notebooks/paper/winner_plots.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n"
     ]
    }
   ],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload\n",
    "\n",
    "from retro_branching.utils import get_most_recent_checkpoint_foldername, PlotAesthetics\n",
    "\n",
    "import glob\n",
    "import gzip\n",
    "import pickle\n",
    "from collections import defaultdict\n",
    "import copy\n",
    "import numpy as np\n",
    "import scipy.stats as st\n",
    "import math\n",
    "from ordered_set import OrderedSet\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import pandas as pd\n",
    "from tabulate import tabulate"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Config"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [],
   "source": [
    "# scip_params = 'default'\n",
    "scip_params = 'gasse_2019'\n",
    "\n",
    "\n",
    "# # SET COVER \n",
    "# n_rows=100 n_cols=100\n",
    "agent_to_path = {'IL': '/scratch/datasets/retro_branching/instances/set_covering_n_rows_100_n_cols_100/baselines/gnn_341_checkpoint_120/rl_validator/rl_validator_1/',\n",
    "                 'RL': '/scratch/datasets/retro_branching/instances/set_covering_n_rows_100_n_cols_100/baselines/dqn_gnn_1094_checkpoint_108/rl_validator/rl_validator_1/',}\n",
    "\n",
    "# # n_rows=165 n_cols=230\n",
    "# agent_to_path = {'IL': '/scratch/datasets/retro_branching/instances/set_covering_n_rows_165_n_cols_230/baselines/gnn_356_checkpoint_104/rl_validator/rl_validator_1/',\n",
    "#                  'RL': '/scratch/datasets/retro_branching/instances/set_covering_n_rows_165_n_cols_230/baselines/dqn_gnn_1407_checkpoint_370/rl_validator/rl_validator_1/',}\n",
    "agent_to_path = {\n",
    "                 'IL': f'/scratch/datasets/retro_branching/instances/set_covering_n_rows_165_n_cols_230/scip_{scip_params}/baselines/gnn_356_checkpoint_104/rl_validator/rl_validator_1/',\n",
    "                 'RL': f'/scratch/datasets/retro_branching/instances/set_covering_n_rows_165_n_cols_230/scip_{scip_params}/baselines/dqn_gnn_1491_checkpoint_40/rl_validator/rl_validator_1/',\n",
    "                }\n",
    "\n",
    "# # n_rows=500 n_cols=1000\n",
    "# agent_to_path = {'IL': '/scratch/datasets/retro_branching/instances/set_covering_n_rows_500_n_cols_1000/baselines/gnn_343_checkpoint_233/rl_validator/rl_validator_1/',\n",
    "#                  'RL': '/scratch/datasets/retro_branching/instances/set_covering_n_rows_500_n_cols_1000/baselines/dqn_gnn_1236_checkpoint_457/rl_validator/rl_validator_1/',}\n",
    "# agent_to_path = {'RL': '/scratch/datasets/retro_branching/instances/set_covering_n_rows_500_n_cols_1000/baselines/dqn_gnn_1094_checkpoint_108/rl_validator/rl_validator_1/',\n",
    "#                  'SL': '/scratch/datasets/retro_branching/instances/set_covering_n_rows_500_n_cols_1000/baselines/gnn_341_checkpoint_120/rl_validator/rl_validator_1/'}\n",
    "agent_to_path = {\n",
    "                 'IL': f'/scratch/datasets/retro_branching/instances/set_covering_n_rows_500_n_cols_1000/scip_{scip_params}/baselines/gnn_361_checkpoint_139/rl_validator/rl_validator_1/',\n",
    "#                  'RL': f'/scratch/datasets/retro_branching/instances/set_covering_n_rows_500_n_cols_1000/scip_{scip_params}/baselines/dqn_gnn_1481_checkpoint_129/rl_validator/rl_validator_1/',\n",
    "                 'RL': f'/scratch/datasets/retro_branching/instances/set_covering_n_rows_500_n_cols_1000/scip_{scip_params}/baselines/dqn_gnn_1481_checkpoint_166/rl_validator/rl_validator_1/',\n",
    "                }\n",
    "\n",
    "\n",
    "\n",
    "# # COMBINATORIAL AUCTION\n",
    "# # n_items=10 n_bids=50\n",
    "# agent_to_path = {\n",
    "#                  'RL': '/scratch/datasets/retro_branching/instances/combinatorial_auction_n_items_10_n_bids_50/baselines/dqn_gnn_1379_checkpoint_73/rl_validator/rl_validator_1/',\n",
    "#                  'IL': '/scratch/datasets/retro_branching/instances/combinatorial_auction_n_items_10_n_bids_50/baselines/gnn_347_checkpoint_124/rl_validator/rl_validator_1/',\n",
    "#                 }\n",
    "# # n_rows=500 n_cols=1000 sub-tree ablation\n",
    "# agent_to_path = {'Whole tree': '/scratch/datasets/retro_branching/instances/set_covering_n_rows_500_n_cols_1000/baselines/dqn_gnn_1405_checkpoint_111/rl_validator/rl_validator_1/',\n",
    "#                  'Sub-trees': '/scratch/datasets/retro_branching/instances/set_covering_n_rows_500_n_cols_1000/baselines/dqn_gnn_1236_checkpoint_457/rl_validator/rl_validator_1/'\n",
    "#                 }\n",
    "\n",
    "\n",
    "# # CAPACITATED FACILITY LOCATION\n",
    "# # n_customers=5 n_facilities=5\n",
    "# agent_to_path = {\n",
    "#                  'RL': '/scratch/datasets/retro_branching/instances/capacitated_facility_location_n_customers_5_n_facilities_5/baselines/dqn_gnn_1382_checkpoint_10/rl_validator/rl_validator_1/',\n",
    "#                  'IL': '/scratch/datasets/retro_branching/instances/capacitated_facility_location_n_customers_5_n_facilities_5/baselines/gnn_350_checkpoint_104/rl_validator/rl_validator_1/',\n",
    "#                 }\n",
    "\n",
    "\n",
    "\n",
    "# # MAXIMUM INDEPENDENT SET\n",
    "# # n_nodes=25\n",
    "# agent_to_path = {\n",
    "#                  'RL': '/scratch/datasets/retro_branching/instances/maximum_independent_set_n_nodes_25/baselines/dqn_gnn_1383_checkpoint_101/rl_validator/rl_validator_1/',\n",
    "#                  'IL': '/scratch/datasets/retro_branching/instances/maximum_independent_set_n_nodes_25/baselines/gnn_353_checkpoint_209/rl_validator/rl_validator_1/',\n",
    "#                 }\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "baseline_agent_idx = 0 # 'SL' 'Whole Tree' use as normaliser agent\n",
    "baseline_metric = 'num_nodes' # 'num_nodes' 'calibrated_solve_time' 'lp_iterations'\n",
