feature-encoding.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 0. Introduction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One of the preprocessing steps in machine learning is feature encoding. It is the process of turning categorical data in a dataset into numerical data. It is important that we perform feature encoding because most machine learning algorithms only handle numerical data and not data in text form.\n",
    "\n",
    "In this notebook, we will learn the difference between nominal variables and ordinal variables. In addition, we will explore how [OneHotEncoder](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.OneHotEncoder.html) and [OrdinalEncoder](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.OrdinalEncoder.html#sklearn.preprocessing.OrdinalEncoder) can be used to transform these variables as part of a machine learning pipeline.\n",
    "\n",
    "We will use this pipeline to predict the mean test score of different students. This is a regression problem in machine learning."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#  1. Import libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data wrangling\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "# Data visualisation\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Machine learning\n",
    "from sklearn.preprocessing import OneHotEncoder, OrdinalEncoder\n",
    "from sklearn.compose import make_column_transformer\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.ensemble import GradientBoostingRegressor\n",
    "from sklearn.metrics import mean_absolute_error, mean_squared_error\n",
    "\n",
    "# Remove warnings\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Import and read data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": false
   },
   "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>gender</th>\n",
       "      <th>race/ethnicity</th>\n",
       "      <th>parental level of education</th>\n",
       "      <th>lunch</th>\n",
       "      <th>test preparation course</th>\n",
       "      <th>math score</th>\n",
       "      <th>reading score</th>\n",
       "      <th>writing score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>female</td>\n",
       "      <td>group B</td>\n",
       "      <td>bachelor's degree</td>\n",
       "      <td>standard</td>\n",
       "      <td>none</td>\n",
       "      <td>72</td>\n",
       "      <td>72</td>\n",
       "      <td>74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>female</td>\n",
       "      <td>group C</td>\n",
       "      <td>some college</td>\n",
       "      <td>standard</td>\n",
       "      <td>completed</td>\n",
       "      <td>69</td>\n",
       "      <td>90</td>\n",
       "      <td>88</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>female</td>\n",
       "      <td>group B</td>\n",
       "      <td>master's degree</td>\n",
       "      <td>standard</td>\n",
       "      <td>none</td>\n",
       "      <td>90</td>\n",
       "      <td>95</td>\n",
       "      <td>93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>male</td>\n",
       "      <td>group A</td>\n",
       "      <td>associate's degree</td>\n",
       "      <td>free/reduced</td>\n",
       "      <td>none</td>\n",
       "      <td>47</td>\n",
       "      <td>57</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>male</td>\n",
       "      <td>group C</td>\n",
       "      <td>some college</td>\n",
       "      <td>standard</td>\n",
       "      <td>none</td>\n",
       "      <td>76</td>\n",
       "      <td>78</td>\n",
       "      <td>75</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   gender race/ethnicity parental level of education         lunch  \\\n",
       "0  female        group B           bachelor's degree      standard   \n",
       "1  female        group C                some college      standard   \n",
       "2  female        group B             master's degree      standard   \n",
       "3    male        group A          associate's degree  free/reduced   \n",
       "4    male        group C                some college      standard   \n",
       "\n",
       "  test preparation course  math score  reading score  writing score  \n",
       "0                    none          72             72             74  \n",
       "1               completed          69             90             88  \n",
       "2                    none          90             95             93  \n",
       "3                    none          47             57             44  \n",
       "4                    none          76             78             75  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv(\"C:/Users/Jason Chong/Documents/Kaggle/Exam/data.csv\")\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Data shape:  (1000, 8)\n"
     ]
    }
   ],
   "source": [
    "print(\"Data shape: \", data.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Data description"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The [students performance in exams dataset](https://www.kaggle.com/spscientist/students-performance-in-exams) on Kaggle consists of marks secured by 1,000 students in the math, reading and writing subjects. Along with these test scores, the description of each student such as their gender, race/ethnicity, parental level of education, lunch and test preparation course are also included in the dataset. \n",
    "\n",
    "The goal of this project is to build a machine learning model that can predict students' test score given their description. To reiterate, this is a regression problem in machine learning. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Missing values and data types"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1000 entries, 0 to 999\n",
