Youth_Unemployment_Bayesian_Network.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Youth Unemployment Bayesian Network\n",
    "\n",
    "Matteo Antonio Inajetovic 0000949648"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from pgmpy.factors.discrete import State\n",
    "from pgmpy.factors.discrete import TabularCPD\n",
    "from pgmpy.models import BayesianModel\n",
    "from pgmpy.inference import VariableElimination\n",
    "from pgmpy.sampling import BayesianModelSampling\n",
    "import networkx as nx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPDs:\n",
      "+-----------+-----+\n",
      "| Gender(0) | 0.5 |\n",
      "+-----------+-----+\n",
      "| Gender(1) | 0.5 |\n",
      "+-----------+-----+ \n",
      "\n",
      " +--------+-----+\n",
      "| N/S(0) | 0.5 |\n",
      "+--------+-----+\n",
      "| N/S(1) | 0.5 |\n",
      "+--------+-----+ \n",
      "\n",
      " +-----------------------+--------+--------+\n",
      "| N/S                   | N/S(0) | N/S(1) |\n",
      "+-----------------------+--------+--------+\n",
      "| Economic_Situation(0) | 0.06   | 0.1    |\n",
      "+-----------------------+--------+--------+\n",
      "| Economic_Situation(1) | 0.94   | 0.9    |\n",
      "+-----------------------+--------+--------+ \n",
      "\n",
      " +--------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+\n",
      "| Gender             | Gender(0)             | Gender(0)             | Gender(0)             | Gender(0)             | Gender(1)             | Gender(1)             | Gender(1)             | Gender(1)             |\n",
      "+--------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+\n",
      "| Economic_Situation | Economic_Situation(0) | Economic_Situation(0) | Economic_Situation(1) | Economic_Situation(1) | Economic_Situation(0) | Economic_Situation(0) | Economic_Situation(1) | Economic_Situation(1) |\n",
      "+--------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+\n",
      "| N/S                | N/S(0)                | N/S(1)                | N/S(0)                | N/S(1)                | N/S(0)                | N/S(1)                | N/S(0)                | N/S(1)                |\n",
      "+--------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+\n",
      "| Graduate(0)        | 0.97                  | 0.98                  | 0.89                  | 0.93                  | 0.96                  | 0.97                  | 0.86                  | 0.9                   |\n",
      "+--------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+\n",
      "| Graduate(1)        | 0.03                  | 0.02                  | 0.11                  | 0.07                  | 0.04                  | 0.03                  | 0.14                  | 0.1                   |\n",
      "+--------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+-----------------------+ \n",
      "\n",
      " +---------------+-------------+-------------+-------------+-------------+\n",
      "| Gender        | Gender(0)   | Gender(0)   | Gender(1)   | Gender(1)   |\n",
      "+---------------+-------------+-------------+-------------+-------------+\n",
      "| Graduate      | Graduate(0) | Graduate(1) | Graduate(0) | Graduate(1) |\n",
      "+---------------+-------------+-------------+-------------+-------------+\n",
      "| Unemployed(0) | 0.78        | 0.94        | 0.75        | 0.9         |\n",
      "+---------------+-------------+-------------+-------------+-------------+\n",
      "| Unemployed(1) | 0.22        | 0.06        | 0.25        | 0.1         |\n",
      "+---------------+-------------+-------------+-------------+-------------+ \n",
      "\n",
      "\n",
      "Let's check if the sum of the probabilities for each state is equal to 1 and wheteher the CPDs associated with nodes are consistent with their parents:\n",
      "True or False? True\n"
     ]
    }
   ],
   "source": [
    "yu_net = BayesianModel([('Gender', 'Graduate'), ('Gender', 'Unemployed'), ('N/S','Graduate'),('N/S', 'Economic_Situation'),\n",
    "                        ('Economic_Situation', 'Graduate'), ('Graduate', 'Unemployed')])\n",
    "\n",
    "cpd_gender= TabularCPD('Gender', 2, [[.5], [.5]])\n",
    "cpd_ns= TabularCPD('N/S', 2, [[.5], [.5]])\n",
    "cpd_situation= TabularCPD('Economic_Situation', 2, ([.06, .1],[.94, .9]), evidence=['N/S'], evidence_card= [2]) \n",
    "cpd_graduate= TabularCPD('Graduate', 2, ([.97,  .98, .89, .93, .96, .97, .86, .9],[.03, .02, .11, .07, .04, .03, .14, .10]),\n",
    "                      evidence=['Gender','Economic_Situation','N/S'], evidence_card=[2,2,2]) \n",
    "cpd_unemployed= TabularCPD('Unemployed', 2, ([.78, .94, .75, .9],[.22, .06,.25, .1]),\n",
    "                        evidence= ['Gender', 'Graduate'], evidence_card= [2,2])\n",
    "\n",
    "yu_net.add_cpds(cpd_gender, cpd_ns, cpd_situation, cpd_graduate, cpd_unemployed)\n",
    "print(\"CPDs:\")\n",
    "print(cpd_gender, '\\n\\n', cpd_ns, '\\n\\n', cpd_situation, '\\n\\n', cpd_graduate, '\\n\\n',cpd_unemployed, '\\n\\n')\n",
    "print(\"Let's check if the sum of the probabilities for each state is equal to 1 and wheteher the CPDs associated with nodes are consistent with their parents:\")\n",
    "print(\"True or False?\",yu_net.check_model())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\39345\\Anaconda3\\lib\\site-packages\\networkx\\drawing\\nx_pylab.py:579: MatplotlibDeprecationWarning: \n",
      "The iterable function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use np.iterable instead.\n",
      "  if not cb.iterable(width):\n",
      "C:\\Users\\39345\\Anaconda3\\lib\\site-packages\\networkx\\drawing\\nx_pylab.py:676: MatplotlibDeprecationWarning: \n",
      "The iterable function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use np.iterable instead.\n",
      "  if cb.iterable(node_size):  # many node sizes\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "np.random.seed(14)\n",
    "nx.draw(yu_net, with_labels = True, node_size = 2500, node_color = 'y', node_shape = 'o')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Independence relations between the net's variables:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(Gender _|_ N/S, Economic_Situation)\n",
       "(Gender _|_ Economic_Situation | N/S)\n",
