final.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# FAIKR Module 3 Project\n",
    "\n",
    "*Group members*:\n",
    "- Ciapponi Stefano (mat. 0001030211)\n",
    "- Cortecchia Tommaso (mat. 0001054600)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Aims of the Project:\n",
    "\n",
    "1. Create a **Pipeline** which is able to handle Bayesian Network creation starting from **any Dataset**.\n",
    "2. Build a **Bayesian Network Model** starting from a Dataset downloaded from Kaggle\n",
    "3. Test the Network through queries inspired by the module's content.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Library imports:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from matplotlib import pyplot as plt\n",
    "import seaborn as sns\n",
    "from scipy.stats import chi2_contingency\n",
    "import numpy as np\n",
    "from IPython.display import display"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The Dataset\n",
    "<a href=\"https://www.kaggle.com/datasets/stackoverflow/stack-overflow-2018-developer-survey\"> Dataset Link </a>: https://www.kaggle.com/datasets/stackoverflow/stack-overflow-2018-developer-survey\n",
    "\n",
    "The chosen dataset contains results from a survey Taken by the <a href=\"https://stackoverflow.com/\"> Stack Overflow Community</a> in the year 2018.\n",
    "\n",
    "The Survey originally covered numerous different topics concearning the ICT world, but we were mostly interested in **Data regarding the user profiles**.\n",
    " \n",
    "So we decided to apply some pre-processing to reduce the number of variables, but also to simplify the dataset so that the Bayesian Model built from it would become more manageable and readable.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%capture\n",
    "#Supresses stderr about Mixed-Type Data\n",
    "df = pd.read_csv('dataset/survey_results_public.csv', sep=',')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "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>Respondent</th>\n",
       "      <th>AssessJob1</th>\n",
       "      <th>AssessJob2</th>\n",
       "      <th>AssessJob3</th>\n",
       "      <th>AssessJob4</th>\n",
       "      <th>AssessJob5</th>\n",
       "      <th>AssessJob6</th>\n",
       "      <th>AssessJob7</th>\n",
       "      <th>AssessJob8</th>\n",
       "      <th>AssessJob9</th>\n",
       "      <th>...</th>\n",
       "      <th>JobEmailPriorities6</th>\n",
       "      <th>JobEmailPriorities7</th>\n",
       "      <th>ConvertedSalary</th>\n",
       "      <th>AdsPriorities1</th>\n",
       "      <th>AdsPriorities2</th>\n",
       "      <th>AdsPriorities3</th>\n",
       "      <th>AdsPriorities4</th>\n",
       "      <th>AdsPriorities5</th>\n",
       "      <th>AdsPriorities6</th>\n",
       "      <th>AdsPriorities7</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>98855.000000</td>\n",
       "      <td>66985.000000</td>\n",
       "      <td>66985.000000</td>\n",
       "      <td>66985.000000</td>\n",
       "      <td>66985.000000</td>\n",
       "      <td>66985.000000</td>\n",
       "      <td>66985.000000</td>\n",
       "      <td>66985.000000</td>\n",
       "      <td>66985.000000</td>\n",
       "      <td>66985.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>46213.00000</td>\n",
       "      <td>46213.000000</td>\n",
       "      <td>4.770200e+04</td>\n",
       "      <td>60479.000000</td>\n",
       "      <td>60479.000000</td>\n",
       "      <td>60479.000000</td>\n",
       "      <td>60479.000000</td>\n",
       "      <td>60479.000000</td>\n",
       "      <td>60479.000000</td>\n",
       "      <td>60479.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>50822.971635</td>\n",
       "      <td>6.397089</td>\n",
       "      <td>6.673524</td>\n",
       "      <td>5.906875</td>\n",
       "      <td>4.065791</td>\n",
       "      <td>3.953243</td>\n",
       "      <td>4.407196</td>\n",
       "      <td>5.673181</td>\n",
       "      <td>4.225200</td>\n",
       "      <td>7.640009</td>\n",
       "      <td>...</td>\n",
       "      <td>4.97425</td>\n",
       "      <td>4.836388</td>\n",
       "      <td>9.578086e+04</td>\n",
       "      <td>2.726880</td>\n",
       "      <td>3.805784</td>\n",
       "      <td>3.340945</td>\n",
       "      <td>3.782470</td>\n",
       "      <td>4.383604</td>\n",
       "      <td>5.138809</td>\n",
       "      <td>4.821459</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>29321.650410</td>\n",
       "      <td>2.788428</td>\n",
       "      <td>2.531202</td>\n",
       "      <td>2.642734</td>\n",
       "      <td>2.541196</td>\n",
       "      <td>2.520499</td>\n",
       "      <td>2.502069</td>\n",
       "      <td>2.923998</td>\n",
       "      <td>2.507411</td>\n",
       "      <td>2.407457</td>\n",
       "      <td>...</td>\n",
       "      <td>1.86063</td>\n",
       "      <td>1.659844</td>\n",
       "      <td>2.023482e+05</td>\n",
       "      <td>1.881078</td>\n",
       "      <td>1.821323</td>\n",
       "      <td>1.673485</td>\n",
       "      <td>1.844864</td>\n",
       "      <td>1.931746</td>\n",
       "      <td>1.853249</td>\n",
       "      <td>1.874895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>25443.500000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>4.00000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>2.384400e+04</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>50823.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>5.00000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.507500e+04</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>5.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>76219.500000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>7.00000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>9.300000e+04</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>101592.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>7.00000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>2.000000e+06</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 42 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          Respondent    AssessJob1    AssessJob2    AssessJob3    AssessJob4  \\\n",
       "count   98855.000000  66985.000000  66985.000000  66985.000000  66985.000000   \n",
       "mean    50822.971635      6.397089      6.673524      5.906875      4.065791   \n",
       "std     29321.650410      2.788428      2.531202      2.642734      2.541196   \n",
       "min         1.000000      1.000000      1.000000      1.000000      1.000000   \n",
       "25%     25443.500000      4.000000      5.000000      4.000000      2.000000   \n",
       "50%     50823.000000      7.000000      7.000000      6.000000      4.000000   \n",
       "75%     76219.500000      9.000000      9.000000      8.000000      6.000000   \n",
       "max    101592.000000     10.000000     10.000000     10.000000     10.000000   \n",
       "\n",
       "         AssessJob5    AssessJob6    AssessJob7    AssessJob8    AssessJob9  \\\n",
       "count  66985.000000  66985.000000  66985.000000  66985.000000  66985.000000   \n",
       "mean       3.953243      4.407196      5.673181      4.225200      7.640009   \n",
       "std        2.520499      2.502069      2.923998      2.507411      2.407457   \n",
       "min        1.000000      1.000000      1.000000      1.000000      1.000000   \n",
       "25%        2.000000      2.000000      3.000000      2.000000      6.000000   \n",
       "50%        3.000000      4.000000      6.000000      4.000000      8.000000   \n",
       "75%        6.000000      6.000000      8.000000      6.000000     10.000000   \n",
       "max       10.000000     10.000000     10.000000     10.000000     10.000000   \n",
       "\n",
       "       ...  JobEmailPriorities6  JobEmailPriorities7  ConvertedSalary  \\\n",
       "count  ...          46213.00000         46213.000000     4.770200e+04   \n",
       "mean   ...              4.97425             4.836388     9.578086e+04   \n",
       "std    ...              1.86063             1.659844     2.023482e+05   \n",
       "min    ...              1.00000             1.000000     0.000000e+00   \n",
       "25%    ...              4.00000             4.000000     2.384400e+04   \n",
       "50%    ...              5.00000             5.000000     5.507500e+04   \n",
       "75%    ...              7.00000             6.000000     9.300000e+04   \n",
       "max    ...              7.00000             7.000000     2.000000e+06   \n",
       "\n",
       "       AdsPriorities1  AdsPriorities2  AdsPriorities3  AdsPriorities4  \\\n",
       "count    60479.000000    60479.000000    60479.000000    60479.000000   \n",
       "mean         2.726880        3.805784        3.340945        3.782470   \n",
       "std          1.881078        1.821323        1.673485        1.844864   \n",
       "min          1.000000        1.000000        1.000000        1.000000   \n",
       "25%          1.000000        2.000000        2.000000        2.000000   \n",
       "50%          2.000000        4.000000        3.000000        4.000000   \n",
       "75%          4.000000        5.000000        5.000000        5.000000   \n",
       "max          7.000000        7.000000        7.000000        7.000000   \n",
       "\n",
       "       AdsPriorities5  AdsPriorities6  AdsPriorities7  \n",
       "count    60479.000000    60479.000000    60479.000000  \n",
       "mean         4.383604        5.138809        4.821459  \n",
       "std          1.931746        1.853249        1.874895  \n",
       "min          1.000000        1.000000        1.000000  \n",
       "25%          3.000000        4.000000        3.000000  \n",
       "50%          5.000000        6.000000        5.000000  \n",
       "75%          6.000000        7.000000        7.000000  \n",
       "max          7.000000        7.000000        7.000000  \n",
       "\n",
       "[8 rows x 42 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(df.describe())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Chosen Attributes\n",
    "\n",
    "Most of the Dataset attributes consisted in **answers to open question** which would have been hard to manage and also contained mixed-type data.\n",
    "\n",
    "Among the user's information the following attributes have been chosen:\n",
    "\n",
