io_alergia_bpic17.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "PRISM_PATH = \"/home/paul/Downloads/prism-games-3.2.1-linux64-x86/bin/prism\"  # path to PRISM-games install\n",
    "STRATEGY_PATH = \"/home/paul/Documents/probabilistic_games/adv.tra\" # path to where strategies shall be stored\n",
    "STORE_PATH = \"/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/\" # path to where generated models can be stored\n",
    "QUERY_PATH = \"/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/\" # path to queries\n",
    "OUTPUT_PATH = \"/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/outputs/\" # path to PRISM-games generated output files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from journepy.src.preprocessing.bpic17 import preprocessed_log\n",
    "from journepy.src.alergia_utils import convert_utils\n",
    "from journepy.src.mc_utils.prism_utils import PrismPrinter\n",
    "from journepy.src.mc_utils.prism_utils import PrismQuery\n",
    "\n",
    "import probabilistic_game_utils as pgu \n",
    "\n",
    "from aalpy.learning_algs import run_Alergia\n",
    "from aalpy.utils import save_automaton_to_file\n",
    "from IPython.display import Image\n",
    "\n",
    "\n",
    "import pandas as pd\n",
    "\n",
    "from networkx.drawing.nx_agraph import to_agraph\n",
    "\n",
    "import json\n",
    "\n",
    "import networkx as nx\n",
    "\n",
    "import subprocess\n",
    "\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load Data\n",
    "We load the event log and additionally needed information."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paul/.local/lib/python3.10/site-packages/pm4py/util/dt_parsing/parser.py:77: UserWarning: ISO8601 strings are not fully supported with strpfromiso for Python versions below 3.11\n",
      "  warnings.warn(\n",
      "/home/paul/.local/lib/python3.10/site-packages/tqdm/auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n",
      "parsing log, completed traces :: 100%|██████████| 31509/31509 [00:50<00:00, 625.93it/s]\n"
     ]
    }
   ],
   "source": [
    "filtered_log_before, filtered_log_after = preprocessed_log(\"data/BPI Challenge 2017.xes\") # uses common preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5515\n",
      "6989\n"
     ]
    }
   ],
   "source": [
    "print(len(filtered_log_before))\n",
    "print(len(filtered_log_after))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# change from xes format\n",
    "def parse(s):\n",
    "    return s.replace('(', ')').replace(')', '').replace(' ', '')\n",
    "filtered_log_before_activities = [[parse(e['concept:name']) for e in t] for t in filtered_log_before]\n",
    "filtered_log_after_activities = [[parse(e['concept:name']) for e in t] for t in filtered_log_after]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load actor mapping: maps events to an actor (service provider or user)\n",
    "with open('data/activities.xml') as f:\n",
    "    data = f.read()\n",
    "actors = json.loads(data)\n",
    "actors = {parse(k) : parse(actors[k]) for k in actors}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "O_Returned ['O_Returned']\n",
      "O_Accepted ['O_Accepted']\n",
      "A_Submitted ['A_Submitted']\n",
      "negative ['negative']\n",
      "positive ['positive']\n",
      "A_Accepted ['A_Accepted']\n",
      "A_Concept ['A_Concept']\n",
      "A_Validating ['A_Validating']\n",
      "A_CancelledCUSTOMER ['A_CancelledCUSTOMER']\n",
      "W_CallincompletefilesSHORT ['W_Callincompletefiles']\n",
      "W_CallafteroffersSHORT ['W_Callafteroffers']\n",
      "O_CreateOffer2 ['O_CreateOffer']\n",
      "W_CallincompletefilesLONG ['W_Callincompletefiles']\n",
      "start ['start']\n",
      "W_CallafteroffersLONG ['W_Callafteroffers']\n",
      "W_CallafteroffersSUPERLONG ['W_Callafteroffers']\n",
      "O_Sentmailandonline ['O_Sentmailandonline']\n",
      "W_CallincompletefilesSUPERLONG ['W_Callincompletefiles']\n",
      "A_CreateApplication ['A_CreateApplication']\n",
      "A_Pending ['A_Pending']\n",
      "O_Cancelled ['O_Cancelled']\n",
      "TIMEOUT ['TIMEOUT']\n",
      "O_CreateOffer3 ['O_CreateOffer']\n",
      "O_CreateOffer1 ['O_CreateOffer']\n",
      "O_CreateOffer0 ['O_CreateOffer']\n",
      "O_Sentonlineonly ['O_Sentonlineonly']\n"
     ]
    }
   ],
   "source": [
    "# build action mapping: assigns each event to an actor\n",
    "actions_to_activities = {}\n",
    "actions_observed = set()\n",
    "for trace in filtered_log_before_activities:\n",
    "    actions_observed.update(trace)\n",
    "\n",
    "actors_observed_actions = {}\n",
    "for action in actions_observed:\n",
    "    contained = [a for a in actors if a in action]\n",
    "    print(action, contained)\n",
    "    assert len(contained) == 1# each once - skip enumerating\n",
    "    if actors[contained[0]] == \"company\": # events where company is NOT deterministic\n",
    "        if 'O_Sentonlineonly' in action or 'O_Sentmailandonline' in action: \n",
    "            # sent online vs sent mail seems to be critical interaction - if aggregated together and diff. from others- 90%\n",
    "            actions_to_activities[action] = \"sent\"\n",
    "        elif 'O_CreateOffer0' in action:\n",
    "            actions_to_activities[action] = action\n",
    "        elif 'O_CreateOffer1' in action:\n",
    "            actions_to_activities[action] = action\n",
    "        elif 'O_CreateOffer2' in action:\n",
    "            actions_to_activities[action] = action\n",
    "        elif 'O_CreateOffer3' in action:\n",
    "            actions_to_activities[action] = action\n",
    "        elif 'O_CreateOffer4' in action:\n",
    "            actions_to_activities[action] = action\n",
    "        elif 'O_CreateOffer5' in action:\n",
    "            actions_to_activities[action] = action\n",
    "        elif 'A_Denied' in action:\n",
    "            actions_to_activities[action] = action\n",
    "        elif 'A_Incomplete' in action:\n",
    "            actions_to_activities[action] = action\n",
    "        elif 'A_Validating' in action:\n",
    "            actions_to_activities[action] = action\n",
    "        else:\n",
    "            actions_to_activities[action] = action\n",
    "    else:\n",
    "        if \"A_Submitted\" in action: # event where user is deterministic\n",
    "            actions_to_activities[action] = action\n",
    "        elif \"Callafteroffers\" in action: # event where user is deterministic\n",
    "            actions_to_activities[action] = \"call\"\n",
    "        elif \"Callincomplete\" in action: # event where user is deterministic\n",
    "            actions_to_activities[action] = \"call\"\n",
    "        else: # includes \"negative\" action\n",
    "            actions_to_activities[action] = \"user\"\n",
    "\n",
    "    actors_observed_actions[action] = actors[contained[0]]\n",
    "\n",
    "actors_observed_actions['offer'] = 'company'\n",
    "actors_observed_actions['company'] = 'company'\n",
    "actors_observed_actions['user'] = 'customer'\n",
    "actors_observed_actions['sent'] = 'company'\n",
    "actors_observed_actions['offer_response'] = 'company'\n",
    "actors_observed_actions['callafter'] = 'customer'\n",
    "actors_observed_actions['callincomplete'] = 'customer'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Passive automata learning\n",
    "The event log is transformed into an Input/Output format and environment actions to select the next player are added additionally.\n",
    "Alergia learning automatically selects IO/Alergia if input is in [O, (I,O), ...] format."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# add activities\n",
    "data_before = [[(actions_to_activities[t[i]], t[i]) for i in range(1, len(t))] for t in filtered_log_before_activities]\n",
    "for d in data_before:\n",
    "    d.insert(0, 'start')\n",
    "\n",
    "data_after = [[(actions_to_activities[t[i]], t[i]) for i in range(1, len(t))] for t in filtered_log_after_activities]\n",
    "for d in data_after:\n",
    "    d.insert(0, 'start')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# quantify environment -> becomes MDP\n",
    "data_before_environment = []\n",
    "for trace in data_before:\n",
    "    current = [trace[0]]\n",
    "    for i in range(1, len(trace)):\n",
    "        e = trace[i]\n",
    "        previous_state = \"start\" if i == 1 else trace[i-1][1]\n",
    "        \n",
    "        # encode decision in one step\n",
    "        current.append(('env', actors_observed_actions[e[1]] + previous_state))\n",
    "        current.append(e)\n",
    "    data_before_environment.append(current)\n",
    "\n",
    "data_after_environment = []\n",
    "for trace in data_after:\n",
    "    current = [trace[0]]\n",
    "    for i in range(1, len(trace)):\n",
    "        e = trace[i]\n",
    "        previous_state = \"start\" if i == 1 else trace[i-1][1]\n",
    "        \n",
    "        # encode decision in one step\n",
    "        current.append(('env', actors_observed_actions[e[1]] + previous_state))\n",
    "        current.append(e)\n",
    "    data_after_environment.append(current)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PTA Construction Time:  0.11\n",
      "Alergia Learning Time: 0.09\n",
      "Alergia Learned 95 state automaton.\n",
      "PTA Construction Time:  0.12\n",
      "Alergia Learning Time: 0.22\n",
      "Alergia Learned 131 state automaton.\n"
     ]
    }
   ],
   "source": [
    "model_before_environment = run_Alergia(data_before_environment, automaton_type='mdp', eps=0.8, print_info=True) # 0.1 plot interesting, 0.8 is confirms knowledge, 1.2 : plot align, 2 grows stronger, 1.8 : 2 is worse, grows stronger, gas is bit more interesting\n",
    "model_after_environment = run_Alergia(data_after_environment, automaton_type='mdp', eps=0.8, print_info=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model saved to out/bpic_17-1_environment_actions.dot.\n",
      "Model saved to out/bpic_17-1_environment_actions.png.\n"
     ]
    }
   ],
   "source": [
    "filename = \"out/bpic_17-1_environment_actions\"\n",
    "save_automaton_to_file(model_before_environment, f\"{filename}.dot\")\n",
    "save_automaton_to_file(model_before_environment, f\"{filename}.png\", file_type=\"png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model saved to out/bpic_17-2_environment_actions.dot.\n",
      "Model saved to out/bpic_17-2_environment_actions.png.\n"
     ]
    }
   ],
   "source": [
    "filename = \"out/bpic_17-2_environment_actions\"\n",
    "save_automaton_to_file(model_after_environment, f\"{filename}.dot\")\n",
    "save_automaton_to_file(model_after_environment, f\"{filename}.png\", file_type=\"png\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Convert MDP to Game in Networkx format"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "g_before = convert_utils.mdp_to_nx(model_before_environment, actors_observed_actions)\n",
    "g_after = convert_utils.mdp_to_nx(model_after_environment, actors_observed_actions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "g_before = pgu.add_gas_and_user_count(g_before, data_before_environment)\n",
    "g_after = pgu.add_gas_and_user_count(g_after, data_after_environment)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# test on deterministic cycles:\n",
    "def assert_no_det_cycle(g):\n",
    "    for c in list(nx.simple_cycles(g)):\n",
    "        found = False\n",
    "        for i in range(len(c)):\n",
    "            if g[c[i]][c[(i+1)%len(c)]]['prob_weight'] != 1:\n",
    "                found = True\n",
    "        assert found\n",
    "assert_no_det_cycle(g_before)\n",
    "assert_no_det_cycle(g_after)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model Checking Stochastic User Journey Games\n",
    "The games are written to the prism format and PRISM-games is called."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "printer_before = PrismPrinter(g_before, STORE_PATH, \"bpic_17_1_alergia.prism\")\n",
    "printer_before.write_to_prism()\n",
    "printer_after = PrismPrinter(g_after, STORE_PATH, \"bpic_17_2_alergia.prism\")\n",
    "printer_after.write_to_prism()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "query_before = PrismQuery(g_before, STORE_PATH, \"bpic_17_1_alergia.prism\", PRISM_PATH)\n",
    "model_name_before = STORE_PATH+\"bpic_17_1_alergia.prism\"\n",
    "\n",
    "query_after = PrismQuery(g_after, STORE_PATH, \"bpic_17_2_alergia.prism\", PRISM_PATH)\n",
    "model_name_after = STORE_PATH+\"bpic_17_2_alergia.prism\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8865074490124265\n",
      "0.9166362285809062\n"
     ]
    }
   ],
   "source": [
    "# Query Q1 from Table 1\n",
    "results_file_before = query_before.query(QUERY_PATH+\"pos_alergia.props\", write_parameterized=True)\n",
    "print(results_file_before['q0start'])\n",
    "\n",
    "results_file_after = query_after.query(QUERY_PATH+\"pos_alergia.props\", write_parameterized=True)\n",
    "print(results_file_after['q0start'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PRISM-games\n",
      "===========\n",
      "\n",
      "Version: 3.2.1 (based on PRISM 4.8.dev)\n",
      "Date: Wed Sep 18 11:29:27 CEST 2024\n",
      "Hostname: hp\n",
      "Memory limits: cudd=1g, java(heap)=1g\n",
      "Command line: prism-games /home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/bpic_17_1_alergia.prism /home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/mc_runs.props -const envprob=0\n",
      "\n",
      "Parsing model file \"/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/bpic_17_1_alergia.prism\"...\n",
      "\n",
      "Type:        SMG\n",
      "Modules:     userModule providerModule controll\n",
      "Variables:   gas state negative positive final_gas steps min_gas\n",
      "\n",
      "Parsing properties file \"/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/mc_runs.props\"...\n",
      "\n",
      "3 properties:\n",