       "(Gender _|_ N/S | Economic_Situation)\n",
       "(Unemployed _|_ N/S, Economic_Situation | Gender, Graduate)\n",
       "(Unemployed _|_ Economic_Situation | N/S, Gender, Graduate)\n",
       "(Unemployed _|_ N/S | Gender, Economic_Situation, Graduate)\n",
       "(N/S _|_ Gender)\n",
       "(N/S _|_ Gender | Economic_Situation)\n",
       "(N/S _|_ Unemployed | Gender, Graduate)\n",
       "(N/S _|_ Unemployed | Gender, Economic_Situation, Graduate)\n",
       "(Economic_Situation _|_ Gender)\n",
       "(Economic_Situation _|_ Gender | N/S)\n",
       "(Economic_Situation _|_ Unemployed | Gender, Graduate)\n",
       "(Economic_Situation _|_ Unemployed | N/S, Gender, Graduate)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(\"Independence relations between the net's variables:\")\n",
    "yu_net.get_independencies() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Local independencies:\n",
      " (Gender _|_ N/S, Economic_Situation)\n",
      "(Unemployed _|_ N/S, Economic_Situation | Gender, Graduate)\n",
      "(N/S _|_ Gender)\n",
      "(Economic_Situation _|_ Gender | N/S)\n"
     ]
    }
   ],
   "source": [
    "print('Local independencies:\\n', \n",
    "      yu_net.local_independencies(list(yu_net.nodes())))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Couples of parents in the eventual V-structures present in the net:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{('Economic_Situation', 'Gender'), ('Gender', 'N/S')}"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(\"Couples of parents in the eventual V-structures present in the net:\")\n",
    "yu_net.get_immoralities() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Given some evidences we can see whether a trail is active or not: \n",
      "Is the trail ('Gender', 'N/S') active?\n",
      "False\n",
      "..but, if we add the evidence 'Graduate':\n",
      "True\n",
      "Is the trail ('Economic_Situation', 'Gender') active?\n",
      "False\n",
      "..but, if we add the evidence 'Unemployed':\n",
      "True\n"
     ]
    }
   ],
   "source": [
    "print(\"Given some evidences we can see whether a trail is active or not: \")\n",
    "print(\"Is the trail ('Gender', 'N/S') active?\")\n",
    "print(yu_net.is_active_trail('Gender', 'N/S')) \n",
    "print(\"..but, if we add the evidence 'Graduate':\")\n",
    "print(yu_net.is_active_trail('Gender', 'N/S',observed = ['Graduate']))\n",
    "print(\"Is the trail ('Economic_Situation', 'Gender') active?\")\n",
    "print(yu_net.is_active_trail('Gender', 'Economic_Situation')) \n",
    "print(\"..but, if we add the evidence 'Unemployed':\")\n",
    "print(yu_net.is_active_trail('Gender', 'Economic_Situation',observed = ['Unemployed']))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Nodes belonging to the Markov Blanket of \"Graduate\":\n",
      " ['Gender', 'N/S', 'Economic_Situation', 'Unemployed']\n"
     ]
    }
   ],
   "source": [
    "print('Nodes belonging to the Markov Blanket of \"Graduate\":\\n',\n",
    "      yu_net.get_markov_blanket('Graduate')) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exact Inference - Variable Elimination"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now i'm going to apply an exact inference method to obtain new informations from the net, the **Variable Elimination** algorithm:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding Elimination Order: : : 0it [00:00, ?it/s]\n",
      "0it [00:00, ?it/s]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Map_query execution to find the most probable combination of variables given 'Unemployed' = True:\n",
      "{'Gender': 1, 'N/S': 1, 'Economic_Situation': 1, 'Graduate': 0}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "yu_inference = VariableElimination(yu_net)\n",
    "print(\"Map_query execution to find the most probable combination of variables given 'Unemployed' = True:\")\n",
    "print(yu_inference.map_query(['Gender','N/S','Economic_Situation','Graduate'],{'Unemployed':1}))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "  0%|                                                                                            | 0/4 [00:00<?, ?it/s]\u001b[A\n",
      "Finding Elimination Order: :   0%|                                                               | 0/4 [00:00<?, ?it/s]\u001b[A\n",
      "\n",
      "  0%|                                                                                            | 0/4 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Eliminating: Economic_Situation:   0%|                                                           | 0/4 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Eliminating: N/S:   0%|                                                                          | 0/4 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "P(Unemployed)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[A\u001b[A\n",
      "\n",
      "Eliminating: Graduate:   0%|                                                                     | 0/4 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Eliminating: Gender: 100%|██████████████████████████████████████████████████████████████| 4/4 [00:00<00:00, 161.92it/s]\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---------------+-------------------+\n",
      "| Unemployed    |   phi(Unemployed) |\n",
      "+===============+===================+\n",
      "| Unemployed(0) |            0.7803 |\n",
      "+---------------+-------------------+\n",
      "| Unemployed(1) |            0.2197 |\n",
      "+---------------+-------------------+\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "print('P(Unemployed)')\n",
    "q = yu_inference.query(['Unemployed'])\n",
    "print(q)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Causal Reasoning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's evaluate some effects given some causes:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "P(Unemployed| Gender = True, N/S = True)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "  0%|                                                                                            | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Finding Elimination Order: :   0%|                                                               | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "Eliminating: Economic_Situation:   0%|                                                           | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "Eliminating: Graduate: 100%|████████████████████████████████████████████████████████████| 2/2 [00:00<00:00, 178.69it/s]\u001b[A\u001b[A\u001b[A\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---------------+-------------------+\n",
