notebooks/4. Bayesian Networks.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Bayesian Networks"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook will show you how to query the main class of this package: `thomas.core.BayesianNetwork`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Python version: 3.8.10\n",
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n",
      "\n",
      "available imports:\n",
      "  import os\n",
      "  import logging\n",
      "  import pandas as pd\n",
      "  import numpy as np\n",
      "\n",
      "connect to this kernel with:\n",
      "  jupyter console --existing 4011a28c-a690-46f4-b0c4-c42d9a2f126b\n",
      "\n",
      "Could not create logging directory \"../logs\"\n",
      "Logging to: \"../logs/notebook.log\"\n",
      "Current date/time: 11-06-2021, 21:29\n",
      "Current working directory: \"/Users/melle/software-development/thomas-master/notebooks\"\n"
     ]
    }
   ],
   "source": [
    "%run '_preamble.ipynb'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from thomas.core.models.bn import BayesianNetwork\n",
    "from thomas.core import examples\n",
    "from thomas.core.reader import oobn\n",
    "\n",
    "from IPython.display import display, HTML"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Student example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<BayesianNetwork name='Student'>\n",
       "  <Node RV='I' description='Intelligence' states=['i0', 'i1'] />\n",
       "  <Node RV='S' description='SAT Score' states=['s0', 's1'] />\n",
       "  <Node RV='D' description='Difficulty' states=['d0', 'd1'] />\n",
       "  <Node RV='G' description='Grade' states=['g1', 'g2', 'g3'] />\n",
       "  <Node RV='L' description='Letter' states=['l0', 'l1'] />\n",
       "</BayesianNetwork>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load the Student Network from the examples\n",
    "Gs = examples.get_student_network()\n",
    "Gs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Show the network and its relations\n",
    "Gs.draw()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DiscreteNetworkNode('S', name='S', states=['s0', 's1'], description='SAT Score')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Nodes can be accessed by using the BN as a `dict`:\n",
    "Gs['S']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['s0', 's1']"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The Node-property 'states' will return a list of allowed states.\n",
    "Gs['S'].states"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "            <div>\n",
       "                <div style=\"margin-top:6px\">\n",
       "                    <span><b>P(S|I)</b></span>\n",
       "                    <span style=\"font-style: italic;\">SAT Score</span>\n",
       "                    <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>S</th>\n",
       "      <th>s0</th>\n",
       "      <th>s1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>i0</th>\n",
       "      <td>0.95</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>i1</th>\n",
       "      <td>0.20</td>\n",
       "      <td>0.80</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "                </div>\n",
       "            </div>\n",
       "        "
      ],
      "text/plain": [
       "P(S|I)\n",
       "I   S \n",
       "i0  s0    0.95\n",
       "    s1    0.05\n",
       "i1  s0    0.20\n",
       "    s1    0.80\n",
       "dtype: float64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Accessing individual Nodes/CPTs is easy\n",
    "Gs['S'].cpt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'I': factor(I)\n",
       " I \n",
       " i0    0.7\n",
       " i1    0.3\n",
       " dtype: float64,\n",
       " 'S': factor(S)\n",
       " S \n",
       " s0    0.725\n",
       " s1    0.275\n",
       " dtype: float64,\n",
       " 'D': factor(D)\n",
       " D \n",
       " d0    0.6\n",
       " d1    0.4\n",
       " dtype: float64}"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Querying the network is also simple. To compute the marginals over I, S and D:\n",
    "Gs.compute_marginals(['I', 'S', 'D'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'I': factor(I)\n",
       " I \n",
       " i0    1.0\n",
       " i1    0.0\n",
       " dtype: float64,\n",
       " 'S': factor(S)\n",
       " S \n",
       " s0    0.95\n",
       " s1    0.05\n",
       " dtype: float64,\n",
       " 'D': factor(D)\n",
       " D \n",
       " d0    0.6\n",
       " d1    0.4\n",
       " dtype: float64}"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# To compute the marginals over I, S and D given evidence I=i0:\n",
    "Gs.compute_marginals(['I', 'S', 'D'], {'I':'i0'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.91724138, 0.12727273])"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Complex queries can be run simply by providing the query as string:\n",
    "Gs.P('I=i0|S')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "            <div>\n",
       "                <div style=\"margin-top:6px\">\n",
       "                    <span><b>P(S|I)</b></span>\n",
       "                    <span style=\"font-style: italic;\"></span>\n",
       "                    <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>S</th>\n",
       "      <th>s0</th>\n",
       "      <th>s1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>i0</th>\n",
       "      <td>0.95</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>i1</th>\n",
       "      <td>0.20</td>\n",
       "      <td>0.80</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "                </div>\n",
       "            </div>\n",
       "        "
      ],
      "text/plain": [
       "P(S|I)\n",
       "I   S \n",
       "i0  s0    0.95\n",
       "    s1    0.05\n",
       "i1  s0    0.20\n",
       "    s1    0.80\n",
       "dtype: float64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# This should be equal to Gs['S']\n",
    "Gs.P('S|I')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "            <div>\n",
       "                <div style=\"margin-top:6px\">\n",
       "                    <span><b>P(S|I)</b></span>\n",
       "                    <span style=\"font-style: italic;\">SAT Score</span>\n",
       "                    <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>S</th>\n",
       "      <th>s0</th>\n",
       "      <th>s1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>i0</th>\n",
       "      <td>0.95</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>i1</th>\n",