    "- ***Hobby***: *True if the user considers Coding a Hobby.*\n",
    "- ***OpenSource***: *True if the user considers itself an OpenSource supporter*\n",
    "- ***Country***: *The user's Country of origin*\n",
    "- ***Employment***: *The user's employement type (Full-Time Part-Time...)*\n",
    "- ***Education***: *The user's Formal Education (e.g., Bachelor, Master's Degree, etc.)*\n",
    "- ***Undergrad Major***: *The user's major during it's Undergraduate studies*\n",
    "- ***Job Satisfaction***: *A score between 1 and 7 representing the user's Jobs Satisfaction*\n",
    "- ***Salary***: *The user's Salary Converted in USD*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "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>Hobby</th>\n",
       "      <th>OpenSource</th>\n",
       "      <th>Country</th>\n",
       "      <th>Employment</th>\n",
       "      <th>Education</th>\n",
       "      <th>UndergradMajor</th>\n",
       "      <th>JobSatisfaction</th>\n",
       "      <th>Salary</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Employed full-time</td>\n",
       "      <td>Bachelor’s degree (BA, BS, B.Eng., etc.)</td>\n",
       "      <td>A natural science (ex. biology, chemistry, phy...</td>\n",
       "      <td>Moderately dissatisfied</td>\n",
       "      <td>70841.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>South Africa</td>\n",
       "      <td>Employed full-time</td>\n",
       "      <td>Some college/university study without earning ...</td>\n",
       "      <td>Computer science, computer engineering, or sof...</td>\n",
       "      <td>Slightly satisfied</td>\n",
       "      <td>21426.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Employed full-time</td>\n",
       "      <td>Bachelor’s degree (BA, BS, B.Eng., etc.)</td>\n",
       "      <td>Computer science, computer engineering, or sof...</td>\n",
       "      <td>Moderately satisfied</td>\n",
       "      <td>41671.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United States</td>\n",
       "      <td>Employed full-time</td>\n",
       "      <td>Some college/university study without earning ...</td>\n",
       "      <td>Computer science, computer engineering, or sof...</td>\n",
       "      <td>Slightly satisfied</td>\n",
       "      <td>120000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United States</td>\n",
       "      <td>Employed full-time</td>\n",
       "      <td>Some college/university study without earning ...</td>\n",
       "      <td>Fine arts or performing arts (ex. graphic desi...</td>\n",
       "      <td>Moderately satisfied</td>\n",
       "      <td>250000.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Hobby OpenSource         Country          Employment  \\\n",
       "1   Yes        Yes  United Kingdom  Employed full-time   \n",
       "4   Yes         No    South Africa  Employed full-time   \n",
       "5   Yes         No  United Kingdom  Employed full-time   \n",
       "6   Yes        Yes   United States  Employed full-time   \n",
       "8   Yes        Yes   United States  Employed full-time   \n",
       "\n",
       "                                           Education  \\\n",
       "1           Bachelor’s degree (BA, BS, B.Eng., etc.)   \n",
       "4  Some college/university study without earning ...   \n",
       "5           Bachelor’s degree (BA, BS, B.Eng., etc.)   \n",
       "6  Some college/university study without earning ...   \n",
       "8  Some college/university study without earning ...   \n",
       "\n",
       "                                      UndergradMajor          JobSatisfaction  \\\n",
       "1  A natural science (ex. biology, chemistry, phy...  Moderately dissatisfied   \n",
       "4  Computer science, computer engineering, or sof...       Slightly satisfied   \n",
       "5  Computer science, computer engineering, or sof...     Moderately satisfied   \n",
       "6  Computer science, computer engineering, or sof...       Slightly satisfied   \n",
       "8  Fine arts or performing arts (ex. graphic desi...     Moderately satisfied   \n",
       "\n",
       "     Salary  \n",
       "1   70841.0  \n",
       "4   21426.0  \n",
       "5   41671.0  \n",
       "6  120000.0  \n",
       "8  250000.0  "
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "columns=['Hobby','OpenSource','Country','Employment',\\\n",
    "    'FormalEducation','UndergradMajor','JobSatisfaction','ConvertedSalary']\n",
    "\n",
    "data=df.get(columns).dropna()\n",
    "\n",
    "\n",
    "data = data.rename(columns={'FormalEducation':'Education', 'ConvertedSalary':'Salary'})\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The ***JobSatisfaction*** values have been remapped into an integer scale."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "js_dict = {\n",
    "    'Extremely dissatisfied':0,\n",
    "    'Moderately dissatisfied':1,\n",
    "    'Slightly dissatisfied':2,\n",
    "    'Neither satisfied nor dissatisfied':3,\n",
    "    'Slightly satisfied':4,\n",
    "    'Moderately satisfied':5,\n",
    "    'Extremely satisfied':6,\n",
    "}\n",
    "\n",
    "data['JobSatisfaction']=data['JobSatisfaction'].map(js_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Dataset has been pruned from entries which had a ***Formal Education*** different from a *Bachelor's Degree, a Master's Degree or a Ph.D* to **reduce the complexity** of the Model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.replace('Bachelor’s degree (BA, BS, B.Eng., etc.)',\"Bachelor\")\n",
    "data = data.replace('Master’s degree (MA, MS, M.Eng., MBA, etc.)',\"Master\")\n",
    "data = data.replace('Other doctoral degree (Ph.D, Ed.D., etc.)',\"Ph.D\")\n",
    "data = data[data['Education'].isin(['Bachelor','Master','Ph.D'])]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We considered data concerning only the **three most represented countries** in the Dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "countries=data.groupby('Country').size().sort_values()[-3:].index.tolist()\n",
    "data = data[data['Country'].isin(countries)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The ***Converted Salary*** attribute has been discretized into three ranges to reduce the Dataset Complexity and to Work on **Categorical Values**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,8))\n",
    "data.Salary=data.Salary.map(lambda x: '0-250.000' if x<=250000 else '250.000-1.000.000' if x<=1000000 else '1.000.000+')\n",
    "data.Salary.hist()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The ***Undergrad Major*** Attribute has been Binarized into \"STEM\" and \"NON STEM\" majors."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "stem=['A natural science (ex. biology, chemistry, physics)',\n",
    " 'Computer science, computer engineering, or software engineering',\n",
    " 'Web development or web design',\n",
    " 'Another engineering discipline (ex. civil, electrical, mechanical)',\n",
    " 'Mathematics or statistics',\n",
    " 'Information systems, information technology, or system administration',\n",
    "]\n",
    "\n",
    "not_stem=[ 'A social science (ex. anthropology, psychology, political science)',\n",
    " 'A humanities discipline (ex. literature, history, philosophy)',\n",
    " 'A business discipline (ex. accounting, finance, marketing)',\n",
    " 'Fine arts or performing arts (ex. graphic design, music, studio art)',\n",
    " 'A health science (ex. nursing, pharmacy, radiology)',\n",
    "]\n",
    "\n",
    "data=data[data['UndergradMajor'].isin(stem+not_stem)]\n",
    "data.UndergradMajor=data.UndergradMajor.map(lambda x: 'STEM' if x in stem else 'NOT_STEM')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Content of the Dataset after applying pre-processing:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "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>Hobby</th>\n",
       "      <th>OpenSource</th>\n",
       "      <th>Country</th>\n",
       "      <th>Employment</th>\n",
       "      <th>Education</th>\n",
       "      <th>UndergradMajor</th>\n",
       "      <th>JobSatisfaction</th>\n",
       "      <th>Salary</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Employed full-time</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>STEM</td>\n",
       "      <td>1</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Employed full-time</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>STEM</td>\n",
       "      <td>5</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Employed full-time</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>STEM</td>\n",
       "      <td>4</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>India</td>\n",
       "      <td>Employed full-time</td>\n",
       "      <td>Master</td>\n",
       "      <td>STEM</td>\n",
       "      <td>5</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>United States</td>\n",
       "      <td>Employed full-time</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>NOT_STEM</td>\n",
       "      <td>5</td>\n",
       "      <td>250.000-1.000.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Hobby OpenSource         Country          Employment Education  \\\n",
       "1    Yes        Yes  United Kingdom  Employed full-time  Bachelor   \n",
       "5    Yes         No  United Kingdom  Employed full-time  Bachelor   \n",
       "22    No         No  United Kingdom  Employed full-time  Bachelor   \n",
       "24   Yes         No           India  Employed full-time    Master   \n",
       "27    No         No   United States  Employed full-time  Bachelor   \n",
       "\n",
       "   UndergradMajor  JobSatisfaction             Salary  \n",
       "1            STEM                1          0-250.000  \n",
       "5            STEM                5          0-250.000  \n",
       "22           STEM                4          0-250.000  \n",
       "24           STEM                5          0-250.000  \n",
       "27       NOT_STEM                5  250.000-1.000.000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(data.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Hobby : ['Yes' 'No']\n",
      "OpenSource : ['Yes' 'No']\n",
      "Country : ['United Kingdom' 'India' 'United States']\n",
      "Employment : ['Employed full-time' 'Employed part-time'\n",
      " 'Independent contractor, freelancer, or self-employed']\n",
