical and numerical variables\n",
    "    \n",
    "    - Select the predictor and target variables\n",
    "    \n",
    "    - Create separate train and test sets\n",
    "    \n",
    "    - Feature Scaling\n",
    "    \n",
    "    - Fit the Linear Regression model\n",
    "\n",
    "12.\tPredicting the results\n",
    "    - Predicting the test set results\n",
    "    \n",
    "    - Predicting estimated relative CPU performance values\n",
    "\n",
    "13.\tModel slope and intercept terms\n",
    "\n",
    "14.\tEvaluate model performance\n",
    "    - RMSE (Root Mean Square Error)\n",
    "    \n",
    "    - R2 Score\n",
    "    \n",
    "    - Overfitting or Underfitting\n",
    "    \n",
    "    - Cross validation\n",
    "    \n",
    "    - Residual analysis\n",
    "    \n",
    "    - Normality test (Q-Q Plot)\n",
    "    \n",
    "15.\tConclusion\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Introduction\n",
    "\n",
    "\n",
    "In this project, I build a multiple linear regression model to estimate the relative cpu performance of computer hardware dataset. Relative cpu performance of the computer hardware is described in terms of machine cycle time, main memory, cache memory and minimum and maximum channels as given in the dataset.\n",
    "\n",
    "\n",
    "I discuss the basics and assumptions of linear regression. I also discuss the advantages and disadvantages and common pitfalls of linear regression. I present the implementation in Python programming language using Scikit-learn. Scikit-learn is the popular machine learning library of Python programming language. I also discuss various tools to evaluate the linear regression model performance.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Multiple linear regression intuition\n",
    "\n",
    "\n",
    "Linear Regression is a machine learning algorithm which is used to establish the linear relationship between dependent and one or more independent variables. This technique is applicable for supervised learning regression problems where we try to predict a continuous variable. Linear Regression can be further classified into two types – Simple and Multiple Linear Regression. \n",
    "\n",
    "I have discussed the linear regression intuition in detail in the readme document.\n",
    "\n",
    "In this project, I employ Multiple Linear Regression technique where I have one dependent variable and more than one independent variables.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Independent and dependent variables\n",
    "\n",
    "\n",
    "In this project, I refer Independent variable as Feature variable and Dependent variable as Target variable. These variables are also recognized by different names as follows: -\n",
    "\n",
    "\n",
    "**Independent variable**\n",
    "\n",
    "Independent variable is also called Input variable and is denoted by X. In practical applications, independent variable is also called Feature variable or Predictor variable. We can denote it as: -\n",
    "\n",
    "\n",
    "Independent or Input variable (X) = Feature variable = Predictor variable \n",
    "\n",
    "\n",
    "**Dependent variable**\n",
    "\n",
    "\n",
    "Dependent variable is also called Output variable and is denoted by y. Dependent variable is also called Target variable or Response variable. It can be denoted it as follows: -\n",
    "\n",
    "\n",
    "Dependent or Output variable (y) = Target variable = Response variable\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Assumptions of Linear Regression\n",
    "\n",
    "The Linear Regression model is based on several assumptions which are as follows:-\n",
    "\n",
    "\n",
    "1. Linear relationship\n",
    "\n",
    "2. Multivariate normality\n",
    "\n",
    "3. No or little multi-collinearity\n",
    "\n",
    "4. No auto-correlation in error terms\n",
    "\n",
    "5. Homoscedasticity\n",
    "\n",
    "\n",
    "\n",
    "I have described these assumptions in more detail in readme document.\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Dataset description\n",
    "\n",
    "\n",
    "Now, we should get to know more about the dataset. It is a computer hardware dataset. The dataset consists of information about the computer vendors selling computers, model name of computers and various attributes to estimate the relative performance of CPU.\n",
    "The dataset can be found at the following url –\n",
    "\n",
    "https://archive.ics.uci.edu/ml/datasets/Computer+Hardware\n",
    "\n",
    "\n",
    "The dataset description will help us to know more about the data.\n",
    "\n",
    "\n",
    "**Dataset description** is given as follows:-\n",
    "\n",
    "\n",
    "1. vendor name: 30 \n",
    "      (adviser, amdahl,apollo, basf, bti, burroughs, c.r.d, cambex, cdc, dec, \n",
    "       dg, formation, four-phase, gould, honeywell, hp, ibm, ipl, magnuson, \n",
    "       microdata, nas, ncr, nixdorf, perkin-elmer, prime, siemens, sperry, \n",
    "       sratus, wang)\n",
    "       \n",
    "2. Model Name: many unique symbols\n",
    "\n",
    "3. MYCT: machine cycle time in nanoseconds (integer)\n",
    "\n",
    "4. MMIN: minimum main memory in kilobytes (integer)\n",
    "\n",
    "5. MMAX: maximum main memory in kilobytes (integer)\n",
    "\n",
    "6. CACH: cache memory in kilobytes (integer)\n",
    "\n",
    "7. CHMIN: minimum channels in units (integer)\n",
    "\n",
    "8. CHMAX: maximum channels in units (integer)\n",
    "\n",
    "9. PRP: published relative performance (integer)\n",
    "\n",
    "10. ERP: estimated relative performance from the original article (integer)\n",
    "  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. The problem statement\n",
    "\n",
    "\n",
    "A machine learning model is built with the aim of solving a problem. So, first of all I have to define the problem to be solved in this project.\n",
    "\n",
    "As described earlier, the problem is to estimate the relative CPU performance of computer hardware dataset. Relative CPU performance of the computer hardware is described in terms of machine cycle time, main memory, cache memory and minimum and maximum channels as given in the dataset.\n",
    "\n",
    "So, let's get started. I will start by importing the required Python libraries."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7. Import the Python libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import required Python libraries\n",
    "\n",
    "# to handle datasets\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "# for plotting\n",
    "import matplotlib.pyplot as plt\n",
    "% matplotlib inline\n",
    "import seaborn as sns\n",
    "\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 8. Import the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import the dataset\n",
    "\n",
    "filename = \"c:/datasets/machine.data.csv\"\n",
    "\n",
    "df = pd.read_csv(filename, header = None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 9. Exploratory Data Analysis\n",
    "\n",
    "\n",
    "Now, I will perform Exploratory Data Analysis. It provides useful insights into the dataset which is important for further analysis.\n",
    "\n",
    "First of all, we should check the dimensions of the dataframe as follows:-"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of dataframe df: (209, 10)\n"
     ]
    }
   ],
   "source": [
    "# view the dimensions of dataframe df\n",
    "\n",
    "print(\"Shape of dataframe df: {}\".format(df.shape))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that there are 209 rows and 10 columns in the dataset. Next, we should get an insight about the dataset.\n",
    "\n",
    "The **df.head()** function helps us to visualize the first 5 rows of the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>adviser</td>\n",
       "      <td>32/60</td>\n",
       "      <td>125</td>\n",
       "      <td>256</td>\n",
       "      <td>6000</td>\n",
       "      <td>256</td>\n",
       "      <td>16</td>\n",
       "      <td>128</td>\n",
       "      <td>198</td>\n",
       "      <td>199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>amdahl</td>\n",
       "      <td>470v/7</td>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>32000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>269</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>amdahl</td>\n",
       "      <td>470v/7a</td>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>32000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>220</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>amdahl</td>\n",
       "      <td>470v/7b</td>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>32000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>172</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>amdahl</td>\n",
       "      <td>470v/7c</td>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>16000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>16</td>\n",
       "      <td>132</td>\n",
       "      <td>132</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         0        1    2     3      4    5   6    7    8    9\n",
       "0  adviser    32/60  125   256   6000  256  16  128  198  199\n",
       "1   amdahl   470v/7   29  8000  32000   32   8   32  269  253\n",
       "2   amdahl  470v/7a   29  8000  32000   32   8   32  220  253\n",
       "3   amdahl  470v/7b   29  8000  32000   32   8   32  172  253\n",
       "4   amdahl  470v/7c   29  8000  16000   32   8   16  132  132"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# view the top five rows of dataframe df\n",
    "\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that the column names are from 0 to 9. They should be descriptive. So, we should rename them as follows:-"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# rename columns of dataframe df\n",
    "\n",
    "col_names = ['Vendor Name','Model Name', 'MYCT', 'MMIN', 'MMAX', 'CACH','CHMIN', 'CHMAX', 'PRP', 'ERP' ]\n",
    "\n",
    "df.columns = col_names"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We should now check that the columns have appropriate names."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Vendor Name</th>\n",
       "      <th>Model Name</th>\n",
       "      <th>MYCT</th>\n",
       "      <th>MMIN</th>\n",
       "      <th>MMAX</th>\n",
       "      <th>CACH</th>\n",
       "      <th>CHMIN</th>\n",
       "      <th>CHMAX</th>\n",
       "      <th>PRP</th>\n",
       "      <th>ERP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>adviser</td>\n",
       "      <td>32/60</td>\n",
       "      <td>125</td>\n",
       "      <td>256</td>\n",
       "      <td>6000</td>\n",
       "      <td>256</td>\n",
       "      <td>16</td>\n",