      "| Unemployed    |   phi(Unemployed) |\n",
      "+===============+===================+\n",
      "| Unemployed(0) |            0.7640 |\n",
      "+---------------+-------------------+\n",
      "| Unemployed(1) |            0.2361 |\n",
      "+---------------+-------------------+\n",
      "P(Unemployed| Graduate = True, Economic Situation = True )\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "Finding Elimination Order: :   0%|                                                               | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "Finding Elimination Order: : 100%|███████████████████████████████████████████████████████| 4/4 [00:40<00:00, 10.05s/it]\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "Finding Elimination Order: : 100%|███████████████████████████████████████████████████████| 2/2 [01:20<00:00, 40.07s/it]\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "  0%|                                                                                            | 0/2 [00:00<?, ?it/s]\u001b[A\n",
      "Eliminating: N/S:   0%|                                                                          | 0/2 [00:00<?, ?it/s]\u001b[A\n",
      "Eliminating: Gender: 100%|███████████████████████████████████████████████████████████████| 2/2 [00:00<00:00, 52.15it/s]\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+---------------+-------------------+\n",
      "| Unemployed    |   phi(Unemployed) |\n",
      "+===============+===================+\n",
      "| Unemployed(0) |            0.9172 |\n",
      "+---------------+-------------------+\n",
      "| Unemployed(1) |            0.0828 |\n",
      "+---------------+-------------------+\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "print('P(Unemployed| Gender = True, N/S = True)')\n",
    "q1_a = yu_inference.query(['Unemployed'],{'Gender':1,'N/S':1})\n",
    "print(q1_a)\n",
    "print('P(Unemployed| Graduate = True, Economic Situation = True )')\n",
    "q1_b = yu_inference.query(['Unemployed'],{'Graduate':1, 'Economic_Situation':1})\n",
    "print(q1_b)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evidential Reasoning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's evaluate some causes given the effects:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With the first query, we'll evaluate the probability for a graduated person to be a female or a male."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/3 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Finding Elimination Order: :   0%|                                                               | 0/3 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/3 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Eliminating: Unemployed:   0%|                                                                   | 0/3 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Eliminating: Economic_Situation:   0%|                                                           | 0/3 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Eliminating: N/S: 100%|██████████████████████████████████████████████████████████████████| 3/3 [00:00<00:00, 86.70it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "P(Gender| Graduate = True\n",
      "+-----------+---------------+\n",
      "| Gender    |   phi(Gender) |\n",
      "+===========+===============+\n",
      "| Gender(0) |        0.4285 |\n",
      "+-----------+---------------+\n",
      "| Gender(1) |        0.5715 |\n",
      "+-----------+---------------+\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "print('P(Gender| Graduate = True')\n",
    "q2 = yu_inference.query(['Gender'],{'Graduate':1})\n",
    "print(q2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we'll see the probability for a person to be graduated, given that he's employed as evidence:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/3 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Finding Elimination Order: :   0%|                                                               | 0/3 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/3 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Eliminating: Economic_Situation:   0%|                                                           | 0/3 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Eliminating: N/S:   0%|                                                                          | 0/3 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Eliminating: Gender: 100%|██████████████████████████████████████████████████████████████| 3/3 [00:00<00:00, 250.63it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "P(Graduated| Unemployed = False )\n",
      "+-------------+-----------------+\n",
      "| Graduate    |   phi(Graduate) |\n",
      "+=============+=================+\n",
      "| Graduate(0) |          0.8833 |\n",
      "+-------------+-----------------+\n",
      "| Graduate(1) |          0.1167 |\n",
      "+-------------+-----------------+\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "print('P(Graduated| Unemployed = False )')\n",
    "q3 = yu_inference.query(['Graduate'],{'Unemployed':0})\n",
    "print(q3) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Intercausal Reasoning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Queries about mutual causes of a given effect:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the following query, we can see how given a graduated person, it has more probability to be a woman rather than a man and to come from North Italy rather than South Italy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Finding Elimination Order: :   0%|                                                               | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Eliminating: Economic_Situation:   0%|                                                           | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Eliminating: Unemployed: 100%|██████████████████████████████████████████████████████████| 2/2 [00:00<00:00, 334.35it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "P(Gender,N/S| Graduate = True)\n",