      "Education : ['Bachelor' 'Master' 'Ph.D']\n",
      "UndergradMajor : ['STEM' 'NOT_STEM']\n",
      "JobSatisfaction : [1 5 4 0 6 2 3]\n",
      "Salary : ['0-250.000' '250.000-1.000.000' '1.000.000+']\n"
     ]
    }
   ],
   "source": [
    "for col in data.columns:\n",
    "    \n",
    "    print(col, \":\", data[col].unique())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bayesian Network Creation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We decided to test the Dataset Attribute independence through a $Chi^2 Test$ since we are working with Categorical Values.\n",
    "\n",
    "Each attribute's indipendence has been tested with all the others and a Contingency matrix has been Created. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*Function that computes Tests on all the Dataset Attributes independences and creates a Contingency matrix*:\n",
    "\n",
    "- For Readability We decided to set all the the test outpus to 1 if their p values are greater or equal to $\\alpha$."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Chi-Square Test:\n",
    "\n",
    "The Chi-square test of independence is a statistical hypothesis test used to determine whether two categorical or nominal variables are likely to be related or not.\n",
    "\n",
    "A test of independence assesses whether observations consisting of measures on two variables, expressed in a contingency table, are independent of each other.\n",
    "\n",
    "The test Consists in the following steps:\n",
    "\n",
    "1. Defining an Hypothesis:\n",
    "     - `Null Hypothesis (H0)`: Two variables are independent.\n",
    "     - `Alternate Hypothesis (H1)`: Two variables are not independent.\n",
    "2. Calculating a **Contingency Table** for both Attributes tested:\n",
    "     - Table showing the distribution of one variable in rows and another in columns.\n",
    "3. Find the Expected Value\n",
    "     - $A$ and $B$ are independent iff $P(A \\cap B)=P(A)* P(B)$\n",
    "4. Calculate the $p-value$:\n",
    "     - $\\chi^2=\\Sigma\\frac{(O_i-E_i)^2}{E_i}$\n",
    "     - Where:\n",
    "       - $E_i$ is the expected value computed in the previous step.\n",
    "       - $O_i$ is the number of observations of type $i$\n",
    "5. Accept or Reject The Hypothesis:\n",
    "     - Decide $\\alpha$ = Significance level of the Test\n",
    "     - if: $p-value>\\alpha$ `H0` is accepted.\n",
    "     - Otherwise: `H1` is accepted\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "def chi2_contingency_mat(data,alpha=None):\n",
    "    s=data.columns.size\n",
    "    a = 0\n",
    "    b = 0\n",
    "    mat=np.zeros((s,s))\n",
    "    for i in data.columns:\n",
    "        for j in data.columns:\n",
    "            contigency_pct = pd.crosstab(data[i], data[j])\n",
    "            c, p, dof, expected = chi2_contingency(contigency_pct)\n",
    "            mat[a][b]=p\n",
    "            b=b+1\n",
    "        a = a+1\n",
    "        b=0\n",
    "    if alpha:\n",
    "        mat[mat>=alpha]=1\n",
    "        mat[mat<alpha]=0\n",
    "    return mat\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$Chi^2$ Contingency matrix plot:\n",
    "\n",
    "- $\\alpha$ has been set to $5\\times 10^{-7}$ to take in consideration only **Highly Dependent** attributes and take less risk in misjudging the results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "chi2Mat=chi2_contingency_mat(data,5e-7)\n",
    "labels = list(data.columns)\n",
    "plt.figure(figsize=(10,8))\n",
    "sns.heatmap(chi2Mat,annot=True, xticklabels=labels, yticklabels=labels,cmap=\"YlGnBu\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see the *Employment* column could make the **BN** too complicated and also doesn't really seem signicative enough, so we decided to drop it and recompute the Contingency matrix."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "data = data.drop('Employment', axis=1)\n",
    "chi2Mat=chi2_contingency_mat(data,5e-3)\n",
    "labels = list(data.columns)\n",
    "plt.figure(figsize=(10,8))\n",
    "sns.heatmap(chi2Mat,annot=True, xticklabels=labels, yticklabels=labels,cmap=\"YlGnBu\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Graph Plotting"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Function used to Compute a Graph Table starting from a computed $\\chi^2$ matrix."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_graph_table(data, chi2Mat):\n",
    "\n",
    "    graph_table = pd.DataFrame()\n",
    "\n",
    "\n",
    "    source = []\n",
    "    target=[]\n",
    "\n",
    "    a=1\n",
    "    for i in range(data.columns.size):\n",
    "        for j in range(a, data.columns.size):\n",
    "            if chi2Mat[i][j]==0:\n",
    "                source.append(data.columns[i])\n",
    "                target.append(data.columns[j])\n",
    "        a=a+1\n",
    "\n",
    "    type=['Unidirected' for i in range(len(source))]\n",
    "    weight=[None for i in range(len(source))]\n",
    "    graph_table['Source']=source\n",
    "    graph_table['Target']=target\n",
    "    graph_table['Type']=type\n",
    "    graph_table['weight']=weight\n",
    "    return graph_table\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Computed Graph Table:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "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>Source</th>\n",
       "      <th>Target</th>\n",
       "      <th>Type</th>\n",
       "      <th>weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Hobby</td>\n",
       "      <td>OpenSource</td>\n",
       "      <td>Unidirected</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>OpenSource</td>\n",
       "      <td>Country</td>\n",
       "      <td>Unidirected</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>OpenSource</td>\n",
       "      <td>Education</td>\n",
       "      <td>Unidirected</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>OpenSource</td>\n",
       "      <td>JobSatisfaction</td>\n",
       "      <td>Unidirected</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Country</td>\n",
       "      <td>Education</td>\n",
       "      <td>Unidirected</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Country</td>\n",
       "      <td>UndergradMajor</td>\n",
       "      <td>Unidirected</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Country</td>\n",
       "      <td>JobSatisfaction</td>\n",
       "      <td>Unidirected</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Country</td>\n",
       "      <td>Salary</td>\n",
       "      <td>Unidirected</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Education</td>\n",
       "      <td>JobSatisfaction</td>\n",
       "      <td>Unidirected</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Education</td>\n",
       "      <td>Salary</td>\n",
       "      <td>Unidirected</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>UndergradMajor</td>\n",
       "      <td>JobSatisfaction</td>\n",
       "      <td>Unidirected</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Source           Target         Type weight\n",
       "0            Hobby       OpenSource  Unidirected   None\n",
       "1       OpenSource          Country  Unidirected   None\n",
       "2       OpenSource        Education  Unidirected   None\n",
       "3       OpenSource  JobSatisfaction  Unidirected   None\n",
       "4          Country        Education  Unidirected   None\n",
       "5          Country   UndergradMajor  Unidirected   None\n",
       "6          Country  JobSatisfaction  Unidirected   None\n",
       "7          Country           Salary  Unidirected   None\n",
       "8        Education  JobSatisfaction  Unidirected   None\n",
       "9        Education           Salary  Unidirected   None\n",
       "10  UndergradMajor  JobSatisfaction  Unidirected   None"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "graph_table = compute_graph_table(data, chi2Mat)\n",
    "graph_table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import networkx as nx\n",
    "\n",
    "G = nx.from_pandas_edgelist(graph_table, source='Source', target='Target',create_using=nx.DiGraph())\n",
    "pos={'Hobby':(0.5,10),\n",
    "     'JobSatisfaction':(2,7),\n",
    "     'OpenSource':(0.5,7),\n",
    "     'Education':(0.5,3),\n",
    "     'Country':(2,3),\n",
    "     'Salary':(1.25,1),\n",
    "     'UndergradMajor':(3,5)\n",
    " }\n",
    "plt.figure(figsize=(10,8))\n",
    "\n",
    "nx.draw(G, arrows=True, with_labels=True, node_size=7000, arrowsize=20, pos=pos, font_size=10, font_color='white')\n",
    "# print(G.edges())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Markov Blanket Plotting"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Function to compute a Markov Blanked given a Graph and a Node.\n",
    "- Function that Plots the Markov Blanket of a Node:\n",
    "  - **Red**: Node We are Interested in.\n",
    "  - **Green**: Markov Blanket Set of Nodes.\n",
    "  - **Blue**: Set of nodes which are not part of the Markov Blanket.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Function to extract the Markov Blanket of a Node from the Graph\n",
    "def get_mb(graph, node):\n",
    "    mb = set()\n",
    "    parents = [i for i in graph.predecessors(node)]\n",
    "    mb.update(parents)\n",
    "    sons = [i for i in graph.successors(node)]\n",
    "    mb.update(sons)\n",
    "    sons_parents=[]\n",
    "    for s in sons:\n",
    "        sons_parents.extend([i for i in graph.predecessors(s) if i != node])\n",
    "\n",
    "    mb.update(sons_parents)\n",
    "    return mb\n",
    "\n",
    "#Markov Blanket Plotting Function\n",
    "def plot_mb(G, mb, node):\n",
    "    pos={'Hobby':(0.5,10),\n",
    "     'JobSatisfaction':(2,7),\n",
    "     'OpenSource':(0.5,7),\n",
    "     'Education':(0.5,3),\n",
    "     'Country':(2,3),\n",
    "     'Salary':(1.25,1),\n",
    "     'UndergradMajor':(3,5)\n",
    "    }\n",
    "    plt.figure(figsize=(10,8))\n",
    "    color_map=[]\n",
    "    for n in G:\n",
    "        if n in mb:\n",
    "            color_map.append('green')\n",
    "        elif n == node:\n",
    "            color_map.append('red')\n",
    "        else: \n",
    "            color_map.append('blue') \n",
    "    nx.draw(G, arrows=True,node_color=color_map, with_labels=True, node_size=7000, arrowsize=20, pos=pos, font_size=10, font_color='white')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*Markov Blanket of the OpenSource Node*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mb=get_mb(G, 'OpenSource')\n",