      "(1) <<providerPlayer,userPlayer>>R{\"gas_neg\"}min=? [ F (positive|negative) ]\n",
      "(2) <<providerPlayer>>R{\"gas_neg\"}min=? [ F (positive|negative) ]\n",
      "(3) <<providerPlayer>>R{\"gas_pos\"}max=? [ F (positive|negative) ]\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "\n",
      "Model checking: <<providerPlayer,userPlayer>>R{\"gas_neg\"}min=? [ F (positive|negative) ]\n",
      "Model constants: envprob=0\n",
      "\n",
      "Switching to explicit engine, which supports SMGs...\n",
      "\n",
      "Building model...\n",
      "Model constants: envprob=0\n",
      "\n",
      "Computing reachable states... 265 states\n",
      "Reachable states exploration and model construction done in 0.074 secs.\n",
      "Sorting reachable states list...\n",
      "\n",
      "Time for model construction: 0.133 seconds.\n",
      "\n",
      "Type:        SMG\n",
      "States:      265 (1 initial)\n",
      "Transitions: 342\n",
      "Choices:     292\n",
      "Max/avg:     5/1.10\n",
      "Building reward structure...\n",
      "\n",
      "Starting expected reachability...\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 25 iterations and 0.023 seconds.\n",
      "target=2, inf=0, rest=263\n",
      "Computing the upper bound where 0.24 is used instead of 0.0\n",
      "Starting value iteration (minmax)...\n",
      "Value iteration (minmax) took 69 iterations and 0.029 seconds.\n",
      "Computed an over-approximation of the solution (in 0 seconds), this will now be used to get the solution\n",
      "Starting value iteration (minmax)...\n",
      "Value iteration (minmax) took 69 iterations and 0.016 seconds.\n",
      "Expected reachability took 0.08 seconds.\n",
      "\n",
      "Value in the initial state: 33.10846818642869\n",
      "\n",
      "Time for model checking: 0.109 seconds.\n",
      "\n",
      "Result: 33.10846818642869 (+/- 1.3081069516438385E-4 estimated; rel err 3.950973945028473E-6)\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "\n",
      "Model checking: <<providerPlayer>>R{\"gas_neg\"}min=? [ F (positive|negative) ]\n",
      "Model constants: envprob=0\n",
      "Building reward structure...\n",
      "\n",
      "Starting expected reachability...\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 25 iterations and 0.007 seconds.\n",
      "target=2, inf=0, rest=263\n",
      "Computing the upper bound where 0.24 is used instead of 0.0\n",
      "Starting value iteration (minmax)...\n",
      "Value iteration (minmax) took 225 iterations and 0.029 seconds.\n",
      "Computed an over-approximation of the solution (in 0 seconds), this will now be used to get the solution\n",
      "Starting value iteration (minmax)...\n",
      "Value iteration (minmax) took 248 iterations and 0.038 seconds.\n",
      "Expected reachability took 0.074 seconds.\n",
      "\n",
      "Value in the initial state: 37.347331303929145\n",
      "\n",
      "Time for model checking: 0.077 seconds.\n",
      "\n",
      "Result: 37.347331303929145 (+/- 2.932157845703593E-4 estimated; rel err 7.851050512396622E-6)\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "\n",
      "Model checking: <<providerPlayer>>R{\"gas_pos\"}max=? [ F (positive|negative) ]\n",
      "Model constants: envprob=0\n",
      "Building reward structure...\n",
      "\n",
      "Starting expected reachability...\n",
      "Starting Prob1 (minmax)...\n",
      "Prob1 (minmax) took 25 iterations and 0.009 seconds.\n",
      "target=2, inf=0, rest=263\n",
      "Computing the upper bound where 0.15 is used instead of 0.0\n",
      "Starting value iteration (maxmin)...\n",
      "Value iteration (maxmin) took 77 iterations and 0.009 seconds.\n",
      "Computed an over-approximation of the solution (in 0 seconds), this will now be used to get the solution\n",
      "Starting value iteration (maxmin)...\n",
      "Value iteration (maxmin) took 68 iterations and 0.005 seconds.\n",
      "Expected reachability took 0.023 seconds.\n",
      "\n",
      "Value in the initial state: 67.78734807562431\n",
      "\n",
      "Time for model checking: 0.03 seconds.\n",
      "\n",
      "Result: 67.78734807562431 (+/- 6.4199544297594E-4 estimated; rel err 9.470726635591687E-6)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "CompletedProcess(args=['/home/paul/Downloads/prism-games-3.2.1-linux64-x86/bin/prism', '/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/bpic_17_1_alergia.prism', '/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/mc_runs.props', '-const', 'envprob=0'], returncode=0)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Query Q2, Q3, and Q4 from Table 1 - BPIC'17-1\n",
    "subprocess.run([PRISM_PATH, model_name_before, QUERY_PATH+\"mc_runs.props\", \"-const\", \"envprob=0\"]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PRISM-games\n",
      "===========\n",
      "\n",
      "Version: 3.2.1 (based on PRISM 4.8.dev)\n",
      "Date: Wed Sep 18 11:29:28 CEST 2024\n",
      "Hostname: hp\n",
      "Memory limits: cudd=1g, java(heap)=1g\n",
      "Command line: prism-games /home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/bpic_17_2_alergia.prism /home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/mc_runs.props -const envprob=0\n",
      "\n",
      "Parsing model file \"/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/bpic_17_2_alergia.prism\"...\n",
      "\n",
      "Type:        SMG\n",
      "Modules:     userModule providerModule controll\n",
      "Variables:   gas state negative positive final_gas steps min_gas\n",
      "\n",
      "Parsing properties file \"/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/mc_runs.props\"...\n",
      "\n",
      "3 properties:\n",
      "(1) <<providerPlayer,userPlayer>>R{\"gas_neg\"}min=? [ F (positive|negative) ]\n",
      "(2) <<providerPlayer>>R{\"gas_neg\"}min=? [ F (positive|negative) ]\n",
      "(3) <<providerPlayer>>R{\"gas_pos\"}max=? [ F (positive|negative) ]\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "\n",
      "Model checking: <<providerPlayer,userPlayer>>R{\"gas_neg\"}min=? [ F (positive|negative) ]\n",
      "Model constants: envprob=0\n",
      "\n",
      "Switching to explicit engine, which supports SMGs...\n",
      "\n",
      "Building model...\n",
      "Model constants: envprob=0\n",
      "\n",
      "Computing reachable states... 373 states\n",
      "Reachable states exploration and model construction done in 0.109 secs.\n",
      "Sorting reachable states list...\n",
      "\n",
      "Time for model construction: 0.175 seconds.\n",
      "\n",
      "Type:        SMG\n",
      "States:      373 (1 initial)\n",
      "Transitions: 486\n",
      "Choices:     407\n",
      "Max/avg:     4/1.09\n",
      "Building reward structure...\n",
      "\n",
      "Starting expected reachability...\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 25 iterations and 0.035 seconds.\n",
      "target=2, inf=0, rest=371\n",
      "Computing the upper bound where 0.24 is used instead of 0.0\n",
      "Starting value iteration (minmax)...\n",
      "Value iteration (minmax) took 53 iterations and 0.037 seconds.\n",
      "Computed an over-approximation of the solution (in 0 seconds), this will now be used to get the solution\n",
      "Starting value iteration (minmax)...\n",
      "Value iteration (minmax) took 57 iterations and 0.028 seconds.\n",
      "Expected reachability took 0.117 seconds.\n",
      "\n",
      "Value in the initial state: 33.86562504311371\n",
      "\n",
      "Time for model checking: 0.144 seconds.\n",
      "\n",
      "Result: 33.86562504311371 (+/- 3.1032456604407344E-8 estimated; rel err 9.163408785427845E-10)\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "\n",
      "Model checking: <<providerPlayer>>R{\"gas_neg\"}min=? [ F (positive|negative) ]\n",
      "Model constants: envprob=0\n",
      "Building reward structure...\n",
      "\n",
      "Starting expected reachability...\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 25 iterations and 0.013 seconds.\n",
      "target=2, inf=0, rest=371\n",
      "Computing the upper bound where 0.24 is used instead of 0.0\n",
      "Starting value iteration (minmax)...\n",
      "Value iteration (minmax) took 149 iterations and 0.038 seconds.\n",
      "Computed an over-approximation of the solution (in 0 seconds), this will now be used to get the solution\n",
      "Starting value iteration (minmax)...\n",
      "Value iteration (minmax) took 168 iterations and 0.039 seconds.\n",
      "Expected reachability took 0.091 seconds.\n",
      "\n",
      "Value in the initial state: 36.06797265976299\n",
      "\n",
      "Time for model checking: 0.098 seconds.\n",
      "\n",
      "Result: 36.06797265976299 (+/- 3.5644837623641213E-4 estimated; rel err 9.882683997763529E-6)\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "\n",
      "Model checking: <<providerPlayer>>R{\"gas_pos\"}max=? [ F (positive|negative) ]\n",
      "Model constants: envprob=0\n",
      "Building reward structure...\n",
      "\n",
      "Starting expected reachability...\n",
      "Starting Prob1 (minmax)...\n",
      "Prob1 (minmax) took 36 iterations and 0.032 seconds.\n",
      "target=2, inf=0, rest=371\n",
      "Computing the upper bound where 0.15 is used instead of 0.0\n",
      "Starting value iteration (maxmin)...\n",
      "Value iteration (maxmin) took 2227 iterations and 0.245 seconds.\n",
      "Computed an over-approximation of the solution (in 0 seconds), this will now be used to get the solution\n",
      "Starting value iteration (maxmin)...\n",
      "Value iteration (maxmin) took 285 iterations and 0.027 seconds.\n",
      "Expected reachability took 0.305 seconds.\n",
      "\n",
      "Value in the initial state: 68.07028486327721\n",
      "\n",
      "Time for model checking: 0.312 seconds.\n",
      "\n",
      "Result: 68.07028486327721 (exact floating point)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "CompletedProcess(args=['/home/paul/Downloads/prism-games-3.2.1-linux64-x86/bin/prism', '/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/bpic_17_2_alergia.prism', '/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/mc_runs.props', '-const', 'envprob=0'], returncode=0)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Query Q2, Q3, and Q4 from Table 1 - BPIC'17-2\n",
    "subprocess.run([PRISM_PATH, model_name_after, QUERY_PATH+\"mc_runs.props\", \"-const\", \"envprob=0\"]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# run Activity experiment and produce Fig. 7\n",
    "# remove \"stdout=subprocess.DEVNULL\" to print output again\n",
    "file_name = OUTPUT_PATH+\"succ_prop_cond-bpic-17-1.txt\"\n",
    "subprocess.run([PRISM_PATH, STORE_PATH+\"bpic_17_1_alergia.prism\",\n",
    "                 QUERY_PATH+\"pos_alergia.props\",\n",
    "                   \"-const\", \"envprob=-0.95:0.05:0.95\", \"-exportresults\", file_name+\":dataframe\"], stdout=subprocess.DEVNULL) \n",
    "\n",
    "df_visual = pd.read_csv(file_name)\n",
    "plt.plot(df_visual['envprob'], df_visual['Result'], label = \"BPIC'17-1\", linewidth=3)\n",
    "\n",
    "file_name = OUTPUT_PATH+\"succ_prop_cond-bpic-17-2.txt\"\n",
    "subprocess.run([PRISM_PATH, STORE_PATH+\"bpic_17_2_alergia.prism\",\n",
    "                 QUERY_PATH+\"pos_alergia.props\",\n",
    "                   \"-const\", \"envprob=-0.95:0.05:0.95\", \"-exportresults\", file_name+\":dataframe\"], stdout=subprocess.DEVNULL) \n",
    "\n",
    "df_visual = pd.read_csv(file_name)\n",
    "plt.plot(df_visual['envprob'], df_visual['Result'], label = \"BPIC'17-2\", linewidth=3)\n",
    "\n",
    "plt.vlines(x=0, ymin=0, ymax = 1, linewidth=2, color = 'grey', linestyles='--')\n",
    "plt.text(-1, 0.05, 'Service Provider', fontsize = 18)\n",
    "plt.text(0.8, 0.05, 'User', fontsize = 18)\n",
    "plt.legend(fontsize=16)\n",
    "plt.xlabel(\"Scaled activity\", fontsize=22)\n",
    "plt.ylabel(\"Success probability\", fontsize=22)\n",
    "plt.xticks(fontsize=15)\n",
    "plt.yticks(fontsize=20)\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"out/bpic_activity_comparison.png\", dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# run gas upper and lower bound under limited steps\n",
    "PrismPrinter(g_before, STORE_PATH, \"bpic_17_1_alergia.prism\").write_to_prism(write_extended_parameterized=True)\n",
    "file_name_pos = OUTPUT_PATH+\"steps_gas_pos_bound-bpic-17-1.txt\"\n",
    "subprocess.run([PRISM_PATH, model_name_before, \n",
    "                QUERY_PATH+\"reward_props.props\", \"-prop\", \"3\",\n",
    "                \"-const\", \"m0=0,m1=0:1:80,m2=0,\", \"-exportresults\", file_name_pos+\":dataframe\"], stdout=subprocess.DEVNULL) \n",
    "file_name_neg = OUTPUT_PATH+\"steps_gas_neg_bound-bpic-17-1.txt\"\n",
    "subprocess.run([PRISM_PATH, model_name_before, \n",
    "                QUERY_PATH+\"reward_props.props\", \"-prop\", \"4\",\n",
    "                \"-const\", \"m0=0,m1=0:1:80,m2=0,\", \"-exportresults\", file_name_neg+\":dataframe\"], stdout=subprocess.DEVNULL) \n",
    "\n",
    "df_visual = pd.read_csv(file_name_pos)\n",
    "plt.plot(df_visual['m1']/4, df_visual['Result'], label=\"BPIC'17-1 max pos\", c = \"blue\")\n",
    "df_visual = pd.read_csv(file_name_neg)\n",
    "plt.plot(df_visual['m1']/4, df_visual['Result'], label=\"BPIC'17-1 min neg\", c = \"orange\")\n",
    "\n",
    "PrismPrinter(g_after, STORE_PATH, \"bpic_17_2_alergia.prism\").write_to_prism(write_extended_parameterized=True)\n",
    "file_name_pos = OUTPUT_PATH+\"steps_gas_pos_bound-bpic-17-2.txt\"\n",
    "subprocess.run([PRISM_PATH, model_name_after, \n",
    "                QUERY_PATH+\"reward_props.props\", \"-prop\", \"3\",\n",
    "                \"-const\", \"m0=0,m1=0:1:80,m2=0,\", \"-exportresults\", file_name_pos+\":dataframe\"], stdout=subprocess.DEVNULL) \n",
    "file_name_neg = OUTPUT_PATH+\"steps_gas_neg_bound-bpic-17-2.txt\"\n",
    "subprocess.run([PRISM_PATH, model_name_after, \n",