       "      <td>128</td>\n",
       "      <td>198</td>\n",
       "      <td>199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>amdahl</td>\n",
       "      <td>470v/7</td>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>32000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>269</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>amdahl</td>\n",
       "      <td>470v/7a</td>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>32000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>220</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>amdahl</td>\n",
       "      <td>470v/7b</td>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>32000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>172</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>amdahl</td>\n",
       "      <td>470v/7c</td>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>16000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>16</td>\n",
       "      <td>132</td>\n",
       "      <td>132</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Vendor Name Model Name  MYCT  MMIN   MMAX  CACH  CHMIN  CHMAX  PRP  ERP\n",
       "0     adviser      32/60   125   256   6000   256     16    128  198  199\n",
       "1      amdahl     470v/7    29  8000  32000    32      8     32  269  253\n",
       "2      amdahl    470v/7a    29  8000  32000    32      8     32  220  253\n",
       "3      amdahl    470v/7b    29  8000  32000    32      8     32  172  253\n",
       "4      amdahl    470v/7c    29  8000  16000    32      8     16  132  132"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# view the top five rows of dataframe with column names renamed\n",
    "\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Explore types of variables\n",
    "\n",
    "\n",
    "In this section, I will explore the types of variables in the dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, let's view a concise summary of the dataframe with **df.info()** method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 209 entries, 0 to 208\n",
      "Data columns (total 10 columns):\n",
      "Vendor Name    209 non-null object\n",
      "Model Name     209 non-null object\n",
      "MYCT           209 non-null int64\n",
      "MMIN           209 non-null int64\n",
      "MMAX           209 non-null int64\n",
      "CACH           209 non-null int64\n",
      "CHMIN          209 non-null int64\n",
      "CHMAX          209 non-null int64\n",
      "PRP            209 non-null int64\n",
      "ERP            209 non-null int64\n",
      "dtypes: int64(8), object(2)\n",
      "memory usage: 16.4+ KB\n"
     ]
    }
   ],
   "source": [
    "# view dataframe summary\n",
    "\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that there are categorical and numerical variables in the dataset. Numerical variables have data types int64 and categorical variables are those of type object.\n",
    "\n",
    "First, let's explore the categorical variables."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 2 categorical variables\n"
     ]
    }
   ],
   "source": [
    "# find categorical variables\n",
    "\n",
    "categorical = [col for col in df.columns if df[col].dtype=='O']\n",
    "print('There are {} categorical variables'.format(len(categorical)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Vendor Name', 'Model Name']\n"
     ]
    }
   ],
   "source": [
    "# view the categorical variables\n",
    "\n",
    "print(categorical)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So, there are two categorical variables - **Vendor Name** and **Model Name** in the dataset. \n",
    "\n",
    "Let's explore more about them. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Vendor Name</th>\n",
       "      <th>Model Name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>adviser</td>\n",
       "      <td>32/60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>amdahl</td>\n",
       "      <td>470v/7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>amdahl</td>\n",
       "      <td>470v/7a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>amdahl</td>\n",
       "      <td>470v/7b</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>amdahl</td>\n",
       "      <td>470v/7c</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Vendor Name Model Name\n",
       "0     adviser      32/60\n",
       "1      amdahl     470v/7\n",
       "2      amdahl    470v/7a\n",
       "3      amdahl    470v/7b\n",
       "4      amdahl    470v/7c"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# view the top five rows of categorical variables\n",
    "\n",
    "df[categorical].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ibm             32\n",
       "nas             19\n",
       "ncr             13\n",
       "honeywell       13\n",
       "sperry          13\n",
       "siemens         12\n",
       "amdahl           9\n",
       "cdc              9\n",
       "burroughs        8\n",
       "harris           7\n",
       "dg               7\n",
       "hp               7\n",
       "ipl              6\n",
       "c.r.d            6\n",
       "dec              6\n",
       "magnuson         6\n",
       "formation        5\n",
       "cambex           5\n",
       "prime            5\n",
       "nixdorf          3\n",
       "gould            3\n",
       "perkin-elmer     3\n",
       "apollo           2\n",
       "wang             2\n",
       "bti              2\n",
       "basf             2\n",
       "microdata        1\n",
       "adviser          1\n",
       "four-phase       1\n",
       "sratus           1\n",
       "Name: Vendor Name, dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# exploring the categories in Vendor Name column\n",
    "\n",
    "df['Vendor Name'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "ibm is the most frequent category in the **Vendor Name** column.\n",
    "\n",
    "Next, let's explore the **Model Name** column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of unique Model Names:  209\n",
      "Number  of instances of models:  209\n"
     ]
    }
   ],
   "source": [
    "print('Number of unique Model Names: ', len(df['Model Name'].unique()))\n",
    "print('Number  of instances of models: ', len(df))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that **Model Name** is a unique identifier for each of the computer models. Thus this is not a variable that we can use to predict the estimated relative performance of computer models. So, we should not use this column for model building.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's explore the numerical variables."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 8 numerical variables\n"
     ]
    }
   ],
   "source": [
    "# find numerical variables\n",
    "\n",
    "numerical = [col for col in df.columns if df[col].dtype!='O']\n",
    "print('There are {} numerical variables'.format(len(numerical)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['MYCT', 'MMIN', 'MMAX', 'CACH', 'CHMIN', 'CHMAX', 'PRP', 'ERP']\n"
     ]
    }
   ],
   "source": [
    "# view numerical variables\n",
    "\n",
    "print(numerical)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So, there are eight numerical variables in the dataset. Let's explore more about them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>MYCT</th>\n",
       "      <th>MMIN</th>\n",
       "      <th>MMAX</th>\n",
       "      <th>CACH</th>\n",
       "      <th>CHMIN</th>\n",
       "      <th>CHMAX</th>\n",
       "      <th>PRP</th>\n",
       "      <th>ERP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>125</td>\n",
       "      <td>256</td>\n",
       "      <td>6000</td>\n",
       "      <td>256</td>\n",
       "      <td>16</td>\n",
       "      <td>128</td>\n",
       "      <td>198</td>\n",
       "      <td>199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>32000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>269</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>32000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>220</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>32000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>172</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>16000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>16</td>\n",
       "      <td>132</td>\n",
       "      <td>132</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   MYCT  MMIN   MMAX  CACH  CHMIN  CHMAX  PRP  ERP\n",
       "0   125   256   6000   256     16    128  198  199\n",
       "1    29  8000  32000    32      8     32  269  253\n",
       "2    29  8000  32000    32      8     32  220  253\n",
       "3    29  8000  32000    32      8     32  172  253\n",
       "4    29  8000  16000    32      8     16  132  132"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# view the top 5 rows of numerical variables\n",
    "\n",
    "df[numerical].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that we have eight numerical variables in the dataset. All the eight numerical variables are of discrete type.\n",
    "\n",
    "On closer inspection, we find that **PRP** is a redundant column in the dataframe. It denotes **published relative performance**. Our target is to predict **estimated relative performance**. So, we should delete **PRP** from the dataframe."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Summary : types of variables**\n",
    "\n",
    "\n",
    "- There are 2 categorical variables and 8 numerical variables.\n",
    "\n",
    "- The 2 categorical variables, **Vendor Name** and **Model Name** are 2 non-predictive attributes as given in the dataset description. So, I do not use them for model building.\n",
    "\n",
    "- All of the 8 numerical variables are of discrete type.\n",
    "\n",
    "- Out of the 8 numerical variables, **PRP** is the linear regression's guess. It is redundant column. I do not use it for model building.\n",
    "\n",
    "- **ERP** (estimated relative performance is the goal field). It is the target variable."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Estimate correlation coefficients\n",
    "\n",
    "\n",
    "Our dataset is very small. So, we can compute the standard correlation coefficient (also called Pearson's r) between every pair of attributes. \n",
    "\n",
    "We can compute it using the `df.corr()` method as follows:-"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>MYCT</th>\n",
       "      <th>MMIN</th>\n",
       "      <th>MMAX</th>\n",
       "      <th>CACH</th>\n",
       "      <th>CHMIN</th>\n",
       "      <th>CHMAX</th>\n",
       "      <th>PRP</th>\n",
       "      <th>ERP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MYCT</th>\n",