    "plot_mb(G, mb, 'OpenSource')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bayesian Network Model (PGMPY library)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This function computes the Graph Edges using the Upper triangular part of the $\\chi^2$ matrix as an adjacency Table."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "def getEdges(mat,names=[]):\n",
    "    result=[]\n",
    "    l,_=mat.shape\n",
    "    for i in range(l): \n",
    "        for j in range(i+1,l):\n",
    "            if mat[i,j]==0:\n",
    "                if len(names)>0:\n",
    "                    result.append((names[i],names[j]))\n",
    "                else: result.append((i,j))\n",
    "    return result    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Maximum likelihood\n",
    "Model Creation using a **Maximum Likelihood Estimator** to compute the Probability Tables.\n",
    "\n",
    "- **Maximum likelihood estimation (MLE)** is a method of estimating the parameters of an assumed probability distribution, given some observed data. \n",
    " \n",
    "- This is achieved by maximizing a likelihood function so that, under the assumed statistical model, the observed data is most probable.\n",
    "\n",
    "Given the **Likelihood** of a function:\n",
    "\n",
    "$L_n(\\theta)=L_n(\\theta; y) = f_n(y;\\theta)$\n",
    "\n",
    "where $y =  \\{ y_1, y_2,...\\}$ is a set of random variables.\n",
    "\n",
    "**MLE** aims at maximizing $\\hat{\\theta}$ using the following formula:\n",
    "\n",
    "$\\hat{\\theta} = arg_{\\theta \\in \\Theta} max \\hat{L_n}(\\theta;y)$\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------------+----------+\n",
      "| Hobby(No)  | 0.199923 |\n",
      "+------------+----------+\n",
      "| Hobby(Yes) | 0.800077 |\n",
      "+------------+----------+\n",
      "+-----------------+--------------------+---------------------+\n",
      "| Hobby           | Hobby(No)          | Hobby(Yes)          |\n",
      "+-----------------+--------------------+---------------------+\n",
      "| OpenSource(No)  | 0.7966101694915254 | 0.49800223749400674 |\n",
      "+-----------------+--------------------+---------------------+\n",
      "| OpenSource(Yes) | 0.2033898305084746 | 0.5019977625059933  |\n",
      "+-----------------+--------------------+---------------------+\n",
      "+-------------------------+---------------------+---------------------+\n",
      "| OpenSource              | OpenSource(No)      | OpenSource(Yes)     |\n",
      "+-------------------------+---------------------+---------------------+\n",
      "| Country(India)          | 0.20440215522182736 | 0.24486845909222318 |\n",
      "+-------------------------+---------------------+---------------------+\n",
      "| Country(United Kingdom) | 0.17872291642783447 | 0.16536571263370917 |\n",
      "+-------------------------+---------------------+---------------------+\n",
      "| Country(United States)  | 0.6168749283503382  | 0.5897658282740676  |\n",
      "+-------------------------+---------------------+---------------------+\n",
      "+---------------------+-----+------------------------+\n",
      "| Country             | ... | Country(United States) |\n",
      "+---------------------+-----+------------------------+\n",
      "| OpenSource          | ... | OpenSource(Yes)        |\n",
      "+---------------------+-----+------------------------+\n",
      "| Education(Bachelor) | ... | 0.7134803921568628     |\n",
      "+---------------------+-----+------------------------+\n",
      "| Education(Master)   | ... | 0.24215686274509804    |\n",
      "+---------------------+-----+------------------------+\n",
      "| Education(Ph.D)     | ... | 0.04436274509803922    |\n",
      "+---------------------+-----+------------------------+\n",
      "+--------------------+-----+------------------------+\n",
      "| Country            | ... | Country(United States) |\n",
      "+--------------------+-----+------------------------+\n",
      "| Education          | ... | Education(Ph.D)        |\n",
      "+--------------------+-----+------------------------+\n",
      "| OpenSource         | ... | OpenSource(Yes)        |\n",
      "+--------------------+-----+------------------------+\n",
      "| UndergradMajor     | ... | UndergradMajor(STEM)   |\n",
      "+--------------------+-----+------------------------+\n",
      "| JobSatisfaction(0) | ... | 0.05806451612903226    |\n",
      "+--------------------+-----+------------------------+\n",
      "| JobSatisfaction(1) | ... | 0.05806451612903226    |\n",
      "+--------------------+-----+------------------------+\n",
      "| JobSatisfaction(2) | ... | 0.08387096774193549    |\n",
      "+--------------------+-----+------------------------+\n",
      "| JobSatisfaction(3) | ... | 0.025806451612903226   |\n",
      "+--------------------+-----+------------------------+\n",
      "| JobSatisfaction(4) | ... | 0.08387096774193549    |\n",
      "+--------------------+-----+------------------------+\n",
      "| JobSatisfaction(5) | ... | 0.3741935483870968     |\n",
      "+--------------------+-----+------------------------+\n",
      "| JobSatisfaction(6) | ... | 0.3161290322580645     |\n",
      "+--------------------+-----+------------------------+\n",
      "+--------------------------+-----+------------------------+\n",
      "| Country                  | ... | Country(United States) |\n",
      "+--------------------------+-----+------------------------+\n",
      "| UndergradMajor(NOT_STEM) | ... | 0.15759433463693057    |\n",
      "+--------------------------+-----+------------------------+\n",
      "| UndergradMajor(STEM)     | ... | 0.8424056653630695     |\n",
      "+--------------------------+-----+------------------------+\n",
      "+---------------------------+-----+------------------------+\n",
      "| Country                   | ... | Country(United States) |\n",
      "+---------------------------+-----+------------------------+\n",
      "| Education                 | ... | Education(Ph.D)        |\n",
      "+---------------------------+-----+------------------------+\n",
      "| Salary(0-250.000)         | ... | 0.9244712990936556     |\n",
      "+---------------------------+-----+------------------------+\n",
      "| Salary(1.000.000+)        | ... | 0.027190332326283987   |\n",
      "+---------------------------+-----+------------------------+\n",
      "| Salary(250.000-1.000.000) | ... | 0.04833836858006042    |\n",
      "+---------------------------+-----+------------------------+\n"
     ]
    }
   ],
   "source": [
    "from pgmpy.models import BayesianNetwork\n",
    "from pgmpy.estimators import MaximumLikelihoodEstimator\n",
    "\n",
    "edges=getEdges(chi2Mat,names=data.columns)\n",
    "model= BayesianNetwork(edges)\n",
    "model.fit(data, estimator=MaximumLikelihoodEstimator)\n",
    "for cpd in model.get_cpds():\n",
    "    print(cpd)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model Independencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(Education ⟂ Hobby | OpenSource)\n",
       "(Education ⟂ UndergradMajor | Country)\n",
       "(Education ⟂ Hobby | OpenSource, JobSatisfaction)\n",
       "(Education ⟂ Hobby | OpenSource, UndergradMajor)\n",
       "(Education ⟂ Hobby | OpenSource, Salary)\n",
       "(Education ⟂ Hobby, UndergradMajor | OpenSource, Country)\n",
       "(Education ⟂ UndergradMajor | Hobby, Country)\n",
       "(Education ⟂ UndergradMajor | Salary, Country)\n",
       "(Education ⟂ UndergradMajor | OpenSource, Hobby, Country)\n",
       "(Education ⟂ Hobby | OpenSource, JobSatisfaction, UndergradMajor)\n",
       "(Education ⟂ Hobby | OpenSource, Salary, JobSatisfaction)\n",
       "(Education ⟂ Hobby | OpenSource, JobSatisfaction, Country)\n",
       "(Education ⟂ Hobby | OpenSource, Salary, UndergradMajor)\n",
       "(Education ⟂ Hobby | OpenSource, Country, UndergradMajor)\n",
       "(Education ⟂ Hobby, UndergradMajor | OpenSource, Salary, Country)\n",
       "(Education ⟂ UndergradMajor | Salary, Hobby, Country)\n",
       "(Education ⟂ UndergradMajor | OpenSource, Salary, Hobby, Country)\n",
       "(Education ⟂ Hobby | OpenSource, Salary, JobSatisfaction, UndergradMajor)\n",
       "(Education ⟂ Hobby | OpenSource, JobSatisfaction, Country, UndergradMajor)\n",
       "(Education ⟂ Hobby | OpenSource, Salary, JobSatisfaction, Country)\n",
       "(Education ⟂ Hobby | OpenSource, Salary, Country, UndergradMajor)\n",
       "(Education ⟂ Hobby | OpenSource, JobSatisfaction, Country, UndergradMajor, Salary)\n",
       "(OpenSource ⟂ UndergradMajor | Country)\n",
       "(OpenSource ⟂ Salary, UndergradMajor | Education, Country)\n",
       "(OpenSource ⟂ UndergradMajor | Hobby, Country)\n",
       "(OpenSource ⟂ UndergradMajor | Salary, Country)\n",
       "(OpenSource ⟂ Salary, UndergradMajor | Education, Hobby, Country)\n",
       "(OpenSource ⟂ Salary | Education, JobSatisfaction, Country)\n",
       "(OpenSource ⟂ Salary | Education, Country, UndergradMajor)\n",
       "(OpenSource ⟂ UndergradMajor | Education, Salary, Country)\n",
       "(OpenSource ⟂ UndergradMajor | Salary, Hobby, Country)\n",
       "(OpenSource ⟂ Salary | Education, Hobby, Country, JobSatisfaction)\n",
       "(OpenSource ⟂ Salary | Education, Hobby, Country, UndergradMajor)\n",
       "(OpenSource ⟂ UndergradMajor | Education, Salary, Hobby, Country)\n",
       "(OpenSource ⟂ Salary | Education, JobSatisfaction, Country, UndergradMajor)\n",
       "(OpenSource ⟂ Salary | Education, Hobby, JobSatisfaction, Country, UndergradMajor)\n",
       "(JobSatisfaction ⟂ Hobby | OpenSource)\n",
       "(JobSatisfaction ⟂ Hobby | Education, OpenSource)\n",
       "(JobSatisfaction ⟂ Salary | Education, Country)\n",
       "(JobSatisfaction ⟂ Hobby | OpenSource, UndergradMajor)\n",
       "(JobSatisfaction ⟂ Hobby | OpenSource, Salary)\n",
       "(JobSatisfaction ⟂ Hobby | OpenSource, Country)\n",
       "(JobSatisfaction ⟂ Hobby | Education, OpenSource, UndergradMajor)\n",
       "(JobSatisfaction ⟂ Hobby | Education, OpenSource, Salary)\n",
       "(JobSatisfaction ⟂ Salary, Hobby | Education, OpenSource, Country)\n",
       "(JobSatisfaction ⟂ Salary | Education, Hobby, Country)\n",
