ClusterPC_simulation_2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ClusterPC simulation 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook runs a simulation comparing PC with Cluster-PC. It also uses different graph and distribtuion methods and shows their effect on the algorithms. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No GPU automatically detected. Setting SETTINGS.GPU to 0, and SETTINGS.NJOBS to cpu_count.\n",
      "2023-11-28 20:04:57,353 - c:\\Users\\JanMarco\\MLProjects\\Masterthesis\\masterthesis\\mthesis\\lib\\site-packages\\rpy2\\situation.py[line:41] - INFO: cffi mode is CFFI_MODE.ANY\n",
      "2023-11-28 20:04:57,361 - c:\\Users\\JanMarco\\MLProjects\\Masterthesis\\masterthesis\\mthesis\\lib\\site-packages\\rpy2\\situation.py[line:218] - INFO: R home found: C:\\Program Files\\R\\R-4.3.1\n",
      "2023-11-28 20:04:57,370 - c:\\Users\\JanMarco\\MLProjects\\Masterthesis\\masterthesis\\mthesis\\lib\\site-packages\\rpy2\\rinterface_lib\\embedded.py[line:20] - INFO: Default options to initialize R: rpy2, --quiet, --no-save\n",
      "IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n"
     ]
    }
   ],
   "source": [
    "# Importing dependencies\n",
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "import itertools\n",
    "import yaml\n",
    "import os\n",
    "import pickle\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from clustercausal.experiments.ExperimentRunner import ExperimentRunner\n",
    "from clustercausal.utils.Utils import *\n",
    "from clustercausal.experiments.Utils import load_experiment_folder"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The cell below executes a simulation according to the specified yaml file. To run several yaml files, it is recommended to modify the run_gridsearch.py file and run the simulation in the terminal with \n",
    "\n",
    ">```python clustercausal\\experiments\\run_gridsearch.py```\n",
    "\n",
    "Depending on the number of experiments, the simulation can take a while. Simulation 2 takes around 35 minutes on a standard laptop. A sample yaml of all possible configurations can be found in ```clustercausal\\experiments\\configs\\cluster_pc_all_param_configs.yaml```. It is not recommended to run that file, it is for reference only. \n",
    "\n",
    "To run the experiment, uncomment the code in the cell below. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# config_path = 'clustercausal\\experiments\\configs\\cluster_pc_simulation_2.yaml'\n",
    "# config_path = config_path.replace('\\\\', '/')\n",
    "# experiment = ExperimentRunner(config_path)\n",
    "# experiment.run_gridsearch_experiment()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluating the simulation\n",
    "\n",
    "First load the data. Make sure that the _results folder is empty, otherwise old experiments will also be loaded. Consider moving old experiments to a new folder _results_storage. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Can take a longer time depending on the numer of experiments, ca. 40sec for the 'mass_simulation' experiment\n",
    "# This loads all experiment folder present in _results\n",
    "data = load_experiment_folder('clustercausal\\experiments\\_results')\n",
    "\n",
    "# If one wants to load only specific simulation folders, put these in a separate folder named e.g. my_experiment and load\n",
    "# data = load_experiment_folder('clustercausal\\experiments\\_results\\my_experiment')\n",
    "# where my_experiment contains the simulations of interest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# data_all also contains runs with n_clusters = 1\n",
    "data_all = data.copy()\n",
    "data = data[data['n_clusters'] != 1]\n",
    "# To look only at clusterings where at least two clusters were non-adjacent\n",
    "data_nonadj_clusters = data[data['cluster_connectivity'] != 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <th>base_adj_false_negative</th>\n",
       "      <th>base_adj_false_positive</th>\n",
       "      <th>base_adj_precision</th>\n",
       "      <th>base_adj_recall</th>\n",
       "      <th>base_adj_true_negative</th>\n",
       "      <th>base_adj_true_positive</th>\n",
       "      <th>base_arrow_f1_score</th>\n",
       "      <th>base_arrow_f1_score_ce</th>\n",
       "      <th>base_arrow_false_negative</th>\n",
       "      <th>...</th>\n",
       "      <th>alpha</th>\n",
       "      <th>cluster_connectivity</th>\n",
       "      <th>empty_graph_shd</th>\n",
       "      <th>n_clusters</th>\n",
       "      <th>n_edges</th>\n",
       "      <th>n_nodes</th>\n",
       "      <th>noise_scale</th>\n",
       "      <th>sample_size</th>\n",
       "      <th>true_sid_lower</th>\n",
       "      <th>true_sid_upper</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1078.000000</td>\n",
       "      <td>1078.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.0</td>\n",
       "      <td>1080.0</td>\n",
       "      <td>1080.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.558015</td>\n",
       "      <td>33.237037</td>\n",
       "      <td>2.963889</td>\n",
       "      <td>0.863080</td>\n",
       "      <td>0.463783</td>\n",
       "      <td>51.439815</td>\n",
       "      <td>17.359259</td>\n",
       "      <td>0.343881</td>\n",
       "      <td>0.632246</td>\n",
       "      <td>40.306481</td>\n",
       "      <td>...</td>\n",
       "      <td>0.182000</td>\n",
       "      <td>0.878259</td>\n",
       "      <td>50.596296</td>\n",
       "      <td>3.500000</td>\n",
       "      <td>43.750000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.197564</td>\n",
       "      <td>24.538131</td>\n",
       "      <td>2.952135</td>\n",
       "      <td>0.119324</td>\n",
       "      <td>0.253751</td>\n",
       "      <td>25.915885</td>\n",
       "      <td>3.956273</td>\n",
       "      <td>0.163435</td>\n",
       "      <td>0.171687</td>\n",
       "      <td>24.603653</td>\n",
       "      <td>...</td>\n",
       "      <td>0.178677</td>\n",