       "      <td>1.0000</td>\n",
       "      <td>-0.3356</td>\n",
       "      <td>-0.3786</td>\n",
       "      <td>-0.3210</td>\n",
       "      <td>-0.3011</td>\n",
       "      <td>-0.2505</td>\n",
       "      <td>-0.3071</td>\n",
       "      <td>-0.2884</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MMIN</th>\n",
       "      <td>-0.3356</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.7582</td>\n",
       "      <td>0.5347</td>\n",
       "      <td>0.5172</td>\n",
       "      <td>0.2669</td>\n",
       "      <td>0.7949</td>\n",
       "      <td>0.8193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MMAX</th>\n",
       "      <td>-0.3786</td>\n",
       "      <td>0.7582</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.5380</td>\n",
       "      <td>0.5605</td>\n",
       "      <td>0.5272</td>\n",
       "      <td>0.8630</td>\n",
       "      <td>0.9012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CACH</th>\n",
       "      <td>-0.3210</td>\n",
       "      <td>0.5347</td>\n",
       "      <td>0.5380</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.5822</td>\n",
       "      <td>0.4878</td>\n",
       "      <td>0.6626</td>\n",
       "      <td>0.6486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CHMIN</th>\n",
       "      <td>-0.3011</td>\n",
       "      <td>0.5172</td>\n",
       "      <td>0.5605</td>\n",
       "      <td>0.5822</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.5483</td>\n",
       "      <td>0.6089</td>\n",
       "      <td>0.6106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CHMAX</th>\n",
       "      <td>-0.2505</td>\n",
       "      <td>0.2669</td>\n",
       "      <td>0.5272</td>\n",
       "      <td>0.4878</td>\n",
       "      <td>0.5483</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.6052</td>\n",
       "      <td>0.5922</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PRP</th>\n",
       "      <td>-0.3071</td>\n",
       "      <td>0.7949</td>\n",
       "      <td>0.8630</td>\n",
       "      <td>0.6626</td>\n",
       "      <td>0.6089</td>\n",
       "      <td>0.6052</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.9665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ERP</th>\n",
       "      <td>-0.2884</td>\n",
       "      <td>0.8193</td>\n",
       "      <td>0.9012</td>\n",
       "      <td>0.6486</td>\n",
       "      <td>0.6106</td>\n",
       "      <td>0.5922</td>\n",
       "      <td>0.9665</td>\n",
       "      <td>1.0000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         MYCT    MMIN    MMAX    CACH   CHMIN   CHMAX     PRP     ERP\n",
       "MYCT   1.0000 -0.3356 -0.3786 -0.3210 -0.3011 -0.2505 -0.3071 -0.2884\n",
       "MMIN  -0.3356  1.0000  0.7582  0.5347  0.5172  0.2669  0.7949  0.8193\n",
       "MMAX  -0.3786  0.7582  1.0000  0.5380  0.5605  0.5272  0.8630  0.9012\n",
       "CACH  -0.3210  0.5347  0.5380  1.0000  0.5822  0.4878  0.6626  0.6486\n",
       "CHMIN -0.3011  0.5172  0.5605  0.5822  1.0000  0.5483  0.6089  0.6106\n",
       "CHMAX -0.2505  0.2669  0.5272  0.4878  0.5483  1.0000  0.6052  0.5922\n",
       "PRP   -0.3071  0.7949  0.8630  0.6626  0.6089  0.6052  1.0000  0.9665\n",
       "ERP   -0.2884  0.8193  0.9012  0.6486  0.6106  0.5922  0.9665  1.0000"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# estimate correlation coefficients\n",
    "\n",
    "pd.options.display.float_format = '{:,.4f}'.format\n",
    "corr_matrix = df.corr()\n",
    "corr_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ERP      1.0000\n",
       "PRP      0.9665\n",
       "MMAX     0.9012\n",
       "MMIN     0.8193\n",
       "CACH     0.6486\n",
       "CHMIN    0.6106\n",
       "CHMAX    0.5922\n",
       "MYCT    -0.2884\n",
       "Name: ERP, dtype: float64"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_matrix['ERP'].sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Interpretation of correlation coefficient**\n",
    "\n",
    "The correlation coefficient ranges from -1 to +1. \n",
    "\n",
    "When it is close to +1, this signifies that there is a strong positive correlation. So, we can see that there is a strong positive correlation between `ERP` and `MMAX`. \n",
    "\n",
    "\n",
    "When it is clsoe to -1, it means that there is a strong negative correlation. So, there is a small negative correlation between `ERP` and `MYCT`.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Correlation heat map"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(16,10))\n",
    "plt.title('Correlation of Attributes with ERP')\n",
    "a = sns.heatmap(corr_matrix, square=True, annot=True, fmt='.2f', linecolor='white')\n",
    "a.set_xticklabels(a.get_xticklabels(), rotation=90)\n",
    "a.set_yticklabels(a.get_yticklabels(), rotation=30)           \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that **ERP** is positively correlated with **MMIN**, **MMAX**, **CACH**, **CHMIN** and **CHMAX**.\n",
    "\n",
    "Also, there is a strong positive correlation between **ERP** and **MMIN** and also between **ERP** and **MMAX**."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 10. Detect problems within variables"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Detect missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Vendor Name    0\n",
       "Model Name     0\n",
       "MYCT           0\n",
       "MMIN           0\n",
       "MMAX           0\n",
       "CACH           0\n",
       "CHMIN          0\n",
       "CHMAX          0\n",
       "PRP            0\n",
       "ERP            0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# let's visualise the number of missing values\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can confirm that there are no missing values in the dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Outliers in discrete variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>MYCT</th>\n",
       "      <th>MMIN</th>\n",
       "      <th>MMAX</th>\n",
       "      <th>CACH</th>\n",
       "      <th>CHMIN</th>\n",
       "      <th>CHMAX</th>\n",
       "      <th>PRP</th>\n",
       "      <th>ERP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>209.0000</td>\n",
       "      <td>209.0000</td>\n",
       "      <td>209.0000</td>\n",
       "      <td>209.0000</td>\n",
       "      <td>209.0000</td>\n",
       "      <td>209.0000</td>\n",
       "      <td>209.0000</td>\n",
       "      <td>209.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>203.8230</td>\n",
       "      <td>2,867.9809</td>\n",
       "      <td>11,796.1531</td>\n",
       "      <td>25.2057</td>\n",
       "      <td>4.6986</td>\n",
       "      <td>18.2679</td>\n",
       "      <td>105.6220</td>\n",
       "      <td>99.3301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>260.2629</td>\n",
       "      <td>3,878.7428</td>\n",
       "      <td>11,726.5644</td>\n",
       "      <td>40.6287</td>\n",
       "      <td>6.8163</td>\n",
       "      <td>25.9973</td>\n",
       "      <td>160.8307</td>\n",
       "      <td>154.7571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>17.0000</td>\n",
       "      <td>64.0000</td>\n",
       "      <td>64.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>6.0000</td>\n",
       "      <td>15.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>50.0000</td>\n",
       "      <td>768.0000</td>\n",
       "      <td>4,000.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>5.0000</td>\n",
       "      <td>27.0000</td>\n",
       "      <td>28.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>110.0000</td>\n",
       "      <td>2,000.0000</td>\n",
       "      <td>8,000.0000</td>\n",
       "      <td>8.0000</td>\n",
       "      <td>2.0000</td>\n",
       "      <td>8.0000</td>\n",
       "      <td>50.0000</td>\n",
       "      <td>45.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>225.0000</td>\n",
       "      <td>4,000.0000</td>\n",
       "      <td>16,000.0000</td>\n",
       "      <td>32.0000</td>\n",
       "      <td>6.0000</td>\n",
       "      <td>24.0000</td>\n",
       "      <td>113.0000</td>\n",
       "      <td>101.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1,500.0000</td>\n",
       "      <td>32,000.0000</td>\n",
       "      <td>64,000.0000</td>\n",
       "      <td>256.0000</td>\n",
       "      <td>52.0000</td>\n",
       "      <td>176.0000</td>\n",
       "      <td>1,150.0000</td>\n",
       "      <td>1,238.0000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            MYCT        MMIN        MMAX     CACH    CHMIN    CHMAX  \\\n",
       "count   209.0000    209.0000    209.0000 209.0000 209.0000 209.0000   \n",
       "mean    203.8230  2,867.9809 11,796.1531  25.2057   4.6986  18.2679   \n",
       "std     260.2629  3,878.7428 11,726.5644  40.6287   6.8163  25.9973   \n",
       "min      17.0000     64.0000     64.0000   0.0000   0.0000   0.0000   \n",
       "25%      50.0000    768.0000  4,000.0000   0.0000   1.0000   5.0000   \n",
       "50%     110.0000  2,000.0000  8,000.0000   8.0000   2.0000   8.0000   \n",
       "75%     225.0000  4,000.0000 16,000.0000  32.0000   6.0000  24.0000   \n",
       "max   1,500.0000 32,000.0000 64,000.0000 256.0000  52.0000 176.0000   \n",
       "\n",
       "             PRP        ERP  \n",
       "count   209.0000   209.0000  \n",
       "mean    105.6220    99.3301  \n",
       "std     160.8307   154.7571  \n",
       "min       6.0000    15.0000  \n",
       "25%      27.0000    28.0000  \n",
       "50%      50.0000    45.0000  \n",
       "75%     113.0000   101.0000  \n",
       "max   1,150.0000 1,238.0000  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# let's view the summary statistics of the dataset\n",
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50     0.1196\n",
      "140    0.0431\n",
      "300    0.0383\n",
      "26     0.0383\n",
      "38     0.0335\n",
      "320    0.0335\n",
      "56     0.0335\n",
      "180    0.0335\n",
      "800    0.0287\n",
      "75     0.0287\n",
      "105    0.0287\n",
      "200    0.0287\n",
      "143    0.0239\n",
      "900    0.0239\n",
      "160    0.0239\n",
      "400    0.0191\n",
      "60     0.0191\n",
      "29     0.0191\n",
      "25     0.0191\n",
      "23     0.0191\n",
      "110    0.0191\n",
      "92     0.0144\n",
      "100    0.0144\n",
      "250    0.0144\n",
      "115    0.0144\n",
      "125    0.0144\n",
      "30     0.0144\n",
      "480    0.0144\n",
      "225    0.0144\n",
      "330    0.0144\n",
      "810    0.0096\n",
      "1500   0.0096\n",
      "72     0.0096\n",
      "40     0.0096\n",
      "57     0.0096\n",
      "59     0.0096\n",
      "17     0.0096\n",
      "133    0.0096\n",