    "use_percentages = True # if True, normalise w/ percentages, else normalised w/ fractions\n",
    "\n",
    "verbose = True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Baseline agent: IL | Competitor agent: RL\n"
     ]
    }
   ],
   "source": [
    "baseline_agent = list(agent_to_path.keys())[baseline_agent_idx]\n",
    "competitor_agent = list(agent_to_path.keys())[abs(1-baseline_agent_idx)]\n",
    "\n",
    "print(f'Baseline agent: {baseline_agent} | Competitor agent: {competitor_agent}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Core scripts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Num evaluation instances solved in IL agent log file: 100\n",
      "Num evaluation instances solved in RL agent log file: 100\n"
     ]
    }
   ],
   "source": [
    "# load agent logs\n",
    "agent_to_log = {}\n",
    "agent_name_to_display_name = {}\n",
    "for display_name, path in agent_to_path.items():\n",
    "    path += get_most_recent_checkpoint_foldername(path)\n",
    "    with gzip.open(*glob.glob(path+'/*log.pkl'), 'rb') as f:\n",
    "        log = pickle.load(f)\n",
    "    agent_name = log['agent_names'][0]\n",
    "    agent_to_log[agent_name] = log\n",
    "    agent_name_to_display_name[agent_name] = display_name\n",
    "    \n",
    "    num_instances = len(log[agent_name][\"num_nodes\"])\n",
    "    print(f'Num evaluation instances solved in {display_name} agent log file: {num_instances}')\n",
    "display_name_to_agent_name = {val: key for key, val in agent_name_to_display_name.items()}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get stats for each agent\n",
    "agent_to_stats = {agent: defaultdict(list) for agent in agent_to_log.keys()}\n",
    "for agent, log in agent_to_log.items():\n",
    "    if 'calibrated' in baseline_metric:\n",
    "        agent_to_stats[agent][baseline_metric] = [np.array(elapsed_times).sum() for elapsed_times in log[agent][f'elapsed_{baseline_metric}']]\n",
    "    else:\n",
    "        agent_to_stats[agent][baseline_metric] = [np.array(np.abs(elapsed_times)).sum() for elapsed_times in log[agent][baseline_metric]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [],
   "source": [
    "# count # wins (in terms of solve time) for each agent\n",
    "for idx in range(num_instances):\n",
    "    baseline_performance = agent_to_stats[display_name_to_agent_name[baseline_agent]][baseline_metric][idx]\n",
    "    for agent in agent_to_stats.keys():\n",
    "        performance = agent_to_stats[agent][baseline_metric][idx] / baseline_performance\n",
    "        if use_percentages:\n",
    "            performance = int(performance * 100)\n",
    "        agent_to_stats[agent][f'normalised_{baseline_metric}'].append(performance)\n",
    "        agent_to_stats[agent]['Instance'].append(idx+1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+-------------+------------------------+------------+---------+\n",
      "|   num_nodes |   normalised_num_nodes |   Instance | Agent   |\n",
      "|-------------+------------------------+------------+---------|\n",
      "|          28 |                    100 |          1 | IL      |\n",
      "|          14 |                    100 |          2 | IL      |\n",
      "|           6 |                    100 |          3 | IL      |\n",
      "|          48 |                    100 |          4 | IL      |\n",
      "|           4 |                    100 |          5 | IL      |\n",
      "|          18 |                    100 |          6 | IL      |\n",
      "|          10 |                    100 |          7 | IL      |\n",
      "|          56 |                    100 |          8 | IL      |\n",
      "|          37 |                    100 |          9 | IL      |\n",
      "|          16 |                    100 |         10 | IL      |\n",
      "|          58 |                    100 |         11 | IL      |\n",
      "|           2 |                    100 |         12 | IL      |\n",
      "|          44 |                    100 |         13 | IL      |\n",
      "|           6 |                    100 |         14 | IL      |\n",
      "|          10 |                    100 |         15 | IL      |\n",
      "|          48 |                    100 |         16 | IL      |\n",
      "|          21 |                    100 |         17 | IL      |\n",
      "|          12 |                    100 |         18 | IL      |\n",
      "|          12 |                    100 |         19 | IL      |\n",
      "|          65 |                    100 |         20 | IL      |\n",
      "|          38 |                    100 |         21 | IL      |\n",
      "|          78 |                    100 |         22 | IL      |\n",
      "|          65 |                    100 |         23 | IL      |\n",
      "|         120 |                    100 |         24 | IL      |\n",
      "|          50 |                    100 |         25 | IL      |\n",
      "|          12 |                    100 |         26 | IL      |\n",
      "|          48 |                    100 |         27 | IL      |\n",
      "|          18 |                    100 |         28 | IL      |\n",
      "|          16 |                    100 |         29 | IL      |\n",
      "|           4 |                    100 |         30 | IL      |\n",
      "|          20 |                    100 |         31 | IL      |\n",
      "|           8 |                    100 |         32 | IL      |\n",
      "|          48 |                    100 |         33 | IL      |\n",
      "|          24 |                    100 |         34 | IL      |\n",
      "|           9 |                    100 |         35 | IL      |\n",
      "|           6 |                    100 |         36 | IL      |\n",
      "|          58 |                    100 |         37 | IL      |\n",
      "|         138 |                    100 |         38 | IL      |\n",