      "Data columns (total 8 columns):\n",
      " #   Column                       Non-Null Count  Dtype \n",
      "---  ------                       --------------  ----- \n",
      " 0   gender                       1000 non-null   object\n",
      " 1   race/ethnicity               1000 non-null   object\n",
      " 2   parental level of education  1000 non-null   object\n",
      " 3   lunch                        1000 non-null   object\n",
      " 4   test preparation course      1000 non-null   object\n",
      " 5   math score                   1000 non-null   int64 \n",
      " 6   reading score                1000 non-null   int64 \n",
      " 7   writing score                1000 non-null   int64 \n",
      "dtypes: int64(3), object(5)\n",
      "memory usage: 62.6+ KB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hooray, no missing values!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "object    5\n",
       "int64     3\n",
       "dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.dtypes.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have 5 categorical variables and 3 numerical variables.\n",
    "\n",
    "As it turns out, all the predictor variables are categorical variables and all the target variables are numerical variables. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 5. Create target variable (mean score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For simplicity purposes, let's take the average of the 3 test scores so that we are left with only one target variable to predict. "
   ]
  },
  {
   "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>math score</th>\n",
       "      <th>reading score</th>\n",
       "      <th>writing score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>72</td>\n",
       "      <td>72</td>\n",
       "      <td>74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>69</td>\n",
       "      <td>90</td>\n",
       "      <td>88</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>90</td>\n",
       "      <td>95</td>\n",
       "      <td>93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>47</td>\n",
       "      <td>57</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>76</td>\n",
       "      <td>78</td>\n",
       "      <td>75</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   math score  reading score  writing score\n",
       "0          72             72             74\n",
       "1          69             90             88\n",
       "2          90             95             93\n",
       "3          47             57             44\n",
       "4          76             78             75"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score = data[['math score', 'reading score', 'writing score']]\n",
    "score.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x12eb8488b08>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Correlation between test scores\n",
    "\n",
    "correlation = score.corr()\n",
    "sns.heatmap(correlation, annot = True, fmt = '.2f', cmap = 'coolwarm')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since the 3 test scores are highly positively correlated with each other, it is appropriate for us to take the average. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>gender</th>\n",
       "      <th>race/ethnicity</th>\n",
       "      <th>parental level of education</th>\n",
       "      <th>lunch</th>\n",
       "      <th>test preparation course</th>\n",
       "      <th>mean score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>female</td>\n",
       "      <td>group B</td>\n",
       "      <td>bachelor's degree</td>\n",
       "      <td>standard</td>\n",
       "      <td>none</td>\n",
       "      <td>72.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>female</td>\n",
       "      <td>group C</td>\n",
       "      <td>some college</td>\n",
       "      <td>standard</td>\n",
       "      <td>completed</td>\n",
       "      <td>82.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>female</td>\n",
       "      <td>group B</td>\n",
       "      <td>master's degree</td>\n",
       "      <td>standard</td>\n",
       "      <td>none</td>\n",
       "      <td>92.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>male</td>\n",
       "      <td>group A</td>\n",
       "      <td>associate's degree</td>\n",
       "      <td>free/reduced</td>\n",
       "      <td>none</td>\n",
       "      <td>49.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>male</td>\n",
       "      <td>group C</td>\n",
       "      <td>some college</td>\n",
       "      <td>standard</td>\n",
       "      <td>none</td>\n",
       "      <td>76.333333</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   gender race/ethnicity parental level of education         lunch  \\\n",
       "0  female        group B           bachelor's degree      standard   \n",
       "1  female        group C                some college      standard   \n",
       "2  female        group B             master's degree      standard   \n",
       "3    male        group A          associate's degree  free/reduced   \n",
       "4    male        group C                some college      standard   \n",
       "\n",
       "  test preparation course  mean score  \n",
       "0                    none   72.666667  \n",
       "1               completed   82.333333  \n",
       "2                    none   92.666667  \n",
       "3                    none   49.333333  \n",
       "4                    none   76.333333  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calculate average of all test results\n",
    "data['mean score'] = data.mean(axis = 1)\n",