    "                QUERY_PATH+\"reward_props.props\", \"-prop\", \"4\",\n",
    "                \"-const\", \"m0=0,m1=0:1:80,m2=0,\", \"-exportresults\", file_name_neg+\":dataframe\"], stdout=subprocess.DEVNULL) \n",
    "\n",
    "df_visual = pd.read_csv(file_name_pos)\n",
    "plt.plot(df_visual['m1']/4, df_visual['Result'], label=\"BPIC'17-2 max pos\", c = \"blue\", linestyle=\"dashed\")\n",
    "df_visual = pd.read_csv(file_name_neg)\n",
    "plt.plot(df_visual['m1']/4, df_visual['Result'], label=\"BPIC'17-2 min neg\", c = \"orange\", linestyle=\"dashed\")\n",
    "\n",
    "plt.legend(fontsize=14)\n",
    "plt.xlabel(\"Steps S\", fontsize=18)\n",
    "plt.ylabel(\"Accumulated weight\", fontsize=18)\n",
    "plt.xticks(fontsize=15)\n",
    "plt.yticks(fontsize=18)\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"out/bpic_17_steps.png\", dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Induced and Reduced Model\n",
    "We generate a strategy for Q1 and build the induced Markov chain. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "query_before = PrismQuery(g_before, STORE_PATH, \"bpic_17_1_alergia.prism\", PRISM_PATH)\n",
    "query_after = PrismQuery(g_after, STORE_PATH, \"bpic_17_2_alergia.prism\", PRISM_PATH)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "skipped ['q19O_Sentmailandonline']\n",
      "skipped ['q20O_Sentonlineonly']\n",
      "skipped ['q28customerO_Sentonlineonly']\n",
      "skipped ['q27companyO_Sentonlineonly']\n",
      "skipped ['q35O_CreateOffer1']\n",
      "skipped ['q42O_Sentmailandonline']\n",
      "skipped ['q37W_CallafteroffersSHORT']\n",
      "skipped ['q36W_CallafteroffersLONG']\n",
      "skipped ['q38W_CallafteroffersSUPERLONG']\n",
      "skipped ['q40W_CallafteroffersSHORT']\n",
      "skipped ['q59O_Sentmailandonline']\n",
      "skipped ['q45companyO_CreateOffer1']\n",
      "skipped ['q46companyW_CallafteroffersLONG']\n",
      "skipped ['q47customerW_CallafteroffersLONG']\n",
      "skipped ['q48companyW_CallafteroffersSHORT']\n",
      "skipped ['q49customerW_CallafteroffersSHORT']\n",
      "skipped ['q51customerW_CallafteroffersSUPERLONG']\n",
      "skipped ['q50companyW_CallafteroffersSUPERLONG']\n",
      "skipped ['q53customerW_CallafteroffersSHORT']\n",
      "skipped ['q56customerO_Sentmailandonline']\n",
      "skipped ['q55companyO_Sentmailandonline']\n",
      "skipped ['q57O_Sentmailandonline']\n",
      "skipped ['q83O_CreateOffer2']\n",
      "skipped ['q60A_Validating']\n",
      "skipped ['q61O_Cancelled']\n",
      "skipped ['q63W_CallafteroffersSHORT']\n",
      "skipped ['q62W_CallafteroffersLONG']\n",
      "skipped ['q66companyA_Validating']\n",
      "skipped ['q67customerO_Cancelled']\n",
      "skipped ['q69companyW_CallafteroffersSHORT']\n",
      "skipped ['q74O_Returned']\n",
      "skipped ['q81customerO_Returned']\n",
      "skipped ['q85W_CallincompletefilesSHORT']\n",
      "skipped ['q93W_CallincompletefilesSUPERLONG']\n",
      "skipped ['q88companyO_CreateOffer2']\n",
      "skipped ['q90companyW_CallincompletefilesSHORT']\n",
      "skipped ['q94companyW_CallincompletefilesSUPERLONG']\n"
     ]
    }
   ],
   "source": [
    "strategy_before = query_before.get_strategy(QUERY_PATH+\"pos_alergia.props\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "skipped ['q19O_Sentmailandonline']\n",
      "skipped ['q20O_Sentonlineonly']\n",
      "skipped ['q27customerO_Sentmailandonline']\n",
      "skipped ['q28customerO_Sentonlineonly']\n",
      "skipped ['q46companyO_Sentonlineonly']\n",
      "skipped ['q33O_Sentmailandonline']\n",
      "skipped ['q34O_Sentonlineonly']\n",
      "skipped ['q62O_Sentmailandonline']\n",
      "skipped ['q37W_CallafteroffersSHORT']\n",
      "skipped ['q36W_CallafteroffersLONG']\n",
      "skipped ['q38W_CallafteroffersSUPERLONG']\n",
      "skipped ['q41O_CreateOffer1']\n",
      "skipped ['q43O_Sentmailandonline']\n",
      "skipped ['q45customerO_Sentmailandonline']\n",
      "skipped ['q49companyW_CallafteroffersLONG']\n",
      "skipped ['q50customerW_CallafteroffersLONG']\n",
      "skipped ['q51companyW_CallafteroffersSHORT']\n",
      "skipped ['q52customerW_CallafteroffersSHORT']\n",
      "skipped ['q54customerW_CallafteroffersSUPERLONG']\n",
      "skipped ['q53companyW_CallafteroffersSUPERLONG']\n",
      "skipped ['q56companyO_CreateOffer1']\n",
      "skipped ['q60customerO_Sentmailandonline']\n",
      "skipped ['q59companyO_Sentmailandonline']\n",
      "skipped ['q101O_Sentmailandonline']\n",
      "skipped ['q63W_CallafteroffersSHORT']\n",
      "skipped ['q67O_CreateOffer1']\n",
      "skipped ['q68A_Validating']\n",
      "skipped ['q69O_Cancelled']\n",
      "skipped ['q70A_Validating']\n",
      "skipped ['q71O_Sentmailandonline']\n",
      "skipped ['q72O_Sentmailandonline']\n",
      "skipped ['q74W_CallafteroffersSHORT']\n",
      "skipped ['q73W_CallafteroffersLONG']\n",
      "skipped ['q76customerO_Sentmailandonline']\n",
      "skipped ['q106customerW_CallafteroffersSHORT']\n",
      "skipped ['q81companyO_CreateOffer1']\n",
      "skipped ['q83customerO_CreateOffer1']\n",
      "skipped ['q82companyA_Validating']\n",
      "skipped ['q84customerO_Cancelled']\n",
      "skipped ['q85companyA_Validating']\n",
      "skipped ['q86customerO_Sentmailandonline']\n",
      "skipped ['q88customerO_Sentmailandonline']\n",
      "skipped ['q89companyW_CallafteroffersSHORT']\n",
      "skipped ['q91W_CallafteroffersSHORT']\n",
      "skipped ['q92W_CallafteroffersSHORT']\n",
      "skipped ['q97O_Returned']\n",
      "skipped ['q98O_Returned']\n",
      "skipped ['q100W_CallafteroffersSHORT']\n",
      "skipped ['q103W_CallafteroffersSHORT']\n",
      "skipped ['q104A_Validating']\n",
      "skipped ['q105companyW_CallafteroffersSHORT']\n",
      "skipped ['q124companyW_CallafteroffersSHORT']\n",
      "skipped ['q111customerO_Returned']\n",
      "skipped ['q112customerO_Returned']\n",
      "skipped ['q114customerW_CallafteroffersSHORT']\n",
      "skipped ['q115companyA_Validating']\n",
      "skipped ['q116A_Validating']\n",
      "skipped ['q118W_CallincompletefilesSUPERLONG']\n",
      "skipped ['q119O_Cancelled']\n",
      "skipped ['q120O_Returned']\n",
      "skipped ['q121companyA_Validating']\n",
      "skipped ['q123customerW_CallincompletefilesSUPERLONG']\n",
      "skipped ['q125customerO_Cancelled']\n",
      "skipped ['q126customerO_Returned']\n",
      "skipped ['q127O_Returned']\n",
      "skipped ['q129customerO_Returned']\n",
      "skipped ['q130W_CallincompletefilesSHORT']\n"
     ]
    }
   ],
   "source": [
    "strategy_after = query_after.get_strategy(QUERY_PATH+\"pos_alergia.props\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "def can_be_merged(g, results_file):\n",
    "    for s in g.nodes():\n",
    "        reachable_values = [round(results_file[t],4) for t in g[s]]\n",
    "        if round(results_file[s],4) in reachable_values:\n",
    "        #if len(set(reachable_values)) == 1 and reachable_values[0] == round(results_file[s],2):\n",
    "            return s \n",
    "    return None\n",
    "\n",
    "\"\"\"\n",
    "NOTE: One positive and one negative node is kept and all remaining from positive/negative cluster are merged into them.\n",
    "\"\"\"\n",
    "def reduce_graph(g, results_file):\n",
    "    neg_cluster = []\n",
    "    pos_cluster = []\n",
    "    print(\"size start\", len(g.nodes()))\n",
    "    s = can_be_merged(g, results_file)\n",
    "    while(s != None):\n",
    "        for t in g[s]:\n",
    "            if round(results_file[t],4) != round(results_file[s],4):\n",
    "                continue\n",
    "            g = nx.contracted_nodes(g, s, t, self_loops = False)\n",
    "        s = can_be_merged(g, results_file)\n",
    "\n",
    "    for s in g:\n",
    "        if results_file[s] == 0:\n",
    "            neg_cluster.append(s)\n",
    "        if results_file[s] == 1:\n",
    "            pos_cluster.append(s)\n",
    "    for s in pos_cluster[1:]:\n",
    "        g = nx.contracted_nodes(g, pos_cluster[0], s, self_loops=False)\n",
    "    for s in neg_cluster[1:]:\n",
    "        g = nx.contracted_nodes(g, neg_cluster[0], s, self_loops=False)\n",
    "\n",
    "    g.remove_edges_from(nx.selfloop_edges(g))\n",
    "\n",
    "    print(\"size reduced\", len(g.nodes()))\n",
    "    return g"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "from  matplotlib.colors import LinearSegmentedColormap # for color map\n",
    "from matplotlib.colors import rgb2hex\n",
    "\n",
    "def compute_color_map(g, results_file):\n",
    "    c = [\"#D55E00\", \"#F0E442\", \"#009E73\"]\n",
    "    v = [0,0.5,1]\n",
    "    l = list(zip(v,c))\n",
    "    cmap=LinearSegmentedColormap.from_list('rg',l, N=256)\n",
    "    s = cmap(0.23)\n",
    "    map = {}\n",
    "    for s in g.nodes():\n",
    "        map[s] = rgb2hex(cmap(results_file[s])) # have to convert to hex color\n",
    "    return map\n",
    "\n",
    "def draw_dfg(g, name, names={}, layout = \"sfdp\", color_map = []):\n",
    "    \"\"\"\n",
    "    Helper function to draw Networkx graphs.\n",
    "    \"\"\"\n",
    "    scaling = 10\n",
    "    # build graph with variable thickness\n",
    "    #scaling = 1/np.mean(list(nx.get_edge_attributes(g,'edge_weight').values()))\n",
    "\n",
    "    A = to_agraph(g)\n",
    "\n",
    "    edge_weights = nx.get_edge_attributes(g,'edge_weight')\n",
    "    for e in edge_weights:\n",
    "        e = A.get_edge(e[0], e[1])\n",
    "        e.attr[\"penwidth\"] = edge_weights[e]*scaling\n",
    "        e.attr[\"fontsize\"] = \"20\"\n",
    "    for e in g.edges:\n",
    "        edge = A.get_edge(e[0], e[1])\n",
    "        if 'controllable' in g[e[0]][e[1]]:\n",
    "            if not g[e[0]][e[1]]['controllable']:\n",
    "                edge.attr[\"style\"] = \"dotted\"\n",
    "                #edge.attr[\"label\"] =  str(g[e[0]][e[1]][\"prob_weight\"])\n",
    "        #A.add_edge(e[0], e[1], penwidth = edge_weights[e]*scaling)\n",
    "\n",
    "    #A.graph_attr.update(size=\"7.75,10.25\")\n",
    "\n",
    "    for n in A.nodes():\n",
    "        if n in names:\n",
    "            new = names[n]\n",
    "            if isinstance(names[n], float): \n",
    "                new = round(names[n], 2)\n",
    "            n.attr['label'] = new\n",
    "            #if new == 1:\n",
    "            #    n.attr['label'] = \"pos\"\n",
    "            #elif new == 0:\n",
    "            #    n.attr['label'] = \"neg\"\n",
    "            #else:\n",
    "            #    n.attr[\"label\"] = \"\" # uncomment to print state names\n",
    "        if n in color_map:\n",
    "            n.attr['color'] = color_map[n]\n",
    "    \n",
    "        n.attr['fontsize'] = 120\n",
    "        n.attr['penwidth'] = 30\n",
    "        n.attr['height'] = 3\n",
    "        n.attr['width'] = 3\n",
    "\n",
    "    for e in A.edges():\n",
    "        e.attr['penwidth'] = 20\n",
    "        e.attr[\"fontsize\"] = 120\n",
    "        e.attr[\"label\"] = str(round(g[e[0]][e[1]][\"prob_weight\"],2))\n",
    "        e.attr[\"color\"] = \"black\"\n",
    "\n",
    "        if g[e[0]][e[1]]['gas'] > 0:\n",
    "            e.attr[\"color\"] =\"darkgreen\"\n",
    "        if g[e[0]][e[1]]['gas'] < 0:\n",
    "            e.attr[\"color\"] =\"red\"\n",
    "             \n",
    "    A.write(name.split(\".\")[0]+\".dot\")\n",
    "    A.layout(layout)\n",
    "    A.draw(name)\n",
    "    print(\"Plotted\", name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Plotted out/before_clustered.png\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/paul/.local/lib/python3.10/site-packages/pygraphviz/agraph.py:1407: RuntimeWarning: neato: graph is too large for cairo-renderer bitmaps. Scaling by 0.997959 to fit\n",
      "\n",
      "  warnings.warn(b\"\".join(errors).decode(self.encoding), RuntimeWarning)\n"
     ]
    }
   ],
   "source": [
    "def get_probs_file(results_file, g, printer):\n",
    "    isomophism = nx.vf2pp_isomorphism(printer.g, g, node_label=None)\n",
    "    parsed_results_file = {isomophism[r] : results_file[r] for r in results_file}\n",
    "    return parsed_results_file\n",
    "\n",
    "color_map = compute_color_map(g_before, get_probs_file(results_file_before, g_before, printer_before))\n",
    "\n",
    "extended_naming = []\n",
    "draw_dfg(g_before, \"out/before_clustered.png\", names=extended_naming, layout = \"dot\", color_map=color_map)\n",
    "#draw_dfg(g_before, \"out/alergia_bpic_17-1_probs.png\", get_probs_file(results_file, g_before), layout = \"dot\") # plots graph with marked probabilities, but takes some time to render"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "size start 95\n",
      "size reduced 32\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Plotted out/alergia_bpic_17-1_probs_reduced.png\n",
      "size start 131\n",
      "size reduced 47\n",
      "Plotted out/alergia_bpic_17-2_probs_reduced.png\n"
     ]
    }
   ],
   "source": [
    "# Produces Figure 6 (1) and (2)\n",
    "import copy\n",
    "def plot_reduction(g, name, results_file, layout = \"sdf\"):\n",
    "    g = copy.deepcopy(g)\n",
    "    g = reduce_graph(g, results_file)\n",
    "    color_map = compute_color_map(g, results_file)\n",
    "    draw_dfg(g, name, names = results_file, layout = layout, color_map=color_map)\n",
    "plot_reduction(g_before, \"out/alergia_bpic_17-1_probs_reduced.png\", get_probs_file(results_file_before, g_before, printer_before), layout = \"dot\")\n",
    "plot_reduction(g_after, \"out/alergia_bpic_17-2_probs_reduced.png\", get_probs_file(results_file_after, g_after, printer_after), layout = \"dot\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Constrained steps and parameterized transitions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PRISM-games\n",
      "===========\n",
      "\n",
      "Version: 3.2.1 (based on PRISM 4.8.dev)\n",
      "Date: Wed Sep 18 11:31:06 CEST 2024\n",
      "Hostname: hp\n",