      "|          16 |                    100 |         39 | IL      |\n",
      "|          42 |                    100 |         40 | IL      |\n",
      "|         114 |                    100 |         41 | IL      |\n",
      "|          12 |                    100 |         42 | IL      |\n",
      "|           6 |                    100 |         43 | IL      |\n",
      "|          10 |                    100 |         44 | IL      |\n",
      "|         128 |                    100 |         45 | IL      |\n",
      "|           4 |                    100 |         46 | IL      |\n",
      "|          18 |                    100 |         47 | IL      |\n",
      "|           6 |                    100 |         48 | IL      |\n",
      "|          56 |                    100 |         49 | IL      |\n",
      "|          28 |                    100 |         50 | IL      |\n",
      "|          72 |                    100 |         51 | IL      |\n",
      "|         102 |                    100 |         52 | IL      |\n",
      "|          56 |                    100 |         53 | IL      |\n",
      "|          50 |                    100 |         54 | IL      |\n",
      "|          48 |                    100 |         55 | IL      |\n",
      "|         186 |                    100 |         56 | IL      |\n",
      "|          52 |                    100 |         57 | IL      |\n",
      "|           8 |                    100 |         58 | IL      |\n",
      "|          65 |                    100 |         59 | IL      |\n",
      "|          12 |                    100 |         60 | IL      |\n",
      "|         166 |                    100 |         61 | IL      |\n",
      "|          20 |                    100 |         62 | IL      |\n",
      "|          90 |                    100 |         63 | IL      |\n",
      "|          56 |                    100 |         64 | IL      |\n",
      "|          20 |                    100 |         65 | IL      |\n",
      "|          68 |                    100 |         66 | IL      |\n",
      "|          39 |                    100 |         67 | IL      |\n",
      "|          34 |                    100 |         68 | IL      |\n",
      "|          20 |                    100 |         69 | IL      |\n",
      "|          14 |                    100 |         70 | IL      |\n",
      "|          10 |                    100 |         71 | IL      |\n",
      "|         100 |                    100 |         72 | IL      |\n",
      "|         220 |                    100 |         73 | IL      |\n",
      "|          44 |                    100 |         74 | IL      |\n",
      "|          52 |                    100 |         75 | IL      |\n",
      "|          40 |                    100 |         76 | IL      |\n",
      "|           8 |                    100 |         77 | IL      |\n",
      "|           2 |                    100 |         78 | IL      |\n",
      "|          18 |                    100 |         79 | IL      |\n",
      "|          92 |                    100 |         80 | IL      |\n",
      "|          52 |                    100 |         81 | IL      |\n",
      "|          64 |                    100 |         82 | IL      |\n",
      "|          10 |                    100 |         83 | IL      |\n",
      "|          84 |                    100 |         84 | IL      |\n",
      "|          62 |                    100 |         85 | IL      |\n",
      "|           6 |                    100 |         86 | IL      |\n",
      "|          18 |                    100 |         87 | IL      |\n",
      "|          40 |                    100 |         88 | IL      |\n",
      "|          22 |                    100 |         89 | IL      |\n",
      "|         129 |                    100 |         90 | IL      |\n",
      "|          10 |                    100 |         91 | IL      |\n",
      "|          16 |                    100 |         92 | IL      |\n",
      "|          10 |                    100 |         93 | IL      |\n",
      "|         104 |                    100 |         94 | IL      |\n",
      "|           6 |                    100 |         95 | IL      |\n",
      "|          16 |                    100 |         96 | IL      |\n",
      "|          20 |                    100 |         97 | IL      |\n",
      "|          14 |                    100 |         98 | IL      |\n",
      "|          16 |                    100 |         99 | IL      |\n",
      "|          34 |                    100 |        100 | IL      |\n",
      "|          42 |                    150 |          1 | RL      |\n",
      "|          14 |                    100 |          2 | RL      |\n",
      "|           6 |                    100 |          3 | RL      |\n",
      "|          52 |                    108 |          4 | RL      |\n",
      "|           6 |                    150 |          5 | RL      |\n",
      "|          15 |                     83 |          6 | RL      |\n",
      "|          20 |                    200 |          7 | RL      |\n",
      "|          68 |                    121 |          8 | RL      |\n",
      "|          51 |                    137 |          9 | RL      |\n",
      "|          18 |                    112 |         10 | RL      |\n",
      "|          50 |                     86 |         11 | RL      |\n",
      "|           2 |                    100 |         12 | RL      |\n",
      "|          69 |                    156 |         13 | RL      |\n",
      "|           6 |                    100 |         14 | RL      |\n",
      "|          14 |                    140 |         15 | RL      |\n",
      "|          54 |                    112 |         16 | RL      |\n",