    "\n",
    "# Drop math score, reading score and writing score\n",
    "data = data.drop(['math score', 'reading score', 'writing score'], axis = 1)\n",
    "\n",
    "# Show first 5 rows of new dataframe\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean score data type:  float64\n"
     ]
    }
   ],
   "source": [
    "# Mean score data type\n",
    "\n",
    "print(\"Mean score data type: \", data['mean score'].dtype)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 6. Exploratory data analysis (EDA)\n",
    "\n",
    "Exploratory data analysis is the process of analysing and visualising the variables in a dataset. This step is not necessary for feature encoding but I personally like to examine my data first before performing any machine learning. So, feel free to skip past this section if you are only interested to learn about feature encoding. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 6.1 Predictor variables\n",
    "\n",
    "The predictor variables in the dataset are:\n",
    "\n",
    "- Gender\n",
    "- Race/ethnicity\n",
    "- Parental level of education\n",
    "- Lunch\n",
    "- Test preparation course\n",
    "\n",
    "In this section, we will explore how these different features influence the outcome of a student's test score. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6.1.1 Gender"
   ]
  },
  {
   "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>gender</th>\n",
       "      <th>mean score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>female</td>\n",
       "      <td>69.569498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>male</td>\n",
       "      <td>65.837483</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   gender  mean score\n",
       "0  female   69.569498\n",
       "1    male   65.837483"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mean score by gender \n",
    "\n",
    "data[['gender', 'mean score']].groupby('gender', as_index = False).mean().sort_values(by = 'mean score', ascending = False, ignore_index = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Mean Score by Gender')"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Mean score by gender barplot\n",
    "\n",
    "plt.figure(figsize = (5, 5))\n",
    "sns.barplot(x = 'gender', y = 'mean score', data = data)\n",
    "plt.title('Mean Score by Gender')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Female students perform better than male students on average. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6.1.2 Race/ethnicity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>race/ethnicity</th>\n",
       "      <th>mean score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group E</td>\n",
       "      <td>72.752381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>group D</td>\n",
       "      <td>69.179389</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>group C</td>\n",
       "      <td>67.131661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>group B</td>\n",
       "      <td>65.468421</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>group A</td>\n",
       "      <td>62.992509</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  race/ethnicity  mean score\n",
       "0        group E   72.752381\n",
       "1        group D   69.179389\n",
       "2        group C   67.131661\n",
       "3        group B   65.468421\n",
       "4        group A   62.992509"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mean score by race/ethnicity\n",
    "\n",
    "data[['race/ethnicity', 'mean score']].groupby('race/ethnicity', as_index = False).mean().sort_values(by = 'mean score', ascending = False, ignore_index = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Mean Score by Race/Ethnicity')"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Mean score by race/ethnicity barplot\n",
    "\n",
    "plt.figure(figsize = (6, 6))\n",
    "sns.barplot(x = 'race/ethnicity', y = 'mean score', data = data)\n",
    "plt.title('Mean Score by Race/Ethnicity')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There seems to be an increasing trend in mean test scores as we move from group A to group E i.e. group A students perform the worst while group E students perform the best.\n",
    "\n",
    "However, in the absence of any further information, it is hard for us to draw any conclusion. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6.1.3 Parental level of education"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>parental level of education</th>\n",
       "      <th>mean score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>master's degree</td>\n",
       "      <td>73.598870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>bachelor's degree</td>\n",
       "      <td>71.923729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>associate's degree</td>\n",
       "      <td>69.569069</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>some college</td>\n",
       "      <td>68.476401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>some high school</td>\n",
       "      <td>65.108007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>high school</td>\n",
       "      <td>63.096939</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  parental level of education  mean score\n",
       "0             master's degree   73.598870\n",
       "1           bachelor's degree   71.923729\n",
       "2          associate's degree   69.569069\n",
       "3                some college   68.476401\n",
       "4            some high school   65.108007\n",
       "5                 high school   63.096939"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mean score by parental level of education \n",
    "\n",