      "Memory limits: cudd=1g, java(heap)=1g\n",
      "Command line: prism-games /home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/bpic_17_1_alergia.prism /home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/exp_values.props\n",
      "\n",
      "Parsing model file \"/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/bpic_17_1_alergia.prism\"...\n",
      "\n",
      "Type:        SMG\n",
      "Modules:     userModule providerModule controll\n",
      "Variables:   gas state negative positive final_gas steps min_gas\n",
      "\n",
      "Parsing properties file \"/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/exp_values.props\"...\n",
      "\n",
      "3 properties:\n",
      "(1) <<providerPlayer>>R{\"steps\"}max=? [ Fc positive ]\n",
      "(2) <<providerPlayer>>R{\"gas_neg\"}max=? [ Fc positive ]\n",
      "(3) <<providerPlayer>>R{\"gas_pos\"}max=? [ Fc positive ]\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "\n",
      "Model checking: <<providerPlayer>>R{\"steps\"}max=? [ Fc positive ]\n",
      "\n",
      "Switching to explicit engine, which supports SMGs...\n",
      "\n",
      "Building model...\n",
      "\n",
      "Computing reachable states... 265 states\n",
      "Reachable states exploration and model construction done in 0.09 secs.\n",
      "Sorting reachable states list...\n",
      "\n",
      "Time for model construction: 0.134 seconds.\n",
      "\n",
      "Type:        SMG\n",
      "States:      265 (1 initial)\n",
      "Transitions: 342\n",
      "Choices:     292\n",
      "Max/avg:     5/1.10\n",
      "Building reward structure...\n",
      "\n",
      "Starting expected reachability...\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 8 iterations and 0.038 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 8 iterations and 0.005 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 16 iterations and 0.008 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 28 iterations and 0.013 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 47 iterations and 0.02 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 12 iterations and 0.002 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Starting Prob0 (maxmin)...\n",
      "target=1, inf=0, rest=264\n",
      "Starting value iteration (maxmin)...\n",
      "Value iteration (maxmin) took 2085 iterations and 0.282 seconds.\n",
      "Expected reachability took 0.39 seconds.\n",
      "\n",
      "Value in the initial state: 61.71192798322763\n",
      "\n",
      "Time for model checking: 0.414 seconds.\n",
      "\n",
      "Result: 61.71192798322763 (+/- 6.15018094033184E-4 estimated; rel err 9.96595170710493E-6)\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "\n",
      "Model checking: <<providerPlayer>>R{\"gas_neg\"}max=? [ Fc positive ]\n",
      "Building reward structure...\n",
      "\n",
      "Starting expected reachability...\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 23 iterations and 0.005 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 33 iterations and 0.006 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 31 iterations and 0.006 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 31 iterations and 0.005 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 8 iterations and 0.001 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 12 iterations and 0.002 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Starting Prob0 (maxmin)...\n",
      "target=1, inf=0, rest=264\n",
      "Starting value iteration (maxmin)...\n",
      "Value iteration (maxmin) took 2065 iterations and 0.139 seconds.\n",
      "Expected reachability took 0.164 seconds.\n",
      "\n",
      "Value in the initial state: 1080.6729137522418\n",
      "\n",
      "Time for model checking: 0.167 seconds.\n",
      "\n",
      "Result: 1080.6729137522418 (+/- 0.010693656602336385 estimated; rel err 9.895368400792586E-6)\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "\n",
      "Model checking: <<providerPlayer>>R{\"gas_pos\"}max=? [ Fc positive ]\n",
      "Building reward structure...\n",
      "\n",
      "Starting expected reachability...\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 37 iterations and 0.007 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 44 iterations and 0.006 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 35 iterations and 0.006 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Starting Prob0 (maxmin)...\n",
      "target=1, inf=0, rest=264\n",
      "Starting value iteration (maxmin)...\n",
      "Value iteration (maxmin) took 77 iterations and 0.006 seconds.\n",
      "Expected reachability took 0.028 seconds.\n",
      "\n",
      "Value in the initial state: 67.78719170892771\n",
      "\n",
      "Time for model checking: 0.031 seconds.\n",
      "\n",
      "Result: 67.78719170892771 (+/- 6.173528212524061E-4 estimated; rel err 9.107219309265168E-6)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "CompletedProcess(args=['/home/paul/Downloads/prism-games-3.2.1-linux64-x86/bin/prism', '/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/bpic_17_1_alergia.prism', '/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/exp_values.props'], returncode=0)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# estimate max counts before\n",
    "subprocess.run([PRISM_PATH, model_name_before, QUERY_PATH+\"exp_values.props\"]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PRISM-games\n",
      "===========\n",
      "\n",
      "Version: 3.2.1 (based on PRISM 4.8.dev)\n",
      "Date: Wed Sep 18 11:31:08 CEST 2024\n",
      "Hostname: hp\n",
      "Memory limits: cudd=1g, java(heap)=1g\n",
      "Command line: prism-games /home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/bpic_17_2_alergia.prism /home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/exp_values.props\n",
      "\n",
      "Parsing model file \"/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/bpic_17_2_alergia.prism\"...\n",
      "\n",
      "Type:        SMG\n",
      "Modules:     userModule providerModule controll\n",
      "Variables:   gas state negative positive final_gas steps min_gas\n",
      "\n",
      "Parsing properties file \"/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/exp_values.props\"...\n",
      "\n",
      "3 properties:\n",
      "(1) <<providerPlayer>>R{\"steps\"}max=? [ Fc positive ]\n",
      "(2) <<providerPlayer>>R{\"gas_neg\"}max=? [ Fc positive ]\n",
      "(3) <<providerPlayer>>R{\"gas_pos\"}max=? [ Fc positive ]\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "\n",
      "Model checking: <<providerPlayer>>R{\"steps\"}max=? [ Fc positive ]\n",
      "\n",
      "Switching to explicit engine, which supports SMGs...\n",
      "\n",
      "Building model...\n",
      "\n",
      "Computing reachable states... 373 states\n",
      "Reachable states exploration and model construction done in 0.114 secs.\n",
      "Sorting reachable states list...\n",
      "\n",
      "Time for model construction: 0.165 seconds.\n",
      "\n",
      "Type:        SMG\n",
      "States:      373 (1 initial)\n",
      "Transitions: 486\n",
      "Choices:     407\n",
      "Max/avg:     4/1.09\n",
      "Building reward structure...\n",
      "\n",
      "Starting expected reachability...\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 8 iterations and 0.029 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 8 iterations and 0.006 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 16 iterations and 0.006 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 25 iterations and 0.015 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 33 iterations and 0.017 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 44 iterations and 0.013 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 56 iterations and 0.012 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 64 iterations and 0.011 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 62 iterations and 0.015 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Starting Prob0 (maxmin)...\n",
      "target=1, inf=0, rest=372\n",
      "Starting value iteration (maxmin)...\n",
      "Value iteration (maxmin) took 2585 iterations and 0.371 seconds.\n",
      "Expected reachability took 0.517 seconds.\n",
      "\n",
      "Value in the initial state: 75.8204080577497\n",
      "\n",
      "Time for model checking: 0.54 seconds.\n",
      "\n",
      "Result: 75.8204080577497 (+/- 7.53630407328735E-4 estimated; rel err 9.939677543738905E-6)\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "\n",
      "Model checking: <<providerPlayer>>R{\"gas_neg\"}max=? [ Fc positive ]\n",
      "Building reward structure...\n",
      "\n",
      "Starting expected reachability...\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 36 iterations and 0.007 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 50 iterations and 0.01 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 20 iterations and 0.003 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 41 iterations and 0.008 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 45 iterations and 0.011 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 57 iterations and 0.014 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 56 iterations and 0.018 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 62 iterations and 0.012 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Starting Prob0 (maxmin)...\n",
      "target=1, inf=0, rest=372\n",
      "Starting value iteration (maxmin)...\n",
      "Value iteration (maxmin) took 2597 iterations and 0.233 seconds.\n",
      "Expected reachability took 0.324 seconds.\n",
      "\n",
      "Value in the initial state: 1322.6761737352444\n",
      "\n",
      "Time for model checking: 0.329 seconds.\n",
      "\n",
      "Result: 1322.6761737352444 (+/- 0.013225796009583712 estimated; rel err 9.999269868326118E-6)\n",
      "\n",
      "---------------------------------------------------------------------\n",
      "\n",
      "Model checking: <<providerPlayer>>R{\"gas_pos\"}max=? [ Fc positive ]\n",
      "Building reward structure...\n",
      "\n",
      "Starting expected reachability...\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 77 iterations and 0.007 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 88 iterations and 0.011 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 45 iterations and 0.004 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Prob1 (maxmin) took 2 iterations and 0.001 seconds.\n",
      "Starting Prob1 (maxmin)...\n",
      "Starting Prob0 (maxmin)...\n",
      "target=1, inf=0, rest=372\n",
      "Starting value iteration (maxmin)...\n",
      "Value iteration (maxmin) took 69 iterations and 0.007 seconds.\n",
      "Expected reachability took 0.031 seconds.\n",
      "\n",
      "Value in the initial state: 68.07028486327721\n",
      "\n",
      "Time for model checking: 0.036 seconds.\n",
      "\n",
      "Result: 68.07028486327721 (exact floating point)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "CompletedProcess(args=['/home/paul/Downloads/prism-games-3.2.1-linux64-x86/bin/prism', '/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/bpic_17_2_alergia.prism', '/home/paul/Downloads/prism-games-3.2.1-linux64-x86/generated/queries/exp_values.props'], returncode=0)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# estimate max counts after\n",
    "subprocess.run([PRISM_PATH, model_name_after, QUERY_PATH+\"exp_values.props\"]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "steps_max_before = 40#int(62/2)\n",
    "max_gas_before = int(68/2)\n",
    "min_gas_before = 50 #int(1080/10)\n",
    "\n",
    "steps_max_after = 40 #int(76/2)\n",
    "max_gas_after = int(68/2)\n",
    "min_gas_after = 50 #int(1323/17)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.886508297569364"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Query Q1 on bounded model BPIC'17-1\n",
    "query = PrismQuery(g_before, STORE_PATH, \"bpic_17_1_alergia_param.prism\", PRISM_PATH)\n",
    "results_file = query.query(QUERY_PATH+\"pos_alergia.props\", \n",
    "                           write_attributes=True, write_parameterized=True, envprob=0, \n",
    "                           steps_max=steps_max_before, min_gas=-min_gas_before, max_gas=max_gas_before)\n",
    "results_file['q0start']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.916636228569082"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Query Q1 on bounded model BPIC'17-2\n",
    "query = PrismQuery(g_after, STORE_PATH, \"bpic_17_2_alergia_param.prism\", PRISM_PATH)\n",
    "results_file = query.query(QUERY_PATH+\"pos_alergia.props\", \n",
    "                           write_attributes=True, write_parameterized=True, envprob=0, \n",
    "                           steps_max=steps_max_after, min_gas=-max_gas_after, max_gas=max_gas_after)\n",
    "results_file['q0start']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# experiment over gas (m0), steps (m1), and min_gas (m2)\n",
    "# Takes some time to execute\n",
    "PrismPrinter(g_before, STORE_PATH, \"bpic_17_1_alergia_param.prism\").write_to_prism(write_extended_parameterized=True, \n",
    "                                                                                   write_attributes=True, steps_max=steps_max_before, min_gas=-min_gas_before, max_gas=max_gas_before)\n",
    "# file_name = OUTPUT_PATH+\"bounded_steps_gas_min_gas_bpic_17-1.txt\"\n",
    "# subprocess.run([PRISM_PATH, STORE_PATH+\"bpic_17_1_alergia_param.prism\", \n",
    "#                 QUERY_PATH+\"bounded_props.props\",\n",
    "#                 \"-const\", \"m0=32:1:36,m1=12:10:32,m2=-35:5:-15,\", \"-exportresults\", file_name+\":dataframe\"])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "PrismPrinter(g_after, STORE_PATH, \"bpic_17_2_alergia_param.prism\").write_to_prism(write_extended_parameterized=True,\n",