      "|          24 |                    114 |         17 | RL      |\n",
      "|          16 |                    133 |         18 | RL      |\n",
      "|          14 |                    116 |         19 | RL      |\n",
      "|         104 |                    160 |         20 | RL      |\n",
      "|          38 |                    100 |         21 | RL      |\n",
      "|         148 |                    189 |         22 | RL      |\n",
      "|          60 |                     92 |         23 | RL      |\n",
      "|         142 |                    118 |         24 | RL      |\n",
      "|          57 |                    113 |         25 | RL      |\n",
      "|          10 |                     83 |         26 | RL      |\n",
      "|          50 |                    104 |         27 | RL      |\n",
      "|          20 |                    111 |         28 | RL      |\n",
      "|          16 |                    100 |         29 | RL      |\n",
      "|          10 |                    250 |         30 | RL      |\n",
      "|          15 |                     75 |         31 | RL      |\n",
      "|           8 |                    100 |         32 | RL      |\n",
      "|          62 |                    129 |         33 | RL      |\n",
      "|          20 |                     83 |         34 | RL      |\n",
      "|           7 |                     77 |         35 | RL      |\n",
      "|          10 |                    166 |         36 | RL      |\n",
      "|          58 |                    100 |         37 | RL      |\n",
      "|         183 |                    132 |         38 | RL      |\n",
      "|          14 |                     87 |         39 | RL      |\n",
      "|          60 |                    142 |         40 | RL      |\n",
      "|         156 |                    136 |         41 | RL      |\n",
      "|          16 |                    133 |         42 | RL      |\n",
      "|           6 |                    100 |         43 | RL      |\n",
      "|          10 |                    100 |         44 | RL      |\n",
      "|         161 |                    125 |         45 | RL      |\n",
      "|           4 |                    100 |         46 | RL      |\n",
      "|          22 |                    122 |         47 | RL      |\n",
      "|          12 |                    200 |         48 | RL      |\n",
      "|          52 |                     92 |         49 | RL      |\n",
      "|          26 |                     92 |         50 | RL      |\n",
      "|          68 |                     94 |         51 | RL      |\n",
      "|          98 |                     96 |         52 | RL      |\n",
      "|          82 |                    146 |         53 | RL      |\n",
      "|          70 |                    140 |         54 | RL      |\n",
      "|          68 |                    141 |         55 | RL      |\n",
      "|         220 |                    118 |         56 | RL      |\n",
      "|          58 |                    111 |         57 | RL      |\n",
      "|           8 |                    100 |         58 | RL      |\n",
      "|          69 |                    106 |         59 | RL      |\n",
      "|           8 |                     66 |         60 | RL      |\n",
      "|         203 |                    122 |         61 | RL      |\n",
      "|          16 |                     80 |         62 | RL      |\n",
      "|         146 |                    162 |         63 | RL      |\n",
      "|          62 |                    110 |         64 | RL      |\n",
      "|          24 |                    120 |         65 | RL      |\n",
      "|          88 |                    129 |         66 | RL      |\n",
      "|          53 |                    135 |         67 | RL      |\n",
      "|          34 |                    100 |         68 | RL      |\n",
      "|          10 |                     50 |         69 | RL      |\n",
      "|          14 |                    100 |         70 | RL      |\n",
      "|           8 |                     80 |         71 | RL      |\n",
      "|         104 |                    104 |         72 | RL      |\n",
      "|         258 |                    117 |         73 | RL      |\n",
      "|          44 |                    100 |         74 | RL      |\n",
      "|          48 |                     92 |         75 | RL      |\n",
      "|          40 |                    100 |         76 | RL      |\n",
      "|           8 |                    100 |         77 | RL      |\n",
      "|           4 |                    200 |         78 | RL      |\n",
      "|          16 |                     88 |         79 | RL      |\n",
      "|          84 |                     91 |         80 | RL      |\n",
      "|          58 |                    111 |         81 | RL      |\n",
      "|          80 |                    125 |         82 | RL      |\n",
      "|          12 |                    120 |         83 | RL      |\n",
      "|          92 |                    109 |         84 | RL      |\n",
      "|          52 |                     83 |         85 | RL      |\n",
      "|           6 |                    100 |         86 | RL      |\n",
      "|          16 |                     88 |         87 | RL      |\n",
      "|          46 |                    114 |         88 | RL      |\n",
      "|          26 |                    118 |         89 | RL      |\n",
      "|         214 |                    165 |         90 | RL      |\n",
      "|          10 |                    100 |         91 | RL      |\n",
      "|          20 |                    125 |         92 | RL      |\n",
      "|          14 |                    140 |         93 | RL      |\n",
      "|         177 |                    170 |         94 | RL      |\n",