      "+-----------+--------+-------------------+\n",
      "| Gender    | N/S    |   phi(Gender,N/S) |\n",
      "+===========+========+===================+\n",
      "| Gender(0) | N/S(0) |            0.2649 |\n",
      "+-----------+--------+-------------------+\n",
      "| Gender(0) | N/S(1) |            0.1636 |\n",
      "+-----------+--------+-------------------+\n",
      "| Gender(1) | N/S(0) |            0.3374 |\n",
      "+-----------+--------+-------------------+\n",
      "| Gender(1) | N/S(1) |            0.2341 |\n",
      "+-----------+--------+-------------------+\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "print('P(Gender,N/S| Graduate = True)')\n",
    "q4 = yu_inference.query(['Gender','N/S'],{'Graduate':1})\n",
    "print(q4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The sequent query tells us that an unemployed person is more likely to be nongraduate than graduate and female rather than male."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Finding Elimination Order: :   0%|                                                               | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Eliminating: Economic_Situation:   0%|                                                           | 0/2 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "Eliminating: N/S: 100%|█████████████████████████████████████████████████████████████████| 2/2 [00:00<00:00, 286.02it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "P(Gender,Graduate| Unemployed = True)\n",
      "+-----------+-------------+------------------------+\n",
      "| Gender    | Graduate    |   phi(Gender,Graduate) |\n",
      "+===========+=============+========================+\n",
      "| Gender(0) | Graduate(0) |                 0.4581 |\n",
      "+-----------+-------------+------------------------+\n",
      "| Gender(0) | Graduate(1) |                 0.0116 |\n",
      "+-----------+-------------+------------------------+\n",
      "| Gender(1) | Graduate(0) |                 0.5044 |\n",
      "+-----------+-------------+------------------------+\n",
      "| Gender(1) | Graduate(1) |                 0.0258 |\n",
      "+-----------+-------------+------------------------+\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "print('P(Gender,Graduate| Unemployed = True)')\n",
    "q5 = yu_inference.query(['Gender','Graduate'],{'Unemployed':1})\n",
    "print(q5) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Approximate Inference"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Variable elimination, in the worst case, can even have exponential time and space complexity.\n",
    "Thus, in large and multi connected networks, it is better to consider approximate inference methods rather than the exact ones. \n",
    "Let's see the behaviour of **Likelihood Weighting** and **Rejection sampling** with our net:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Rather than Rejection Sampling, Likelihood Weighting only generates conistent events (w.r.t. the evidences) and uses all the samples generated. Therefore, it can be much more efficient. Nevertheless, increasing the number of evidence variables, its performance can get worse."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Finding Elimination Order: :   0%|                                                               | 0/2 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/2 [00:00<?, ?it/s]\u001b[A\n",
      "Eliminating: Economic_Situation:   0%|                                                           | 0/2 [00:00<?, ?it/s]\u001b[A\n",
      "Eliminating: Graduate: 100%|█████████████████████████████████████████████████████████████| 2/2 [00:00<00:00, 78.78it/s]\u001b[A\n"
     ]
    }
   ],
   "source": [
    "inference   = BayesianModelSampling(yu_net)\n",
    "evidence    = [State('Gender', 1),State('N/S', 1)]\n",
    "punemployed = yu_inference.query(['Unemployed'],{'Gender':1,'N/S':1}).values[1] #reference value\n",
    "\n",
    "def prob_LW(samples,variable):\n",
    "    return round(np.sum(np.dot(samples[variable],samples['_weight']))/np.sum(samples['_weight']),2)\n",
    "\n",
    "def absolute_error(exact_value,approx_value):\n",
    "    return np.absolute(exact_value-approx_value)\n",
    "\n",
    "def run_experiment(sample_size):\n",
    "    # Sample\n",
    "    samples_LW = inference.likelihood_weighted_sample(evidence = evidence, size=sample_size, return_type='recarray')\n",
    "    samples_RS = inference.rejection_sample(evidence=evidence, size=sample_size, return_type='recarray')\n",
    "\n",
    "    # Statistics\n",
    "    punemployed_LW = prob_LW(samples_LW,'Unemployed')\n",
    "    punemployed_RS = np.recarray.mean(samples_RS['Unemployed'], axis=0)\n",
    "    \n",
    "    \n",
    "    # Return results\n",
    "    return np.array([(sample_size, \n",
    "                      punemployed_RS, absolute_error(punemployed,punemployed_RS),\n",
    "                      punemployed_LW, absolute_error(punemployed,punemployed_LW))],\n",
    "                    dtype=[('sample_size', '<i8'), \n",
    "                           ('punemployed_RS', '<f8'),('e_unemployed_RS','<f8'),\n",
    "                           ('punemployed_LW', '<f8'),('e_unemployed_LW','<f8')])\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "  0%|                                                                                           | 0/10 [00:00<?, ?it/s]\u001b[A\n",
      "\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 69.12it/s]\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:00<00:00, 39.46it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 34.86it/s]\u001b[A\u001b[A\n",
      "\n",
      "27it [00:00, 106.92it/s]                                                                                               \u001b[A\n",
      "\n",
      "  0%|                                                                                           | 0/16 [00:00<?, ?it/s]\u001b[A\n",