       "(JobSatisfaction ⟂ Salary | Education, Country, UndergradMajor)\n",
       "(JobSatisfaction ⟂ Hobby | OpenSource, Salary, UndergradMajor)\n",
       "(JobSatisfaction ⟂ Hobby | OpenSource, Country, UndergradMajor)\n",
       "(JobSatisfaction ⟂ Hobby | OpenSource, Salary, Country)\n",
       "(JobSatisfaction ⟂ Salary | Education, OpenSource, Hobby, Country)\n",
       "(JobSatisfaction ⟂ Hobby | Education, OpenSource, Salary, UndergradMajor)\n",
       "(JobSatisfaction ⟂ Salary, Hobby | Education, OpenSource, Country, UndergradMajor)\n",
       "(JobSatisfaction ⟂ Hobby | Education, OpenSource, Salary, Country)\n",
       "(JobSatisfaction ⟂ Salary | Education, Hobby, Country, UndergradMajor)\n",
       "(JobSatisfaction ⟂ Hobby | OpenSource, Salary, Country, UndergradMajor)\n",
       "(JobSatisfaction ⟂ Salary | Education, OpenSource, Hobby, Country, UndergradMajor)\n",
       "(JobSatisfaction ⟂ Hobby | Education, OpenSource, Country, UndergradMajor, Salary)\n",
       "(Hobby ⟂ Education, JobSatisfaction, UndergradMajor, Salary, Country | OpenSource)\n",
       "(Hobby ⟂ UndergradMajor | Country)\n",
       "(Hobby ⟂ UndergradMajor, Salary, JobSatisfaction, Country | Education, OpenSource)\n",
       "(Hobby ⟂ Salary, UndergradMajor | Education, Country)\n",
       "(Hobby ⟂ Education, Salary, UndergradMajor, Country | OpenSource, JobSatisfaction)\n",
       "(Hobby ⟂ Education, Salary, JobSatisfaction, Country | OpenSource, UndergradMajor)\n",
       "(Hobby ⟂ Education, JobSatisfaction, UndergradMajor, Country | OpenSource, Salary)\n",
       "(Hobby ⟂ Education, Salary, JobSatisfaction, UndergradMajor | OpenSource, Country)\n",
       "(Hobby ⟂ UndergradMajor | Salary, Country)\n",
       "(Hobby ⟂ UndergradMajor, Salary, Country | Education, OpenSource, JobSatisfaction)\n",
       "(Hobby ⟂ Salary, JobSatisfaction, Country | Education, OpenSource, UndergradMajor)\n",
       "(Hobby ⟂ UndergradMajor, JobSatisfaction, Country | Education, OpenSource, Salary)\n",
       "(Hobby ⟂ Salary, JobSatisfaction, UndergradMajor | Education, OpenSource, Country)\n",
       "(Hobby ⟂ Salary | Education, JobSatisfaction, Country)\n",
       "(Hobby ⟂ Salary | Education, Country, UndergradMajor)\n",
       "(Hobby ⟂ UndergradMajor | Education, Salary, Country)\n",
       "(Hobby ⟂ Education, Salary, Country | OpenSource, JobSatisfaction, UndergradMajor)\n",
       "(Hobby ⟂ Education, UndergradMajor, Country | OpenSource, Salary, JobSatisfaction)\n",
       "(Hobby ⟂ Education, Salary, UndergradMajor | OpenSource, JobSatisfaction, Country)\n",
       "(Hobby ⟂ Education, JobSatisfaction, Country | OpenSource, Salary, UndergradMajor)\n",
       "(Hobby ⟂ Education, Salary, JobSatisfaction | OpenSource, Country, UndergradMajor)\n",
       "(Hobby ⟂ Education, JobSatisfaction, UndergradMajor | OpenSource, Salary, Country)\n",
       "(Hobby ⟂ Salary, Country | Education, OpenSource, JobSatisfaction, UndergradMajor)\n",
       "(Hobby ⟂ UndergradMajor, Country | Education, OpenSource, JobSatisfaction, Salary)\n",
       "(Hobby ⟂ Salary, UndergradMajor | Education, OpenSource, JobSatisfaction, Country)\n",
       "(Hobby ⟂ JobSatisfaction, Country | Education, OpenSource, Salary, UndergradMajor)\n",
       "(Hobby ⟂ Salary, JobSatisfaction | Education, OpenSource, Country, UndergradMajor)\n",
       "(Hobby ⟂ JobSatisfaction, UndergradMajor | Education, OpenSource, Salary, Country)\n",
       "(Hobby ⟂ Salary | Education, JobSatisfaction, Country, UndergradMajor)\n",
       "(Hobby ⟂ Education, Country | OpenSource, Salary, JobSatisfaction, UndergradMajor)\n",
       "(Hobby ⟂ Education, Salary | OpenSource, JobSatisfaction, Country, UndergradMajor)\n",
       "(Hobby ⟂ Education, UndergradMajor | OpenSource, Salary, JobSatisfaction, Country)\n",
       "(Hobby ⟂ Education, JobSatisfaction | OpenSource, Salary, Country, UndergradMajor)\n",
       "(Hobby ⟂ Country | Education, OpenSource, JobSatisfaction, UndergradMajor, Salary)\n",
       "(Hobby ⟂ Salary | Education, OpenSource, JobSatisfaction, Country, UndergradMajor)\n",
       "(Hobby ⟂ UndergradMajor | Education, OpenSource, JobSatisfaction, Country, Salary)\n",
       "(Hobby ⟂ JobSatisfaction | Education, OpenSource, Country, UndergradMajor, Salary)\n",
       "(Hobby ⟂ Education | OpenSource, JobSatisfaction, Country, UndergradMajor, Salary)\n",
       "(Country ⟂ Hobby | OpenSource)\n",
       "(Country ⟂ Hobby | Education, OpenSource)\n",
       "(Country ⟂ Hobby | OpenSource, JobSatisfaction)\n",
       "(Country ⟂ Hobby | OpenSource, UndergradMajor)\n",
       "(Country ⟂ Hobby | OpenSource, Salary)\n",
       "(Country ⟂ Hobby | Education, OpenSource, JobSatisfaction)\n",
       "(Country ⟂ Hobby | Education, OpenSource, UndergradMajor)\n",
       "(Country ⟂ Hobby | Education, OpenSource, Salary)\n",
       "(Country ⟂ Hobby | OpenSource, JobSatisfaction, UndergradMajor)\n",
       "(Country ⟂ Hobby | OpenSource, Salary, JobSatisfaction)\n",
       "(Country ⟂ Hobby | OpenSource, Salary, UndergradMajor)\n",
       "(Country ⟂ Hobby | Education, OpenSource, JobSatisfaction, UndergradMajor)\n",
       "(Country ⟂ Hobby | Education, OpenSource, JobSatisfaction, Salary)\n",
       "(Country ⟂ Hobby | Education, OpenSource, Salary, UndergradMajor)\n",
       "(Country ⟂ Hobby | OpenSource, Salary, JobSatisfaction, UndergradMajor)\n",
       "(Country ⟂ Hobby | Education, OpenSource, JobSatisfaction, UndergradMajor, Salary)\n",
       "(UndergradMajor ⟂ Hobby | OpenSource)\n",
       "(UndergradMajor ⟂ Education, OpenSource, Hobby, Salary | Country)\n",
       "(UndergradMajor ⟂ Hobby | Education, OpenSource)\n",
       "(UndergradMajor ⟂ OpenSource, Salary, Hobby | Education, Country)\n",
       "(UndergradMajor ⟂ Hobby | OpenSource, JobSatisfaction)\n",
       "(UndergradMajor ⟂ Education, Salary, Hobby | OpenSource, Country)\n",
       "(UndergradMajor ⟂ Hobby | OpenSource, Salary)\n",
       "(UndergradMajor ⟂ Education, OpenSource, Salary | Hobby, Country)\n",
       "(UndergradMajor ⟂ Education, OpenSource, Hobby | Salary, Country)\n",
       "(UndergradMajor ⟂ Hobby | Education, OpenSource, JobSatisfaction)\n",
       "(UndergradMajor ⟂ Salary, Hobby | Education, OpenSource, Country)\n",
       "(UndergradMajor ⟂ Hobby | Education, OpenSource, Salary)\n",
       "(UndergradMajor ⟂ OpenSource, Salary | Education, Hobby, Country)\n",
       "(UndergradMajor ⟂ Salary | Education, JobSatisfaction, Country)\n",
       "(UndergradMajor ⟂ OpenSource, Hobby | Education, Salary, Country)\n",
       "(UndergradMajor ⟂ Education, Salary | OpenSource, Hobby, Country)\n",
       "(UndergradMajor ⟂ Hobby | OpenSource, JobSatisfaction, Country)\n",
       "(UndergradMajor ⟂ Hobby | OpenSource, Salary, JobSatisfaction)\n",
       "(UndergradMajor ⟂ Education, Hobby | OpenSource, Salary, Country)\n",
       "(UndergradMajor ⟂ Education, OpenSource | Salary, Hobby, Country)\n",
       "(UndergradMajor ⟂ Salary | Education, OpenSource, Hobby, Country)\n",
       "(UndergradMajor ⟂ Salary, Hobby | Education, OpenSource, JobSatisfaction, Country)\n",
       "(UndergradMajor ⟂ Hobby | Education, OpenSource, JobSatisfaction, Salary)\n",
       "(UndergradMajor ⟂ Hobby | Education, OpenSource, Salary, Country)\n",
       "(UndergradMajor ⟂ Salary | Education, Hobby, Country, JobSatisfaction)\n",
       "(UndergradMajor ⟂ OpenSource | Education, Salary, Hobby, Country)\n",
       "(UndergradMajor ⟂ Education | OpenSource, Salary, Hobby, Country)\n",
       "(UndergradMajor ⟂ Hobby | OpenSource, Salary, JobSatisfaction, Country)\n",
       "(UndergradMajor ⟂ Salary | Education, OpenSource, Hobby, JobSatisfaction, Country)\n",
       "(UndergradMajor ⟂ Hobby | Education, OpenSource, JobSatisfaction, Country, Salary)\n",
       "(Salary ⟂ Hobby | OpenSource)\n",
       "(Salary ⟂ UndergradMajor | Country)\n",
       "(Salary ⟂ Hobby | Education, OpenSource)\n",
       "(Salary ⟂ OpenSource, JobSatisfaction, Hobby, UndergradMajor | Education, Country)\n",
       "(Salary ⟂ Hobby | OpenSource, JobSatisfaction)\n",
       "(Salary ⟂ Hobby, UndergradMajor | OpenSource, Country)\n",
       "(Salary ⟂ Hobby | OpenSource, UndergradMajor)\n",
       "(Salary ⟂ UndergradMajor | Hobby, Country)\n",
       "(Salary ⟂ Hobby | Education, OpenSource, JobSatisfaction)\n",
       "(Salary ⟂ JobSatisfaction, Hobby, UndergradMajor | Education, OpenSource, Country)\n",
       "(Salary ⟂ Hobby | Education, OpenSource, UndergradMajor)\n",
       "(Salary ⟂ OpenSource, JobSatisfaction, UndergradMajor | Education, Hobby, Country)\n",
       "(Salary ⟂ OpenSource, Hobby, UndergradMajor | Education, JobSatisfaction, Country)\n",
       "(Salary ⟂ OpenSource, Hobby, JobSatisfaction | Education, UndergradMajor, Country)\n",
       "(Salary ⟂ UndergradMajor | OpenSource, Hobby, Country)\n",
       "(Salary ⟂ Hobby | OpenSource, JobSatisfaction, Country)\n",
       "(Salary ⟂ Hobby | OpenSource, JobSatisfaction, UndergradMajor)\n",
       "(Salary ⟂ Hobby | OpenSource, UndergradMajor, Country)\n",
       "(Salary ⟂ JobSatisfaction, UndergradMajor | Education, OpenSource, Hobby, Country)\n",
       "(Salary ⟂ Hobby, UndergradMajor | Education, OpenSource, JobSatisfaction, Country)\n",
       "(Salary ⟂ Hobby | Education, OpenSource, JobSatisfaction, UndergradMajor)\n",
       "(Salary ⟂ Hobby, JobSatisfaction | Education, OpenSource, UndergradMajor, Country)\n",
       "(Salary ⟂ OpenSource, UndergradMajor | Education, Hobby, Country, JobSatisfaction)\n",
       "(Salary ⟂ OpenSource, JobSatisfaction | Education, Hobby, UndergradMajor, Country)\n",
       "(Salary ⟂ OpenSource, Hobby | Education, JobSatisfaction, UndergradMajor, Country)\n",
       "(Salary ⟂ Hobby | OpenSource, JobSatisfaction, UndergradMajor, Country)\n",
       "(Salary ⟂ UndergradMajor | Education, OpenSource, Hobby, JobSatisfaction, Country)\n",
       "(Salary ⟂ JobSatisfaction | Education, OpenSource, Hobby, UndergradMajor, Country)\n",
       "(Salary ⟂ Hobby | Education, OpenSource, JobSatisfaction, UndergradMajor, Country)\n",
       "(Salary ⟂ OpenSource | Education, Hobby, JobSatisfaction, UndergradMajor, Country)"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.get_independencies()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Flows of probabilistic influence\n",
    "**Definition (active two-edge trail):**\n",
    "If influence can flow from X to Y via Z, the trail X $\\rightleftharpoons$ Z $\\rightleftharpoons$ Y is active.\n",