      "|           8 |                    133 |         95 | RL      |\n",
      "|          18 |                    112 |         96 | RL      |\n",
      "|          28 |                    140 |         97 | RL      |\n",
      "|          12 |                     85 |         98 | RL      |\n",
      "|          22 |                    137 |         99 | RL      |\n",
      "|          28 |                     82 |        100 | RL      |\n",
      "+-------------+------------------------+------------+---------+\n"
     ]
    }
   ],
   "source": [
    "# Create dataframes (from above dict) for data manipulation and seaborn plotting (use display_name for legends)\n",
    "_df = defaultdict(list)\n",
    "for agent in agent_to_stats.keys():\n",
    "    for stat in agent_to_stats[agent].keys():\n",
    "        try:\n",
    "            for el in agent_to_stats[agent][stat]:\n",
    "                _df[stat].append(el)\n",
    "        except TypeError:\n",
    "            # non-iterable, is already an element\n",
    "            _df[stat].append(agent_to_stats[agent][stat])\n",
    "\n",
    "for agent in agent_to_stats.keys():\n",
    "    for _ in range(num_instances):\n",
    "        _df['Agent'].append(agent_name_to_display_name[agent])\n",
    "                        \n",
    "df = pd.DataFrame(_df)\n",
    "if verbose:\n",
    "    print(tabulate(df, headers='keys', tablefmt='psql', showindex=False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [],
   "source": [
    "def filter_rows_by_values(df, col, values):\n",
    "    return df[df[col].isin(values)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     num_nodes  normalised_num_nodes  Instance Agent\n",
      "129       10.0                   250        30    RL\n",
      "177        4.0                   200        78    RL\n",
      "147       12.0                   200        48    RL\n",
      "106       20.0                   200         7    RL\n",
      "121      148.0                   189        22    RL\n",
      "..         ...                   ...       ...   ...\n",
      "170        8.0                    80        71    RL\n",
      "134        7.0                    77        35    RL\n",
      "130       15.0                    75        31    RL\n",
      "159        8.0                    66        60    RL\n",
      "168       10.0                    50        69    RL\n",
      "\n",
      "[100 rows x 4 columns]\n",
      "129    30\n",
      "177    78\n",
      "147    48\n",
      "106     7\n",
      "121    22\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177    78\n",
      "147    48\n",
      "106     7\n",
      "121    22\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147    48\n",
      "106     7\n",
      "121    22\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     7\n",
      "121    22\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121    22\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    71\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    96\n",
      "134    35\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    96\n",
      "134    97\n",
      "130    31\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    96\n",
      "134    97\n",
      "130    98\n",
      "159    60\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "129     1\n",
      "177     2\n",
      "147     3\n",
      "106     4\n",
      "121     5\n",
      "       ..\n",
      "170    96\n",
      "134    97\n",
      "130    98\n",
      "159    99\n",
      "168    69\n",
      "Name: Instance, Length: 100, dtype: int64\n",
      "+----+---------+-------------+------------------------+------------+---------+\n",
      "|    |   index |   num_nodes |   normalised_num_nodes |   Instance | Agent   |\n",
      "|----+---------+-------------+------------------------+------------+---------|\n",
      "|  0 |     129 |          10 |                    250 |          1 | RL      |\n",
      "|  1 |     177 |           4 |                    200 |          2 | RL      |\n",
      "|  2 |     147 |          12 |                    200 |          3 | RL      |\n",
      "|  3 |     106 |          20 |                    200 |          4 | RL      |\n",
      "|  4 |     121 |         148 |                    189 |          5 | RL      |\n",
      "|  5 |     193 |         177 |                    170 |          6 | RL      |\n",
      "|  6 |     135 |          10 |                    166 |          7 | RL      |\n",
      "|  7 |     189 |         214 |                    165 |          8 | RL      |\n",
      "|  8 |     162 |         146 |                    162 |          9 | RL      |\n",
      "|  9 |     119 |         104 |                    160 |         10 | RL      |\n",
      "| 10 |     112 |          69 |                    156 |         11 | RL      |\n",
      "| 11 |     100 |          42 |                    150 |         12 | RL      |\n",
      "| 12 |     104 |           6 |                    150 |         13 | RL      |\n",
      "| 13 |     152 |          82 |                    146 |         14 | RL      |\n",
      "| 14 |     139 |          60 |                    142 |         15 | RL      |\n",
      "| 15 |     154 |          68 |                    141 |         16 | RL      |\n",
      "| 16 |     114 |          14 |                    140 |         17 | RL      |\n",
      "| 17 |     153 |          70 |                    140 |         18 | RL      |\n",
      "| 18 |     196 |          28 |                    140 |         19 | RL      |\n",
      "| 19 |     192 |          14 |                    140 |         20 | RL      |\n",
      "| 20 |     108 |          51 |                    137 |         21 | RL      |\n",
      "| 21 |     198 |          22 |                    137 |         22 | RL      |\n",