      "\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Unemployed: 100%|██████████████████████████████████████████████████| 5/5 [00:00<00:00, 147.70it/s]\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Unemployed: 100%|██████████████████████████████████████████████████| 5/5 [00:00<00:00, 147.13it/s]\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\u001b[A\n",
      "\n",
      "Finding Elimination Order: : 100%|███████████████████████████████████████████████████████| 2/2 [00:30<00:00, 15.45s/it]\u001b[A\u001b[A\n",
      "\n",
      "\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:40<00:45, 15.02s/it]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:40<00:45, 15.02s/it]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Gender:  60%|█████████████████████████████████                      | 3/5 [00:40<00:27, 13.52s/it]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Graduate:  60%|███████████████████████████████▊                     | 3/5 [00:40<00:27, 13.52s/it]\u001b[A\u001b[A\n",
      "\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:40<00:00,  8.02s/it]\u001b[A\u001b[A\n",
      "\n",
      "23it [00:40,  1.75s/it]                                                                                                \u001b[A\n",
      "  0%|                                                                                           | 0/26 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 67.78it/s]\u001b[A\n",
      "\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 72.97it/s]\u001b[A\n",
      " 92%|██████████████████████████████████████████████████████████████████████████▊      | 24/26 [00:00<00:00, 142.08it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 80.77it/s]\u001b[A\n",
      "27it [00:00, 111.67it/s]                                                                                               \n",
      "  0%|                                                                                           | 0/42 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 76.63it/s]\u001b[A\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Unemployed:   0%|                                                           | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 38.15it/s]\u001b[A\n",
      "45it [00:00, 194.00it/s]                                                                                               \n",
      "  0%|                                                                                           | 0/69 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 67.62it/s]\u001b[A\n",
      "\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:00<00:00, 38.13it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 30.42it/s]\u001b[A\n",
      "82it [00:00, 305.58it/s]                                                                                               \n",
      "  0%|                                                                                          | 0/112 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:00<00:00, 39.17it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 33.26it/s]\u001b[A\n",
      " 33%|██████████████████████████▍                                                     | 37/112 [00:00<00:00, 214.30it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:00, 19.90it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:00, 19.90it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:00, 19.90it/s]\u001b[A\n",
      "Generating for node: Unemployed:  40%|████████████████████▍                              | 2/5 [00:00<00:00, 19.90it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 17.45it/s]\u001b[A\n",
      "172it [00:00, 365.86it/s]                                                                                              \n",
      "  0%|                                                                                          | 0/183 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:00<00:00, 30.69it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 24.45it/s]\u001b[A\n",
      " 34%|███████████████████████████                                                     | 62/183 [00:00<00:00, 285.98it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:00, 16.17it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:00, 16.17it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:00, 16.17it/s]\u001b[A\n",
      "Generating for node: Unemployed:  40%|████████████████████▍                              | 2/5 [00:00<00:00, 16.17it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 18.64it/s]\u001b[A\n",
      "254it [00:00, 508.90it/s]                                                                                              \n",
      "  0%|                                                                                          | 0/297 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Unemployed:   0%|                                                           | 0/5 [00:00<?, ?it/s]\u001b[A\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 43.49it/s]\u001b[A\n",
      " 37%|█████████████████████████████▌                                                 | 111/297 [00:00<00:00, 924.90it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:  60%|█████████████████████████████████                      | 3/5 [00:00<00:00, 28.52it/s]\u001b[A\n",
      "Generating for node: Graduate:  60%|███████████████████████████████▊                     | 3/5 [00:00<00:00, 28.52it/s]\u001b[A\n",
      "Generating for node: Unemployed:  60%|██████████████████████████████▌                    | 3/5 [00:00<00:00, 28.52it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 18.36it/s]\u001b[A\n",
      "396it [00:00, 975.58it/s]                                                                                              \n",
      "  0%|                                                                                          | 0/483 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Gender:   0%|                                                               | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:   0%|                                                             | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:00<00:00, 34.17it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 25.28it/s]\u001b[A\n",