    "data[['parental level of education', 'mean score']].groupby('parental level of education', as_index = False).mean().sort_values(by = 'mean score', ascending = False, ignore_index = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 1, 2, 3, 4, 5]), <a list of 6 Text xticklabel objects>)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Mean score by parental level of education barplot\n",
    "\n",
    "plt.figure(figsize = (6, 6))\n",
    "sns.barplot(x = 'parental level of education', y = 'mean score', data = data)\n",
    "plt.title('Mean Score by Parental Level of Education')\n",
    "plt.xticks(rotation = 90)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For the most part, the results are aligned with what we would expect i.e. as parental level of education increases, mean test score also increases. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6.1.4 Lunch"
   ]
  },
  {
   "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>lunch</th>\n",
       "      <th>mean score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>standard</td>\n",
       "      <td>70.837209</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>free/reduced</td>\n",
       "      <td>62.199061</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          lunch  mean score\n",
       "0      standard   70.837209\n",
       "1  free/reduced   62.199061"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mean score by lunch\n",
    "\n",
    "data[['lunch', 'mean score']].groupby('lunch', as_index = False).mean().sort_values(by = 'mean score', ascending = False, ignore_index = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Mean Score by Lunch')"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Mean score by lunch barplot\n",
    "\n",
    "plt.figure(figsize = (5, 5))\n",
    "sns.barplot(x = 'lunch', y = 'mean score', data = data)\n",
    "plt.title('Mean Score by Lunch')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Students with standard lunch perform better than those with free/reduced lunch. \n",
    "\n",
    "Lunch can be seen as a proxy for the financial background of the students. Therefore, it makes sense that students that financially better off do better in school compared to those that are of financially worse off, on average. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6.1.5 Test preparation course"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>test preparation course</th>\n",
       "      <th>mean score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>completed</td>\n",
       "      <td>72.669460</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>none</td>\n",
       "      <td>65.038941</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  test preparation course  mean score\n",
       "0               completed   72.669460\n",
       "1                    none   65.038941"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mean score by test preparation course \n",
    "\n",
    "data[['test preparation course', 'mean score']].groupby('test preparation course', as_index = False).mean().sort_values(by = 'mean score', ascending = False, ignore_index = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Mean Score by Test Preparation Course')"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Mean score by test preparation course barplot\n",
    "\n",
    "plt.figure(figsize = (5, 5))\n",
    "sns.barplot(x = 'test preparation course', y = 'mean score', data = data)\n",
    "plt.title('Mean Score by Test Preparation Course')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This should come at no surprise at all. The more prepared students are for their exam, the better they will perform. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 6.2 Target variable (mean score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    1000.000000\n",
       "mean       67.770667\n",
       "std        14.257326\n",
       "min         9.000000\n",
       "25%        58.333333\n",
       "50%        68.333333\n",
       "75%        77.666667\n",
       "max       100.000000\n",
       "Name: mean score, dtype: float64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mean score summary statistics\n",
    "\n",
    "data['mean score'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Distribution of Mean Score')"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Distribution of mean score \n",
    "\n",
    "plt.figure(figsize = (8, 5))\n",
    "sns.distplot(data['mean score'], label = 'Skewness: {:.2f}'.format(data['mean score'].skew()))\n",
    "plt.legend(loc = 'best')\n",
    "plt.title('Distribution of Mean Score')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 7. Feature encoding\n",
    "\n",
    "Now that we have fully explored the variables in the dataset, we can move on to getting the dataset reading for modelling. More specifically, we want to turn the categorical data in our dataset into numerical data. This process is otherwise known as feature encoding.\n",
    "\n",
    "But before we dive into feature encoding, it is important that we first contrast the difference between a nominal variable and an ordinal variable. This is so that we can match the right encoder to the right variable. \n",
    "\n",
    "Nominal variable is a categorical variable where its data does not have a logical ordering. Some examples of nominal data are:\n",
    "- Male and female\n",
    "- Location A, location B and location C\n",
    "\n",
    "Ordinal variable, on the other hand, is also a categorical variable but its data follows a logical ordering. Some example of ordinal data include:\n",
    "- Small, medium and large\n",
    "- Bad, neutral and good\n",
    "- Children, adults and seniors\n",
    "- Low, medium and high income\n",