      "| 22 |     140 |         156 |                    136 |         23 | RL      |\n",
      "| 23 |     166 |          53 |                    135 |         24 | RL      |\n",
      "| 24 |     117 |          16 |                    133 |         25 | RL      |\n",
      "| 25 |     141 |          16 |                    133 |         26 | RL      |\n",
      "| 26 |     194 |           8 |                    133 |         27 | RL      |\n",
      "| 27 |     137 |         183 |                    132 |         28 | RL      |\n",
      "| 28 |     132 |          62 |                    129 |         29 | RL      |\n",
      "| 29 |     165 |          88 |                    129 |         30 | RL      |\n",
      "| 30 |     144 |         161 |                    125 |         31 | RL      |\n",
      "| 31 |     181 |          80 |                    125 |         32 | RL      |\n",
      "| 32 |     191 |          20 |                    125 |         33 | RL      |\n",
      "| 33 |     160 |         203 |                    122 |         34 | RL      |\n",
      "| 34 |     146 |          22 |                    122 |         35 | RL      |\n",
      "| 35 |     107 |          68 |                    121 |         36 | RL      |\n",
      "| 36 |     164 |          24 |                    120 |         37 | RL      |\n",
      "| 37 |     182 |          12 |                    120 |         38 | RL      |\n",
      "| 38 |     123 |         142 |                    118 |         39 | RL      |\n",
      "| 39 |     155 |         220 |                    118 |         40 | RL      |\n",
      "| 40 |     188 |          26 |                    118 |         41 | RL      |\n",
      "| 41 |     172 |         258 |                    117 |         42 | RL      |\n",
      "| 42 |     118 |          14 |                    116 |         43 | RL      |\n",
      "| 43 |     116 |          24 |                    114 |         44 | RL      |\n",
      "| 44 |     187 |          46 |                    114 |         45 | RL      |\n",
      "| 45 |     124 |          57 |                    113 |         46 | RL      |\n",
      "| 46 |     109 |          18 |                    112 |         47 | RL      |\n",
      "| 47 |     115 |          54 |                    112 |         48 | RL      |\n",
      "| 48 |     195 |          18 |                    112 |         49 | RL      |\n",
      "| 49 |     180 |          58 |                    111 |         50 | RL      |\n",
      "| 50 |     127 |          20 |                    111 |         51 | RL      |\n",
      "| 51 |     156 |          58 |                    111 |         52 | RL      |\n",
      "| 52 |     163 |          62 |                    110 |         53 | RL      |\n",
      "| 53 |     183 |          92 |                    109 |         54 | RL      |\n",
      "| 54 |     103 |          52 |                    108 |         55 | RL      |\n",
      "| 55 |     158 |          69 |                    106 |         56 | RL      |\n",
      "| 56 |     126 |          50 |                    104 |         57 | RL      |\n",
      "| 57 |     171 |         104 |                    104 |         58 | RL      |\n",
      "| 58 |     175 |          40 |                    100 |         59 | RL      |\n",
      "| 59 |     176 |           8 |                    100 |         60 | RL      |\n",
      "| 60 |     167 |          34 |                    100 |         61 | RL      |\n",
      "| 61 |     173 |          44 |                    100 |         62 | RL      |\n",
      "| 62 |     169 |          14 |                    100 |         63 | RL      |\n",
      "| 63 |     111 |           2 |                    100 |         64 | RL      |\n",
      "| 64 |     102 |           6 |                    100 |         65 | RL      |\n",
      "| 65 |     185 |           6 |                    100 |         66 | RL      |\n",
      "| 66 |     157 |           8 |                    100 |         67 | RL      |\n",
      "| 67 |     101 |          14 |                    100 |         68 | RL      |\n",
      "| 68 |     113 |           6 |                    100 |         69 | RL      |\n",
      "| 69 |     143 |          10 |                    100 |         70 | RL      |\n",
      "| 70 |     142 |           6 |                    100 |         71 | RL      |\n",
      "| 71 |     190 |          10 |                    100 |         72 | RL      |\n",
      "| 72 |     136 |          58 |                    100 |         73 | RL      |\n",
      "| 73 |     131 |           8 |                    100 |         74 | RL      |\n",
      "| 74 |     128 |          16 |                    100 |         75 | RL      |\n",
      "| 75 |     120 |          38 |                    100 |         76 | RL      |\n",
      "| 76 |     145 |           4 |                    100 |         77 | RL      |\n",
      "| 77 |     151 |          98 |                     96 |         78 | RL      |\n",
      "| 78 |     150 |          68 |                     94 |         79 | RL      |\n",
      "| 79 |     148 |          52 |                     92 |         80 | RL      |\n",
      "| 80 |     149 |          26 |                     92 |         81 | RL      |\n",
      "| 81 |     174 |          48 |                     92 |         82 | RL      |\n",
      "| 82 |     122 |          60 |                     92 |         83 | RL      |\n",
      "| 83 |     179 |          84 |                     91 |         84 | RL      |\n",
      "| 84 |     186 |          16 |                     88 |         85 | RL      |\n",
      "| 85 |     178 |          16 |                     88 |         86 | RL      |\n",
      "| 86 |     138 |          14 |                     87 |         87 | RL      |\n",
      "| 87 |     110 |          50 |                     86 |         88 | RL      |\n",