      " 35%|███████████████████████████▉                                                   | 171/483 [00:00<00:00, 846.74it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:00, 10.59it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:00, 10.59it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:00, 10.59it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:00<00:00, 10.67it/s]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:00<00:00, 10.67it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00,  9.17it/s]\u001b[A\n",
      "677it [00:00, 881.36it/s]                                                                                              \n",
      "  0%|                                                                                          | 0/784 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:00, 19.22it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:00, 19.22it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:00, 19.22it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:00<00:00, 17.73it/s]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:00<00:00, 17.73it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 13.16it/s]\u001b[A\n",
      " 39%|██████████████████████████████▊                                                | 306/784 [00:00<00:00, 769.06it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:00, 10.25it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:00, 10.25it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:00, 10.25it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:00<00:00,  9.95it/s]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:00<00:00,  9.95it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00,  8.19it/s]\u001b[A\n",
      "1032it [00:01, 1000.57it/s]                                                                                            \n",
      "  0%|                                                                                         | 0/1274 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:00, 17.29it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:00, 17.29it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:00, 17.29it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:00<00:00, 15.27it/s]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:00<00:00, 15.27it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00, 11.12it/s]\u001b[A\n",
      " 35%|███████████████████████████                                                   | 442/1274 [00:00<00:00, 941.33it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:00,  5.61it/s]\u001b[A\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:00,  5.61it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:00,  5.61it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:00<00:00,  5.41it/s]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:00<00:00,  5.41it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:01<00:00,  4.46it/s]\u001b[A\n",
      "1770it [00:01, 1104.09it/s]                                                                                            \n",
      "  0%|                                                                                         | 0/2069 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:00,  9.27it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:00,  9.27it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:00,  9.27it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:00<00:00,  9.24it/s]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:00<00:00,  9.24it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:00<00:00,  7.73it/s]\u001b[A\n",
      " 36%|████████████████████████████                                                 | 754/2069 [00:00<00:01, 1153.03it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:00,  3.83it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:00,  3.83it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:00,  3.83it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:01<00:00,  3.79it/s]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:01<00:00,  3.79it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:01<00:00,  3.12it/s]\u001b[A\n",
      "2773it [00:02, 1222.89it/s]                                                                                            \n",
      "  0%|                                                                                         | 0/3359 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:00,  6.08it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:00,  6.08it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:00,  6.08it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:00<00:00,  5.96it/s]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:00<00:00,  5.96it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:01<00:00,  4.87it/s]\u001b[A\n",
      " 37%|███████████████████████████▉                                                | 1235/3359 [00:01<00:01, 1194.34it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:01,  2.42it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:01,  2.42it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:01,  2.42it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:01<00:00,  2.42it/s]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:01<00:00,  2.42it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:02<00:00,  2.01it/s]\u001b[A\n",
      "4436it [00:03, 1254.53it/s]                                                                                            \n",
      "  0%|                                                                                         | 0/5455 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:00,  3.78it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:00,  3.78it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:00,  3.78it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:01<00:00,  3.77it/s]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:01<00:00,  3.77it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:01<00:00,  3.07it/s]\u001b[A\n",
      " 37%|████████████████████████████▍                                               | 2040/5455 [00:01<00:02, 1248.48it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:01<00:02,  1.47it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:01<00:02,  1.47it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:01<00:02,  1.47it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:02<00:00,  1.49it/s]\u001b[A\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:02<00:00,  1.49it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:03<00:00,  1.25it/s]\u001b[A\n",