    "\n",
    "As we will see in this section, nominal variables are encoded using OneHotEncoder while ordinal variables are encoded using OrdinalEncoder."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>gender</th>\n",
       "      <th>race/ethnicity</th>\n",
       "      <th>parental level of education</th>\n",
       "      <th>lunch</th>\n",
       "      <th>test preparation course</th>\n",
       "      <th>mean score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>female</td>\n",
       "      <td>group B</td>\n",
       "      <td>bachelor's degree</td>\n",
       "      <td>standard</td>\n",
       "      <td>none</td>\n",
       "      <td>72.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>female</td>\n",
       "      <td>group C</td>\n",
       "      <td>some college</td>\n",
       "      <td>standard</td>\n",
       "      <td>completed</td>\n",
       "      <td>82.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>female</td>\n",
       "      <td>group B</td>\n",
       "      <td>master's degree</td>\n",
       "      <td>standard</td>\n",
       "      <td>none</td>\n",
       "      <td>92.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>male</td>\n",
       "      <td>group A</td>\n",
       "      <td>associate's degree</td>\n",
       "      <td>free/reduced</td>\n",
       "      <td>none</td>\n",
       "      <td>49.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>male</td>\n",
       "      <td>group C</td>\n",
       "      <td>some college</td>\n",
       "      <td>standard</td>\n",
       "      <td>none</td>\n",
       "      <td>76.333333</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   gender race/ethnicity parental level of education         lunch  \\\n",
       "0  female        group B           bachelor's degree      standard   \n",
       "1  female        group C                some college      standard   \n",
       "2  female        group B             master's degree      standard   \n",
       "3    male        group A          associate's degree  free/reduced   \n",
       "4    male        group C                some college      standard   \n",
       "\n",
       "  test preparation course  mean score  \n",
       "0                    none   72.666667  \n",
       "1               completed   82.333333  \n",
       "2                    none   92.666667  \n",
       "3                    none   49.333333  \n",
       "4                    none   76.333333  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 7.1 Encoding nominal variables"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The nominal variables in our dataset include:\n",
    "\n",
    "- Gender\n",
    "- Race/ethnicity\n",
    "- Lunch\n",
    "- Test prepation course\n",
    "\n",
    "In this section, we will explore two different ways to encode nominal variables, one using the Scikit-learn OneHotEncoder while the other using the [Pandas get_dummies](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.get_dummies.html). "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7.1.1 Scikit-learn OneHotEncoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Instantiate OneHotEncoder\n",
    "\n",
    "ohe = OneHotEncoder(sparse = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1., 0.],\n",
       "       [1., 0.],\n",
       "       [1., 0.],\n",
       "       [0., 1.],\n",
       "       [0., 1.]])"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Apply OneHotEncoder to the gender column \n",
    "\n",
    "ohe.fit_transform(data[['gender']])[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    female\n",
       "1    female\n",
       "2    female\n",
       "3      male\n",
       "4      male\n",
       "Name: gender, dtype: object"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The first 5 rows the gender column for comparison\n",
    "\n",
    "data['gender'].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array(['female', 'male'], dtype=object)]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Gender categories in OneHotEncoder \n",
    "\n",
    "ohe.categories_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see, OneHotEncoder has created two columns to represent the gender feature in our dataframe, one for female and one for male.\n",
    "\n",
    "Female students will receive a value of 1 in the female column and 0 in the male column whereas male students will receive a value of 0 in the female column and 1 in the male column. \n",
    "\n",
    "But most importantly, OneHotEncoder has successfully transformed what was originally a categorical variable into a numerical variable. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7.1.3 Pandas get_dummies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>female</th>\n",
       "      <th>male</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   female  male\n",
       "0       1     0\n",
       "1       1     0\n",
       "2       1     0\n",
       "3       0     1\n",
       "4       0     1"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Pandas get_dummies approach\n",
    "\n",
    "pd.get_dummies(data['gender']).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Although both approaches give the same result, OneHotEncoder is generally preferred over get_dummies due to the following reasons:\n",
    "\n",
    "- Under OneHotEncoder, our original dataframe remains the same size and therefore easier to manage and explore.\n",
    "- OneHotEncoder can be incorporated as part of a pipeline in Scikit-learn whereas get_dummies require a more manual approach to encoding. This makes OneHotEncoder more efficient at transforming both the training set as well as the test set. \n",