    "                                                                                  write_attributes=True, steps_max=steps_max_after, min_gas=-min_gas_after, max_gas=max_gas_after)\n",
    "# file_name = OUTPUT_PATH+\"bounded_steps_gas_min_gas_bpic_17-2.txt\"\n",
    "# subprocess.run([PRISM_PATH, STORE_PATH+\"bpic_17_2_alergia_param.prism\", \n",
    "#                 QUERY_PATH+\"bounded_props.props\",\n",
    "#                 \"-const\", \"m0=32:1:36,m1=12:10:32,m2=-35:5:-15,\", \"-exportresults\", file_name+\":dataframe\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(0.12, 0.77, 0.85) [(32, -35), (33, -35), (34, -35)]\n",
      "34 -35\n",
      "(0.12, 0.33, 0.35) [(32, -30), (32, -25), (33, -30), (33, -25), (34, -30), (34, -25)]\n",
      "34 -25\n",
      "(0.01, 0.03, 0.03) [(32, -20), (33, -20), (34, -20)]\n",
      "34 -20\n",
      "(0.0, 0.0, 0.0) [(32, -15), (33, -15), (34, -15), (35, -35), (35, -30), (35, -25), (35, -20), (35, -15), (36, -35), (36, -30), (36, -25), (36, -20), (36, -15)]\n",
      "36 -15\n",
      "(0.19, 0.87, 0.89) [(32, -35), (33, -35), (34, -35)]\n",
      "34 -35\n",
      "(0.19, 0.35, 0.36) [(32, -30), (33, -30), (34, -30)]\n",
      "34 -30\n",
      "(0.19, 0.34, 0.35) [(32, -25), (33, -25), (34, -25)]\n",
      "34 -25\n",
      "(0.02, 0.04, 0.04) [(32, -20), (33, -20), (34, -20)]\n",
      "34 -20\n",
      "(0.0, 0.0, 0.0) [(32, -15), (33, -15), (34, -15), (35, -35), (35, -30), (35, -25), (35, -20), (35, -15), (36, -35), (36, -30), (36, -25), (36, -20), (36, -15)]\n",
      "36 -15\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# bounded reachability plot for BPIC'17\n",
    "def plot_df(df, dashed = False):\n",
    "    df_visual_grouped = df.groupby(['m0','m2'])\n",
    "    result_dict = {}\n",
    "    for g in df_visual_grouped.groups.keys():\n",
    "        r = tuple([round(h,2) for h in df_visual_grouped.get_group(g)['Result'].values])\n",
    "        if r in result_dict:\n",
    "            result_dict[r].append(g)\n",
    "        else:\n",
    "            result_dict[r] = [g]\n",
    "\n",
    "    for k in result_dict:\n",
    "        g = result_dict[k]\n",
    "        max_m0 = max(h[0] for h in result_dict[k])\n",
    "        max_m2 = max(h[1] for h in result_dict[k])\n",
    "        print(k, result_dict[k])\n",
    "        print(max_m0, max_m2)\n",
    "        assert((max_m0, max_m2) in result_dict[k])\n",
    "        if len(set(k)) != 1:\n",
    "            if dashed:\n",
    "                plt.plot(df_visual_grouped.get_group((max_m0, max_m2))['m1'], k, label = str((max_m0, max_m2)), linestyle=\"dashed\")\n",
    "            else:\n",
    "                plt.plot(df_visual_grouped.get_group((max_m0, max_m2))['m1'], k, label = str((max_m0, max_m2)))\n",
    "\n",
    "file_name = OUTPUT_PATH+\"bounded_steps_gas_min_gas_bpic_17-1.txt\"\n",
    "df_visual = pd.read_csv(file_name)\n",
    "plot_df(df_visual)\n",
    "\n",
    "file_name = OUTPUT_PATH+\"bounded_steps_gas_min_gas_bpic_17-2.txt\"\n",
    "df_visual = pd.read_csv(file_name)\n",
    "plot_df(df_visual, dashed=True)\n",
    "\n",
    "plt.legend(fontsize=10)\n",
    "plt.xlabel(\"Steps\", fontsize=18)\n",
    "plt.ylabel(\"Success probability\", fontsize=18)\n",
    "plt.xticks(fontsize=18)\n",
    "plt.yticks(fontsize=18)\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"out/bpic_bounded.png\", dpi=300)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Improvement recommendation ranking"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "skipped ['q19O_Sentmailandonline']\n",
      "skipped ['q20O_Sentonlineonly']\n",
      "skipped ['q28customerO_Sentonlineonly']\n",
      "skipped ['q27companyO_Sentonlineonly']\n",
      "skipped ['q35O_CreateOffer1']\n",
      "skipped ['q42O_Sentmailandonline']\n",
      "skipped ['q37W_CallafteroffersSHORT']\n",
      "skipped ['q36W_CallafteroffersLONG']\n",
      "skipped ['q38W_CallafteroffersSUPERLONG']\n",
      "skipped ['q40W_CallafteroffersSHORT']\n",
      "skipped ['q59O_Sentmailandonline']\n",
      "skipped ['q45companyO_CreateOffer1']\n",
      "skipped ['q46companyW_CallafteroffersLONG']\n",
      "skipped ['q47customerW_CallafteroffersLONG']\n",
      "skipped ['q48companyW_CallafteroffersSHORT']\n",
      "skipped ['q49customerW_CallafteroffersSHORT']\n",
      "skipped ['q51customerW_CallafteroffersSUPERLONG']\n",
      "skipped ['q50companyW_CallafteroffersSUPERLONG']\n",
      "skipped ['q53customerW_CallafteroffersSHORT']\n",
      "skipped ['q56customerO_Sentmailandonline']\n",
      "skipped ['q55companyO_Sentmailandonline']\n",
      "skipped ['q57O_Sentmailandonline']\n",
      "skipped ['q83O_CreateOffer2']\n",
      "skipped ['q60A_Validating']\n",
      "skipped ['q61O_Cancelled']\n",
      "skipped ['q63W_CallafteroffersSHORT']\n",
      "skipped ['q62W_CallafteroffersLONG']\n",
      "skipped ['q66companyA_Validating']\n",
      "skipped ['q67customerO_Cancelled']\n",
      "skipped ['q69companyW_CallafteroffersSHORT']\n",
      "skipped ['q74O_Returned']\n",
      "skipped ['q81customerO_Returned']\n",
      "skipped ['q85W_CallincompletefilesSHORT']\n",
      "skipped ['q93W_CallincompletefilesSUPERLONG']\n",
      "skipped ['q88companyO_CreateOffer2']\n",
      "skipped ['q90companyW_CallincompletefilesSHORT']\n",
      "skipped ['q94companyW_CallincompletefilesSUPERLONG']\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'q0start': 'env',\n",
       " 'q1companystart': 'A_CreateApplication',\n",
       " 'q2A_CreateApplication': 'env',\n",
       " 'q4customerA_CreateApplication': 'A_Submitted',\n",
       " 'q3companyA_CreateApplication': 'A_Concept',\n",
       " 'q5A_Concept': 'env',\n",
       " 'q6A_Submitted': 'env',\n",
       " 'q7companyA_Concept': 'A_Accepted',\n",
       " 'q8companyA_Submitted': 'A_Concept',\n",
       " 'q9A_Accepted': 'env',\n",
       " 'q10A_Concept': 'env',\n",
       " 'q11companyA_Accepted': 'O_CreateOffer0',\n",
       " 'q12companyA_Concept': 'A_Accepted',\n",
       " 'q13O_CreateOffer0': 'env',\n",
       " 'q14A_Accepted': 'env',\n",
       " 'q15companyO_CreateOffer0': 'O_CreateOffer1',\n",
       " 'q16customerO_CreateOffer0': 'user',\n",
       " 'q17companyA_Accepted': 'O_CreateOffer0',\n",
       " 'q18O_CreateOffer1': 'env',\n",
       " 'q21A_CancelledCUSTOMER': 'env',\n",
       " 'q22O_Cancelled': 'env',\n",
       " 'q23O_CreateOffer0': 'env',\n",
       " 'q24companyO_CreateOffer1': 'sent',\n",
       " 'q54customerO_CreateOffer1': 'user',\n",
       " 'q26customerO_Sentmailandonline': 'user',\n",
       " 'q25companyO_Sentmailandonline': 'A_Validating',\n",
       " 'q29customerA_CancelledCUSTOMER': 'user',\n",
       " 'q30companyO_Cancelled': 'O_CreateOffer2',\n",
       " 'q68customerO_Cancelled': 'user',\n",
       " 'q31companyO_CreateOffer0': 'O_CreateOffer1',\n",
       " 'q33O_Sentmailandonline': 'env',\n",
       " 'q32O_CreateOffer2': 'env',\n",
       " 'q34A_Validating': 'env',\n",
       " 'q82O_CreateOffer3': 'env',\n",
       " 'q39TIMEOUT': 'env',\n",
       " 'q86O_Cancelled': 'env',\n",
       " 'q43companyO_CreateOffer2': 'O_CreateOffer3',\n",
       " 'q44companyA_Validating': 'O_Returned',\n",
       " 'q52customerTIMEOUT': 'user',\n",
       " 'q58O_Returned': 'env',\n",
       " 'q65customerO_Returned': 'call',\n",
       " 'q64companyO_Returned': 'A_Validating',\n",
       " 'q70A_Validating': 'env',\n",
       " 'q73O_Accepted': 'env',\n",
       " 'q72W_CallincompletefilesSHORT': 'env',\n",
       " 'q71W_CallincompletefilesLONG': 'env',\n",
       " 'q75customerA_Validating': 'call',\n",
       " 'q76companyW_CallincompletefilesLONG': 'A_Validating',\n",
       " 'q77customerW_CallincompletefilesLONG': 'user',\n",
       " 'q79customerW_CallincompletefilesSHORT': 'user',\n",
       " 'q78companyW_CallincompletefilesSHORT': 'A_Validating',\n",
       " 'q80companyO_Accepted': 'A_Pending',\n",
       " 'q84A_Pending': 'env',\n",
       " 'q87companyO_CreateOffer3': 'sent',\n",
       " 'q89customerA_Pending': 'user',\n",
       " 'q91customerO_Cancelled': 'user'}"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "strategy_before = query_before.get_strategy(QUERY_PATH+\"pos_alergia.props\")\n",
    "strategy_before"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "skipped ['q19O_Sentmailandonline']\n",
      "skipped ['q20O_Sentonlineonly']\n",
      "skipped ['q27customerO_Sentmailandonline']\n",
      "skipped ['q28customerO_Sentonlineonly']\n",
      "skipped ['q46companyO_Sentonlineonly']\n",
      "skipped ['q33O_Sentmailandonline']\n",
      "skipped ['q34O_Sentonlineonly']\n",
      "skipped ['q62O_Sentmailandonline']\n",
      "skipped ['q37W_CallafteroffersSHORT']\n",
      "skipped ['q36W_CallafteroffersLONG']\n",
      "skipped ['q38W_CallafteroffersSUPERLONG']\n",
      "skipped ['q41O_CreateOffer1']\n",
      "skipped ['q43O_Sentmailandonline']\n",
      "skipped ['q45customerO_Sentmailandonline']\n",
      "skipped ['q49companyW_CallafteroffersLONG']\n",
      "skipped ['q50customerW_CallafteroffersLONG']\n",
      "skipped ['q51companyW_CallafteroffersSHORT']\n",
      "skipped ['q52customerW_CallafteroffersSHORT']\n",
      "skipped ['q54customerW_CallafteroffersSUPERLONG']\n",
      "skipped ['q53companyW_CallafteroffersSUPERLONG']\n",
      "skipped ['q56companyO_CreateOffer1']\n",
      "skipped ['q60customerO_Sentmailandonline']\n",
      "skipped ['q59companyO_Sentmailandonline']\n",
      "skipped ['q101O_Sentmailandonline']\n",
      "skipped ['q63W_CallafteroffersSHORT']\n",
      "skipped ['q67O_CreateOffer1']\n",
      "skipped ['q68A_Validating']\n",
      "skipped ['q69O_Cancelled']\n",
      "skipped ['q70A_Validating']\n",
      "skipped ['q71O_Sentmailandonline']\n",
      "skipped ['q72O_Sentmailandonline']\n",
      "skipped ['q74W_CallafteroffersSHORT']\n",
      "skipped ['q73W_CallafteroffersLONG']\n",
      "skipped ['q76customerO_Sentmailandonline']\n",
      "skipped ['q106customerW_CallafteroffersSHORT']\n",
      "skipped ['q81companyO_CreateOffer1']\n",
      "skipped ['q83customerO_CreateOffer1']\n",
      "skipped ['q82companyA_Validating']\n",
      "skipped ['q84customerO_Cancelled']\n",
      "skipped ['q85companyA_Validating']\n",
      "skipped ['q86customerO_Sentmailandonline']\n",
      "skipped ['q88customerO_Sentmailandonline']\n",
      "skipped ['q89companyW_CallafteroffersSHORT']\n",
      "skipped ['q91W_CallafteroffersSHORT']\n",
      "skipped ['q92W_CallafteroffersSHORT']\n",
      "skipped ['q97O_Returned']\n",
      "skipped ['q98O_Returned']\n",
      "skipped ['q100W_CallafteroffersSHORT']\n",
      "skipped ['q103W_CallafteroffersSHORT']\n",
      "skipped ['q104A_Validating']\n",
      "skipped ['q105companyW_CallafteroffersSHORT']\n",
      "skipped ['q124companyW_CallafteroffersSHORT']\n",
      "skipped ['q111customerO_Returned']\n",
      "skipped ['q112customerO_Returned']\n",
      "skipped ['q114customerW_CallafteroffersSHORT']\n",
      "skipped ['q115companyA_Validating']\n",
      "skipped ['q116A_Validating']\n",
      "skipped ['q118W_CallincompletefilesSUPERLONG']\n",
      "skipped ['q119O_Cancelled']\n",
      "skipped ['q120O_Returned']\n",
      "skipped ['q121companyA_Validating']\n",
      "skipped ['q123customerW_CallincompletefilesSUPERLONG']\n",
      "skipped ['q125customerO_Cancelled']\n",
      "skipped ['q126customerO_Returned']\n",
      "skipped ['q127O_Returned']\n",
      "skipped ['q129customerO_Returned']\n",
      "skipped ['q130W_CallincompletefilesSHORT']\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'q0start': 'env',\n",
       " 'q1companystart': 'A_CreateApplication',\n",
       " 'q2A_CreateApplication': 'env',\n",
       " 'q4customerA_CreateApplication': 'A_Submitted',\n",
       " 'q3companyA_CreateApplication': 'A_Concept',\n",
       " 'q5A_Concept': 'env',\n",
       " 'q6A_Submitted': 'env',\n",
       " 'q7companyA_Concept': 'A_Accepted',\n",
       " 'q8companyA_Submitted': 'A_Concept',\n",
       " 'q9A_Accepted': 'env',\n",
       " 'q10A_Concept': 'env',\n",
       " 'q11companyA_Accepted': 'O_CreateOffer0',\n",
       " 'q12companyA_Concept': 'A_Accepted',\n",
       " 'q13O_CreateOffer0': 'env',\n",
       " 'q14A_Accepted': 'env',\n",
       " 'q15companyO_CreateOffer0': 'O_CreateOffer1',\n",
       " 'q16customerO_CreateOffer0': 'user',\n",
       " 'q17companyA_Accepted': 'O_CreateOffer0',\n",
       " 'q18O_CreateOffer1': 'env',\n",
       " 'q22O_Cancelled': 'env',\n",
       " 'q21A_CancelledCUSTOMER': 'env',\n",
       " 'q39TIMEOUT': 'env',\n",
       " 'q23O_CreateOffer0': 'env',\n",
       " 'q24companyO_CreateOffer1': 'O_CreateOffer2',\n",
       " 'q25customerO_CreateOffer1': 'user',\n",
       " 'q26companyO_Sentmailandonline': 'A_Validating',\n",
       " 'q29customerA_CancelledCUSTOMER': 'user',\n",
       " 'q30companyO_Cancelled': 'sent',\n",
       " 'q31companyO_CreateOffer0': 'O_CreateOffer1',\n",
       " 'q32O_CreateOffer2': 'env',\n",
       " 'q35A_Validating': 'env',\n",
       " 'q64O_Sentmailandonline': 'env',\n",
       " 'q42O_CreateOffer1': 'env',\n",
       " 'q44companyO_CreateOffer2': 'O_CreateOffer3',\n",
       " 'q87customerO_CreateOffer2': 'user',\n",
       " 'q47companyA_Validating': 'O_Returned',\n",
       " 'q48customerA_Validating': 'call',\n",
       " 'q55customerTIMEOUT': 'user',\n",
       " 'q57companyO_CreateOffer1': 'O_CreateOffer2',\n",
       " 'q58customerO_CreateOffer1': 'user',\n",
       " 'q61O_CreateOffer3': 'env',\n",
       " 'q65O_Returned': 'env',\n",
       " 'q66W_CallincompletefilesSHORT': 'env',\n",
       " 'q99W_CallincompletefilesLONG': 'env',\n",
       " 'q96O_Accepted': 'env',\n",
       " 'q75companyO_CreateOffer3': 'sent',\n",
       " 'q77customerO_Sentmailandonline': 'user',\n",
       " 'q79customerO_Returned': 'call',\n",
       " 'q78companyO_Returned': 'A_Validating',\n",
       " 'q80companyW_CallincompletefilesSHORT': 'A_Validating',\n",
       " 'q109customerW_CallincompletefilesSHORT': 'user',\n",
       " 'q90O_Sentmailandonline': 'env',\n",