      "7160it [00:05, 1271.84it/s]                                                                                            \n",
      "  0%|                                                                                         | 0/8858 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:00<00:01,  2.40it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:00<00:01,  2.40it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:00<00:01,  2.40it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:01<00:00,  2.39it/s]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:01<00:00,  2.39it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:02<00:00,  1.98it/s]\u001b[A\n",
      " 37%|████████████████████████████▍                                               | 3316/8858 [00:02<00:04, 1308.26it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:02<00:03,  1.03s/it]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:02<00:03,  1.03s/it]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:02<00:03,  1.03s/it]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:04<00:01,  1.03s/it]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:04<00:01,  1.03s/it]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:06<00:00,  1.23s/it]\u001b[A\n",
      "11512it [00:08, 1319.13it/s]                                                                                           \n",
      "  0%|                                                                                        | 0/14384 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:01<00:02,  1.47it/s]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:01<00:02,  1.47it/s]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:01<00:02,  1.47it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:02<00:00,  1.47it/s]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:02<00:00,  1.47it/s]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:04<00:00,  1.22it/s]\u001b[A\n",
      " 38%|████████████████████████████▊                                              | 5516/14384 [00:04<00:06, 1345.73it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:03<00:05,  1.74s/it]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:03<00:05,  1.74s/it]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:03<00:05,  1.74s/it]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:06<00:01,  1.71s/it]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:06<00:01,  1.71s/it]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:10<00:00,  2.03s/it]\u001b[A\n",
      "18362it [00:14, 1285.78it/s]                                                                                           \n",
      "  0%|                                                                                        | 0/23357 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:02<00:03,  1.15s/it]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:02<00:03,  1.15s/it]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:02<00:03,  1.15s/it]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:04<00:01,  1.14s/it]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:04<00:01,  1.14s/it]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:06<00:00,  1.35s/it]\u001b[A\n",
      " 38%|████████████████████████████▍                                              | 8871/23357 [00:06<00:11, 1309.66it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:05<00:07,  2.65s/it]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:05<00:07,  2.65s/it]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:05<00:07,  2.65s/it]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:10<00:02,  2.65s/it]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:10<00:02,  2.65s/it]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:15<00:00,  3.18s/it]\u001b[A\n",
      "30355it [00:22, 1336.92it/s]                                                                                           \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|                                                                                        | 0/37926 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:03<00:05,  1.77s/it]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:03<00:05,  1.77s/it]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:03<00:05,  1.77s/it]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:07<00:01,  1.77s/it]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:07<00:01,  1.77s/it]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:10<00:00,  2.12s/it]\u001b[A\n",
      " 38%|████████████████████████████▏                                             | 14427/37926 [00:10<00:17, 1358.80it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:08<00:13,  4.44s/it]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:08<00:13,  4.44s/it]\u001b[A\n",
      " 38%|████████████████████████████▏                                             | 14427/37926 [00:26<00:17, 1358.80it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:17<00:04,  4.41s/it]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:17<00:04,  4.41s/it]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:26<00:00,  5.24s/it]\u001b[A\n",
      "49278it [00:36, 1337.30it/s]                                                                                           \n",
      "  0%|                                                                                        | 0/61584 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:06<00:09,  3.03s/it]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:06<00:09,  3.03s/it]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:06<00:09,  3.03s/it]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:11<00:02,  2.99s/it]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:11<00:02,  2.99s/it]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:17<00:00,  3.51s/it]\u001b[A\n",
      " 38%|███████████████████████████▊                                              | 23131/61584 [00:17<00:29, 1317.28it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:14<00:21,  7.01s/it]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:14<00:21,  7.01s/it]\u001b[A\n",