      "1100   0.0096\n",
      "240    0.0096\n",
      "700    0.0096\n",
      "64     0.0048\n",
      "220    0.0048\n",
      "203    0.0048\n",
      "185    0.0048\n",
      "175    0.0048\n",
      "167    0.0048\n",
      "35     0.0048\n",
      "150    0.0048\n",
      "116    0.0048\n",
      "124    0.0048\n",
      "70     0.0048\n",
      "48     0.0048\n",
      "112    0.0048\n",
      "52     0.0048\n",
      "98     0.0048\n",
      "350    0.0048\n",
      "600    0.0048\n",
      "84     0.0048\n",
      "90     0.0048\n",
      "Name: MYCT, dtype: float64\n",
      "\n",
      "2000    0.2584\n",
      "1000    0.1818\n",
      "4000    0.1053\n",
      "512     0.1053\n",
      "8000    0.0957\n",
      "256     0.0622\n",
      "768     0.0478\n",
      "16000   0.0335\n",
      "262     0.0096\n",
      "3100    0.0096\n",
      "5240    0.0096\n",
      "1310    0.0096\n",
      "2620    0.0096\n",
      "384     0.0096\n",
      "32000   0.0048\n",
      "1500    0.0048\n",
      "524     0.0048\n",
      "64      0.0048\n",
      "192     0.0048\n",
      "96      0.0048\n",
      "3000    0.0048\n",
      "500     0.0048\n",
      "5000    0.0048\n",
      "128     0.0048\n",
      "2300    0.0048\n",
      "Name: MMIN, dtype: float64\n",
      "\n",
      "8000    0.2057\n",
      "16000   0.1675\n",
      "4000    0.1579\n",
      "32000   0.1100\n",
      "2000    0.0813\n",
      "12000   0.0478\n",
      "1000    0.0335\n",
      "6000    0.0287\n",
      "3000    0.0239\n",
      "5000    0.0239\n",
      "24000   0.0191\n",
      "64000   0.0191\n",
      "6200    0.0144\n",
      "2620    0.0096\n",
      "10480   0.0096\n",
      "512     0.0096\n",
      "20970   0.0096\n",
      "768     0.0048\n",
      "64      0.0048\n",
      "6300    0.0048\n",
      "3500    0.0048\n",
      "1500    0.0048\n",
      "4500    0.0048\n",
      "Name: MMAX, dtype: float64\n",
      "\n",
      "0     0.3301\n",
      "8     0.1483\n",
      "32    0.1100\n",
      "64    0.0957\n",
      "16    0.0670\n",
      "4     0.0383\n",
      "24    0.0335\n",
      "128   0.0287\n",
      "6     0.0239\n",
      "2     0.0191\n",
      "30    0.0191\n",
      "1     0.0096\n",
      "9     0.0096\n",
      "256   0.0096\n",
      "48    0.0096\n",
      "65    0.0096\n",
      "112   0.0096\n",
      "131   0.0096\n",
      "12    0.0048\n",
      "142   0.0048\n",
      "96    0.0048\n",
      "160   0.0048\n",
      "Name: CACH, dtype: float64\n",
      "\n",
      "1    0.4498\n",
      "3    0.1340\n",
      "8    0.0861\n",
      "6    0.0766\n",
      "12   0.0526\n",
      "16   0.0478\n",
      "4    0.0383\n",
      "5    0.0335\n",
      "2    0.0287\n",
      "0    0.0239\n",
      "52   0.0096\n",
      "32   0.0048\n",
      "26   0.0048\n",
      "24   0.0048\n",
      "7    0.0048\n",
      "Name: CHMIN, dtype: float64\n",
      "\n",
      "6     0.1435\n",
      "24    0.1148\n",
      "8     0.0957\n",
      "32    0.0718\n",
      "5     0.0622\n",
      "16    0.0574\n",
      "2     0.0526\n",
      "4     0.0526\n",
      "1     0.0478\n",
      "3     0.0431\n",
      "12    0.0383\n",
      "20    0.0239\n",
      "0     0.0239\n",
      "64    0.0239\n",
      "10    0.0191\n",
      "14    0.0191\n",
      "38    0.0144\n",
      "54    0.0144\n",
      "7     0.0096\n",
      "176   0.0096\n",
      "128   0.0096\n",
      "104   0.0096\n",
      "13    0.0048\n",
      "15    0.0048\n",
      "26    0.0048\n",
      "28    0.0048\n",
      "31    0.0048\n",
      "48    0.0048\n",
      "52    0.0048\n",
      "112   0.0048\n",
      "19    0.0048\n",
      "Name: CHMAX, dtype: float64\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# outlies in discrete variables\n",
    "\n",
    "for var in ['MYCT', 'MMIN', 'MMAX', 'CACH', 'CHMIN', 'CHMAX']:\n",
    "    print(df[var].value_counts() / np.float(len(df)))\n",
    "    print()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# detect outliers in discrete variables\n",
    "\n",
    "for var in ['MYCT', 'MMIN', 'MMAX', 'CACH', 'CHMIN', 'CHMAX']:\n",
    "    plt.figure(figsize=(16,10))\n",
    "    (df.groupby(var)[var].count() / np.float(len(df))).plot.bar()\n",
    "    plt.ylabel('Percentage of observations per label')\n",
    "    plt.title(var)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the above plot, we can see that the discrete variables show values that are shared by a tiny proportion of variable values\n",
    "in the dataset. For linear regression modeling, this does not cause any problem."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Number of labels: cardinality\n",
    "\n",
    "Now, I will examine the categorical variable **Vendor Name**. First I will determine whether it show high cardinality. This is a high number of labels."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot the categorical variable\n",
    "\n",
    "plt.figure(figsize=(12,8))\n",
    "(df['Vendor Name'].value_counts()).plot.bar()\n",
    "plt.title('Number of categories in Vendor Name variable')\n",
    "plt.xlabel('Vendor Name')\n",
    "plt.ylabel('Number of different categories')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that the **Vendor Name** variable, contain only a few labels. So, we do not have to deal with high cardinality."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 11. Linear Regression Modeling\n",
    "\n",
    "\n",
    "Now, I discuss the most important part of this project which is the Linear Regression model building. \n",
    "\n",
    "First of all, I will divide the dataset into categorical and numerical variables as follows:-"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Divide the dataset into categorical and numerical variables\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cat = df.iloc[:,:2]\n",
    "\n",
    "df_num = df.iloc[:, 2:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>MYCT</th>\n",
       "      <th>MMIN</th>\n",
       "      <th>MMAX</th>\n",
       "      <th>CACH</th>\n",
       "      <th>CHMIN</th>\n",
       "      <th>CHMAX</th>\n",
       "      <th>PRP</th>\n",
       "      <th>ERP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>125</td>\n",
       "      <td>256</td>\n",
       "      <td>6000</td>\n",
       "      <td>256</td>\n",
       "      <td>16</td>\n",
       "      <td>128</td>\n",
       "      <td>198</td>\n",
       "      <td>199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>32000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>269</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>32000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>220</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>32000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>32</td>\n",
       "      <td>172</td>\n",
       "      <td>253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>29</td>\n",
       "      <td>8000</td>\n",
       "      <td>16000</td>\n",
       "      <td>32</td>\n",
       "      <td>8</td>\n",
       "      <td>16</td>\n",
       "      <td>132</td>\n",
       "      <td>132</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   MYCT  MMIN   MMAX  CACH  CHMIN  CHMAX  PRP  ERP\n",
       "0   125   256   6000   256     16    128  198  199\n",
       "1    29  8000  32000    32      8     32  269  253\n",
       "2    29  8000  32000    32      8     32  220  253\n",
       "3    29  8000  32000    32      8     32  172  253\n",
       "4    29  8000  16000    32      8     16  132  132"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_num.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Select the predictor and target variables\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df_num.iloc[:,0:6]\n",
    "\n",
    "y = df_num.iloc[:,-1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create separate train and test sets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X,y,test_size = 0.3,random_state = 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### View the dimensions of X_train, X_test, y_train, y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((146, 6), (146,))"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((63, 6), (63,))"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "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>MYCT</th>\n",
       "      <th>MMIN</th>\n",
       "      <th>MMAX</th>\n",
       "      <th>CACH</th>\n",
       "      <th>CHMIN</th>\n",
       "      <th>CHMAX</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>800</td>\n",
       "      <td>256</td>\n",
       "      <td>8000</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>320</td>\n",
       "      <td>128</td>\n",
       "      <td>6000</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>25</td>\n",
       "      <td>1310</td>\n",
       "      <td>2620</td>\n",
       "      <td>131</td>\n",
       "      <td>12</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>800</td>\n",
       "      <td>256</td>\n",
       "      <td>8000</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>220</td>\n",
       "      <td>1000</td>\n",
       "      <td>8000</td>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    MYCT  MMIN  MMAX  CACH  CHMIN  CHMAX\n",
       "61   800   256  8000     0      1      4\n",
       "24   320   128  6000     0      1     12\n",
       "30    25  1310  2620   131     12     24\n",
       "60   800   256  8000     0      1      4\n",
       "56   220  1000  8000    16      1      2"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# let's inspect the training dataframe\n",
    "\n",
    "X_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>MYCT</th>\n",
       "      <th>MMIN</th>\n",
       "      <th>MMAX</th>\n",
       "      <th>CACH</th>\n",
       "      <th>CHMIN</th>\n",
       "      <th>CHMAX</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>146.0000</td>\n",
       "      <td>146.0000</td>\n",
       "      <td>146.0000</td>\n",
       "      <td>146.0000</td>\n",
       "      <td>146.0000</td>\n",
       "      <td>146.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>205.8082</td>\n",
       "      <td>2,799.9726</td>\n",
       "      <td>11,741.2055</td>\n",
       "      <td>25.5685</td>\n",
       "      <td>4.5479</td>\n",
       "      <td>19.2397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>249.6152</td>\n",
       "      <td>3,865.5077</td>\n",
       "      <td>11,879.6456</td>\n",
       "      <td>41.6903</td>\n",
       "      <td>6.5770</td>\n",
       "      <td>28.8810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>17.0000</td>\n",
       "      <td>64.0000</td>\n",
       "      <td>64.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>50.0000</td>\n",