      "| 88 |     197 |          12 |                     85 |         89 | RL      |\n",
      "| 89 |     184 |          52 |                     83 |         90 | RL      |\n",
      "| 90 |     133 |          20 |                     83 |         91 | RL      |\n",
      "| 91 |     125 |          10 |                     83 |         92 | RL      |\n",
      "| 92 |     105 |          15 |                     83 |         93 | RL      |\n",
      "| 93 |     199 |          28 |                     82 |         94 | RL      |\n",
      "| 94 |     161 |          16 |                     80 |         95 | RL      |\n",
      "| 95 |     170 |           8 |                     80 |         96 | RL      |\n",
      "| 96 |     134 |           7 |                     77 |         97 | RL      |\n",
      "| 97 |     130 |          15 |                     75 |         98 | RL      |\n",
      "| 98 |     159 |           8 |                     66 |         99 | RL      |\n",
      "| 99 |     168 |          10 |                     50 |        100 | RL      |\n",
      "+----+---------+-------------+------------------------+------------+---------+\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/scratch/zciccwf/py36/envs/rlgnn/lib/python3.7/site-packages/pandas/core/indexing.py:1637: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  self._setitem_single_block(indexer, value, name)\n"
     ]
    }
   ],
   "source": [
    "# sort df in order of normalised solve time\n",
    "sorted_agent_normalised_df = filter_rows_by_values(df, 'Agent', [competitor_agent]).sort_values(f'normalised_{baseline_metric}', axis=0, ascending=False)\n",
    "print(sorted_agent_normalised_df)\n",
    "\n",
    "# relabel instances so order is plotted correctly by seaborn\n",
    "for idx in range(num_instances):\n",
    "    print(sorted_agent_normalised_df['Instance'])\n",
    "    sorted_agent_normalised_df['Instance'].iloc[idx] = idx+1\n",
    "    \n",
    "sorted_agent_normalised_df = sorted_agent_normalised_df.reset_index()\n",
    "\n",
    "if verbose:\n",
    "    print(tabulate(sorted_agent_normalised_df, headers='keys', tablefmt='psql', showindex=True))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [],
   "source": [
    "# config\n",
    "context = 'paper'\n",
    "style = 'whitegrid'\n",
    "font_scale = 2\n",
    "palette = 'pastel'\n",
    "# tick_freq = 5\n",
    "ytick_separation = 5\n",
    "ylabels = OrderedSet([str(lab) for lab in range(5, 105, ytick_separation)])\n",
    "axvline_linewidth = 3\n",
    "axvline_color = 'grey'\n",
    "# fig_size = (10, 20)\n",
    "# fig_size = (5, 15)\n",
    "# fig_size = (7.5, 15)\n",
    "fig_size = (7, 15)\n",
    "# fig_size = (7, 50)\n",
    "\n",
    "# height_scaling_factor = 1\n",
    "height_scaling_factor = 0.85\n",
    "# height_scaling_factor = 1.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [],
   "source": [
    "aesthetics = PlotAesthetics()\n",
    "aesthetics.set_icml_paper_plot_aesthetics()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 975x1218.75 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 975x1218.75 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%autoreload\n",
    "\n",
    "# sns.set_theme(font_scale=font_scale,\n",
    "#               context=context,\n",
    "#               style=style,\n",
    "#               palette=palette)\n",
    "\n",
    "if baseline_metric == 'calibrated_solve_time':\n",
    "    xlabel = 'Relative Calibrated Solve Time'\n",
    "elif baseline_metric == 'num_nodes':\n",
    "    xlabel = 'Relative Nodes'\n",
    "elif baseline_metric == 'lp_iterations':\n",
    "    xlabel = 'Relative LP Iterations'\n",
    "else:\n",
    "    raise Exception(f'Unrecognised baseline_metric {baseline_metric}, implement xlabel for this.')\n",
    "    \n",
    "if use_percentages:\n",
    "    xlabel += ' (\\%)'\n",
    "    axvline_x = 100\n",
    "else:\n",
    "    axvline_x = 1\n",
    "\n",
    "f, ax = plt.subplots(figsize=aesthetics.get_winner_bar_fig_size())\n",
    "g = sns.barplot(data=filter_rows_by_values(df, 'Agent', [competitor_agent]), \n",
    "                x=f'normalised_{baseline_metric}', \n",
    "                y='Instance', \n",
    "                hue='Agent', \n",
    "                orient='h')\n",
    "g.set_xlabel(xlabel)\n",
    "g.set_ylabel('Evaluation MILPs')\n",
    "plt.axvline(x=axvline_x, linestyle='--', label=baseline_agent, alpha=1, color=axvline_color, linewidth=axvline_linewidth)\n",
    "for label in ax.yaxis.get_ticklabels():\n",
    "    if label.get_text() in ylabels:\n",
    "        label.set_visible(True)\n",
    "    else:\n",
    "        label.set_visible(False) \n",
    "ax.legend(ncol=1, frameon=True, loc='lower right')\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.tick_params(axis='both', which='major', pad=2)\n",
    "ax.xaxis.labelpad = 2\n",
    "ax.yaxis.labelpad = 2\n",
    "plt.gcf().patch.set_alpha(0.0)\n",
    "plt.show()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "f, ax = plt.subplots(figsize=aesthetics.get_winner_bar_fig_size())\n",
    "g = sns.barplot(data=sorted_agent_normalised_df, \n",
    "                x=f'normalised_{baseline_metric}',\n",
    "                y='Instance', \n",
    "                hue='Agent', \n",
    "                orient='h')\n",
    "g.set_xlabel(xlabel)\n",
    "g.set_ylabel('Evaluation MILPs')\n",
    "prev_instance = None\n",
    "for idx, instance in enumerate(sorted_agent_normalised_df['Instance']):\n",
    "    if sorted_agent_normalised_df[f'normalised_{baseline_metric}'][idx] <= axvline_x:\n",
    "        # all values below this y-axis value better than or equal to baseline\n",
    "        if prev_instance is not None:\n",
    "            instance = prev_instance\n",
    "        break\n",
    "    prev_instance = instance\n",