    "\n",
    "For influence to flow from nodes $X_1$ to $X_n$, it needs to flow through every single\n",
    "node on the trail. \n",
    "This is true if and only if every two-edge trail $X_{i−1}\\rightleftharpoons X_i \\rightleftharpoons X_{i+1}$ along\n",
    "the trail allows influence to flow.\n",
    "\n",
    "**Definition (active trail):**  let $Z$ be a subset of observed variables.\n",
    "The trail $X_{i−1} \\rightleftharpoons X_i \\rightleftharpoons X_{i+1}$ is active given $Z$ if\n",
    "- $\\forall X_{i−1} \\rightarrow X_i \\leftarrow X_{i+1}$, $X_i$ or one of its descendants are in $Z$\n",
    "- no other node along the trail is in $Z$\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'Education', 'OpenSource', 'JobSatisfaction', 'Hobby', 'Country', 'UndergradMajor', 'Salary'}\n",
      "{'Education', 'JobSatisfaction', 'Country', 'UndergradMajor', 'Salary'}\n"
     ]
    }
   ],
   "source": [
    "print(model.active_trail_nodes(['JobSatisfaction'])['JobSatisfaction'])\n",
    "print(model.active_trail_nodes(['JobSatisfaction'],['OpenSource'])['JobSatisfaction'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can see how the presence of *OpenSource* in the evidence 'blocks' the trail between *JobSatisfaction* and *Hobby*."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Markov Blanket Notes\n",
    "All information about a random variable in a **Bayesian network** is contained within this set of nodes (parents, children, and parents of children). \n",
    "\n",
    "If we observe all of these variables (giving an **evidence**), then our node is independent of all other nodes within the network."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_mb(G, model.get_markov_blanket('Salary'), 'Salary')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see asking the BN the probability of having a certain *Salary* given its **Markov Blanket** is the same as adding evidence to other nodes that aren't part of it.\n",
    "\n",
    "This means that ***Salary*** is independent to other nodes given its **Markov Blanket**."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Variable Elimination\n",
    "**Variable Elimination** is an exact inference algorithm which consists in\n",
    "computing a probability by repeatedly applying two basic factor operations: product and marginalization.\n",
    "\n",
    "- In a Bayesian network, a *factor* correspond to a conditional probability distribution.\n",
    "\n",
    "- **Pointwise product** of factors $f_1$ and $f_2$ :\n",
    "$$f_1(x_1,..,x_j,y_1,..,y_k)\\times f_2(y_1,..,y_k,z_1,..,z_l)=f(x_1,..,x_j,y_1,..,y_k,z_1,..,z_l)$$\n",
    "\n",
    "- **Summing out** a variable from a product of factors:\n",
    "  1. move any constant factors outside the summation\n",
    "  2. add up submatrices in pointwise product of remaining factors\n",
    "$$\\begin{align*}\n",
    "    \\sum_{X} f_1\\times...\\times f_k=&f_1\\times...\\times f_i\\sum_{X} f_{i+1}\\times...\\times f_k\\\\ \n",
    "    =&f_1\\times...\\times f_i\\times f_{\\bar{X}}\n",
    "\\end{align*}$$\n",
    "\n",
    "The VE algorithm loops over the variables of the network and eliminates them following an ordering $O$. For each varibale $X_i$ (ordered according to \n",
    "$O$):\n",
    "1. multiply all factors $\\phi_i$ containing $X_i$\n",
    "2. sum out $X_i$ to obtain a new factor $\\tau$\n",
    "3. replace the factors $\\phi_i$ with $\\tau$\n",
    "\n",
    "**Any** ordering yields a valid algorithm. However, different orderings may **drammatically alter the running time of the VE algorithm**, and the search for the best ordering is a **NP_hard**  problem. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we will show two different queries made on the BN putting in evidence what an **irrelevant variable** is. A variable is irrelevant to the query if the **sum over it is equal to 1**.\n",
    "\n",
    " ***Th.*** $Y$ is irrelevant if **d-separated** from $X$ by $E$.\n",
    "\n",
    "In the previous picture, we have shown the Markov Blanket of *Salary*, which is formed by its ancestors *Country* and *Education*. Now we'll make two different queries:\n",
    "- *P(Salary|Country='India',Education='Bachelor')*\n",
    "- *P(Salary|Country='India',Education='Bachelor',OpenSource='Yes')*\n",
    "\n",
    "And we are expecting the outcomes to be equal, since *OpenSource* is d-separated from *Salary* by *Country* and *Education*, thus irrelevant to the query."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%capture\n",
    "from pgmpy.inference import VariableElimination\n",
    "inference = VariableElimination(model)\n",
    "\n",
    "pSalary1 = inference.query(['Salary'],{'Country':'India','Education':'Bachelor'})\n",
    "\n",
    "pSalary2 = inference.query(['Salary'],{'Country':'India','Education':'Bachelor','OpenSource':'Yes'})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "P(Salary | Country=India, Education=Bachelor)\n",
      "+---------------------------+---------------+\n",
      "| Salary                    |   phi(Salary) |\n",
      "+===========================+===============+\n",
      "| Salary(0-250.000)         |        0.9836 |\n",
      "+---------------------------+---------------+\n",
      "| Salary(1.000.000+)        |        0.0008 |\n",
      "+---------------------------+---------------+\n",
      "| Salary(250.000-1.000.000) |        0.0156 |\n",
      "+---------------------------+---------------+\n",
      "\n",
      "P(Salary | Country=India, Education=Bachelor, OpenSource=Yes)\n",
      "+---------------------------+---------------+\n",
      "| Salary                    |   phi(Salary) |\n",
      "+===========================+===============+\n",
      "| Salary(0-250.000)         |        0.9836 |\n",
      "+---------------------------+---------------+\n",
      "| Salary(1.000.000+)        |        0.0008 |\n",
      "+---------------------------+---------------+\n",
      "| Salary(250.000-1.000.000) |        0.0156 |\n",
      "+---------------------------+---------------+\n"
     ]
    }
   ],
   "source": [
    "print(\"P(Salary | Country=India, Education=Bachelor)\")\n",
    "print(pSalary1)\n",
    "print(\"\\nP(Salary | Country=India, Education=Bachelor, OpenSource=Yes)\")\n",
    "print(pSalary2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As expected, the two CPDs are **exactly the same**."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comparison between Exact and Approximate Methods\n",
    "\n",
    "We wil now make a comparison between Exact Inference Methods and Approximate Inference ones. \n",
    "\n",
    "First of all, we define a function that, given in input our model, an evidence, a sample size and the exact probabilities (computed using *Variable Elimination*), returns as results probabilities and absolutes error w.r.t. the exact computed on the samples obtained from two different sampling methods: \n",
    "- **Rejection Sampling:** randomly generates samples rejecting the ones where the evidence is false\n",
    "- **Likelihood Weighting:** in addition to every sample, produces a weight representing the probability that a sample would not be rejected\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Now switch to sampling methods\n",
    "from pgmpy.factors.discrete import State\n",
    "from pgmpy.sampling import BayesianModelSampling\n",
    "\n",
    "def run_experiment(model,sample_size,evidence,p_exact):\n",
    "    # Sample\n",
    "    def prob_LW(samples,variable):\n",
    "        result={}\n",
    "        values=samples[variable].unique()\n",
    "        for value in values:\n",
    "            result[value]=round(np.sum(samples[samples[variable]==value]['_weight'])/np.sum(samples['_weight']),2)\n",
    "        return result\n",
    "\n",
    "    def prob_RS(samples,variable):\n",
    "        result={}\n",
    "        values=samples[variable].unique()\n",
    "        tot=len(samples[variable])\n",
    "        for value in values:\n",
    "            result[value]=len(samples[samples[variable]==value])/tot\n",
    "        return result\n",
    "\n",
    "    def absolute_error(exact_value,approx_value):\n",
    "        return np.absolute(exact_value-approx_value)\n",
    "\n",
    "    evidence2 = [State(key,val) for key,val in evidence.items()]\n",
    "    inference_sampling=BayesianModelSampling(model)\n",
    "    samples_LW = inference_sampling.likelihood_weighted_sample(evidence = evidence2, size=sample_size)\n",
    "    samples_RS = inference_sampling.rejection_sample(evidence=evidence2, size=sample_size)\n",
    "    \n",
    "    ## Statistics  \n",
    "    variables=[str(node) for node in model.nodes if str(node) not in evidence.keys()]\n",
    "    p_LW={}\n",
    "    p_RS={}\n",
    "    absolute_error_LW={}\n",
    "    absolute_error_RS={}\n",
    "    \n",
    "    for variable in variables:\n",
    "\n",
    "        p_LW[variable]=prob_LW(samples_LW,variable)\n",
    "        p_RS[variable]=prob_RS(samples_RS,variable)\n",
    "        \n",
    "        values=samples_LW[variable].unique()\n",
    "        # exec('absolute_error_LW[variable]={value:absolute_error(p_exact[variable].get_value(%s=value),p_LW[variable][value]) for value in values}'%variable,locals())\n",
    "        absolute_error_LW[variable]={value:absolute_error(p_exact[variable].get_value(**{variable:value}),p_LW[variable][value]) for value in values}\n",
    "        \n",
    "        values=samples_RS[variable].unique()\n",
    "        # exec('absolute_error_RS[variable]={value:absolute_error(p_exact[variable].get_value(%s=value),p_RS[variable][value]) for value in values}'%variable,locals())\n",
    "        absolute_error_RS[variable]={value:absolute_error(p_exact[variable].get_value(**{variable:value}),p_RS[variable][value]) for value in values}\n",
    "\n",
    "    # Return results\n",
    "    return p_LW,p_RS,absolute_error_LW,absolute_error_RS"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we run a number of experiments using an increasing sample size."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%capture\n",
    "evidence = {'Country':'India','Education':'Bachelor'}\n",
    "starting_size_=1 # from 10 sample points\n",
    "final_size=5     # to 10^5 sample points\n",
    "experiments=20 # 8 experiments\n",
    "result=[]\n",
    "inference_exact=VariableElimination(model)\n",