       "      <td>512.0000</td>\n",
       "      <td>4,000.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>5.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>115.5000</td>\n",
       "      <td>2,000.0000</td>\n",
       "      <td>8,000.0000</td>\n",
       "      <td>8.0000</td>\n",
       "      <td>1.5000</td>\n",
       "      <td>8.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>240.0000</td>\n",
       "      <td>4,000.0000</td>\n",
       "      <td>16,000.0000</td>\n",
       "      <td>32.0000</td>\n",
       "      <td>6.0000</td>\n",
       "      <td>24.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1,500.0000</td>\n",
       "      <td>32,000.0000</td>\n",
       "      <td>64,000.0000</td>\n",
       "      <td>256.0000</td>\n",
       "      <td>52.0000</td>\n",
       "      <td>176.0000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            MYCT        MMIN        MMAX     CACH    CHMIN    CHMAX\n",
       "count   146.0000    146.0000    146.0000 146.0000 146.0000 146.0000\n",
       "mean    205.8082  2,799.9726 11,741.2055  25.5685   4.5479  19.2397\n",
       "std     249.6152  3,865.5077 11,879.6456  41.6903   6.5770  28.8810\n",
       "min      17.0000     64.0000     64.0000   0.0000   0.0000   0.0000\n",
       "25%      50.0000    512.0000  4,000.0000   0.0000   1.0000   5.0000\n",
       "50%     115.5000  2,000.0000  8,000.0000   8.0000   1.5000   8.0000\n",
       "75%     240.0000  4,000.0000 16,000.0000  32.0000   6.0000  24.0000\n",
       "max   1,500.0000 32,000.0000 64,000.0000 256.0000  52.0000 176.0000"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Scaling\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Feature Scaling - I use the StandardScaler from sklearn\n",
    "\n",
    "# import the StandardScaler class from preprocessing library\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "# instantiate an object scaler\n",
    "scaler = StandardScaler()\n",
    "\n",
    "# fit the scaler to the training set and then transform it\n",
    "X_train = scaler.fit_transform(X_train)\n",
    "\n",
    "# transform the test set\n",
    "X_test = scaler.transform(X_test)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The scaler is now ready, we can use it in a machine learning algorithm when required."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fit the Linear Regression model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False)"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit the linear regression model\n",
    "\n",
    "# import the LinearRegression class from linear_model library\n",
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "# instantiate an object lr\n",
    "lr = LinearRegression()\n",
    "\n",
    "\n",
    "# Train the model using the training sets\n",
    "lr.fit(X_train, y_train)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 12. Predicting the results\n",
    "\n",
    "\n",
    "I have built the linear regression model. Now it is time to predict the results."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Predicting the test set results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Predict on the test data set\n",
    "y_pred = lr.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Predicting estimated relative CPU performance values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 53.25899879,  -7.30914167,  85.61134478, 333.46353054,\n",
       "        88.17105392])"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#print(\"Predicted ERP - estimated relative performance for the first five values\")\n",
    "\n",
    "lr.predict(X_test)[0:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 13. Model slope and intercept terms"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The slope parameters(w) are also called weights or coefficients. They are stored in the **coef_** attribute.\n",
    "\n",
    "The offset or intercept(b) is stored in the **intercept_** attribute.\n",
    "\n",
    "So, the model slope is given by **lr.coef_** and model intercept term is given by **lr.intercept_**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of coefficients: 6\n",
      "Estimated coefficients: [17.70202595 59.11241774 78.35042681 16.53981449 -0.35410978 38.97256261]\n",
      "Estimated intercept: 100.0\n"
     ]
    }
   ],
   "source": [
    "print(\"Number of coefficients:\", len(lr.coef_))\n",
    "\n",
    "print(\"Estimated coefficients: {}\".format(lr.coef_))\n",
    "\n",
    "print(\"Estimated intercept: {}\".format(lr.intercept_))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I constructed a dataframe that contains features and estimated coefficients. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "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>Estimated Coefficients</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Features</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>17.7020</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>59.1124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>78.3504</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>16.5398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.3541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>38.9726</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Estimated Coefficients\n",
       "Features                        \n",
       "0                        17.7020\n",
       "1                        59.1124\n",
       "2                        78.3504\n",
       "3                        16.5398\n",
       "4                        -0.3541\n",
       "5                        38.9726"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset = list(zip(pd.DataFrame(X_train).columns, lr.coef_))\n",
    "\n",
    "pd.DataFrame(data = dataset, columns = ['Features', 'Estimated Coefficients']).set_index('Features')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 14. Evaluate model performance\n",
    "\n",
    "\n",
    "I have built the linear regression model and use it to predict the results. Now, it is the time to evaluate the model performance. We want to understand the outcome of our model and we want to know whether the performance is acceptable or not. \n",
    "For regression problems, there are several ways to evaluate the model performance. These are listed below:-\n",
    "\n",
    "\n",
    "•\tRMSE (Root Mean Square Error)\n",
    "\n",
    "•\tR2 Score\n",
    "\n",
    "•\tOverfitting Vs Underfitting\n",
    "\n",
    "•\tCross validation\n",
    "\n",
    "•\tResidual analysis\n",
    "\n",
    "•\tNormality test\n",
    "\n",
    "I have described these measures in following sections:-\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### i.\tRMSE\n",
    "\n",
    "RMSE stands for **Root Mean Square Error**. RMSE is the standard deviation of the residuals. RMSE gives us the standard deviation of the unexplained variance by the model. It can be calculated by taking square root of Mean Squared Error.\n",
    "\n",
    "\n",
    "RMSE is an absolute measure of fit. It gives us how spread the residuals are, given by the standard deviation of the residuals. The more concentrated the data is around the regression line, the lower the residuals and hence lower the standard deviation of residuals. It results in lower values of RMSE. So, lower values of RMSE indicate better fit of data. \n",
    "\n",
    "RMSE can be calculated as follows:-"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE value : 37.99\n"
     ]
    }
   ],
   "source": [
    "# RMSE(Root Mean Square Error)\n",
    "\n",
    "from sklearn.metrics import mean_squared_error\n",
    "mse = mean_squared_error(y_test, y_pred)\n",
    "rmse = np.sqrt(mse)\n",
    "print(\"RMSE value : {:.2f}\".format(rmse))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Interpretation\n",
    "\n",
    "The RMSE value has been found to be 37.99. It means the standard deviation for our prediction is 37.99. So, sometimes we expect the predictions to be off by more than 37.99 and other times we expect less than 37.99. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ii. R2 Score\n",
    "\n",
    "R2 Score is another metric to evaluate performance of a regression model. It is also called **Coefficient of Determination**. It gives us an idea of goodness of fit for the linear regression models. It indicates the percentage of variance that is explained by the model. \n",
    "\n",
    "\n",
    "**R2 Score = Explained Variation/Total Variation**\n",
    "\n",
    "\n",
    "\n",
    "Mathematically, we have\n",
    "\n",
    "\n",
    "\n",
    "$$R^2=1-\\frac{SS_{res}}{SS_{tot}}$$\n",
    "\n",
    "\n",
    "The total sum of squares, $SS_{tot}=\\sum_i(y_i-\\bar{y})^2$\n",
    "\n",
    "The regression sum of squares (explained sum of squares), $SS_{reg}=\\sum_i(f_i-\\bar{y})^2$\n",
    "\n",
    "The sum of squares of residuals (residual sum of squares), $SS_{res}=\\sum_i(y_i-f_i)^2 = \\sum_ie^2_i$\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "In general, the higher the R2 Score value, the better the model fits the data. Usually, its value ranges from 0 to 1. So, we want its value to be as close to 1. Its value can become negative if our model is wrong.\n",
    "\n",
    "\n",
    "\n",
    "R2 score value can be found as follows:-\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R2 Score value: 0.92\n"
     ]
    }
   ],
   "source": [
    "# R2 Score\n",
    "\n",
    "from sklearn.metrics import r2_score\n",
    "print(\"R2 Score value: {:.2f}\".format(r2_score(y_test, y_pred)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Interpretation\n",
    "\n",
    "\n",
    "In business decisions, the benchmark for the R2 score value is 0.7. It means if R2 score value >= 0.7, then the model is good enough to deploy on unseen data whereas if R2 score value < 0.7, then the model is not good enough to deploy. \n",
    "\n",
    "Our R2 score value has been found to be 0.92. It means that this model explains 92% of the variance in our dependent variable. So, the R2 score value confirms that the model is good enough to deploy because it provides good fit to the data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### iii. Overfitting Vs Underfitting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training set score: 0.91\n"
     ]
    }
   ],
   "source": [
    "# Evaluating training set performance\n",
    "\n",
    "print(\"Training set score: {:.2f}\".format(lr.score(X_train, y_train)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test set score: 0.92\n"
     ]
    }
   ],
   "source": [
    "# Evaluating test set performance\n",
    "\n",
    "print(\"Test set score: {:.2f}\".format(lr.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Interpretation \n",