       " 'q93A_Validating': 'env',\n",
       " 'q95W_CallincompletefilesSHORT': 'env',\n",
       " 'q94W_CallincompletefilesLONG': 'env',\n",
       " 'q102O_Cancelled': 'env',\n",
       " 'q108customerW_CallincompletefilesLONG': 'user',\n",
       " 'q107companyW_CallincompletefilesLONG': 'A_Validating',\n",
       " 'q110companyO_Accepted': 'A_Pending',\n",
       " 'q113customerO_Cancelled': 'user',\n",
       " 'q117A_Pending': 'env',\n",
       " 'q122customerA_Pending': 'user'}"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "strategy_after = query_after.get_strategy(QUERY_PATH+\"pos_alergia.props\")\n",
    "strategy_after"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## BPIC'17 Sankey Diagram"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total 128.30570000000003\n",
      "(35.636199999999974, 'q33: O_Sentmailandonline') 0.2777444805647759\n",
      "(34.08, 'q79: customerW_CallincompletefilesSHORT') 0.26561563515884323\n",
      "(16.15219999999989, 'q23: O_CreateOffer0') 0.12588840558135678\n",
      "(12.139899999999983, 'q13: O_CreateOffer0') 0.09461699675072877\n",
      "(11.4897, 'q16: customerO_CreateOffer0') 0.08954941206820895\n",
      "(7.3456, 'q77: customerW_CallincompletefilesLONG') 0.057250769061701845\n",
      "(3.299900000000002, 'q22: O_Cancelled') 0.02571904443839986\n",
      "(2.3569, 'q68: customerO_Cancelled') 0.018369409932684204\n",
      "(2.0507000000001665, 'q2: A_CreateApplication') 0.015982922036980166\n",
      "(1.7941999999999942, 'q71: W_CallincompletefilesLONG') 0.013983790275880134\n",
      "(1.2792000000000088, 'q18: O_CreateOffer1') 0.009969938981666508\n",
      "(0.3160000000000107, 'q65: customerO_Returned') 0.0024628679785856015\n",
      "(0.26600000000000845, 'q72: W_CallincompletefilesSHORT') 0.002073173678176483\n",
      "(0.09920000000000329, 'q75: customerA_Validating') 0.000773153492011682\n",
      "[17.818199999999973, 17.818, 17.044799999999995, 17.0352, 8.09139999999989, 8.0608, 6.1038999999999835, 6.036, 5.745, 5.7447, 3.6736, 3.672, 1.6500000000000015, 1.6499000000000001, 1.1785, 1.1784, 1.0602000000002756, 0.9904999999998909, 0.9023999999999945, 0.8917999999999997, 0.6420000000000003, 0.6372000000000084, 0.15999999999999903, 0.15600000000001168, 0.13639999999999874, 0.1296000000000097, 0.05119999999999969, 0.048000000000003595, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0, 0.0, 0.0, 0, 0, 0, 0.0, 0, 0.0, 0.0, 0.0, 0.0, 0.0, 0, 0, 0.0, 0.0, 0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0]\n"
     ]
    }
   ],
   "source": [
    "def transform_strategy(strategy, g, printer):\n",
    "    \"\"\" \n",
    "    Adjusts the strategy naming from PRISM node naming to original node naming in g\n",
    "    \"\"\"\n",
    "    isomorphism = nx.vf2pp_isomorphism(printer.g, g, node_label=None)\n",
    "    strategy_isomorphism = {k[len(isomorphism[k].split(\": \")[0]):] : \": \".join(isomorphism[k].split(\": \")[1:]) for k in isomorphism}\n",
    "    parsed_strategy = {isomorphism[r] : strategy_isomorphism[strategy[r]] if strategy[r] not in [\"env\", \"user\", \"company\", \"sent\", \"call\"] else strategy[r] for r in strategy}\n",
    "    assert(len(parsed_strategy)==len(strategy))\n",
    "    return parsed_strategy\n",
    "\n",
    "def lost_users(g, results_file, strategy):\n",
    "    lost_users_dict = {}\n",
    "    total_lost_users_dict = {}\n",
    "    for s in strategy:\n",
    "        assert s in g\n",
    "        next_states = [t for t in g[s] if g[s][t]['action'] == strategy[s]]\n",
    "\n",
    "        assert(next_states)\n",
    "        total_lost_users = 0\n",
    "        for t in next_states:\n",
    "            action_outcome_cost = len(g[s][t]['trace_indices']) * abs(round(results_file[s],4)-round(results_file[t],4))\n",
    "            lost_users_dict[(s,t)] = action_outcome_cost\n",
    "\n",
    "            total_lost_users += action_outcome_cost\n",
    "                \n",
    "        if total_lost_users != 0:\n",
    "            total_lost_users_dict[s] = total_lost_users\n",
    "    total = sum(total_lost_users_dict.values())\n",
    "    print(\"total\", total)\n",
    "    for k in (sorted( ((v,k) for k,v in total_lost_users_dict.items()), reverse=True)):\n",
    "        print(k, k[0]/total)\n",
    "    print(sorted(lost_users_dict.values(), reverse=True))\n",
    "    return lost_users_dict\n",
    "lost_users_dict_before = lost_users(g_before, get_probs_file(results_file_before, g_before, printer_before), transform_strategy(strategy_before, g_before, printer_before))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dict_keys(['q0: start', 'q1: companystart', 'q2: A_CreateApplication', 'q4: customerA_CreateApplication', 'q3: companyA_CreateApplication', 'q5: A_Concept', 'q6: A_Submitted', 'q7: companyA_Concept', 'q8: companyA_Submitted', 'q9: A_Accepted', 'q10: A_Concept', 'q11: companyA_Accepted', 'q12: companyA_Concept', 'q13: O_CreateOffer0', 'q14: A_Accepted', 'q15: companyO_CreateOffer0', 'q16: customerO_CreateOffer0', 'q17: companyA_Accepted', 'q18: O_CreateOffer1', 'q21: A_CancelledCUSTOMER', 'q22: O_Cancelled', 'q23: O_CreateOffer0', 'q24: companyO_CreateOffer1', 'q54: customerO_CreateOffer1', 'q26: customerO_Sentmailandonline', 'q25: companyO_Sentmailandonline', 'q29: customerA_CancelledCUSTOMER', 'q30: companyO_Cancelled', 'q68: customerO_Cancelled', 'q31: companyO_CreateOffer0', 'q33: O_Sentmailandonline', 'q32: O_CreateOffer2', 'q34: A_Validating', 'q82: O_CreateOffer3', 'q39: TIMEOUT', 'q86: O_Cancelled', 'q43: companyO_CreateOffer2', 'q44: companyA_Validating', 'q52: customerTIMEOUT', 'q58: O_Returned', 'q65: customerO_Returned', 'q64: companyO_Returned', 'q70: A_Validating', 'q73: O_Accepted', 'q72: W_CallincompletefilesSHORT', 'q71: W_CallincompletefilesLONG', 'q75: customerA_Validating', 'q76: companyW_CallincompletefilesLONG', 'q77: customerW_CallincompletefilesLONG', 'q79: customerW_CallincompletefilesSHORT', 'q78: companyW_CallincompletefilesSHORT', 'q80: companyO_Accepted', 'q84: A_Pending', 'q87: companyO_CreateOffer3', 'q89: customerA_Pending', 'q91: customerO_Cancelled'])\n",
      "size start 56\n",
      "size reduced 15\n",
      "reduction mapping {'q0: start': ['q1: companystart', 'q2: A_CreateApplication'], 'q4: customerA_CreateApplication': ['q6: A_Submitted', 'q8: companyA_Submitted', 'q10: A_Concept', 'q12: companyA_Concept', 'q14: A_Accepted', 'q17: companyA_Accepted', 'q23: O_CreateOffer0'], 'q3: companyA_CreateApplication': ['q5: A_Concept', 'q7: companyA_Concept', 'q9: A_Accepted', 'q11: companyA_Accepted', 'q13: O_CreateOffer0'], 'q54: customerO_CreateOffer1': ['q22: O_Cancelled'], 'q26: customerO_Sentmailandonline': ['q21: A_CancelledCUSTOMER', 'q29: customerA_CancelledCUSTOMER', 'q39: TIMEOUT', 'q86: O_Cancelled', 'q52: customerTIMEOUT', 'q91: customerO_Cancelled'], 'q30: companyO_Cancelled': ['q24: companyO_CreateOffer1', 'q33: O_Sentmailandonline', 'q32: O_CreateOffer2', 'q43: companyO_CreateOffer2', 'q82: O_CreateOffer3', 'q87: companyO_CreateOffer3'], 'q31: companyO_CreateOffer0': ['q15: companyO_CreateOffer0', 'q18: O_CreateOffer1'], 'q73: O_Accepted': ['q80: companyO_Accepted', 'q84: A_Pending', 'q89: customerA_Pending'], 'q78: companyW_CallincompletefilesSHORT': ['q76: companyW_CallincompletefilesLONG', 'q25: companyO_Sentmailandonline', 'q34: A_Validating', 'q44: companyA_Validating', 'q58: O_Returned', 'q65: customerO_Returned', 'q64: companyO_Returned', 'q70: A_Validating', 'q75: customerA_Validating']}\n",
      "Node_list ['q0: start', 'q4: customerA_CreateApplication', 'q3: companyA_CreateApplication', 'q16: customerO_CreateOffer0', 'q54: customerO_CreateOffer1', 'q26: customerO_Sentmailandonline', 'q30: companyO_Cancelled', 'q68: customerO_Cancelled', 'q31: companyO_CreateOffer0', 'q73: O_Accepted', 'q72: W_CallincompletefilesSHORT', 'q71: W_CallincompletefilesLONG', 'q77: customerW_CallincompletefilesLONG', 'q79: customerW_CallincompletefilesSHORT', 'q78: companyW_CallincompletefilesSHORT']\n",
      "15\n",
      "lost users dict {('q0: start', 'q1: companystart'): 0.0, ('q1: companystart', 'q2: A_CreateApplication'): 0.0, ('q2: A_CreateApplication', 'q4: customerA_CreateApplication'): 1.0602000000002756, ('q2: A_CreateApplication', 'q3: companyA_CreateApplication'): 0.9904999999998909, ('q4: customerA_CreateApplication', 'q6: A_Submitted'): 0.0, ('q3: companyA_CreateApplication', 'q5: A_Concept'): 0.0, ('q5: A_Concept', 'q7: companyA_Concept'): 0.0, ('q6: A_Submitted', 'q8: companyA_Submitted'): 0.0, ('q7: companyA_Concept', 'q9: A_Accepted'): 0.0, ('q8: companyA_Submitted', 'q10: A_Concept'): 0.0, ('q9: A_Accepted', 'q11: companyA_Accepted'): 0.0, ('q10: A_Concept', 'q12: companyA_Concept'): 0.0, ('q11: companyA_Accepted', 'q13: O_CreateOffer0'): 0.0, ('q12: companyA_Concept', 'q14: A_Accepted'): 0.0, ('q13: O_CreateOffer0', 'q15: companyO_CreateOffer0'): 6.1038999999999835, ('q13: O_CreateOffer0', 'q16: customerO_CreateOffer0'): 6.036, ('q14: A_Accepted', 'q17: companyA_Accepted'): 0.0, ('q15: companyO_CreateOffer0', 'q18: O_CreateOffer1'): 0.0, ('q16: customerO_CreateOffer0', 'q21: A_CancelledCUSTOMER'): 5.745, ('q16: customerO_CreateOffer0', 'q22: O_Cancelled'): 5.7447, ('q17: companyA_Accepted', 'q23: O_CreateOffer0'): 0.0, ('q18: O_CreateOffer1', 'q24: companyO_CreateOffer1'): 0.6372000000000084, ('q18: O_CreateOffer1', 'q54: customerO_CreateOffer1'): 0.6420000000000003, ('q21: A_CancelledCUSTOMER', 'q29: customerA_CancelledCUSTOMER'): 0, ('q22: O_Cancelled', 'q30: companyO_Cancelled'): 1.6500000000000015, ('q22: O_Cancelled', 'q68: customerO_Cancelled'): 1.6499000000000001, ('q23: O_CreateOffer0', 'q31: companyO_CreateOffer0'): 8.09139999999989, ('q23: O_CreateOffer0', 'q16: customerO_CreateOffer0'): 8.0608, ('q24: companyO_CreateOffer1', 'q33: O_Sentmailandonline'): 0.0, ('q54: customerO_CreateOffer1', 'q22: O_Cancelled'): 0.0, ('q26: customerO_Sentmailandonline', 'q21: A_CancelledCUSTOMER'): 0, ('q26: customerO_Sentmailandonline', 'q39: TIMEOUT'): 0, ('q26: customerO_Sentmailandonline', 'q86: O_Cancelled'): 0, ('q25: companyO_Sentmailandonline', 'q34: A_Validating'): 0.0, ('q29: customerA_CancelledCUSTOMER', 'q41: negative'): 0, ('q30: companyO_Cancelled', 'q32: O_CreateOffer2'): 0.0, ('q68: customerO_Cancelled', 'q22: O_Cancelled'): 1.1785, ('q68: customerO_Cancelled', 'q21: A_CancelledCUSTOMER'): 1.1784, ('q31: companyO_CreateOffer0', 'q18: O_CreateOffer1'): 0.0, ('q33: O_Sentmailandonline', 'q25: companyO_Sentmailandonline'): 17.818199999999973, ('q33: O_Sentmailandonline', 'q26: customerO_Sentmailandonline'): 17.818, ('q32: O_CreateOffer2', 'q43: companyO_CreateOffer2'): 0.0, ('q34: A_Validating', 'q44: companyA_Validating'): 0.0, ('q82: O_CreateOffer3', 'q87: companyO_CreateOffer3'): 0.0, ('q39: TIMEOUT', 'q52: customerTIMEOUT'): 0, ('q86: O_Cancelled', 'q91: customerO_Cancelled'): 0, ('q43: companyO_CreateOffer2', 'q82: O_CreateOffer3'): 0.0, ('q44: companyA_Validating', 'q58: O_Returned'): 0.0, ('q52: customerTIMEOUT', 'q41: negative'): 0, ('q58: O_Returned', 'q65: customerO_Returned'): 0.0, ('q58: O_Returned', 'q64: companyO_Returned'): 0.0, ('q65: customerO_Returned', 'q72: W_CallincompletefilesSHORT'): 0.15600000000001168, ('q65: customerO_Returned', 'q71: W_CallincompletefilesLONG'): 0.15999999999999903, ('q64: companyO_Returned', 'q70: A_Validating'): 0.0, ('q70: A_Validating', 'q75: customerA_Validating'): 0.0, ('q73: O_Accepted', 'q80: companyO_Accepted'): 0.0, ('q72: W_CallincompletefilesSHORT', 'q79: customerW_CallincompletefilesSHORT'): 0.13639999999999874, ('q72: W_CallincompletefilesSHORT', 'q78: companyW_CallincompletefilesSHORT'): 0.1296000000000097, ('q71: W_CallincompletefilesLONG', 'q76: companyW_CallincompletefilesLONG'): 0.9023999999999945, ('q71: W_CallincompletefilesLONG', 'q77: customerW_CallincompletefilesLONG'): 0.8917999999999997, ('q75: customerA_Validating', 'q72: W_CallincompletefilesSHORT'): 0.048000000000003595, ('q75: customerA_Validating', 'q71: W_CallincompletefilesLONG'): 0.05119999999999969, ('q76: companyW_CallincompletefilesLONG', 'q70: A_Validating'): 0.0, ('q77: customerW_CallincompletefilesLONG', 'q73: O_Accepted'): 3.672, ('q77: customerW_CallincompletefilesLONG', 'q21: A_CancelledCUSTOMER'): 3.6736, ('q79: customerW_CallincompletefilesSHORT', 'q73: O_Accepted'): 17.044799999999995, ('q79: customerW_CallincompletefilesSHORT', 'q21: A_CancelledCUSTOMER'): 17.0352, ('q78: companyW_CallincompletefilesSHORT', 'q70: A_Validating'): 0.0, ('q80: companyO_Accepted', 'q84: A_Pending'): 0.0, ('q84: A_Pending', 'q89: customerA_Pending'): 0.0, ('q87: companyO_CreateOffer3', 'q33: O_Sentmailandonline'): 0.0, ('q89: customerA_Pending', 'q92: positive'): 0.0, ('q91: customerO_Cancelled', 'q39: TIMEOUT'): 0}\n"
     ]
    },
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "link": {
          "arrowlen": 15,
          "source": [
           0,
           0,
           2,
           2,
           3,
           3,
           8,
           8,
           4,
           4,
           1,
           1,
           7,
           7,
           6,
           6,
           14,
           14,
           10,
           10,
           11,
           11,
           14,
           14,
           12,
           12,
           13,
           13
          ],
          "target": [
           1,
           2,
           8,
           3,
           5,
           4,
           6,
           4,
           6,
           7,
           8,
           3,
           4,
           5,
           14,
           5,
           10,
           11,
           13,
           14,
           14,
           12,
           10,
           11,
           9,
           5,
           9,
           5
          ],
          "value": [
           1.0602000000002756,
           0.9904999999998909,
           6.1038999999999835,
           6.036,
           5.745,
           5.7447,
           0.6372000000000084,
           0.6420000000000003,
           1.6500000000000015,
           1.6499000000000001,