       "      <td>0.20</td>\n",
       "      <td>0.80</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "                </div>\n",
       "            </div>\n",
       "        "
      ],
      "text/plain": [
       "P(S|I)\n",
       "I   S \n",
       "i0  s0    0.95\n",
       "    s1    0.05\n",
       "i1  s0    0.20\n",
       "    s1    0.80\n",
       "dtype: float64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Gs.S.cpt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "            <div>\n",
       "                <div style=\"margin-top:6px\">\n",
       "                    <span><b>P(I|S)</b></span>\n",
       "                    <span style=\"font-style: italic;\"></span>\n",
       "                    <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>I</th>\n",
       "      <th>i0</th>\n",
       "      <th>i1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>s0</th>\n",
       "      <td>0.9172</td>\n",
       "      <td>0.0828</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>s1</th>\n",
       "      <td>0.1273</td>\n",
       "      <td>0.8727</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "                </div>\n",
       "            </div>\n",
       "        "
      ],
      "text/plain": [
       "P(I|S)\n",
       "S   I \n",
       "s0  i0    0.9172\n",
       "    i1    0.0828\n",
       "s1  i0    0.1273\n",
       "    i1    0.8727\n",
       "dtype: float64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Alternatively, parameters can be specified separately through \n",
    "# `compute_posterior()`:\n",
    "Gs.compute_posterior(qd=['I'], qv={}, ed=['S'], ev={})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "type(unstacked): <class 'pandas.core.frame.DataFrame'>\n",
      "\n"
     ]
    },
    {
     "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>I</th>\n",
       "      <th>i0</th>\n",
       "      <th>i1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>s0</th>\n",
       "      <td>0.917241</td>\n",
       "      <td>0.082759</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>s1</th>\n",
       "      <td>0.127273</td>\n",
       "      <td>0.872727</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "I         i0        i1\n",
       "S                     \n",
       "s0  0.917241  0.082759\n",
       "s1  0.127273  0.872727"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# A CPT can be converted to a pandas.DataFrame by calling `unstack()` without\n",
    "# parameters. Note that this only works if there are conditioning variables.\n",
    "unstacked = Gs.P('I|S').as_series().unstack()\n",
    "print(f'type(unstacked): {type(unstacked)}')\n",
    "print()\n",
    "display(unstacked)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# The junction tree (jointree) can be inspected by calling `jt.draw()`:\n",
    "Gs.jt.draw()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Importing Bayesian Networks"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`thomas` supports loading Bayesian Networks from different file formats:\n",
    "* `.net`: the Netica file format\n",
    "* `.oobn`: the Hugin Object Oriented Bayesian Network file format\n",
    "* `.json`: thomas' custom, JSON based format.\n",
    "\n",
    "As an example, we'll open an `oobn` file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "from thomas.core.reader import oobn, net"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "factor(death)\n",
       "death      \n",
       "0-30 days      0.0829\n",
       "1-4 months     0.2124\n",
       "4-6 months     0.0988\n",
       "6-12 months    0.1992\n",
       "1-2 years      0.1810\n",
       "> 2 years      0.2257\n",
       "dtype: float64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read a file in OOBN format\n",
    "filename = thomas.core.get_pkg_filename('lungcancer.oobn')\n",
    "bn = oobn.read(filename)\n",
    "\n",
    "bn.P('death').as_factor()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "factor(death)\n",
       "death      \n",
       "0-30 days      0.0829\n",
       "1-4 months     0.2124\n",
       "4-6 months     0.0988\n",
       "6-12 months    0.1992\n",
       "1-2 years      0.1810\n",
       "> 2 years      0.2257\n",
       "dtype: float64"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read a file in NET format\n",
    "filename = thomas.core.get_pkg_filename('lungcancer.net')\n",
    "bn = net.read(filename)\n",
    "\n",
    "bn.P('death').as_factor()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "            <div>\n",
       "                <div style=\"margin-top:6px\">\n",
       "                    <span><b>P(asia)</b></span>\n",
       "                    <span style=\"font-style: italic;\"></span>\n",
       "                    <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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th>asia</th>\n",
       "      <th>yes</th>\n",
       "      <th>no</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.99</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "                </div>\n",
       "            </div>\n",
       "        "
      ],
      "text/plain": [
       "P(asia)\n",
       "asia\n",
       "yes     0.01\n",
       "no      0.99\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Download and read another file in NET format\n",
    "import requests\n",
    "import gzip\n",
    "\n",
    "filename = 'asia.net'\n",
    "\n",
    "if not os.path.exists(filename):\n",
    "    url = 'https://www.bnlearn.com/bnrepository/asia/asia.net.gz'\n",
    "    target_path = 'asia.net.gz'\n",
    "    print(f'Current directory: \"{os.getcwd()}\"')\n",
    "    print(f'Downloading \"{target_path}\" from \"{url}\"')\n",
    "\n",
    "    response = requests.get(url, stream=True)\n",
    "    if response.status_code == 200:\n",
    "        with open(target_path, 'wb') as f:\n",
    "            f.write(response.raw.read())\n",
    "\n",
    "        print(f'Unzipping \"{target_path}\"')\n",
    "        data = gzip.open(target_path)\n",
    "\n",
    "        print(f'Saving \"{filename}\"')\n",
    "        with open(filename, 'wb') as fp:\n",
    "            fp.write(data.read())\n",
    "\n",
    "    else:\n",
    "        print('ERROR: Could not download file!?')\n",
    "\n",
    "bn = net.read(filename)\n",
    "display(bn.P('asia'))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "thomas-jupyter3",
   "language": "python",
   "name": "thomas-jupyter3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}