    "\n",
    "Training set and test set performances are comparable. An R Square value of 0.92 is very good."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### iv. Cross validation\n",
    "\n",
    "\n",
    "Cross-validation is a vital step in evaluating a model. It maximizes the amount of data that is used to train the model. \n",
    "\n",
    "In cross-validation, we split the training data into several subgroups. Then we use each of them in turn to evaluate the model \n",
    "fitted on the remaining portion of the data.\n",
    "\n",
    "It helps us to obtain reliable estimates of the model's generalization performance. So, it helps us to understand how well\n",
    "the model performs on unseen data.\n",
    "\n",
    "We can perform cross validation as follows:-"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import the library\n",
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "# Compute 5-fold cross-validation scores: cv_scores\n",
    "cv_scores = cross_val_score(lr, X, y, cv=5)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.8484 -0.864   0.7149  0.8755  0.7707]\n"
     ]
    }
   ],
   "source": [
    "# print the 5-fold cross-validation scores\n",
    "print(cv_scores.round(4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average 5-Fold CV Score: 0.4691\n"
     ]
    }
   ],
   "source": [
    "# print the avarage 5-fold cross-validation scores\n",
    "print(\"Average 5-Fold CV Score: {}\".format(np.mean(cv_scores).round(4)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Interpretation**\n",
    "\n",
    "There is a large fluctuation in the cross validation scores of the model. \n",
    "\n",
    "The average 5-fold cross validation score is very poor and hence the linear regression model is not a great fit to the data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### v. Residual analysis\n",
    "\n",
    "\n",
    "A linear regression model may not represent the data appropriately. The model may be a poor fit to the data. So, we should validate our model by defining and examining residual plots. The difference between the observed value of the dependent variable (y) and the predicted value (ŷi) is called the **residual** and is denoted by e. The scatter-plot of these residuals is called **residual plot**.\n",
    "\n",
    "\n",
    "If the data points in a residual plot are randomly dispersed around horizontal axis and an approximate zero residual mean, a linear regression model may be appropriate for the data. Otherwise a non-linear model may be more appropriate.\n",
    "\n",
    "\n",
    "Now, I will plot the residual errors. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot for residual error\n",
    "\n",
    "# adjust the figure size\n",
    "plt.figure(figsize=(10,8))\n",
    "\n",
    "# plotting residual errors in training data\n",
    "plt.scatter(lr.predict(X_train), lr.predict(X_train) - y_train, c = 'b', s = 40, label = 'Train data', alpha = 0.5)\n",
    "\n",
    "# plotting residual errors in test data\n",
    "plt.scatter(lr.predict(X_test), lr.predict(X_test) - y_test, c = 'r', s = 40, label = 'Test data')\n",
    "\n",
    "# plotting line for zero residual error\n",
    "plt.hlines(y = 0, xmin = -100, xmax = 1000, linewidth = 3)\n",
    "\n",
    "# plotting legend\n",
    "plt.legend(loc = 'upper right')\n",
    "\n",
    "# plot title\n",
    "plt.title(\"Residual errors plot\")\n",
    "\n",
    "# function to show plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Interpretation of residual plots\n",
    "\n",
    "\n",
    "A regression model that has nicely fit the data will have its residuals display randomness (i.e., lack of any pattern). This comes from the **homoscedasticity** assumption of regression modeling. Typically scatter plots between residuals and predictors are used to confirm the assumption. Any pattern in the scatter-plot, results in a violation of this property and points towards a poor fitting model.\n",
    "\n",
    "Residual errors plot show that the data is randomly scattered around line zero. The plot does not display any pattern in the residuals.  Hence, we can conclude that the Linear Regression model is a good fit to the data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### vi. Normality test (Q-Q Plot)\n",
    "\n",
    "This is a visual or graphical test to check for normality of the data. This test helps us identify outliers and skewness. The test is performed by plotting the data verses theoretical quartiles. The same data is also plotted on a histogram to confirm normality.\n",
    "\n",
    "Any deviation from the straight line in normal plot or skewness/multi-modality in histogram shows that the data does not pass the normality test.\n",
    "\n",
    "\n",
    "Now, I will plot the Q-Q Plot as follows:-\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3AAAAGDCAYAAABqc/JJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzs3XmYHFXZsPH7SUKAgGwJ5kXIghpQFNwigoqGTQERUFzAAUJER2QLiIIa3N+8LohhJwxrgBFERUFAAYGRT1FkEQURNEA29gREQtiSnO+PU2M6k57M0j3T3TP377rm6q5TVaeeGjQ1T50tUkpIkiRJkurfkFoHIEmSJEnqHhM4SZIkSWoQJnCSJEmS1CBM4CRJkiSpQZjASZIkSVKDMIGTJEmSpAZhAqe6FREzI+JrVaprbEQsjoihxXZbRHymGnUX9f06IiZXq74eXPd/I2JhRDze39eWJCkiJkXEgn661uiIuCUinouIk/rjmlI9MoFTTUTEnIh4ofhH+N8RcWtEHBoR//3fZErp0JTSd7pZ1y6rOyalNC+ltG5KaVkVYv9mRFzSof7dU0qzKq27h3GMAY4Ftkop/U8nx6wXESdHxLwigZ1dbI8q9q/yu4uIgyPi9yXbcyLi5fZzSsrvjogUEeM7lH+zKN+2Q/mREXFvRAwvKTs6Iv4SEcN691uQpMGr+Pf6nohYEhGPR8SZEbF+F+esGRHfLZ4LL0TEvyLiixER/RDvhRHxvxVU0QwsBNZLKR3byTW2jYhri78tno6IP0fElGJf2WSz9KVucUyKiCs6HPOWorytQ3lExEMRcV+H8qHFtb/aoeyOiPhiL+9fAkzgVFsfTim9ChgHfA84Hjiv2hcZwMnBOGBRSunJcjuLROlG4E3AbsB6wLuBRcC25c5ZjYeB/Uvq3hpYu8w1AzgQeBro2CJ5BvBvYFpx7GuBbwGHpJSW9jAeSRrUIuJY4PvAl4D1ge2A8cD1EbHGak79KbAzsAfwKvK/2Z8DGqFFaxxwX0opldsZEdsDNwG/A14PjAQ+D+zew+s8Bbw7IkaWlE0G/lnm2PcBrwZeGxHvbC8sXhh/GvhyRLyhKP4ikIAZPYxHWokJnGoupfRsSukq4JPA5Ih4M6z8pi4iRkXE1SVv1P5fRAyJiIuBscCviham4yJifPGW7JCImAfcVFJWmsy9rng79mxEXBkRGxXXWuUNXXtLVUTsBnwV+GRxvb8W+0vf3g2JiBMiYm5EPBkRF7W/ES2JY3Lx9nNhREzr7HcTEesX5z9V1HdCUf8uwA3Aa4o4Lixz+kHF7+YjKaX7UkrLU0pPppS+k1K6tof/mS4u6ms3GbiozHE7AK8BpgL7lba2pZSWA4cAx0TENsA5wJkppbt6GIskDWoRsR75BdiRKaXfpJReSSnNAT4BbA58qpPzdgY+AOybUro3pbQ0pfQn4ABgavFirdx5cyLiKxFxX0Q8ExEXRMRanRz7xuKZ+O+I+HtE7FWUNwNNwHHFc+tXnZz/7oi4vXg23x4R7y7KLyQ/e9rPL9fz5kRgVkrp+ymlhSm7M6X0ibK/yM69DPwS2K+49lDy77a1zLGTgSuBa+nw4jKldC/wI+C8iHgj+e+HT1ejN5AGNxM41Y2U0p+BBeQkoKNji30bA6PJ/wimlNKBwDxya966KaUflJzzfuCNwAc7ueRB5LdjrwGWAqd2I8bfAP8H/KS43lvKHHZw8bMj8FpgXeD0Dse8F9iS/Bb068U/7OWcRn6z+trifg4CpqSUfkt+o/hoEcfBZc7dBfhNSmlxV/fVDX8C1isezEPJyfYlZY6bDPwK+EmxvWfpzpTSA8B3yW9INyP/ASJJ6pl3A2sBK3XzK/69/zU5SStnV+C2lNL8DufdRn7G7ryaazaRn6evA7YATuh4QNHy9yvgenKr1JFAa0RsmVJqISdAPyieWx8uc/5GwDXk5/FIcvJzTUSMLJ5zpef/tsO5I4DtgZ+t5h564iJWvLj8IPB34NEy1/xYEVcrHV5cFv6P3APm/wEnp5TuqVJ8GsRM4FRvHgU2KlP+CrAJMK540/j/OutCUeKbKaXnU0ovdLL/4uIN5PPA14BPFMlJpZqAH6WUHioepl8h/6Ne2vr3rZTSCymlvwJ/BVZJBEsSpa+klJ4r3q6eRO7u0h0jgce6cdwvizel/46IfwNndnJceyvcrsD9wCMd4h0BfBz4cUrpFfJDtNzELv+viO1nKaUXu3UnkqRSo4CFnXQ/f4z8srOz8zp7LqzuPIDTU0rzU0pPA9Mp6VZfYjvyS8vvpZReTindBFzdybHlfAj4V0rp4qJ18FLy82aVZK+MDcl/13b13HtN6TOveO69t+NBKaVbgY0iYkvys69cr5OPAi+RE9argWHFPZTW8zJwG/m5V64FT+oxEzjVm03J46c6OhGYTe7b/1BEfLkbdc3vwf65wBrkh1ulXlPUV1r3MHLLYbvSWSOXkB94HY0Chpepa9NuxrGInPR2ZZ+U0gbtP8BhnRx3MblbzsGUf5B9hNyS2d49sxXYPSL++wdB8WbybHLL4hGdddeRJK3WQmBUJ2O8NyGP4aLoatj+M7Y4r7Pnwn/P60THZ+ZryhzzGmB+0WW+9NjuPrc6Pj97cv4zwHK6fu49WvrMK557v+/k2IuBI8g9an5RZv9k4PIi2XyJ3CK60ovLiNgB2If83DylG/chdckETnWjGPy7KWX+IS1aoI5NKb2W/CbuC0VffsgDgsvpqoVuTMn3seRWvoXA88CIkriGsvJbya7qfZQ80Lq07qXAE12c19HCIqaOdT1S/vBV/Bb4YESs08PrlpVSmkuezGQPOnTbKUwmJ6LzIi9r8FNyUlz65vVrwJPkMXIzycmcJKln/khu+floaWHx7/3u5Ek8KLoatv/MIz8X3hV5FuPS87YlP19uWc01Oz4zHy1zzKPAmCiZUZqVn1s9fX52PL9TKaUl5N/Lvl0d2wMXk19qXlvU/18RsRmwE3BA5BlAHyd3p9wjVsz0vBZ5crYvkidT2TIiDqhifBqkTOBUc5Gnut8TuAy4pFz/8IjYMyJeHxEB/AdYVvxATox605JzQERsVXT9+za5S98y8ixTa0XEh4r+/CcAa5ac9wQwvsMDqtSl5Ik6No+IdVkxZq5HMy0WsVwOTI+IV0XEOOALlB97Vs7F5DemP4+INxSTn4yMiK9GxB49iaXEIcBORbfT/4qITcljJ/YE3lr8vIU8Q9rk4pi3AEcBny26v36T/Huc0stYJGlQSik9Sx5DfFpE7BYRa0Re0uWn5Jd/ZbvqFePGbiQ/F94UeVr77YrjLyrGKXfm8IjYrBin9lVWjHUudRv5JehxRUyTyC9dLyv2d/W8vhbYIiI+FRHDIuKTwFbk7ondcRxwcER8KYoZJCNP/39ZF+eVlVJ6mDz+vNxkYweS/17YkhXPvS3IYwnbX1x+G5ibUrqwSACbgRmlPVOk3jCBUy39KiKeIycZ08iDlTv7Y34C+c3hYvIbtjNTSm3Fvu8CJxR92XuytsrFwIXk7oxrkZOL9gfjYcC55Ld+z5P/QW730+JzUUSUm0Hx/KLuW8gtVi+SB3L3xpHF9R8it0z+uKi/S0V3jl3I4wduICe+fyZ3zbytN8GklB5MKd1RZteBwN0ppetTSo+3/5AHom9TJG/nAdNTSrOLul4APgucGBGjy9QpSepEMWnXV4EfAs+RnzcjgF06vmTrYF/gZuA35OfTH4vvzV1c8sfksV4PFT+rrOdWjPfai9wKuJA8pvqglNL9xSHnAVsVz+tfljl/EflF4LHkYQDHAXumlBZ2EVv7+beSW8V2Ah6KiKeBFlZ07e+xlNLvU0rlWhsnk/8WebzDc28meUbtieTlGZpL6votORk9ubfxSADR9TwQkiRJqmcR8Wlyq9x7iu6S3T1vFnn4wh5FAlbumDnAZzrO/CipNmyBkyRJanAppfPJLXLv7uGpnyH30nh71YOS1CdsgZMkSVKnbIGT6osJnCRJkiQ1CLtQSpIkSVKDMIGTJEmSpAYxrNYBAIwaNSqNHz++ojqef/551lmnKusV19RAuA/voT54D/XBe1jVnXfeuTCl5DpI3VSNZ6Qkqf519/lYFwnc+PHjueOOcktLdV9bWxuTJk2qTkA1NBDuw3uoD95DffAeVhURc6tW2SBQjWekJKn+dff5aBdKSZIkSWoQJnCSJEmS1CBM4CRJkiSpQZjASZIkSVKDMIGTJEmSpAZhAidJkiRJDcIETpIkSZIahAmcJEmSJDUIEzhJkiRJahAmcJIkSZIGhdZWGD8ehgzJn62t9VVfdwzr+0tIkiRJUm21tkJzMyxZkrfnzs3bAE1Nta+vu2yBkyRJkjTgTZu2Itlqt2RJLq+H+rqrywQuIs6PiCcj4t4y+74YESkiRhXbERGnRsTsiPhbRLy9L4KWJEmSpJ6YN69n5f1dX3d1pwvlhcDpwEWlhRExBtgVKA1xd2BC8fMu4Kzis8/d88izHPzla/rjUl2a870P1ToESZIkSSXGjs3dHMuV10N93dVlC1xK6Rbg6TK7ZgDHAamkbG/gopT9CdggIjapSqSSJEmS1EvTp8OIESuXjRiRy+uhvu7q1SQmEbEX8EhK6a8RUbprU2B+yfaCouyxMnU0A80Ao0ePpq2trTeh/NfoteHYrZdWVEe1VHIvixcvrvh3UWveQ33wHuqD9yBJUn1on1hk2rTczXHs2Jxs9XbCkWrX1109TuAiYgQwDfhAud1lylKZMlJKLUALwMSJE9OkSZN6GspKTmu9kpPuqY9JNec0Ter1uW1tbVT6u6g176E+eA/1wXuQJKl+NDVVN8Gqdn3d0ZuM53XA5kB769tmwF0RsS25xW1MybGbAY9WGqQkSZIkqRfLCKSU7kkpvTqlND6lNJ6ctL09pfQ4cBVwUDEb5XbAsymlVbpPSpIkSZJ6rjvLCFwK/BHYMiIWRMQhqzn8WuAhYDZwDnBYVaKUJEmSJHXdhTKltH8X+8eXfE/A4ZWHJUmSJEnqqMddKCVJkiRJtWECJ0mSJEkNwgROkiRJkhqECZwkSZIkNQgTOEmSJElqECZwkiRJktQgTOAkSZIkqUGYwEmSJElSgzCBkyRJkqQGYQInSZIkSQ3CBE6SJEmSGoQJnCRJkiQ1CBM4SZIkSWoQJnCSJEmS1CBM4CRJkiSpQZjASZIkSVKDMIGTJEmSpAZhAidJkiRJDcIETpIkSZIahAmcJEmSJDUIEzhJkiRJahAmcJIkSZLUIEzgJEmSJKlBmMBJkiRJUoMwgZMkqZci4vyIeDIi7i0p+2ZEPBIRdxc/e5Ts+0pEzI6IByLig7WJWpLUyEzgJEnqvQuB3cqUz0gpvbX4uRYgIrYC9gPeVJxzZkQM7bdIJUkDggmcJEm9lFK6BXi6m4fvDVyWUnoppfQwMBvYts+CkyQNSCZwkiRV3xER8beii+WGRdmmwPySYxYUZZIkdZsJnCRJ1XUW8DrgrcBjwElFeZQ5NpWrICKaI+KOiLjjqaee6psoJUkNyQROkqQqSik9kVJallJaDpzDim6SC4AxJYduBjzaSR0tKaWJKaWJG2+8cd8GLElqKCZwkiRVUURsUrL5EaB9hsqrgP0iYs2I2ByYAPy5v+OTJDW2YbUOQJKkRhURlwKTgFERsQD4BjApIt5K7h45B/gcQErp7xFxOXAfsBQ4PKW0rBZxS5IalwmcJEm9lFLav0zxeas5fjowve8ikiQNdHahlCRJkqQGYQInSZIkSQ2iywSuWMPmyYi4t6TsxIi4v1jj5hcRsUHJvq9ExOyIeCAiPthXgUuSJEnSYNOdFrgLgd06lN0AvDmltA3wT+ArABGxFbAf8KbinDMjYmjVopUkSZKkQazLBC6ldAvwdIey61NKS4vNP5HXsgHYG7gspfRSSulhYDYr1r+RJEmSJFWgGrNQfhr4SfF9U3JC125BUbaKiGgGmgFGjx5NW1tbRUGMXhuO3Xpp1wf2g0ruZfHixRX/LmrNe6gP3kN98B4kSVI1VZTARcQ08lo2re1FZQ5L5c5NKbUALQATJ05MkyZNqiQUTmu9kpPuqY9VEeY0Ter1uW1tbVT6u6g176E+eA/1wXuQJEnV1OuMJyImA3sCO6eU2pO0BcCYksM2Ax7tfXiSJEmSpHa9WkYgInYDjgf2SiktKdl1FbBfRKwZEZsDE4A/Vx6mJEmSJKnLFriIuBSYBIyKiAXAN8izTq4J3BARAH9KKR2aUvp7RFwO3EfuWnl4SmlZXwUvSZIkSYNJlwlcSmn/MsXnreb46cD0SoKSJEmSJK2qV10oJUmSJEn9zwROkiRJkhqECZwkSZIkNQgTOEmSJElqECZwkiRJktQgTOAkSZIkqUGYwEmSJElSgzCBkyRJkqQGYQInSZIkSQ3CBE6SJEmSGoQJnCRJkiQ1CBM4SZIkSXWrtRXGj4chQ/Jna2vP9g80w2odgCRJkiSV09oKzc2wZEnenjs3bwM0NXW9fyCyBU6SJElSXZo2bUVy1m7Jklzenf0DkQmcJEmSpLo0b97qy7vaPxCZwEmSJEmqS2PHrr68q/0DkQmcJEmSpLo0fTqMGLFy2YgRubw7+wciEzhJkiRJdampCVpaYNw4iMifLS0rJijpav9A5CyUkiRJkupWU9PqE7Ku9g80tsBJkiRJUoMwgZMkSZKkBmECJ0mSJEkNwgROkiRJkhqECZwkSZIkNQgTOEmSJElqECZwkiRJktQgTOAkSZIkqUGYwEmSJElSgzCBkyRJkqQGYQInSZIkSQ3CBE6SJEmSGoQJnCRJkiQ1CBM4SZIkSWoQJnCSJEmS1CBM4CRJkiTVXGsrjB8PQ4bkz9bWWkdUn7pM4CLi/Ih4MiLuLSnbKCJuiIh/FZ8bFuUREadGxOyI+FtEvL0vg5ckSZLU+FpbobkZ5s6FlPJnc7NJXDndaYG7ENitQ9mXgRtTShOAG4ttgN2BCcVPM3BWdcKUJEmSNFBNmwZLlqxctmRJLtfKukzgUkq3AE93KN4bmFV8nwXsU1J+Ucr+BGwQEZtUK1hJkiRJA8+8eT0rH8yG9fK80SmlxwBSSo9FxKuL8k2B+SXHLSjKHutYQUQ0k1vpGD16NG1tbb0MpQhobTh266UV1VEtldzL4sWLK/5d1Jr3UB+8h/rgPUiS1LWxY3O3yXLlWllvE7jORJmyVO7AlFIL0AIwceLENGnSpIoufFrrlZx0T7Vvp3fmNE3q9bltbW1U+ruoNe+hPngP9cF7kCSpa9On5zFvpd0oR4zI5VpZb2ehfKK9a2Tx+WRRvgAYU3LcZsCjvQ9PkiRJ0kDX1AQtLTBuHETkz5aWXK6V9TaBuwqYXHyfDFxZUn5QMRvldsCz7V0tJUmSJKkzTU0wZw4sX54/Td7K67LPYURcCkwCRkXEAuAbwPeAyyPiEGAe8PHi8GuBPYDZwBJgSh/ELEmSJEmDUpcJXEpp/0527Vzm2AQcXmlQkiRJkqRV9bYLpSRJkiSpn5nASZIkSeozra0wfjwMGZI/W1vLl6l76mPefUmSJEkDTmvryssDzJ0LU6bkmSZffnlFWXNz/u7EJV2zBU6SJElSn5g2beW13QBeeWVF8tZuyZJ8rLpmAidJkiSpT8yb1zfHDmYmcJIkSZL6xNixfXPsYGYCJ0lSL0XE+RHxZETcW1K2UUTcEBH/Kj43LMojIk6NiNkR8beIeHvtIpek/jF9OowYsXLZGmvA8OErl40YkY9V10zgJEnqvQuB3TqUfRm4MaU0Abix2AbYHZhQ/DQDZ/VTjJJUM01N0NIC48bliUvGjYMLLoDzz1+5rKXFCUy6y1koJUnqpZTSLRExvkPx3sCk4vssoA04vii/KKWUgD9FxAYRsUlK6bH+iVaSaqOpqXxyZsLWO7bASZJUXaPbk7Li89VF+abA/JLjFhRlkiR1mwmcJEn9I8qUpbIHRjRHxB0RccdTTz3Vx2FJkhqJCZwkSdX1RERsAlB8PlmULwDGlBy3GfBouQpSSi0ppYkppYkbb7xxnwYrSWosJnCSJFXXVcDk4vtk4MqS8oOK2Si3A551/JskqaecxESSpF6KiEvJE5aMiogFwDeA7wGXR8QhwDzg48Xh1wJ7ALOBJcCUfg9YktTwTOAkSSpRrNs2JqX0t66OTSnt38muncscm4DDKwxPkjTI2YVSkjToRURbRKwXERsBfwUuiIgf1TouSZI6MoGTJAnWTyn9B/gocEFK6R3ALjWOSZKkVZjASZIEw4oZIz8BXF3rYCRJ6owJnCRJ8G3gOuDBlNLtEfFa4F81jkmSpFU4iYkkadBLKf0U+GnJ9kPAvrWLSJKk8myBkyQNehGxRUTcGBH3FtvbRMQJtY5LkqSOTOAkSYJzgK8ArwAUSwjsV9OIJEkqwwROkiQYkVL6c4eypTWJRJKk1TCBkyQJFkbE64AEEBEfAx6rbUiSJK3KSUwkSYLDgRbgDRHxCPAwcEBtQ5IkaVUmcJKkQa+YdXKXiFgHGJJSeq7WMUmSVI4JnCRp0IuIr3fYBiCl9O2aBCRJUidM4CRJgudLvq8F7An8o0axSJLUKRM4SdKgl1I6qXQ7In4IXFWjcCRJ6pSzUEqStKoRwGtrHYQkSR3ZAidJGvQi4h6KJQSAocDGgOPfJEl1xwROkqQ85q3dUuCJlJILeUuS6o4JnCRp0IqIjYqvHZcNWC8iSCk93d8xSZK0OiZwkqTB7E5y18kosy/hODhJUp0xgZMkDVoppc1rHYMkST1RUQIXEccAnyG/pbwHmAJsAlwGbATcBRyYUnq5wjglSepTEbEhMIG8DhwAKaVbaheRJEmr6vUyAhGxKXAUMDGl9GbyrF37Ad8HZqSUJgDPAIdUI1BJkvpKRHwGuAW4DvhW8fnNWsYkSVI5la4DNwxYOyKGkdfMeQzYCfhZsX8WsE+F15Akqa9NBd4JzE0p7Qi8DXiqtiFJkrSqXidwKaVHgB8C88iJ27PkweD/Lpl6eQGwaaVBSpLUx15MKb0IEBFrppTuB7ascUySJK2i12PgirECewObA/8GfgrsXubQVKaMiGgGmgFGjx5NW1tbb0MBYPTacOzW9bFkTyX3snjx4op/F7XmPdQH76E+eA8NY0FEbAD8ErghIp4BHq1xTJIkraKSSUx2AR5OKT0FEBFXAO8GNoiIYUUr3GZ08gBMKbUALQATJ05MkyZNqiAUOK31Sk66pz4m1ZzTNKnX57a1tVHp76LWvIf64D3UB++hMaSUPlJ8/WZE3AysD/ymhiFJklRWJWPg5gHbRcSIiAhgZ+A+4GbgY8Uxk4ErKwtRkqS+ERHXRERTRKzTXpZS+l1K6SpnUJYk1aNKxsDdRp6s5C7yEgJDyC1qxwNfiIjZwEjgvCrEKUlSX2gB9gTmRMRPImKfiBhe66AkSepMRX0OU0rfAL7RofghYNtK6pUkqT+klK4EroyItYG9yD1HZkbEtcClKaUbahqgJEkdVLqMgCRJDS+l9EJK6SfFWLgPkJcRcAycJKnumMBJkga9iBgdEUdGxB/IM1FeD7yjxmFJkrSK+pi2UZKkGoiIzwL7k9d8uwI4LqX0h9pGJUlqGM88A+ecA695DRxwQL9c0hY4SdJg9m7ge8CYlNKRJm+SpG7517/giCNgzBg4/ni4+eZ+u7QtcJKkQSulNKXWMUiSGkRK0NYGM2bA1VfDGmvA/vvDMcfAW97Sb2GYwEmSJElSZ15+GS67LCdud98No0bBCSfAYYfB//xPv4djAidJkiRJHS1cCDNnwhlnwOOPw1Zb5fFuTU2w9to1C8sETpI0aEXERqvbn1J6ur9ikSTVifvug5NPhosvhhdfhN12y90kd90VImodnQmcJGlQuxNIQABjgWeK7xsA84DNaxeaJKnfpATXX5+7SV53Hay1Fhx0EEydmlve6ogJnCRp0EopbQ4QETOBq1JK1xbbuwO71DI2SVI/eOEFaG3Nidt99+Uxbd/5Dhx6aB7rVodM4CRJgnemlA5t30gp/ToivlPLgCRJfejxx+HMM+Gss/JYt7e+FWbNgk9+EtZcs9bRrZYJnCRJsDAiTgAuIXepPABYVNuQJElV99e/5ta2Sy+FV16BD384j297//vrYnxbd5jASZIE+wPfAH5BTuBuKcokSY1u+XK45pqcuN18M6yzDjQ3w1FHwYQJtY6ux4bUOgBJkmotpfR0SmkqsENK6e0ppaOdgVKS8vCw8eNhyJA8JGzUqNV/Hz8+L4/Wfs748bmOmnj++bwEwBveAHvtBbNnww9+APPnw2mnNWTyBrbASZJERLwbOBdYFxgbEW8BPpdSOqy2kUlS7bS25oaqJUvy9qKSjuWdfZ87Nw8rK91ubs7fm5r6LtaVLFgAp58OLS3wzDOw7bZ5Ie6PfhTWWKOfgug7tsBJkgQzgA9SjHtLKf0VeF9NI5KkGps2bUXyVoklS3Jdfe722+FTn4LNN4cTT4Sdd4Y//AH+9Kc8OckASN7AFjhJkgBIKc2PlQewL6tVLJJUD+bNq8+6VrJsGfzyl3l82x/+AOutl8e2HXlk7r85AJnASZIE84tulCkihgNHAf+ocUySVFNjx+YukNWqq6r+8x847zw49VSYMye3up18MkyZkpO4AcwulJIkwaHA4cCmwALgrcW2JA1a06fDiBGV1zNiRK6rKh5+OE/7v9lm8IUvwJgxcMUV8K9/wdSpAz55A1vgJEmDXEQMBQ5MKfXX8HpJagjtk45Mm5a7QG60Ud5++unOv48dC3vsAddem88ZOzYnbxVNYJIS3Hpr7ib5i1/k6S0/8YmcyE2cWEHFjckETpI0qKWUlkXE3uSJTCRJJZqa+nH2yI5eeQV+9rOcuN1+O2y4IRx3HBxxBGy6aY2Cqj0TOEmS4A8RcTrwE+D59sKU0l21C0mSBqlnnslLAJx+el4SYIst4Mwz4aCD8iLcg5wJnCRJ8O7i89slZQnYqQaxSNLg9M9/wimnwIUX5rUHdt4ZZs6E3XfP3SYFmMBJkkRKacdaxyBJg1JK0NYGP/oRXHNNXqutqQmOPhq22abW0dUlU1lJ0qAXEaMj4ryI+HWxvVVEHFLruCRpwHrpJZg1C972NthpJ7iIimwBAAAgAElEQVTtNvja1/LMJ+efb/K2GiZwkiTBhcB1wGuK7X8CR9csGkkaqJ56Cr7zHRg3Dg4+GJYuhXPPzYnbt74Fo0fXOsK6ZwInSRKMSildDiwHSCktBZbVNiRJGkD+/nf47GfzugJf/zq8/e1w/fVwzz1wyCGw1lq1jrBhmMBJkgTPR8RI8sQlRMR2wLO1DUmS+kdrK4wfDxEwbFj+HDUq/wwZkve1tvai4pTgN7+BD34Q3vzmXMnkyXDffXmhuF13zRdTjziJiSRJ8AXgKuB1EfEHYGPgY7UNSZL6XmsrNDfnSR8BlhV9DxYtWnHM3Ln5GOjmmnAvvAAXXwwnnwz/+Adssklezftzn4ORI6sa/2BkAidJGvRSSndFxPuBLYEAHkgpvVLjsCSpz02btiJ5W50lS/Kxq03gHnssr9d21lk5A3zb2+Cii+CTn4Thw6sW82BnAidJGrQi4qOd7NoiIkgpXdGvAUlSP2ptza1r3TVvXic77r4bZsyASy/Nk5LstRcccwy87312kewDJnCSpMHsw8Xnq8mLed9UbO8ItAEmcJIGpPaukz0xdmzJxvLlcPXVOXFra4N11oFDD4WjjoLXv76aoaoDEzhJ0qCVUpoCEBFXA1ullB4rtjcBzqik7oiYAzxHns1yaUppYkRsBPwEGA/MAT6RUnqmkutIUm90t+tkuxEj8jA2Fi+GCy+EU06B2bNhzBg48UT4zGdggw36KlyVcBZKSZJgfHvyVngC2KIK9e6YUnprSmlisf1l4MaU0gTgxmJbkvpdp90hCyNH5p+IvGTbJd+dT9Nfj8sJ25FH5p2XXQYPPQRf/KLJWz+yBU6SJGiLiOuAS8lLCewH3NwH19kbmFR8n0Xupnl8H1xHklZr7Njy49/GjYM5c0oK/vzn3E3yCz/NywLsu28e37b99v0VqjqoKIGLiA2Ac4E3kx94nwYewO4hkqQGklI6IiI+AryvKGpJKf2i0mqB6yMiAWenlFqA0e0tfSmlxyLi1eVOjIhmoBlg7EqDTiSpcq2tuSdkR//tJrl0Kfzylzlxu/VWWG89OPro3PI2bly/x6uVVdoCdwrwm5TSxyJiODAC+Cq5e8j3IuLL5O4hvl2UJNWliBgKXJdS2gWoNGkr9Z6U0qNFknZDRNzf3ROLZK8FYOLEiamKMUka5Dqu+9Zu5Eg487vP8oknzoPXn5qb51772jzWbcoUeNWrahOwVtHrBC4i1iO/qTwYIKX0MvByRNg9RJLUMFJKyyJiSUSsn1J6tor1Plp8PhkRvwC2BZ6IiE2K1rdNgCerdT1J6o5yk5dszkOc8PKpfOLY8+G552CHHfIi3B/+MAwdWptA1alKWuBeCzwFXBARbwHuBKbSze4hkiTVkReBeyLiBuD59sKU0lG9qSwi1gGGpJSeK75/APg2cBUwGfhe8XllpYFLUk+smLwk8V5+zzHMYG+uZPlzQ6Dpk3l82zveUcsQ1YVIqXc9MyJiIvAncheR2yLiFOA/wJEppQ1KjnsmpbRhmfP/279/9OjR77jssst6FUe7J59+lideqKiKqtl60/V7fe7ixYtZd911qxhN//Me6oP3UB+8h1XtuOOOd5bMylgXImJyufKU0qxe1vdaVnTHHAb8OKU0PSJGApcDY4F5wMdTSk+vrq6JEyemO+64ozdhSNIqXj/uFd4173KO5mTeyR0sYiPO5nNcuenh3LZg01qHN6hFRLeej5W0wC0AFqSUbiu2f0Ye79at7iEd+/dPmjSpglDgtNYrOeme+phUc07TpF6f29bWRqW/i1rzHuqD91AfvIeG8RPg9eSJRx5MKb1YSWUppYeAt5QpXwTsXEndktQrTz8NLS387bnTGcEj3M+WHMpZXMRBxIgRtHy/1gGqu3q9DlxK6XFgfkRsWRTtDNzHiu4hYPcQSVIdi4hhEfED8kvJWcAl5GfbDyJijdpGJ0lV8MADcNhhef22r3yFEe94Izd/8Rp2H3sfLXEorx43gpYWaGqqdaDqrkqbrI4EWosZKB8CppCTwssj4hCK7iEVXkOSpL5yIvAqYPOU0nPw30m6flj8TK1hbJLUOynBTTflZQCuuQaGD88Z2tFHwzbbsCPw8Im1DlK9VVECl1K6GyjXT9PuIZKkRrAnsEUqGRCeUvpPRHweuB8TOEmN5KWX4Mc/zjNI/u1vsPHG8I1vwOc/D6NH1zo6VUl9DBqTJKk2Uiozm1extIDrr0lqDE8+CTNnwhln5O9vfjOcdx586lOw1lq1jk5VZgInSRrM7ouIg1JKF5UWRsQB5BY4Sapf996bW9suuSS3vu2xR14GYOedIaLW0amPmMBJkgazw4ErIuLT5PVME/BOYG3gI7UMTJLKWr4crrsuj2+74QZYe22YMgWmToU3vKHW0akfmMBJkgatlNIjwLsiYifgTUAAv04p3VjbyCSpgyVL4OKLc4vb/ffDJpvA//0fNDfDyJG1jk79yAROkjTopZRuAm6qdRyStIpHH81j284+GxYtgre/PSdyn/hEnl1Sg06v14GTJEmS1EfuugsOPBDGj4fvfhd22AF+9zu44w444IBuJW+trfn0CBg2bOXP8ePzfjUeW+AkSZKkerBsGVx9dR7f9rvfwbrr5iUAjjoKXve6HlXV2pp7Vy5ZsqLq0s+5c/N+cBHvRmMLnCRJklRLixfDaafBllvCPvvAww/DD38I8+fDKaf0OHkDmDZtRfLWmSVL8nFqLLbASZIkSbUwb15O3M45B559FrbbLk9M8tGP5r6OFVZdzeNUP0zgJEmSpP70pz/lbpI//3ne3nffvH7bdttV7RJjx+Zukt05To3FLpSSJElSX1u6FH76U9h++/xz3XU5aXvoIfjJT6qavAFMnw4jRqz+mBEj8nFqLLbASZIkSX3l2Wfh3HNzV8m5c/N4tlNPhYMPhle9qs8u2z4xybRp+bJDh+YJTNo/x43LyZsTmDQeEzhJkiSp2h58MCdq55+fJyl5//vzhCR77pmzqH7Q1GSCNhDZhVKSJEmqhpTgllvgIx+BCRPgzDPzrJJ33gltbbD33mWTt9ZWGDUqr8/Wm5+hQ/PnkCGrlrne28BjC5wkSZJUiZdfhssvzxOT3HUXbLQRfOUrcPjh8JrXrPbU1laYMgVeeaX3l1++PH+mtGqZ670NPLbASZIkSb2xaBF897uw+eZw4IF5YbWZM/P6bdOnd5m8AUydWlny1h2u9zaw2AInSZIk9cQDD8DJJ8OsWfDCC7Drrnmikg9+MPdj7KbW1pwD9gfXexs4TOAkSZKkrqQEN96Yu0leey2suSYccAAcfTS8+c29qrI/W8Vc723gMIGTJEmSOvPii/DjH+cWt3vugVe/Gr75Tfj85/P3CvRXq5jrvQ0sJnCSJElSR088AWedlX+efBK23jovCbD//rDWWhVX39qae1suW1Z5qEOG5ElLIlZMZNJe5npvA48JnCRJktTunntya9sll+TZJT/0ITjmGNhpp5whVUFra54ZslzyNmIEtLSYcKlzJnB9YPyXr+n1ucduvZSDKzi/oznf+1DV6pIkSRqQli+H3/wmj2/77W9h7bXhkEPyFJFbbln1y02blmeG7GjoUJM3dc0ETpIkSYPTkiVw0UW5xe2BB/K0/9/9Lnz2szByZJ9ddu7c8uXLl5u8qWsmcJIkSRpcHnkEzjgDzj4bnn4a3vGO3GXy4x+H4cP79NKtrSuPVSvlTJHqDhM4SZIkDQ533pm7Sf7kJ3kA2j775PFt731v1ca3dWXatPLJW4QzRap7ur/SoCRJktRoli2DX/4S3v9+mDgRrrwSDj8cZs+GK66AHXZYJXlrbYVRo3JxtX866z6Zkt0n1T22wEmSJGngee45uOACOOUUeOihPJ/+SSflyUnWX7/T01pbYcoUeOWVfoyVHJ7UHSZwkiRJGjjmzoXTToNzz4Vnn4Xtt4fvfz93lxzW9Z++06b1f/Jm90n1hAmcJEmSGt8f/5jHt11xRd7+2Mfy+LZ3vatH1cyb1wexdcHuk+oJEzhJkiQ1pqVL4ec/z4nbbbflrpFf+AIccUSvp3QcO7bzcWp9xe6T6gknMZEkSVJj+fe/4Yc/hNe9DvbbDxYtyt0mFyyAH/ygx8lb6aQl/Z28DR9u90n1jC1wkiRJagwPPpgnJTn/fHj+eZg0KSduH/oQDB3aqyprNWkJ5LXCTznF7pPqGRM4SZIk1a+U4JZbcjfJq67KE5Hst18e3/a2t1Vc/dSpXSdvQ4fm3ppSPTCBkyRJUv15+eW84PaMGfCXv+Tmqq9+Na/htskmVblEa2vufdmVZcuqcjmpKkzgJEmSVD8WLYKzz4bTT4fHHoM3vjFvH3AAjBhR1UtNm9a943rZO1PqExUncBExFLgDeCSltGdEbA5cBmwE3AUcmFJ6udLrSJIkaQC7/344+WS46CJ44QX4wAfyWLcPfACG9M28e91dMqC5uU8uL/VKNf7fMBX4R8n294EZKaUJwDPAIVW4hiRJkgaalOCGG2CPPXJL24UX5hk97r0XrrsOdtut18lb6cySnf2ktPo6hgyBz38ezjyzVyFIfaKiBC4iNgM+BJxbbAewE/Cz4pBZwD6VXEOSJEkDzIsvwnnnwTbb5Ba2u+6Cb38b5s+Hc86BN72poupbW2Hy5O6Nb+to+HC45JKc3C1bZvKm+lNpF8qTgeOAVxXbI4F/p5Ta5+lZAGxa4TUkSZI0EDzxRM6IzjoLnnoqJ3AXXAD77w9rrlm1y0yd2vuJR171Kqf1V33rdQIXEXsCT6aU7oyISe3FZQ4t2zgdEc1AM8Do0aNpa2vrbSgAjF4bjt268ed3rfZ9VPp77Y3FixfX5LrV5D3UB++hPgyEe5BUY3/7W55N8sc/zrNL7rlnXgZgxx1zX8YqaG3NiVtvWt1KPf10VcKR+kwlLXDvAfaKiD2AtYD1yC1yG0TEsKIVbjPg0XInp5RagBaAiRMnpkmTJlUQCpzWeiUn3dP4k2oeu/XSqt7HnKZJVauru9ra2qj0v2eteQ/1wXuoDwPhHiTVwPLl8Otf58TtxhvzDJKf+UzOsrbYoqqXau8yWY3p/seOrbwOqS/1egxcSukrKaXNUkrjgf2Am1JKTcDNwMeKwyYDV1YcpSRJkhrD88/nLpJbbZVb2u6/H773vTy+7YwzKkreOpuY5IADqpO8DRsG06dXXo/Ul/qiyep44LKI+F/gL8B5fXANSZIk1ZNHHslrt519NjzzDEycmDOuj38c1lij4upbW2HKFHjllSrEWsa668LMmY5/U/2rSgKXUmoD2orvDwHbVqNeSZIk1bk77sjdJC+/PHeb3GefPL7tPe+p2vg2yItuV5K8jRwJCxdWLRypZvpmVURJkiQNXMuWwS9+ATvsAO98J/zqV3DEETB7Nvz85/De91Z1cpJRo2Du3N7XMXw4nHJKVcKRaq7xZ/2QJElS/3juOTj/fDj1VHjoIRg/Hn70I/j0p2H99at+uWp0mxw5Midvdo3UQGECJ0mSpNWbMwdOOw3OPRf+8x9497vhBz+AvffOM3/0kUq6TX7+8y7CrYHJLpSSJElaVUpw6615EpLXvS43Y+2xB9x2G/zhD7Dvvt1O3jqbPbKrn950mxw5Ei65xORNA5ctcJIkSVrhlVfyOLYZM+DPf4YNNoAvfjGPcRszpkdVtbbC5z6XVxboK+PG5QZCabAwgZMkSVKe+v+cc/JSAPPnw+tfn79Pnpzn2O+hvp72H/LkJK7bpsHGLpSSJEmDVGsrvHPD2ZwWR7J4ozFw/PHcNP/1fJirGDL7AeKIw4lXrdvjro/ti2v3ZfI2cmSeT8XJSTTY2AInSZI0ALS2wtSpsGhRd45OvJ/fcQwzuI1fsZRhXMr+zOAY/spb+zrUitltUoOZCZwkSVIdam3NszDOnZtbtFKqvM41eJn9uIxjmMHbuJuFjGQ60ziTw3icTSq/QD+w26QGOxM4SZKkOtPaCs3NsGRJ3q40eRvJQg5lJodzBpvwOH9nKz5LC5dwAC+yduUB9xPXdJNM4CRJkvpFX7SodeUN/IOjOZmDuIi1eZHf8EEO5kKu5wNA9H0AZYwcCQsX1uTS0oBgAidJktSJ0qRr6FBYtqzrz+4kZ32bvCV25QaOYQa78xteYC0u5kBOYSr38aa+vHCXhg/PLWiSes8ETpIkqYyO3RiXLeveZ3+0rJWzFi/QRCtHczJv5u88zmi+xreZyaEsZOPaBFXC7o9SdbiMgCRJqmutrTB+PAwZkj9bW1ctO+ywzrdHjco/nX2PgGHD8mdp2QEHrEje6tloHudbfJ15jOVcPstShjGZCxnHXP6Xr3UreRs5Ei65JCefffWzcKHJm1QNJnCSJPWjiNgtIh6IiNkR8eW+vFa5xKe/6uvq2M72l0vMmptzF8aU8ueUKfDpT69cdtZZnW8vWpR/OvsOK1rPSsvq3Tb8lQs4mLmM4wT+lz+yPTtyE7ts9Bc+cMlkXkprmlxJA5BdKCVJ6icRMRQ4A9gVWADcHhFXpZTuq/a1Onb/mzs3b0Pv/lDvSX1dHdvZ/j/8AWbNWrl85sxVuyT25eLQ9Wr48GLR6v2Xw7XXwowZcNNNMGIETPksTJ3KXhMmsFetA5XU52yBkySp/2wLzE4pPZRSehm4DNi7Ly40bdqq3f+WLMnlfV1fV8d2tr+lZdXyWo0nqycjR8KsM5+n6dkz4Y1vhA9/GB54AL7/fViwAE4/HSZMqHWYkvqJCZwkSf1nU2B+yfaComwlEdEcEXdExB1PPfVUry40b17PyqtZX1fHdra/vRtjvRs6dPWf0cns/EOKv7rGjevBeLP5C1j4mS+z35fGwOGHw/rrw49/DA8/DMcdBxtu2Hc3KqkumcBJktR/yv1pv0obU0qpJaU0MaU0ceONezd74NixPSuvZn1dHdvZ/vYEqKOOCdEaa+Quhf1pxIgVSdfSpav/XL68fDK2bFn+nDOnG91Yb78dPvUp2HxzOPFE2Gkn+P3v4bbbYP/98y9B0qBkAidJUv9ZAIwp2d4MeLQvLjR9ek46So0Ykcv7ur6uju1sf3Nz+fJDD82tVhH584IL8niw0rLPf77z7ZEj809n32FF8liubNy43L2zzyf5WLYMfv5zeO97Ydtt4eqr4cgjYfZs+NnP4D3v6bx5T9Kg4SQmkiT1n9uBCRGxOfAIsB/wqb64UHuyMW1a7rI4dmxOnHqbhPSkvq6OXd3+97yn+zEPmFkT//OfnJGeckpunhs/Pk9S8ulPw3rr1To6SXXGBE6SpH6SUloaEUcA1wFDgfNTSn/vq+s1NVU3yelJfV0d29n+asdc1+bMgVNPhXPPheeey9nrD38I++zTeX9SSYOeCZwkSf0opXQtcG2t41CNpAS33ppb2H7xizyzycc/DsccA+98Z62jk9QATOAkSZL62iuv5HFsM2bkCUo23BC+9CU44gjYbLNaRyepgZjASZIk9ZVnnskzoJx+el6zbcIEOOMMmDwZ1lmn1tFJakAmcJIkSdX2r3/lSUkuuCCvTr7TTnDWWbDHHisWhJOkXjCBkyRJqoaUoK0td5O8+uq8Vtv+++fxbW95S62jkzRAmMBJkiRV4qWX4LLL4OST4e67YdQoOOEEOOww+J//qXV0kgYYEzhJkqTeeOopmDkTzjwTHn8cttoKzjknr4Ow9tq1jk7SAGUCJ0mS1BN//3tubbvkEnjxRdhtt9xNctddIaLW0Uka4EzgJEmSupISXH99Ht923XWw1lpw0EEwdWpueZOkfmICJ0mS1JkXXsgtbSefDPfdl8e0fec7cOiheaybJPUzEzhJkqSOHn88r9c2cyYsXAhvfSvMmgWf/CSsuWato5M0iJnASZIktbv77txN8tJLYelS+PCH8/i297/f8W2S6oIJnCRJGtyWL8/rts2YkddxW2cd+Nzn4KijYMKEWkcnSSsZ0tsTI2JMRNwcEf+IiL9HxNSifKOIuCEi/lV8bli9cCVJkqpk8eLcTXLLLWHvveHBB+EHP4D58+G000zeJNWlXidwwFLg2JTSG4HtgMMjYivgy8CNKaUJwI3FtiRJUn2YPx+OPx7GjIEjjoCNNsoLcT/4IHzpS7Ch754l1a9ed6FMKT0GPFZ8fy4i/gFsCuwNTCoOmwW0AcdXFKUkSVKl/vzn3E3ypz/NywJ89KN5fNv22zu+TVLDqMoYuIgYD7wNuA0YXSR3pJQei4hXV+MakiRJPbZ0Kfzylzlxu/VWWG+9vHbbkUfC+PG1jk6SeixSSpVVELEu8Dtgekrpioj4d0ppg5L9z6SUVumLEBHNQDPA6NGj33HZZZdVFMeTTz/LEy9UVEVdGL02Vb2PrTddv3qVddPixYtZd911+/261eQ91AfvoT5U+x523HHHO1NKE6tW4QA3ceLEdMcdd9Q6jMbz7LNw3nlw6qkwdy5svnlO3KZMyUmcJNWZiOjW87GiFriIWAP4OdCaUrqiKH4iIjYpWt82AZ4sd25KqQVogfxwmjRpUiWhcFrrlZx0T+NPqnns1kureh9zmiZVra7uamtro9L/nrXmPdQH76E+DIR70CDy0EM5aTv/fHjuOdhhh9z6ttdeMHRoraOTpIpVMgtlAOcB/0gp/ahk11XA5OL7ZODK3ocnSZLUhZTg97+HfffNM0eecUZev+322+GWW+AjHzF5kzRgVNLU8x7gQOCeiLi7KPsq8D3g8og4BJgHfLyyECVJksp45ZU8IcmMGXDHHXn2yOOOyzNLbrppraOTpD5RySyUvwc6m7Jp597WK0mStFpPPw0tLXD66fDII7DFFnDmmXDQQXkRbkkawBp/0JgkSRoc/vlPOPlkmDULliyBnXeGs8+G3XeHIZUsbStJjcMETpIk1a+U4Oab4Uc/gmuugeHDoakJjj4attmm1tFJUr8zgZMkSfXnpZfg0kvz+La//Q023hi+/nU47DAYPbrW0UlSzZjASZKk+vHUU3DWWXlM2xNPwJveBOeem1vd1lqr1tFJUs2ZwEmSpNq79948vu2SS3Lr2+67wzHHwC67QHQ2Z5okDT4mcJIkqTaWL4frrsvdJG+4AdZeGw4+GKZOhTe+sdbRSVJdMoGTJEn9a8kSuPhiOOUU+Mc/YJNNYPp0+NznYOTIWkcnSXXNBE6SJPWPxx6DM86AmTNh0SJ429vgoovgk5/Ms0tKkrpkAidJkvrWX/6Su0ledhksXQp77ZXHt73vfY5vk6QeMoGTJEnVt2wZXH11Ttx+9ztYZx049FA46ih4/etrHZ0kNSwTOEmSVD2LF8MFF+TxbQ8+CGPGwIknwmc+AxtsUOvoJKnhmcBJkqTKzZ8Pp50GLS3w7LPwrnfliUn23ReG+eeGJFWL/6JKkqTeu+223E3yZz+DlHLCdswxsP32tY5MkgYkEzj9//buP9iK+rzj+Psjij9QmxgJQ4iCZvwRYgwK1SSOVgNRJIm3piFgqAomY+Mgao0zwaatVmc6xNQ0NcRfKJF0CBqaYkkkihq1caKRHyEiIAlBbBAqSawaq2LBp3/s98b13nPuPRfOuXv2nM9r5s7d8z2H3efZ3bN3H77f3TUzM+ubHTtg0aKscHvsMTjwQLjsMpgxA4YPLzo6M7OW5gLOzMzMavPSS3DbbdlQyWefhcMPz651mzYNDjig6OjMzNqCCzgzMzPr2caNWaE2d252k5KTT4ZvfAM+9SkYMKDo6MzM2ooLODMzM+suAh59NBsmeffdWaE2aVJ2fdvo0UVHZ2bWtlzAmZmZ2VveeAMWLswKtxUr4KCDYOZMmD4dhg0rOjozs7bnAs7MzMzghRfglltg9mzYsgWOOgpuugnOOw/226/o6MzMLHEBZ2Zm1s7Wr8+uZ5s3D157DcaNgzlzYPx42GOPoqMzM7MuXMCZmZm1mwh48MFsmOSSJTBwIEyZkj0K4Nhji47OzMx64ALO+sWImfcUHcIfbZr1iaJDMDMrxuuvw4IFWeG2ejUMHgxXXQUXXQRDhhQdnZmZ1cAFnJmZWavbti27nu3GG7PpY46B22+Hz30O9tmn6OjMzKwPXMCZmZm1sssvzwq37dthwoTsMQBjx4JUdGRmZrYLXMCZmZm1skGDYOrU7Pq2o48uOhozM9tNLuDMzMxa2bXXFh2BmZnVke8PbGZmZmZmVhIu4MzMzMzMzErCQyhbXBG37//SB3cwtYkeG2BmZmZm1ircA2dmZmZmZlYSLuDMzMzMzMxKwgWcmZmZmZlZSbiAMzMzMzMzKwkXcGZmZmZmZiXhu1CaFaTRdwjty91AN836RENj6Yv8einyjqbNuk6K0LkdmmmdmJmZtauG9cBJGi9pvaQNkmY2ajlmZmZmZmbtoiEFnKQBwLeAM4GRwDmSRjZiWWZmZmZmZu2iUUMoTwA2RMRGAEl3Ah3A2gYtz8x2Q9FD9MzMzMysNo0aQjkM+E3u9ebUZmZmZmZmZrtIEVH/mUoTgTMi4gvp9bnACRExI/eZC4EL08ujgPW7udiDgd/t5jyaQSvk4Ryag3NoDs6hu+ERMbiO82tpkn4LPFt0HAVphe/Prmrn3MH5t3P+7Zx7TX8fGzWEcjNwSO71e4Et+Q9ExK3ArfVaoKTlETGmXvMrSivk4Ryag3NoDs7Bdlc7F7vtvO+1c+7g/Ns5/3bOvVaNGkK5DDhC0mGSBgKTgcUNWpaZmZmZmVlbaEgPXETskHQxcB8wAJgbEWsasSwzMzMzM7N20bAHeUfEEmBJo+ZfQd2GYxasFfJwDs3BOTQH52C269p532vn3MH5t3P+7Zx7TRpyExMzMzMzMzOrv0ZdA2dmZmZmZmZ11hIFnKTxktZL2iBpZtHx1ELSIZIekrRO0hpJl6b2qyU9J2lV+plQdKw9kbRJ0uoU6/LUdpCk+yX9Kv1+Z9FxViPpqNy6XiXpZUmXlWE7SJoraZukp3JtFde9Mjek78iTko4vLpXGkrAAAAlOSURBVPK3VMnha5KeTnEukvSO1D5C0mu5bXJzcZG/pUoOVfcfSVem7bBe0hnFRP12VXK4Kxf/JkmrUntTbgdrTdWOB+1C0sR0jvCmpLa4K18Zz+nqqdLxuF1UOze27ko/hFLSAOCXwMfJHl+wDDgnItYWGlgvJA0FhkbESkkHACuAPwc+C7wSEf9UaIA1krQJGBMRv8u1XQe8EBGz0sH3nRHx5aJirFXal54DTgSm0eTbQdIpwCvAdyLimNRWcd2nAmIGMIEsv3+JiBOLir1TlRxOB36cbob0VYCUwwjgh52faxZVcriaCvuPpJHAAuAE4D3AA8CREbGzX4PuolIOXd6/HngpIq5p1u1grana8aDgsPqNpPcDbwK3AFdExPKCQ2qosp7T1VNvx+NWVu3cuJ22f61aoQfuBGBDRGyMiDeAO4GOgmPqVURsjYiVafoPwDpgWLFR1U0HMC9NzyMrTMtgLPDriCjFA3Mj4j+BF7o0V1v3HWR/DCIiHgfekQ6UhaqUQ0QsjYgd6eXjZM+RbFpVtkM1HcCdEbE9Ip4BNpAdwwrVUw6SRPYfSwv6NSgzync8qLeIWBcR64uOox+V8pyunvr4N6WltPi5cV21QgE3DPhN7vVmSrax0/9oHwf8LDVdnIaLzG3m4YdJAEslrZB0YWobEhFbIfsyAu8uLLq+mczbT1LLtB06VVv3Zf2eXAD8KPf6MEk/l/SIpJOLCqpGlfafMm6Hk4HnI+JXubYybQdrHV2PB9Z6yniMtAaocG5sOa1QwKlCW2nGhUraH/g+cFlEvAzcBLwPGAVsBa4vMLxanBQRxwNnAtNT13/pKHvg/FnAwtRUtu3Qm9J9TyR9BdgBzE9NW4FDI+I44HLgu5IOLCq+XlTbf0q3HYBzePt/bJRpO1gJSHpA0lMVfjpyn+l6PGgZteTfRsp4jLQ6q3BubF007Dlw/WgzcEju9XuBLQXF0ieS9iLbQedHxL8DRMTzuffnAD8sKLyaRMSW9HubpEVkwx+elzQ0IramYXrbCg2yNmcCKzvXf9m2Q061dV+q74mk84FPAmMjXagbEduB7Wl6haRfA0cCTXdNSA/7T9m2w57Ap4HRnW1l2g5WDhExrqf3Kx0PWklv+beZUh0jrf4qnRtbd63QA7cMOELSYakXZTKwuOCYepWuK7kdWBcRX8+1569LOhto2rsQSRqULjJF0iDgdLJ4FwPnp4+dD/xHMRH2ydt6Gcq0Hbqotu4XA+cp82GyG1JsLSLA3kgaD3wZOCsiXs21D04XuCPpcOAIYGMxUfash/1nMTBZ0t6SDiPL4Yn+jq8PxgFPR8TmzoYybQcrv2rHA2tZpTyns/qodm5s3ZW+By7dmepi4D5gADA3ItYUHFYtTgLOBVYr3Z4b+BvgHEmjyIYMbAL+qpjwajIEWJR939gT+G5E3CtpGfA9SZ8H/guYWGCMvZK0H9kdr/Lr+rpm3w6SFgCnAgdL2gxcBcyi8rpfQnYHyg3Aq2R32SxclRyuBPYG7k/71uMR8UXgFOAaSTuAncAXI6LwC72r5HBqpf0nItZI+h6wlmw42PSi70AJlXOIiNvpfl0oNOl2sJY1m8rHg7Yg6Wzgm8Bg4B5JqyKiKR4/0gglPqermx6Ox+2g4rlxRCwpMKamVPrHCJiZmZmZmbWLVhhCaWZmZmZm1hZcwJmZmZmZmZWECzgzMzMzM7OScAFnZmZmZmZWEi7gzMzMzMzMSsIFnJmZmVmJSXqXpFXp578lPZemX5S0tp9jGSVpQu71WZJm7uK8Nkk6uH7R9WnZUyW9J/f6Nkkji47LDFzAmZmZmZVaRPw+IkZFxCjgZuCf0/Qo4M16L09ST88RHkX23NHO2BZHxKx6x9APpgJ/LOAi4gsR0a/FsFk1LuDMzMzMWtcASXMkrZG0VNK+AJLeJ+leSSsk/UTS0al9uKQHJT2Zfh+a2u+Q9HVJDwFflTRI0lxJyyT9XFKHpIHANcCk1AM4KfVkzU7zGCJpkaRfpJ+Ppva7UxxrJF3YW0KSpkn6paRHUm6d879D0mdyn3sl/d4/5bJS0mpJHal9hKR1XddPmscYYH7KY19JD0saUyGWv5T0RPrcLZIGpJ87JD2VlvfXu7H9zLpxAWdmZmbWuo4AvhURHwBeBP4itd8KzIiI0cAVwI2pfTbwnYg4FpgP3JCb15HAuIj4EvAV4McR8afAacDXgL2AvwfuSj2Cd3WJ5QbgkYj4EHA8sCa1X5DiGANcIuld1ZKRNBT4B+Ak4OPAyBrWwevA2RFxfIr1ekmqtn4i4t+A5cCUlMdrVWJ5PzAJOCn1eO4EppD1Qg6LiGMi4oPAt2uI0axmPXWBm5mZmVm5PRMRq9L0CmCEpP2BjwIL36pj2Dv9/gjw6TT9r8B1uXktjIidafp04CxJV6TX+wCH9hLLx4DzANJ8Xkrtl0g6O00fQlZU/b7KPE4EHo6I3wJIuoussOyJgH+UdArZkNJhwJD0Xrf108u88sYCo4FlaT3uC2wDfgAcLumbwD3A0j7M06xXLuDMzMzMWtf23PROsiJjD+DF1GvUm8hN/29uWmS9VevzH5Z0Yl+Ck3QqMA74SES8KulhsmKw1pjydpBGl6UetoGpfQowGBgdEf8naVNuGZXWT83hA/Mi4spub0gfAs4ApgOfBS7ow3zNeuQhlGZmZmZtJCJeBp6RNBGyYicVHAA/BSan6SnAo1Vmcx8wo3MooqTjUvsfgAOq/JsHgYvS5wdIOhD4E+B/UvF2NPDhXsL/GXBquvPmXsDE3HubyHrEADrIhnSSlrEtFW+nAcN7WUZveeTz+Yykd6ecDkrXEB4M7BER3wf+jmy4qFnduIAzMzMzaz9TgM9L+gXZtWgdqf0SYJqkJ4FzgUur/PtryQqkJyU9lV4DPASM7LyJSZd/cylwmqTVZMMVPwDcC+yZlnct8HhPQUfEVuBq4DHgAWBl7u05wJ9JeoJsqGVnj+F8YIyk5Snvp3taRnIHcHPnTUyqxLIW+FtgaYr/fmAo2RDNhyWtSvPp1kNntjsUUa0X2szMzMyseUmaCoyJiIuLjsWsv7gHzszMzMzMrCTcA2dmZmZmZlYS7oEzMzMzMzMrCRdwZmZmZmZmJeECzszMzMzMrCRcwJmZmZmZmZWECzgzMzMzM7OScAFnZmZmZmZWEv8PJKmziHOtxNsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1080x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plotting the Q-Q plot\n",
    "\n",
    "import pylab \n",
    "import scipy.stats as stats\n",
    "\n",
    "\n",
    "for var in ['MYCT', 'MMIN', 'MMAX', 'CACH', 'CHMIN', 'CHMAX']:\n",
    "    \n",
    "    plt.figure(figsize=(15,6))\n",
    "\n",
    "    plt.subplot(1, 2, 1)\n",
    "    df[var].hist()\n",
    "    plt.title('Distribution of '+ var)\n",
    "\n",
    "    plt.subplot(1, 2, 2)\n",
    "    stats.probplot(df[var], dist=\"norm\", plot=pylab)\n",
    "    plt.title('Q-Q plot of '+ var)\n",
    "\n",
    "    plt.show() \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Interpretation**\n",
    "\n",
    "\n",
    "From the distribution plots, we can see that all the above variables are positively skewed. The Q-Q plot of all the variables confirm that the variables are not normally distributed.\n",
    "\n",
    "Hence, the variables do not pass the normality test."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 15. Conclusion\n",
    "\n",
    "\n",
    "\n",
    "I carry out residual analysis to check for homoscedasticity assumption. Residual errors plot show that the data is randomly scattered around line zero. The plot does not display any pattern in the residuals.  Hence, we can conclude that the Linear Regression model is a good fit to the data.\n",
    "\n",
    "\n",
    "The r-squared or the coefficient of determination is 0.4691 on an average for 5-fold cross validation. It means that the predictor is only able to explain 46.91% of the variance in the target variable. This indicates that the model is not a good fit to the data.\n",
    "\n",
    "\n",
    "I carry out normality test to check for distribution of the variables. We can see that the variables do not follow the normal distribution. The Q-Q plots confirm the same.\n",
    "\n",
    "\n",
    "So, we can conclude that the linear regression model is unable to model the data to generate decent results. It should be noted that the model is performing equally on both training and testing datasets. It seems like a case where we would need to model this data using methods that can model non-linear relationships. Also variables need to be transformed to satisfy the normality assumption.\n",
    "\n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}