           8.09139999999989,
           8.0608,
           1.1785,
           1.1784,
           17.818199999999973,
           17.818,
           0.15600000000001168,
           0.15999999999999903,
           0.13639999999999874,
           0.1296000000000097,
           0.9023999999999945,
           0.8917999999999997,
           0.048000000000003595,
           0.05119999999999969,
           3.672,
           3.6736,
           17.044799999999995,
           17.0352
          ]
         },
         "node": {
          "align": "right",
          "color": [
           "#37ae68",
           "#35ad68",
           "#37ae68",
           "#eac533",
           "#53b662",
           "#d55e00",
           "#33ad69",
           "#c6d84b",
           "#35ad68",
           "#009e73",
           "#1aa66e",
           "#1ea76d",
           "#26a96b",
           "#18a56e",
           "#1aa66e"
          ],
          "label": [
           "Start",
           "Create Offer 0",
           "Create Offer 0",
           "customerO_CreateOffer0",
           "",
           "unsucc",
           "Sent (mail and online)",
           "customerO_Cancelled",
           "",
           "succ",
           "Call incomplete files",
           "Call incomplete files",
           "customerCall incomplete files",
           "customerCall incomplete files",
           ""
          ],
          "line": {
           "color": "black",
           "width": 0.5
          },
          "pad": 5,
          "thickness": 30
         },
         "type": "sankey"
        }
       ],
       "layout": {
        "font": {
         "size": 30
        },
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "fillpattern": {
             "fillmode": "overlay",
             "size": 10,
             "solidity": 0.2
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import plotly.graph_objects as go\n",
    "\n",
    "def flatten(d, l):\n",
    "      \"\"\"\n",
    "      Helper function to get contracted nodes dict.\n",
    "      \"\"\"\n",
    "      for e in d:\n",
    "            l.append(e)\n",
    "            if 'contraction' in d[e]:\n",
    "                  flatten(d[e]['contraction'], l)\n",
    "      return l\n",
    "\n",
    "def reduced_sankey_diagram(g, results_file, strategy, lost_users_dict):\n",
    "      g = copy.deepcopy(g)\n",
    "      print(strategy.keys())\n",
    "      g = g.subgraph(strategy.keys()) # consider only nodes seen in strategy\n",
    "      g = reduce_graph(g, results_file)\n",
    "      color_map = compute_color_map(g, results_file)\n",
    "\n",
    "      d = nx.get_node_attributes(g, \"contraction\")\n",
    "      reduction_mapping = {}\n",
    "      for k in d:\n",
    "            reduction_mapping[k] = flatten(d[k], [])\n",
    "      print(\"reduction mapping\", reduction_mapping)\n",
    "                             \n",
    "      node_list = list(g.nodes())\n",
    "      print(\"Node_list\", node_list)\n",
    "      print(len(node_list))\n",
    "      node_dict = {node_list[i] : i for i in range(len(node_list))}\n",
    "\n",
    "      print(\"lost users dict\", lost_users_dict)\n",
    "\n",
    "      edge_list = []\n",
    "      value_list = []\n",
    "      for e in [e for e in lost_users_dict if lost_users_dict[e] != 0]:\n",
    "            if e[0] in node_list:\n",
    "                  s = e[0]\n",
    "            else:\n",
    "                  s = None\n",
    "                  for h in reduction_mapping:\n",
    "                        if e[0] in reduction_mapping[h]:\n",
    "                              s = h\n",
    "                  assert s\n",
    "            if e[1] in node_list:\n",
    "                  t = e[1]\n",
    "            else:\n",
    "                  t = None\n",
    "                  for h in reduction_mapping:\n",
    "                        if e[1] in reduction_mapping[h]:\n",
    "                              t = h\n",
    "                  assert t\n",
    "            edge_list.append((s,t))\n",
    "            value_list.append(lost_users_dict[e])\n",
    "\n",
    "      labels = []\n",
    "      \n",
    "      # build label list with right states\n",
    "      for s in node_list:\n",
    "            if s not in reduction_mapping or s in lost_users_dict:\n",
    "                  label = str(s).split(\": \")[1].replace(\"LONG\", \"\").replace(\"SHORT\", \"\").replace(\"W_\", \"\").replace(\"Callincompletefiles\", 'Call incomplete files')\n",
    "                  labels.append(label)\n",
    "            elif any([\"start\" in h for h in reduction_mapping[s]]):\n",
    "                  labels.append(\"Start\")\n",
    "            elif any([\"q30\" in h for h in reduction_mapping[s]]) or \"q30\" in s:\n",
    "                  labels.append(\"Sent (mail and online)\")\n",
    "            elif any([\"q23\" in h for h in reduction_mapping[s]]) or \"q23\" in s:\n",
    "                  labels.append(\"Create Offer 0\")\n",
    "            elif any([\"q13\" in h for h in reduction_mapping[s]]) or \"q13\" in s:\n",
    "                  labels.append(\"Create Offer 0\")\n",
    "            elif results_file[s] == 1:\n",
    "                  labels.append(\"succ\")\n",
    "            elif results_file[s] == 0:\n",
    "                  labels.append(\"unsucc\")\n",
    "            else:\n",
    "                  # s is in reduction_mapping and not in lost_users_dict\n",
    "                  labels.append(\"\")\n",
    "                  #labels.append(\"C\"+str(s).split(\": \")[0])\n",
    "\n",
    "      fig = go.Figure(data=[go.Sankey(\n",
    "      node = dict(\n",
    "          pad = 5,\n",
    "          thickness = 30,\n",
    "          line = dict(color = \"black\", width = 0.5),\n",
    "          label = labels,\n",
    "          color = [color_map[s] for s in node_list],\n",
    "          align = \"right\"\n",
    "      ),\n",
    "      link = dict(\n",
    "          source = [node_dict[e[0]] for e in edge_list],\n",
    "          target = [node_dict[e[1]] for e in edge_list],\n",
    "          value = value_list,\n",
    "          arrowlen=15,\n",
    "      ))])\n",
    "      fig.update_layout(\n",
    "        font=dict(size = 30)\n",
    "    )\n",
    "      fig.write_image(\"out/before_sankey.png\")\n",
    "      fig.show()\n",
    "\n",
    "\n",
    "reduced_sankey_diagram(g_before, get_probs_file(results_file_before, g_before, printer_before), transform_strategy(strategy_before, g_before, printer_before), lost_users_dict_before)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total 252.42059999999964\n",
      "(124.98580000000001, 'q64: O_Sentmailandonline') 0.49514896961658517\n",
      "(31.464500000000047, 'q109: customerW_CallincompletefilesSHORT') 0.1246510783985146\n",
      "(27.15919999999985, 'q13: O_CreateOffer0') 0.10759502195938005\n",
      "(22.321199999999756, 'q23: O_CreateOffer0') 0.08842859893368366\n",
      "(12.212500000000002, 'q42: O_CreateOffer1') 0.04838155047567441\n",
      "(12.014899999999995, 'q2: A_CreateApplication') 0.04759873005610482\n",
      "(7.652000000000003, 'q108: customerW_CallincompletefilesLONG') 0.030314483049323288\n",
      "(3.3796, 'q16: customerO_CreateOffer0') 0.013388764625391132\n",
      "(3.0944999999999983, 'q18: O_CreateOffer1') 0.012259300548370468\n",
      "(2.2316000000000056, 'q99: W_CallincompletefilesLONG') 0.008840799839632775\n",
      "(2.1107000000000014, 'q32: O_CreateOffer2') 0.00836183734608033\n",
      "(1.2427000000000041, 'q79: customerO_Returned') 0.004923132264165468\n",
      "(1.0996999999999535, 'q95: W_CallincompletefilesSHORT') 0.00435661748684519\n",
      "(0.6542000000000012, 'q94: W_CallincompletefilesLONG') 0.0025917060651943708\n",
      "(0.44560000000000766, 'q48: customerA_Validating') 0.0017653075858309834\n",
      "(0.24869999999998993, 'q66: W_CallincompletefilesSHORT') 0.000985260315520961\n",
      "(0.09329999999999139, 'q35: A_Validating') 0.0003696211798878203\n",
      "(0.009899999999999909, 'q65: O_Returned') 3.92202538144665e-05\n",
      "[62.500600000000006, 62.4852, 15.7488, 15.715700000000048, 13.57999999999985, 13.5792, 11.337499999999759, 10.983699999999999, 6.1128, 6.102500000000141, 6.099700000000002, 5.912399999999854, 3.8280000000000034, 3.824, 1.69, 1.6192000000000002, 1.5489, 1.5455999999999983, 1.1179999999999994, 1.1136000000000061, 1.0557000000000012, 1.055, 0.6270000000000042, 0.6157, 0.5620999999999381, 0.5376000000000154, 0.32760000000000145, 0.3265999999999998, 0.22480000000000644, 0.22080000000000122, 0.12480000000000357, 0.12389999999998635, 0.07829999999999138, 0.0704, 0.015000000000000013, 0.009899999999999909, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0, 0, 0.0, 0.0, 0.0, 0, 0.0, 0.0, 0.0, 0.0, 0.0, 0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0, 0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]\n"
     ]
    }
   ],
   "source": [
    "lost_users_dict_after = lost_users(g_after, get_probs_file(results_file_after, g_after, printer_after), transform_strategy(strategy_after, g_after, printer_after))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dict_keys(['q0: start', 'q1: companystart', 'q2: A_CreateApplication', 'q4: customerA_CreateApplication', 'q3: companyA_CreateApplication', 'q5: A_Concept', 'q6: A_Submitted', 'q7: companyA_Concept', 'q8: companyA_Submitted', 'q9: A_Accepted', 'q10: A_Concept', 'q11: companyA_Accepted', 'q12: companyA_Concept', 'q13: O_CreateOffer0', 'q14: A_Accepted', 'q15: companyO_CreateOffer0', 'q16: customerO_CreateOffer0', 'q17: companyA_Accepted', 'q18: O_CreateOffer1', 'q22: O_Cancelled', 'q21: A_CancelledCUSTOMER', 'q39: TIMEOUT', 'q23: O_CreateOffer0', 'q24: companyO_CreateOffer1', 'q25: customerO_CreateOffer1', 'q26: companyO_Sentmailandonline', 'q29: customerA_CancelledCUSTOMER', 'q30: companyO_Cancelled', 'q31: companyO_CreateOffer0', 'q32: O_CreateOffer2', 'q35: A_Validating', 'q64: O_Sentmailandonline', 'q42: O_CreateOffer1', 'q44: companyO_CreateOffer2', 'q87: customerO_CreateOffer2', 'q47: companyA_Validating', 'q48: customerA_Validating', 'q55: customerTIMEOUT', 'q57: companyO_CreateOffer1', 'q58: customerO_CreateOffer1', 'q61: O_CreateOffer3', 'q65: O_Returned', 'q66: W_CallincompletefilesSHORT', 'q99: W_CallincompletefilesLONG', 'q96: O_Accepted', 'q75: companyO_CreateOffer3', 'q77: customerO_Sentmailandonline', 'q79: customerO_Returned', 'q78: companyO_Returned', 'q80: companyW_CallincompletefilesSHORT', 'q109: customerW_CallincompletefilesSHORT', 'q90: O_Sentmailandonline', 'q93: A_Validating', 'q95: W_CallincompletefilesSHORT', 'q94: W_CallincompletefilesLONG', 'q102: O_Cancelled', 'q108: customerW_CallincompletefilesLONG', 'q107: companyW_CallincompletefilesLONG', 'q110: companyO_Accepted', 'q113: customerO_Cancelled', 'q117: A_Pending', 'q122: customerA_Pending'])\n",
      "size start 62\n",
      "size reduced 19\n",
      "reduction mapping {'q31: companyO_CreateOffer0': ['q42: O_CreateOffer1'], 'q47: companyA_Validating': ['q65: O_Returned', 'q79: customerO_Returned'], 'q4: customerA_CreateApplication': ['q6: A_Submitted', 'q8: companyA_Submitted', 'q10: A_Concept', 'q12: companyA_Concept', 'q14: A_Accepted', 'q17: companyA_Accepted', 'q23: O_CreateOffer0'], 'q0: start': ['q1: companystart', 'q2: A_CreateApplication'], 'q44: companyO_CreateOffer2': ['q61: O_CreateOffer3', 'q75: companyO_CreateOffer3', 'q90: O_Sentmailandonline', 'q26: companyO_Sentmailandonline', 'q35: A_Validating'], 'q15: companyO_CreateOffer0': ['q18: O_CreateOffer1'], 'q3: companyA_CreateApplication': ['q5: A_Concept', 'q7: companyA_Concept', 'q9: A_Accepted', 'q11: companyA_Accepted', 'q13: O_CreateOffer0'], 'q78: companyO_Returned': ['q107: companyW_CallincompletefilesLONG', 'q80: companyW_CallincompletefilesSHORT', 'q93: A_Validating', 'q48: customerA_Validating'], 'q24: companyO_CreateOffer1': ['q57: companyO_CreateOffer1', 'q32: O_CreateOffer2'], 'q77: customerO_Sentmailandonline': ['q39: TIMEOUT', 'q55: customerTIMEOUT', 'q21: A_CancelledCUSTOMER', 'q29: customerA_CancelledCUSTOMER'], 'q96: O_Accepted': ['q110: companyO_Accepted', 'q117: A_Pending', 'q122: customerA_Pending'], 'q87: customerO_CreateOffer2': ['q102: O_Cancelled', 'q113: customerO_Cancelled', 'q58: customerO_CreateOffer1', 'q25: customerO_CreateOffer1', 'q22: O_Cancelled', 'q30: companyO_Cancelled', 'q64: O_Sentmailandonline']}\n",
      "Node_list ['q108: customerW_CallincompletefilesLONG', 'q31: companyO_CreateOffer0', 'q47: companyA_Validating', 'q4: customerA_CreateApplication', 'q16: customerO_CreateOffer0', 'q0: start', 'q66: W_CallincompletefilesSHORT', 'q44: companyO_CreateOffer2', 'q15: companyO_CreateOffer0', 'q94: W_CallincompletefilesLONG', 'q3: companyA_CreateApplication', 'q78: companyO_Returned', 'q24: companyO_CreateOffer1', 'q77: customerO_Sentmailandonline', 'q99: W_CallincompletefilesLONG', 'q109: customerW_CallincompletefilesSHORT', 'q96: O_Accepted', 'q87: customerO_CreateOffer2', 'q95: W_CallincompletefilesSHORT']\n",
      "19\n",