      " 38%|███████████████████████████▊                                              | 23131/61584 [00:36<00:29, 1317.28it/s]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:28<00:07,  7.09s/it]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:28<00:07,  7.09s/it]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:43<00:00,  8.63s/it]\u001b[A\n",
      "80391it [01:00, 1323.41it/s]                                                                                           \n",
      "  0%|                                                                                       | 0/100000 [00:00<?, ?it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:   0%|                                                   | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:09<00:14,  4.87s/it]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:09<00:14,  4.87s/it]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:09<00:14,  4.87s/it]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:19<00:04,  4.81s/it]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:19<00:04,  4.81s/it]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [00:28<00:00,  5.64s/it]\u001b[A\n",
      " 37%|███████████████████████████▎                                             | 37382/100000 [00:28<00:47, 1325.56it/s]\n",
      "  0%|                                                                                            | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      "Generating for node: N/S:   0%|                                                                  | 0/5 [00:00<?, ?it/s]\u001b[A\n",
      " 37%|███████████████████████████▎                                             | 37382/100000 [00:42<00:47, 1325.56it/s]\u001b[A\n",
      "Generating for node: Economic_Situation:  40%|█████████████████▏                         | 2/5 [00:24<00:36, 12.15s/it]\u001b[A\n",
      "Generating for node: Gender:  40%|██████████████████████                                 | 2/5 [00:24<00:36, 12.15s/it]\u001b[A\n",
      "Generating for node: Graduate:  40%|█████████████████████▏                               | 2/5 [00:24<00:36, 12.15s/it]\u001b[A\n",
      "Generating for node: Graduate:  80%|██████████████████████████████████████████▍          | 4/5 [00:48<00:12, 12.12s/it]\u001b[A\n",
      "Generating for node: Unemployed:  80%|████████████████████████████████████████▊          | 4/5 [00:48<00:12, 12.12s/it]\u001b[A\n",
      "Generating for node: Unemployed: 100%|███████████████████████████████████████████████████| 5/5 [01:12<00:00, 14.56s/it]\u001b[A\n",
      "131608it [01:41, 1302.96it/s]                                                                                          "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(    10, 0.5       , 2.63950000e-01, 0.2 , 0.03605)\n",
      " (    16, 0.375     , 1.38950000e-01, 0.31, 0.07395)\n",
      " (    26, 0.30769231, 7.16423077e-02, 0.23, 0.00605)\n",
      " (    42, 0.30952381, 7.34738095e-02, 0.31, 0.07395)\n",
      " (    69, 0.24637681, 1.03268116e-02, 0.28, 0.04395)\n",
      " (   112, 0.28571429, 4.96642857e-02, 0.25, 0.01395)\n",
      " (   183, 0.28415301, 4.81030055e-02, 0.22, 0.01605)\n",
      " (   297, 0.21885522, 1.71947811e-02, 0.25, 0.01395)\n",
      " (   483, 0.24223602, 6.18602484e-03, 0.23, 0.00605)\n",
      " (   784, 0.20790816, 2.81418367e-02, 0.24, 0.00395)\n",
      " (  1274, 0.23783359, 1.78359498e-03, 0.23, 0.00605)\n",
      " (  2069, 0.23151281, 4.53719188e-03, 0.22, 0.01605)\n",
      " (  3359, 0.24382257, 7.77256624e-03, 0.24, 0.00395)\n",
      " (  5455, 0.23849679, 2.44679193e-03, 0.24, 0.00395)\n",
      " (  8858, 0.23142922, 4.62078347e-03, 0.24, 0.00395)\n",
      " ( 14384, 0.23651279, 4.62791991e-04, 0.23, 0.00605)\n",
      " ( 23357, 0.23620328, 1.53279531e-04, 0.24, 0.00395)\n",
      " ( 37926, 0.2399673 , 3.91730475e-03, 0.24, 0.00395)\n",
      " ( 61584, 0.23584048, 2.09521954e-04, 0.24, 0.00395)\n",
      " (100000, 0.23584   , 2.10000000e-04, 0.24, 0.00395)]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "results = np.array([],dtype=[('sample_size', '<i8'), \n",
    "                             ('punemployed_RS', '<f8'),('e_unemployed_RS','<f8'),\n",
    "                             ('punemployed_LW', '<f8'),('e_unemployed_LW','<f8')])\n",
    "starting_size_ = 1\n",
    "final_size     = 5  \n",
    "experiments    = 20\n",
    "\n",
    "#sample_size from 10 to 10^5 in 20 experiments:\n",
    "for size in np.logspace(starting_size_, final_size, num=experiments, dtype='<i8'): #default base = 10.0\n",
    "    results=np.append(results,run_experiment(size))\n",
    "    \n",
    "print(results)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Comparison between the two implemented approximate inference methods:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20,5))\n",
    "plt.subplot(121)\n",
    "plt.ylim(.0,.5)\n",
    "plt.grid()\n",
    "plt.title('P(Unemployed=true| Gender=true, N/S=true)')\n",
    "LWplot, = plt.semilogx(results['sample_size'],results['punemployed_LW'],'b',label=\"Likelihood Weighting\")\n",
    "RSplot, = plt.semilogx(results['sample_size'],results['punemployed_RS'],'g',label=\"Rejection Sampling\")\n",
    "VEplot, = plt.semilogx(results['sample_size'],punemployed*np.ones(results.size),'r',label=\"reference value\")\n",
    "plt.legend(handles=[LWplot,RSplot,VEplot])\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Increasing the number of samples, the estimate probability value becomes closer to the reference value."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x2757079b6c8>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20,5))\n",
    "plt.subplot(121)\n",
    "plt.ylim(0,.2)\n",
    "plt.grid()\n",
    "plt.title('Absolute error on P(Unemployed=true| Gender=true, N/S=true)')\n",
    "ELWplot, = plt.semilogx(results['sample_size'],results['e_unemployed_LW'],'b',label=\"Likelihood Weighting\")\n",
    "ERSplot, = plt.semilogx(results['sample_size'],results['e_unemployed_RS'],'g',label=\"Rejection Sampling\")\n",
    "plt.legend(handles=[ELWplot,ERSplot])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}