    "\n",
    "variables=[str(node) for node in model.nodes if str(node) not in evidence.keys()]\n",
    "p_exact={variable:inference_exact.query([variable],evidence) for variable in variables}\n",
    "\n",
    "for size in np.logspace(starting_size_, final_size, num=experiments, dtype='<i8'):\n",
    "    p_LW,p_RS,absolute_error_LW,absolute_error_RS=run_experiment(model,size,evidence,p_exact)\n",
    "    result.append({\n",
    "        'sample_size':size,\n",
    "        'p_exact':p_exact,\n",
    "        'p_LW':p_LW,\n",
    "        'p_RS':p_RS,\n",
    "        'e_LW':absolute_error_LW,\n",
    "        'e_RS':absolute_error_RS\n",
    "    })"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "These two function simply plot respectively the probability and the error wrt the exact probability for a given variable assuming a certain value.    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_prob(results,p_exact,evidence,**variables):\n",
    "\n",
    "    sample_size=[r['sample_size'] for r in results]\n",
    "    \n",
    "    for var,val in variables.items():\n",
    "        p_LW=[r['p_LW'][var][val] for r in results]\n",
    "        p_RS=[r['p_RS'][var][val] for r in results]\n",
    "        p=p_exact[var].get_value(**{var:val})\n",
    "\n",
    "        plt.figure(figsize=(20,5))\n",
    "        plt.subplot()\n",
    "        plt.ylim(0,1)\n",
    "        plt.title('P({}={}|{}'.format(var,val,','.join([str(k)+\"=\"+str(v) for k,v in evidence.items()])))\n",
    "        LWCplot, = plt.semilogx(sample_size,p_LW,'b',label=\"Likelihood Weighting\")\n",
    "        RSCplot, = plt.semilogx(sample_size,p_RS,'g',label=\"Rejection Sampling\")\n",
    "        VECplot, = plt.semilogx(sample_size,p*np.ones(len(results)),'r',label=\"reference value\")\n",
    "        plt.legend(handles=[LWCplot,RSCplot,VECplot])\n",
    "        plt.show()\n",
    "\n",
    "def plot_error(results,evidence,**variables):\n",
    "    # evidence=p_exact['Hobby'].get_evidence()\n",
    "    # print(evidence)\n",
    "    sample_size=[r['sample_size'] for r in results]\n",
    "    \n",
    "    for var,val in variables.items():\n",
    "        e_LW=[r['e_LW'][var][val] for r in results]\n",
    "        e_RS=[r['e_RS'][var][val] for r in results]\n",
    "\n",
    "        plt.figure(figsize=(20,5))\n",
    "        plt.subplot()\n",
    "        plt.ylim(0,np.max(e_LW+e_RS)+0.1*np.max(e_LW+e_RS))\n",
    "        plt.title('Absolute error on P({}={}|{}'.format(var,val,','.join([str(k)+\"=\"+str(v) for k,v in evidence.items()])))\n",
    "        LWCplot, = plt.semilogx(sample_size,e_LW,'b',label=\"Likelihood Weighting\")\n",
    "        RSCplot, = plt.semilogx(sample_size,e_RS,'g',label=\"Rejection Sampling\")\n",
    "        plt.legend(handles=[LWCplot,RSCplot])\n",
    "        plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_prob(result,p_exact,evidence,OpenSource='Yes',Hobby='Yes')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_error(result,evidence,OpenSource='Yes',Hobby='Yes')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As expected, as the sample size increases, the accuracy of the probability computed using sampling methods approximate the exact one."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Final Considerations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### To conclude we'll show some Interesting Queries that can be formulated using the BN we built.\n",
    "*Given a **Country** what's the probability of having a **NON-STEM Degree**?*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding Elimination Order: : : 0it [00:00, ?it/s]\n",
      "0it [00:00, ?it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "India\n",
      "+--------------------------+-----------------------+\n",
      "| UndergradMajor           |   phi(UndergradMajor) |\n",
      "+==========================+=======================+\n",
      "| UndergradMajor(NOT_STEM) |                0.0129 |\n",
      "+--------------------------+-----------------------+\n",
      "| UndergradMajor(STEM)     |                0.9871 |\n",
      "+--------------------------+-----------------------+\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding Elimination Order: : : 0it [00:00, ?it/s]\n",
      "0it [00:00, ?it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "United Kingdom\n",
      "+--------------------------+-----------------------+\n",
      "| UndergradMajor           |   phi(UndergradMajor) |\n",
      "+==========================+=======================+\n",
      "| UndergradMajor(NOT_STEM) |                0.0844 |\n",
      "+--------------------------+-----------------------+\n",
      "| UndergradMajor(STEM)     |                0.9156 |\n",
      "+--------------------------+-----------------------+\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding Elimination Order: : : 0it [00:00, ?it/s]\n",
      "0it [00:00, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "United States\n",
      "+--------------------------+-----------------------+\n",
      "| UndergradMajor           |   phi(UndergradMajor) |\n",
      "+==========================+=======================+\n",
      "| UndergradMajor(NOT_STEM) |                0.1576 |\n",
      "+--------------------------+-----------------------+\n",
      "| UndergradMajor(STEM)     |                0.8424 |\n",
      "+--------------------------+-----------------------+\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# %%capture\n",
    "\n",
    "for country in ['India', 'United Kingdom', 'United States']:\n",
    "    \n",
    "    p_country_um = inference.query(['UndergradMajor'], {'Country':country})\n",
    "    print(country)\n",
    "    print(p_country_um)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There's a certain **bias towards STEM degrees** since the Dataset comes from a Stack Overflow Dataset and the Website is usually aimed at people which deal with IT concepts.\n",
    "\n",
    "As we can see **Indian entries** are more prone to having pursued a **STEM degree**.\n",
    "\n",
    "On the other hand, entries from the **USA** are more inclined to **Humanities Majors compared to the other countries.**\n",
    "\n",
    "This point is quite interesting because it shows that people from the **United States** with a **NON-STEM Background** are more **engaged in programming** (Given The Dataset Website of Reference) than people from the other two countries."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*Given a **Country** what's the probability distribution of Bachelor's, Master's degres and Ph.Ds?*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "India\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding Elimination Order: : 100%|██████████| 2/2 [00:00<?, ?it/s]\n",
      "Eliminating: OpenSource: 100%|██████████| 2/2 [00:00<00:00, 235.08it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---------------------+------------------+\n",
      "| Education           |   phi(Education) |\n",
      "+=====================+==================+\n",
      "| Education(Bachelor) |           0.7363 |\n",
      "+---------------------+------------------+\n",
      "| Education(Master)   |           0.2614 |\n",
      "+---------------------+------------------+\n",
      "| Education(Ph.D)     |           0.0023 |\n",
      "+---------------------+------------------+\n",
      "United Kingdom\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding Elimination Order: : 100%|██████████| 2/2 [00:00<00:00, 4032.98it/s]\n",
      "Eliminating: OpenSource: 100%|██████████| 2/2 [00:00<?, ?it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---------------------+------------------+\n",
      "| Education           |   phi(Education) |\n",
      "+=====================+==================+\n",
      "| Education(Bachelor) |           0.6337 |\n",
      "+---------------------+------------------+\n",
      "| Education(Master)   |           0.2986 |\n",
      "+---------------------+------------------+\n",
      "| Education(Ph.D)     |           0.0677 |\n",
      "+---------------------+------------------+\n",
      "United States\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding Elimination Order: : 100%|██████████| 2/2 [00:00<?, ?it/s]\n",
      "Eliminating: OpenSource: 100%|██████████| 2/2 [00:00<00:00, 227.05it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---------------------+------------------+\n",
      "| Education           |   phi(Education) |\n",
      "+=====================+==================+\n",
      "| Education(Bachelor) |           0.7406 |\n",
      "+---------------------+------------------+\n",
      "| Education(Master)   |           0.2244 |\n",
      "+---------------------+------------------+\n",
      "| Education(Ph.D)     |           0.0350 |\n",
      "+---------------------+------------------+\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "for country in ['India', 'United Kingdom', 'United States']:\n",
    "    print(country)\n",
    "    p_country_ed = inference.query(['Education'], {'Country':country})\n",
    "    print(p_country_ed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "People from the **UK** are more likely to have a **Master's degree** while in the US most entries own Just a Bachelor's.\n",
    "\n",
    "Also **UK** has the highest number of **Ph.D** entries."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*Job Satisfaction Index comparison between People with a Lower Income based on wether they consider Coding a Hobby or not.*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%capture\n",
    "p_js_h_y = inference.query(['JobSatisfaction'], {'Hobby':'Yes', 'Salary':'0-250.000'})\n",
    "p_js_h_n = inference.query(['JobSatisfaction'], {'Hobby':'No', 'Salary':'0-250.000'})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*P(JobSatisfaction|Hobby=Yes,Salary=0-250.000)*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+--------------------+------------------------+\n",
      "| JobSatisfaction    |   phi(JobSatisfaction) |\n",
      "+====================+========================+\n",
      "| JobSatisfaction(0) |                 0.0334 |\n",
      "+--------------------+------------------------+\n",
      "| JobSatisfaction(1) |                 0.0929 |\n",
      "+--------------------+------------------------+\n",