      "dict {('q0: start', 'q1: companystart'): 0.0, ('q1: companystart', 'q2: A_CreateApplication'): 0.0, ('q2: A_CreateApplication', 'q4: customerA_CreateApplication'): 5.912399999999854, ('q2: A_CreateApplication', 'q3: companyA_CreateApplication'): 6.102500000000141, ('q4: customerA_CreateApplication', 'q6: A_Submitted'): 0.0, ('q3: companyA_CreateApplication', 'q5: A_Concept'): 0.0, ('q5: A_Concept', 'q7: companyA_Concept'): 0.0, ('q6: A_Submitted', 'q8: companyA_Submitted'): 0.0, ('q7: companyA_Concept', 'q9: A_Accepted'): 0.0, ('q8: companyA_Submitted', 'q10: A_Concept'): 0.0, ('q9: A_Accepted', 'q11: companyA_Accepted'): 0.0, ('q10: A_Concept', 'q12: companyA_Concept'): 0.0, ('q11: companyA_Accepted', 'q13: O_CreateOffer0'): 0.0, ('q12: companyA_Concept', 'q14: A_Accepted'): 0.0, ('q13: O_CreateOffer0', 'q15: companyO_CreateOffer0'): 13.57999999999985, ('q13: O_CreateOffer0', 'q16: customerO_CreateOffer0'): 13.5792, ('q14: A_Accepted', 'q17: companyA_Accepted'): 0.0, ('q15: companyO_CreateOffer0', 'q18: O_CreateOffer1'): 0.0, ('q16: customerO_CreateOffer0', 'q22: O_Cancelled'): 1.69, ('q16: customerO_CreateOffer0', 'q21: A_CancelledCUSTOMER'): 1.6192000000000002, ('q16: customerO_CreateOffer0', 'q39: TIMEOUT'): 0.0704, ('q17: companyA_Accepted', 'q23: O_CreateOffer0'): 0.0, ('q18: O_CreateOffer1', 'q24: companyO_CreateOffer1'): 1.5455999999999983, ('q18: O_CreateOffer1', 'q25: customerO_CreateOffer1'): 1.5489, ('q22: O_Cancelled', 'q30: companyO_Cancelled'): 0.0, ('q21: A_CancelledCUSTOMER', 'q29: customerA_CancelledCUSTOMER'): 0, ('q39: TIMEOUT', 'q55: customerTIMEOUT'): 0, ('q23: O_CreateOffer0', 'q31: companyO_CreateOffer0'): 11.337499999999759, ('q23: O_CreateOffer0', 'q16: customerO_CreateOffer0'): 10.983699999999999, ('q24: companyO_CreateOffer1', 'q32: O_CreateOffer2'): 0.0, ('q25: customerO_CreateOffer1', 'q22: O_Cancelled'): 0.0, ('q26: companyO_Sentmailandonline', 'q35: A_Validating'): 0.0, ('q29: customerA_CancelledCUSTOMER', 'q40: negative'): 0, ('q30: companyO_Cancelled', 'q64: O_Sentmailandonline'): 0.0, ('q31: companyO_CreateOffer0', 'q42: O_CreateOffer1'): 0.0, ('q32: O_CreateOffer2', 'q44: companyO_CreateOffer2'): 1.0557000000000012, ('q32: O_CreateOffer2', 'q87: customerO_CreateOffer2'): 1.055, ('q35: A_Validating', 'q47: companyA_Validating'): 0.07829999999999138, ('q35: A_Validating', 'q48: customerA_Validating'): 0.015000000000000013, ('q64: O_Sentmailandonline', 'q77: customerO_Sentmailandonline'): 62.4852, ('q64: O_Sentmailandonline', 'q26: companyO_Sentmailandonline'): 62.500600000000006, ('q42: O_CreateOffer1', 'q57: companyO_CreateOffer1'): 6.099700000000002, ('q42: O_CreateOffer1', 'q58: customerO_CreateOffer1'): 6.1128, ('q44: companyO_CreateOffer2', 'q61: O_CreateOffer3'): 0.0, ('q87: customerO_CreateOffer2', 'q102: O_Cancelled'): 0.0, ('q47: companyA_Validating', 'q65: O_Returned'): 0.0, ('q48: customerA_Validating', 'q66: W_CallincompletefilesSHORT'): 0.22480000000000644, ('q48: customerA_Validating', 'q99: W_CallincompletefilesLONG'): 0.22080000000000122, ('q55: customerTIMEOUT', 'q40: negative'): 0, ('q57: companyO_CreateOffer1', 'q32: O_CreateOffer2'): 0.0, ('q58: customerO_CreateOffer1', 'q22: O_Cancelled'): 0.0, ('q61: O_CreateOffer3', 'q75: companyO_CreateOffer3'): 0.0, ('q65: O_Returned', 'q79: customerO_Returned'): 0.0, ('q65: O_Returned', 'q78: companyO_Returned'): 0.009899999999999909, ('q66: W_CallincompletefilesSHORT', 'q80: companyW_CallincompletefilesSHORT'): 0.12480000000000357, ('q66: W_CallincompletefilesSHORT', 'q109: customerW_CallincompletefilesSHORT'): 0.12389999999998635, ('q99: W_CallincompletefilesLONG', 'q107: companyW_CallincompletefilesLONG'): 1.1136000000000061, ('q99: W_CallincompletefilesLONG', 'q108: customerW_CallincompletefilesLONG'): 1.1179999999999994, ('q96: O_Accepted', 'q110: companyO_Accepted'): 0.0, ('q75: companyO_CreateOffer3', 'q90: O_Sentmailandonline'): 0.0, ('q77: customerO_Sentmailandonline', 'q39: TIMEOUT'): 0, ('q77: customerO_Sentmailandonline', 'q21: A_CancelledCUSTOMER'): 0, ('q79: customerO_Returned', 'q95: W_CallincompletefilesSHORT'): 0.6270000000000042, ('q79: customerO_Returned', 'q94: W_CallincompletefilesLONG'): 0.6157, ('q78: companyO_Returned', 'q93: A_Validating'): 0.0, ('q80: companyW_CallincompletefilesSHORT', 'q93: A_Validating'): 0.0, ('q109: customerW_CallincompletefilesSHORT', 'q96: O_Accepted'): 15.715700000000048, ('q109: customerW_CallincompletefilesSHORT', 'q21: A_CancelledCUSTOMER'): 15.7488, ('q90: O_Sentmailandonline', 'q26: companyO_Sentmailandonline'): 0.0, ('q93: A_Validating', 'q48: customerA_Validating'): 0.0, ('q95: W_CallincompletefilesSHORT', 'q80: companyW_CallincompletefilesSHORT'): 0.5376000000000154, ('q95: W_CallincompletefilesSHORT', 'q109: customerW_CallincompletefilesSHORT'): 0.5620999999999381, ('q94: W_CallincompletefilesLONG', 'q108: customerW_CallincompletefilesLONG'): 0.32760000000000145, ('q94: W_CallincompletefilesLONG', 'q107: companyW_CallincompletefilesLONG'): 0.3265999999999998, ('q102: O_Cancelled', 'q113: customerO_Cancelled'): 0.0, ('q108: customerW_CallincompletefilesLONG', 'q96: O_Accepted'): 3.8280000000000034, ('q108: customerW_CallincompletefilesLONG', 'q21: A_CancelledCUSTOMER'): 3.824, ('q107: companyW_CallincompletefilesLONG', 'q93: A_Validating'): 0.0, ('q110: companyO_Accepted', 'q117: A_Pending'): 0.0, ('q113: customerO_Cancelled', 'q22: O_Cancelled'): 0.0, ('q117: A_Pending', 'q122: customerA_Pending'): 0.0, ('q122: customerA_Pending', 'q128: positive'): 0.0}\n"
     ]
    },
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "plotlyServerURL": "https://plot.ly"
       },
       "data": [
        {
         "link": {
          "arrowlen": 15,
          "source": [
           5,
           5,
           10,
           10,
           4,
           4,
           4,
           8,
           8,
           3,
           3,
           12,
           12,
           7,
           7,
           17,
           17,
           1,
           1,
           11,
           11,
           2,
           6,
           6,
           14,
           14,
           2,
           2,
           15,
           15,
           18,
           18,
           9,
           9,
           0,
           0
          ],
          "target": [
           3,
           10,
           8,
           4,
           17,
           13,
           13,
           12,
           17,
           1,
           4,
           7,
           17,
           2,
           11,
           13,
           7,
           12,
           17,
           6,
           14,
           11,
           11,
           15,
           11,
           0,
           18,
           9,
           16,
           13,
           11,
           15,
           0,
           11,
           16,
           13
          ],
          "value": [
           5.912399999999854,
           6.102500000000141,
           13.57999999999985,
           13.5792,
           1.69,
           1.6192000000000002,
           0.0704,
           1.5455999999999983,
           1.5489,
           11.337499999999759,
           10.983699999999999,
           1.0557000000000012,
           1.055,
           0.07829999999999138,
           0.015000000000000013,
           62.4852,
           62.500600000000006,
           6.099700000000002,
           6.1128,
           0.22480000000000644,
           0.22080000000000122,
           0.009899999999999909,
           0.12480000000000357,
           0.12389999999998635,
           1.1136000000000061,
           1.1179999999999994,
           0.6270000000000042,
           0.6157,
           15.715700000000048,
           15.7488,
           0.5376000000000154,
           0.5620999999999381,
           0.32760000000000145,
           0.3265999999999998,
           3.8280000000000034,
           3.824
          ]
         },
         "node": {
          "align": "right",
          "color": [
           "#15a46f",
           "#28aa6b",
           "#08a071",
           "#28aa6b",
           "#d97109",
           "#28aa6b",
           "#08a071",
           "#08a071",
           "#24a86c",
           "#0fa270",
           "#26a96b",
           "#08a071",
           "#1ea76d",
           "#d55e00",
           "#0ba171",
           "#08a071",
           "#009e73",
           "#ebcb36",
           "#08a071"
          ],
          "label": [
           "customerCall incomplete files",
           "Create Offer 1",
           "",
           "Create Offer 0",
           "customerO_CreateOffer0",
           "Start",
           "Call incomplete files",
           "",
           "",
           "Call incomplete files",
           "Create Offer 0",
           "",
           "Create Offer 2",
           "unsucc",
           "Call incomplete files",
           "customerCall incomplete files",
           "succ",
           "Sent (mail and online)",
           "Call incomplete files"
          ],
          "line": {
           "color": "black",
           "width": 0.5
          },
          "pad": 10,
          "thickness": 30
         },
         "type": "sankey"
        }
       ],
       "layout": {
        "font": {
         "size": 30
        },
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             },
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "pattern": {
              "fillmode": "overlay",
              "size": 10,
              "solidity": 0.2
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "fillpattern": {
             "fillmode": "overlay",
             "size": 10,
             "solidity": 0.2
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "autotypenumbers": "strict",
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        }
       }
      }
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def reduced_sankey_diagram(g, results_file, strategy, lost_users_dict):\n",
    "      g = copy.deepcopy(g)\n",
    "      print(strategy.keys())\n",
    "      g = g.subgraph(strategy.keys()) # consider only nodes seen in strategy\n",
    "      g = reduce_graph(g, results_file)\n",
    "      color_map = compute_color_map(g, results_file)\n",
    "\n",
    "      d = nx.get_node_attributes(g, \"contraction\")\n",
    "      reduction_mapping = {}\n",
    "      for k in d:\n",
    "            reduction_mapping[k] = flatten(d[k], [])\n",
    "      print(\"reduction mapping\", reduction_mapping)\n",
    "                             \n",
    "      node_list = list(g.nodes())\n",
    "      print(\"Node_list\", node_list)\n",
    "      print(len(node_list))\n",
    "      node_dict = {node_list[i] : i for i in range(len(node_list))}\n",
    "\n",
    "      print(\"dict\", lost_users_dict)\n",
    "\n",
    "      edge_list = []\n",
    "      value_list = []\n",
    "      for e in [e for e in lost_users_dict if lost_users_dict[e] != 0]:\n",
    "            if e[0] in node_list:\n",
    "                  s = e[0]\n",
    "            else:\n",
    "                  s = None\n",
    "                  for h in reduction_mapping:\n",
    "                        if e[0] in reduction_mapping[h]:\n",
    "                              s = h\n",
    "                  assert s\n",
    "            if e[1] in node_list:\n",
    "                  t = e[1]\n",
    "            else:\n",
    "                  t = None\n",
    "                  for h in reduction_mapping:\n",
    "                        if e[1] in reduction_mapping[h]:\n",
    "                              t = h\n",
    "                  assert t\n",
    "            edge_list.append((s,t))\n",
    "            value_list.append(lost_users_dict[e])\n",
    "\n",
    "      labels = []\n",
    "\n",
    "      # build label list with right states\n",
    "      for s in node_list:\n",
    "            if s not in reduction_mapping or s in lost_users_dict:\n",
    "                  label = str(s).split(\": \")[1].replace(\"LONG\", \"\").replace(\"SHORT\", \"\").replace(\"W_\", \"\").replace(\"Callincompletefiles\", 'Call incomplete files')\n",
    "                  labels.append(label)\n",
    "            elif any([\"start\" in h for h in reduction_mapping[s]]):\n",
    "                  labels.append(\"Start\")\n",
    "            elif any([\"q64\" in h for h in reduction_mapping[s]]) or \"q64\" in s:\n",
    "                  labels.append(\"Sent (mail and online)\")\n",
    "            elif any([\"q23\" in h for h in reduction_mapping[s]]) or \"q23\" in s:\n",
    "                  labels.append(\"Create Offer 0\")\n",
    "            elif any([\"q13\" in h for h in reduction_mapping[s]]) or \"q13\" in s:\n",
    "                  labels.append(\"Create Offer 0\")\n",
    "            elif any([\"q42\" in h for h in reduction_mapping[s]]) or \"q42\" in s:\n",
    "                  labels.append(\"Create Offer 1\")\n",
    "            elif any([\"q32\" in h for h in reduction_mapping[s]]) or \"q42\" in s:\n",
    "                  labels.append(\"Create Offer 2\")\n",
    "            elif results_file[s] == 1:\n",
    "                  labels.append(\"succ\")\n",
    "            elif results_file[s] == 0:\n",
    "                  labels.append(\"unsucc\")\n",
    "            else:\n",
    "                  # s is in reduction_mapping and not in lost_users_dict\n",
    "                  labels.append(\"\")\n",
    "                  #labels.append(\"C\"+str(s).split(\": \")[0])\n",
    "\n",
    "      fig = go.Figure(data=[go.Sankey(\n",
    "      node = dict(\n",
    "          pad = 10,\n",
    "          thickness = 30,\n",
    "          line = dict(color = \"black\", width = 0.5),\n",
    "          label = labels,\n",
    "          color = [color_map[s] for s in node_list],\n",
    "          align = \"right\"\n",
    "      ),\n",
    "      link = dict(\n",
    "          source = [node_dict[e[0]] for e in edge_list],\n",
    "          target = [node_dict[e[1]] for e in edge_list],\n",
    "          value = value_list,\n",
    "          arrowlen=15,\n",
    "      ))])\n",
    "      fig.update_layout(\n",
    "        font=dict(size = 30)\n",
    "    )\n",
    "      fig.write_image(\"out/after_sankey.png\")\n",
    "      fig.show()\n",
    "\n",
    "reduced_sankey_diagram(g_after, get_probs_file(results_file_after, g_after, printer_after),  transform_strategy(strategy_after, g_after, printer_after), lost_users_dict_after)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "neg 2907\n",
      "pos 2608\n"
     ]
    }
   ],
   "source": [
    "def total_lost_users(g):\n",
    "    for s in g:\n",
    "        if \"negative\" in s:\n",
    "            print(\"neg\", sum([len(set(g.edges[e]['trace_indices'])) for e in g.in_edges(s)]))\n",
    "        if \"positive\" in s:\n",
    "            print(\"pos\", sum([len(set(g.edges[e]['trace_indices'])) for e in g.in_edges(s)]))\n",
    "\n",
    "total_lost_users(g_before)"
   ]
  }
 ],
 "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": "3.10.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}