    "- Under OneHotEncoder, we can use GridSearch to evaluate and choose preprocessing parameters.\n",
    "- Cross-validation scores are more reliable under OneHotEncoder than get_dummies.\n",
    "\n",
    "We will explore the idea of building a machine learning pipeline later on."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 7.2 Encoding ordinal variable\n",
    "\n",
    "The only ordinal variable in our dataset is the parental level of education feature. \n",
    "\n",
    "In this section, we will again consider two approaches to encoding ordinal variable, one using Scikit-learn OrdinalEncoder while the other using Pandas map method. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7.2.1 Scikit-learn OrdinalEncoder\n",
    "\n",
    "OrdinalEncoder differs from OneHotEncoder such that it assigns incremental values to unique values in an ordinal variable rather than just 0 and 1. \n",
    "\n",
    "This helps machine learning models to recognise an ordinal variable and subsequently use the information that it has learned to make more accurate predictions. \n",
    "\n",
    "Let's now see how we can encode the parental level of education feature using OrdinalEncoder. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[\"bachelor's degree\",\n",
       " 'some college',\n",
       " \"master's degree\",\n",
       " \"associate's degree\",\n",
       " 'high school',\n",
       " 'some high school']"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Unique values in the parental level of education column\n",
    "\n",
    "list(data['parental level of education'].unique())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It is important that we specify the ordering of an ordinal variable. For our parental level of education feature, we want the order to go as follows:\n",
    "\n",
    "1. Some high school\n",
    "2. High school\n",
    "3. Some college\n",
    "4. Associate's degree\n",
    "5. Bachelor's degree\n",
    "6. Master's degree \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Specify the order for the level of education \n",
    "\n",
    "education_categories = ['some high school', 'high school', 'some college', \"associate's degree\", \"bachelor's degree\", \"master's degree\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Instantiate ordinal encoder\n",
    "\n",
    "oe = OrdinalEncoder(categories = [education_categories])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[4.],\n",
       "       [2.],\n",
       "       [5.],\n",
       "       [3.],\n",
       "       [2.]])"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Apply ordinal encoder to parental level of education column\n",
    "\n",
    "oe.fit_transform(data[['parental level of education']])[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0     bachelor's degree\n",
       "1          some college\n",
       "2       master's degree\n",
       "3    associate's degree\n",
       "4          some college\n",
       "Name: parental level of education, dtype: object"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# First 5 rows of parental level of education for comparison \n",
    "\n",
    "data['parental level of education'].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that OrdinalEncoder has assigned the value of 0 to some high school, 1 to high school, 2 to some college, 3 to associate's degree, 4 to bachelor's degree and finally, 5 to master's degree. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7.2.2 Pandas map method\n",
    "\n",
    "Pandas map method is a more manual approach to encoding ordinal variable. Although it replicates the result of the OrdinalEncoder, it is not ideal for encoding ordinal variables with a high number of unique values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    4\n",
       "1    2\n",
       "2    5\n",
       "3    3\n",
       "4    2\n",
       "Name: parental level of education, dtype: int64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Pandas map method \n",
    "\n",
    "data['parental level of education'].map({'some high school': 0, \n",
    "                                         'high school': 1,\n",
    "                                         'some college': 2,\n",
    "                                         \"associate's degree\": 3, \n",
    "                                         \"bachelor's degree\": 4,\n",
    "                                         \"master's degree\": 5}).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 7.3 Make column transformer\n",
    "\n",
    "For the remainder of this notebook, we will only consider OneHotEncoder and OrdinalEncoder as means to encode the categorical variables in our dataset.\n",
    "\n",
    "In this section, we will combine OneHotEncoder and OrdinalEncoder into a single-step column transformer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get predictor variables and target variable from data\n",
    "\n",
    "X = data.drop('mean score', axis = 1)\n",
    "Y = data['mean score']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make column transformer which consists of OneHotEncoder and OrdincalEncoder\n",
    "\n",
    "column_transform = make_column_transformer(\n",
    "    (ohe, ['gender', 'race/ethnicity', 'lunch', 'test preparation course']), \n",
    "    (oe, ['parental level of education']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1., 0., 0., 1., 0., 0., 0., 0., 1., 0., 1., 4.],\n",
       "       [1., 0., 0., 0., 1., 0., 0., 0., 1., 1., 0., 2.],\n",
       "       [1., 0., 0., 1., 0., 0., 0., 0., 1., 0., 1., 5.],\n",
       "       [0., 1., 1., 0., 0., 0., 0., 1., 0., 0., 1., 3.],\n",