    "instance += 1\n",
    "plt.axhline(y=instance, linestyle='--', label=baseline_agent, alpha=1, color=axvline_color, linewidth=axvline_linewidth)\n",
    "ax.text(0.5, instance+3, f'At {baseline_agent}-level or better', transform=ax.get_yaxis_transform())\n",
    "ax.text(0.5, instance-1.5, f'Worse than {baseline_agent}-level', transform=ax.get_yaxis_transform())\n",
    "for label in ax.yaxis.get_ticklabels():\n",
    "    if label.get_text() in ylabels:\n",
    "        label.set_visible(True)\n",
    "    else:\n",
    "        label.set_visible(False)\n",
    "        \n",
    "ax.legend(ncol=1, frameon=True, loc='lower right')\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.tick_params(axis='both', which='major', pad=2)\n",
    "ax.xaxis.labelpad = 2\n",
    "ax.yaxis.labelpad = 2\n",
    "plt.gcf().patch.set_alpha(0.0)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/scratch/zciccwf/py36/envs/rlgnn/lib/python3.7/site-packages/ipykernel_launcher.py:74: UserWarning: FixedFormatter should only be used together with FixedLocator\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "57.9\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 975x1035.94 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%autoreload\n",
    "\n",
    "shift_axis = True\n",
    "barwidth = 6 * 0.25\n",
    "baralpha = 0.2\n",
    "markersize = 7 * 0.25\n",
    "markeralpha = 0.6\n",
    "\n",
    "# shift x-axis values\n",
    "new_sorted_agent_normalised_df = copy.deepcopy(sorted_agent_normalised_df)\n",
    "new_sorted_agent_normalised_df[f'normalised_{baseline_metric}'] = new_sorted_agent_normalised_df[f'normalised_{baseline_metric}'][::-1]\n",
    "if shift_axis:\n",
    "    new_sorted_agent_normalised_df[f'normalised_{baseline_metric}'] = new_sorted_agent_normalised_df[f'normalised_{baseline_metric}'] - 100\n",
    "\n",
    "# init plot\n",
    "figsize = aesthetics.get_winner_bar_fig_size()\n",
    "figsize = list(figsize)\n",
    "figsize[-1] *= height_scaling_factor\n",
    "figsize = tuple(figsize)\n",
    "f, ax = plt.subplots(figsize=figsize)\n",
    "\n",
    "# init placeholders for y-axis\n",
    "my_range = list(range(1,len(new_sorted_agent_normalised_df.index)+1))\n",
    "\n",
    "# draw horizontal bars for each data point\n",
    "plt.hlines(y=my_range, \n",
    "           xmin=0, \n",
    "           xmax=new_sorted_agent_normalised_df[f'normalised_{baseline_metric}'], \n",
    "           color='#007acc', \n",
    "           alpha=baralpha, \n",
    "           linewidth=barwidth)\n",
    "plt.plot(new_sorted_agent_normalised_df[f'normalised_{baseline_metric}'], my_range, \"o\", markersize=markersize, color='#007acc', alpha=markeralpha)\n",
    "# ax.set_xlabel(xlabel, fontsize=15, fontweight='black', color = '#333F4B')\n",
    "ax.set_xlabel(xlabel)\n",
    "ax.set_ylabel('Evaluation MILPs')\n",
    "\n",
    "# draw horizontal line indicating baseline level\n",
    "prev_instance = None\n",
    "for idx, instance in enumerate(sorted_agent_normalised_df['Instance']):\n",
    "    if sorted_agent_normalised_df[f'normalised_{baseline_metric}'][idx] <= axvline_x:\n",
    "        # all values below this y-axis value better than or equal to baseline\n",
    "        if prev_instance is not None:\n",
    "            instance = prev_instance\n",
    "        break\n",
    "    prev_instance = instance\n",
    "instance -= 0.1\n",
    "plt.axhline(y=instance+0.5, linestyle='--', label=baseline_agent, alpha=0.3, color=axvline_color, linewidth=barwidth)\n",
    "# ax.text(0.015, instance+1.25, f'At {baseline_agent}-level\\n or better', transform=ax.get_yaxis_transform())\n",
    "# ax.text(0.5, instance+1.25, f'At {baseline_agent}-level\\n or better', transform=ax.get_yaxis_transform())\n",
    "ax.text(0.4, instance+1.25, f'At {baseline_agent}-level or better', transform=ax.get_yaxis_transform())\n",
    "# ax.text(0.375, instance+1.25, f'At {baseline_agent}-level or better', transform=ax.get_yaxis_transform())\n",
    "# ax.text(0.015, instance-1.25, f'Worse than {baseline_agent}-level', transform=ax.get_yaxis_transform())\n",
    "\n",
    "# draw vertical line marking starting point\n",
    "plt.axvline(x=0, alpha=0.5, color='k')\n",
    "\n",
    "# init yticks\n",
    "plt.yticks(np.arange(0, num_instances+1, ytick_separation))\n",
    "\n",
    "if shift_axis:\n",
    "    # explicitly draw fig to force setting of xticks so can overwrite\n",
    "    f.canvas.draw()\n",
    "    # get axis labels\n",
    "    labels = [label.get_text() for label in ax.xaxis.get_ticklabels()]\n",
    "    # remove any bad minus characters\n",
    "    for idx, label in enumerate(labels):\n",
    "        for i in range(len(label)):\n",
    "            if label[i] == '−':\n",
    "                labels[idx] = '{-' + label[i+1:]\n",
    "    # overwrite axis labels\n",
    "#     new_labels = [str(int(label)+100) for label in labels]\n",
    "    new_labels = [str(float(label[label.index('{')+1:label.index('}')])+100) for label in labels]\n",
    "    \n",
    "    ax.xaxis.set_ticklabels(new_labels)\n",
    "\n",
    "# # overwrite yticks\n",
    "# for label in ax.yaxis.get_ticklabels():\n",
    "#     if label.get_text() in ylabels:\n",
    "#         label.set_visible(True)\n",
    "#     else:\n",
    "#         label.set_visible(False)\n",
    "        \n",
    "# format axes\n",
    "ax.spines['bottom'].set_position(('data', 0))\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.tick_params(axis='both', which='major', pad=2)\n",
    "ax.xaxis.labelpad = 2\n",
    "ax.yaxis.labelpad = 2\n",
    "plt.ylim(bottom=0, top=100)\n",
    "\n",
    "print(instance)\n",
    "\n",
    "plt.gcf().patch.set_alpha(0.0)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "rlgnn",
   "language": "python",
   "name": "rlgnn"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}