      "| JobSatisfaction(2) |                 0.1029 |\n",
      "+--------------------+------------------------+\n",
      "| JobSatisfaction(3) |                 0.0583 |\n",
      "+--------------------+------------------------+\n",
      "| JobSatisfaction(4) |                 0.1336 |\n",
      "+--------------------+------------------------+\n",
      "| JobSatisfaction(5) |                 0.3749 |\n",
      "+--------------------+------------------------+\n",
      "| JobSatisfaction(6) |                 0.2041 |\n",
      "+--------------------+------------------------+\n"
     ]
    }
   ],
   "source": [
    "print(p_js_h_y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*P(JobSatisfaction|Hobby=No,Salary=0-250.000)*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+--------------------+------------------------+\n",
      "| JobSatisfaction    |   phi(JobSatisfaction) |\n",
      "+====================+========================+\n",
      "| JobSatisfaction(0) |                 0.0351 |\n",
      "+--------------------+------------------------+\n",
      "| JobSatisfaction(1) |                 0.0961 |\n",
      "+--------------------+------------------------+\n",
      "| JobSatisfaction(2) |                 0.1085 |\n",
      "+--------------------+------------------------+\n",
      "| JobSatisfaction(3) |                 0.0580 |\n",
      "+--------------------+------------------------+\n",
      "| JobSatisfaction(4) |                 0.1371 |\n",
      "+--------------------+------------------------+\n",
      "| JobSatisfaction(5) |                 0.3732 |\n",
      "+--------------------+------------------------+\n",
      "| JobSatisfaction(6) |                 0.1921 |\n",
      "+--------------------+------------------------+\n"
     ]
    }
   ],
   "source": [
    "print(p_js_h_n)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see results are quite similar.\n",
    "\n",
    "Proportionally entries who **consider Coding a Hobby** are **slightly more probable to be Extremely Satisfied** with their Job."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Augmentation Example\n",
    "\n",
    "This Bayesian Network could be quite useful in a Data Augmentation context.\n",
    "\n",
    "For example we might want to generate $n$ new British Entries which consider Coding a hobby and are moderately satisfied with their job.\n",
    "\n",
    "This can be useful in the context of balancing the number of entries in the Dataset for further Analysis.\n",
    "\n",
    "The Network can be used to generate the non-specified attributes using queries to obtain new parameters.\n",
    "\n",
    "$P(EveryOtherAttribute| Country=United Kingdom, Hobby=Yes)$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%capture\n",
    "inference_sampling=BayesianModelSampling(model)\n",
    "evidence = [State(key, val) for key,val in {'Country':'United Kingdom', 'Hobby':'Yes'}.items()]\n",
    "samples = inference_sampling.rejection_sample(evidence=evidence, size=30)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "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>Hobby</th>\n",
       "      <th>OpenSource</th>\n",
       "      <th>Country</th>\n",
       "      <th>Education</th>\n",
       "      <th>JobSatisfaction</th>\n",
       "      <th>UndergradMajor</th>\n",
       "      <th>Salary</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Master</td>\n",
       "      <td>6</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Master</td>\n",
       "      <td>6</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Master</td>\n",
       "      <td>5</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Ph.D</td>\n",
       "      <td>6</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>1</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>5</td>\n",
       "      <td>NOT_STEM</td>\n",
       "      <td>250.000-1.000.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Master</td>\n",
       "      <td>5</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Master</td>\n",
       "      <td>6</td>\n",
       "      <td>NOT_STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>4</td>\n",
       "      <td>STEM</td>\n",
       "      <td>250.000-1.000.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>0</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Ph.D</td>\n",
       "      <td>5</td>\n",
       "      <td>STEM</td>\n",
       "      <td>250.000-1.000.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>5</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Ph.D</td>\n",
       "      <td>6</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>5</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Master</td>\n",
       "      <td>4</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Ph.D</td>\n",
       "      <td>2</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Master</td>\n",
       "      <td>4</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>2</td>\n",
       "      <td>STEM</td>\n",
       "      <td>250.000-1.000.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>3</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>6</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>5</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>5</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>2</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Master</td>\n",
       "      <td>2</td>\n",
       "      <td>NOT_STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Master</td>\n",
       "      <td>6</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>0</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>5</td>\n",
       "      <td>STEM</td>\n",
       "      <td>250.000-1.000.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>2</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>5</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>United Kingdom</td>\n",
       "      <td>Bachelor</td>\n",
       "      <td>5</td>\n",
       "      <td>STEM</td>\n",
       "      <td>0-250.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Hobby OpenSource         Country Education  JobSatisfaction UndergradMajor  \\\n",
       "0    Yes         No  United Kingdom    Master                6           STEM   \n",
       "1    Yes        Yes  United Kingdom    Master                6           STEM   \n",
       "2    Yes         No  United Kingdom    Master                5           STEM   \n",
       "3    Yes         No  United Kingdom      Ph.D                6           STEM   \n",
       "4    Yes        Yes  United Kingdom  Bachelor                1           STEM   \n",
       "5    Yes         No  United Kingdom  Bachelor                5       NOT_STEM   \n",
       "6    Yes         No  United Kingdom    Master                5           STEM   \n",
       "7    Yes         No  United Kingdom    Master                6       NOT_STEM   \n",
       "8    Yes        Yes  United Kingdom  Bachelor                4           STEM   \n",
       "9    Yes         No  United Kingdom  Bachelor                0           STEM   \n",
       "10   Yes         No  United Kingdom      Ph.D                5           STEM   \n",
       "11   Yes        Yes  United Kingdom  Bachelor                5           STEM   \n",
       "12   Yes         No  United Kingdom      Ph.D                6           STEM   \n",
       "13   Yes         No  United Kingdom  Bachelor                5           STEM   \n",
       "14   Yes        Yes  United Kingdom    Master                4           STEM   \n",
       "15   Yes        Yes  United Kingdom      Ph.D                2           STEM   \n",
       "16   Yes        Yes  United Kingdom    Master                4           STEM   \n",
       "17   Yes        Yes  United Kingdom  Bachelor                2           STEM   \n",
       "18   Yes        Yes  United Kingdom  Bachelor                3           STEM   \n",
       "19   Yes        Yes  United Kingdom  Bachelor                6           STEM   \n",
       "20   Yes         No  United Kingdom  Bachelor                5           STEM   \n",
       "21   Yes         No  United Kingdom  Bachelor                5           STEM   \n",
       "22   Yes        Yes  United Kingdom  Bachelor                2           STEM   \n",
       "23   Yes         No  United Kingdom    Master                2       NOT_STEM   \n",
       "24   Yes         No  United Kingdom    Master                6           STEM   \n",
       "25   Yes         No  United Kingdom  Bachelor                0           STEM   \n",
       "26   Yes         No  United Kingdom  Bachelor                5           STEM   \n",
       "27   Yes         No  United Kingdom  Bachelor                2           STEM   \n",
       "28   Yes        Yes  United Kingdom  Bachelor                5           STEM   \n",
       "29   Yes        Yes  United Kingdom  Bachelor                5           STEM   \n",
       "\n",
       "               Salary  \n",
       "0           0-250.000  \n",
       "1           0-250.000  \n",
       "2           0-250.000  \n",
       "3           0-250.000  \n",
       "4           0-250.000  \n",
       "5   250.000-1.000.000  \n",
       "6           0-250.000  \n",
       "7           0-250.000  \n",
       "8   250.000-1.000.000  \n",
       "9           0-250.000  \n",
       "10  250.000-1.000.000  \n",
       "11          0-250.000  \n",
       "12          0-250.000  \n",
       "13          0-250.000  \n",
       "14          0-250.000  \n",
       "15          0-250.000  \n",
       "16          0-250.000  \n",
       "17  250.000-1.000.000  \n",
       "18          0-250.000  \n",
       "19          0-250.000  \n",
       "20          0-250.000  \n",
       "21          0-250.000  \n",
       "22          0-250.000  \n",
       "23          0-250.000  \n",
       "24          0-250.000  \n",
       "25          0-250.000  \n",
       "26  250.000-1.000.000  \n",
       "27          0-250.000  \n",
       "28          0-250.000  \n",
       "29          0-250.000  "
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "samples"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "c23b91227a54c2af61c5e03a530e77c3a20e29d5fbc858ef43b726817a78fb0f"
  },
  "kernelspec": {
   "display_name": "Python 3.10.2 ('FAIKR_venv')",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.2"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}