       "       [0., 1., 0., 0., 1., 0., 0., 0., 1., 0., 1., 2.]])"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Apply column transformer to predictor variables \n",
    "\n",
    "column_transform.fit_transform(X)[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 8. Build machine learning pipeline\n",
    "\n",
    "A pipeline chains together multiple steps in the machine learning process where the output of each step is used as input to the next step. It is typically used to chain data preprocessing procedures together with modelling into one cohesive workflow. \n",
    "\n",
    "Here, we will build two pipelines that share the same column transformer that we have created above but with a different machine learning model, one using linear regression and the other using gradient boosting. \n",
    "\n",
    "We will then compare the accuracy of the prediction results using mean absolute error (MAE) as well as root mean squared error (RMSE). The model with a lower prediction error is deemed more accurate than the other. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train shape:  (700, 5)\n",
      "Y_train shape:  (700,)\n",
      "X_test shape:  (300, 5)\n",
      "Y_test shape:  (300,)\n"
     ]
    }
   ],
   "source": [
    "# Train test split \n",
    "\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size = 0.3, random_state = 42)\n",
    "\n",
    "print(\"X_train shape: \", X_train.shape)\n",
    "print(\"Y_train shape: \", Y_train.shape)\n",
    "print(\"X_test shape: \", X_test.shape)\n",
    "print(\"Y_test shape: \", Y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Instantiate pipeline with linear regression\n",
    "\n",
    "lm = LinearRegression()\n",
    "lm_pipeline = make_pipeline(column_transform, lm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Instantiate pipeline with gradient boosting\n",
    "\n",
    "gbm = GradientBoostingRegressor()\n",
    "gbm_pipeline = make_pipeline(column_transform, gbm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "First 5 LM predictions:  [70.75915726738543, 66.21057241064891, 71.20605482276784, 56.832675804568915, 78.06510895753397]\n",
      "First 5 GBM predictions:  [70.03246540754012, 72.82747187650256, 68.63467096651154, 58.888117031983654, 76.97875270900471]\n"
     ]
    }
   ],
   "source": [
    "# Fit pipeline to training set and make predictions on test set \n",
    "\n",
    "lm_pipeline.fit(X_train, Y_train)\n",
    "lm_predictions = lm_pipeline.predict(X_test)\n",
    "print(\"First 5 LM predictions: \", list(lm_predictions[:5]))\n",
    "\n",
    "gbm_pipeline.fit(X_train, Y_train)\n",
    "gbm_predictions = gbm_pipeline.predict(X_test)\n",
    "print(\"First 5 GBM predictions: \", list(gbm_predictions[:5]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of LM predictions:  300\n",
      "Number of GBM predictions:  300\n",
      "Size of test set:  300\n"
     ]
    }
   ],
   "source": [
    "# Compare the number of predictions with the size of test set\n",
    "\n",
    "print(\"Number of LM predictions: \", len(lm_predictions))\n",
    "print(\"Number of GBM predictions: \", len(gbm_predictions))\n",
    "print(\"Size of test set: \", len(Y_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The number of predictions matches the size of the test set. We can now proceed to evaluate the accuracy of each model. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LM MAE: 10.49\n",
      "LM RMSE: 13.14\n",
      "GBM MAE: 10.95\n",
      "GBM RMSE: 13.78\n"
     ]
    }
   ],
   "source": [
    "# Calculate mean square error and root mean squared error \n",
    "\n",
    "lm_mae = mean_absolute_error(lm_predictions, Y_test)\n",
    "lm_rmse =  np.sqrt(mean_squared_error(lm_predictions, Y_test))\n",
    "print(\"LM MAE: {:.2f}\".format(round(lm_mae, 2)))\n",
    "print(\"LM RMSE: {:.2f}\".format(round(lm_rmse, 2)))\n",
    "\n",
    "gbm_mae = mean_absolute_error(gbm_predictions, Y_test)\n",
    "gbm_rmse =  np.sqrt(mean_squared_error(gbm_predictions, Y_test))\n",
    "print(\"GBM MAE: {:.2f}\".format(round(gbm_mae, 2)))\n",
    "print(\"GBM RMSE: {:.2f}\".format(round(gbm_rmse, 2)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can conclude that linear regression is slightly more accurate than gradient boosting. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 9. Conclusion "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To summarise, in this notebook, we have learned the difference between a nominal variable and an ordinal variable as well as how to properly encode them using Scikit-learn OneHotEncoder and LabelEncoder.\n",
    "\n",
    "We also discussed the benefits of performing feature encoding using the Scikit-learn library over the Pandas library i.e. the flexibility to chain data preprocessing together with a machine learning model to form a cohesive machine learning pipeline.\n",
    "\n",
    "Finally, we compared the accuracy of two separate pipelines, one with linear regression and the other with gradient boosting, at predicting the student's mean test score. We concluded that linear regression is slighly more accurate than gradient boosting due to the lower mean absolute error (MAE) and root mean squared error (RMSE). "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 10. Follow me on my platforms "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- [Facebook](https://www.facebook.com/chongjason914)\n",
    "- [Instagram](https://www.instagram.com/chongjason914)\n",
    "- [Twitter](https://www.twitter.com/chongjason914)\n",
    "- [LinkedIn](https://www.linkedin.com/in/chongjason914)\n",
    "- [YouTube](https://www.youtube.com/channel/UCQXiCnjatxiAKgWjoUlM-Xg?view_as=subscriber)\n",
    "- [Medium](https://www.medium.com/@chongjason)"
   ]
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}