       "      <td>0.172465</td>\n",
       "      <td>26.367545</td>\n",
       "      <td>1.708616</td>\n",
       "      <td>24.346177</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.231579</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.187500</td>\n",
       "      <td>0.134146</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.037037</td>\n",
       "      <td>0.125000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.166667</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.391641</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.809066</td>\n",
       "      <td>0.253679</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>0.228571</td>\n",
       "      <td>0.514286</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.050000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>26.250000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.507463</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.360000</td>\n",
       "      <td>53.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>0.311111</td>\n",
       "      <td>0.642857</td>\n",
       "      <td>40.500000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>3.500000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.723588</td>\n",
       "      <td>58.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.947368</td>\n",
       "      <td>0.653846</td>\n",
       "      <td>76.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>0.434783</td>\n",
       "      <td>0.756757</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>79.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>57.500000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>89.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 81 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       base_adj_f1_score  base_adj_false_negative  base_adj_false_positive  \\\n",
       "count        1080.000000              1080.000000              1080.000000   \n",
       "mean            0.558015                33.237037                 2.963889   \n",
       "std             0.197564                24.538131                 2.952135   \n",
       "min             0.231579                 0.000000                 0.000000   \n",
       "25%             0.391641                 9.000000                 1.000000   \n",
       "50%             0.507463                31.000000                 2.000000   \n",
       "75%             0.723588                58.000000                 4.000000   \n",
       "max             1.000000                75.000000                26.000000   \n",
       "\n",
       "       base_adj_precision  base_adj_recall  base_adj_true_negative  \\\n",
       "count         1080.000000      1080.000000             1080.000000   \n",
       "mean             0.863080         0.463783               51.439815   \n",
       "std              0.119324         0.253751               25.915885   \n",
       "min              0.187500         0.134146               12.000000   \n",
       "25%              0.809066         0.253679               25.000000   \n",
       "50%              0.888889         0.360000               53.000000   \n",
       "75%              0.947368         0.653846               76.000000   \n",
       "max              1.000000         1.000000               98.000000   \n",
       "\n",
       "       base_adj_true_positive  base_arrow_f1_score  base_arrow_f1_score_ce  \\\n",
       "count             1080.000000          1078.000000             1078.000000   \n",
       "mean                17.359259             0.343881                0.632246   \n",
       "std                  3.956273             0.163435                0.171687   \n",
       "min                  6.000000             0.037037                0.125000   \n",
       "25%                 15.000000             0.228571                0.514286   \n",
       "50%                 17.000000             0.311111                0.642857   \n",
       "75%                 20.000000             0.434783                0.756757   \n",
       "max                 29.000000             0.923077                1.000000   \n",
       "\n",
       "       base_arrow_false_negative  ...        alpha  cluster_connectivity  \\\n",
       "count                1080.000000  ...  1080.000000            900.000000   \n",
       "mean                   40.306481  ...     0.182000              0.878259   \n",
       "std                    24.603653  ...     0.178677              0.172465   \n",
       "min                     1.000000  ...     0.010000              0.166667   \n",
       "25%                    17.000000  ...     0.050000              0.800000   \n",
       "50%                    40.500000  ...     0.100000              1.000000   \n",
       "75%                    64.000000  ...     0.250000              1.000000   \n",
       "max                    84.000000  ...     0.500000              1.000000   \n",
       "\n",
       "       empty_graph_shd   n_clusters      n_edges  n_nodes  noise_scale  \\\n",
       "count      1080.000000  1080.000000  1080.000000   1080.0       1080.0   \n",
       "mean         50.596296     3.500000    43.750000     15.0          1.0   \n",
       "std          26.367545     1.708616    24.346177      0.0          0.0   \n",
       "min           7.000000     1.000000    15.000000     15.0          1.0   \n",
       "25%          26.000000     2.000000    26.250000     15.0          1.0   \n",
       "50%          50.000000     3.500000    40.000000     15.0          1.0   \n",
       "75%          79.000000     5.000000    57.500000     15.0          1.0   \n",
       "max          89.000000     6.000000    80.000000     15.0          1.0   \n",
       "\n",
       "       sample_size  true_sid_lower  true_sid_upper  \n",
       "count       1080.0             0.0             0.0  \n",
       "mean        1000.0             NaN             NaN  \n",
       "std            0.0             NaN             NaN  \n",
       "min         1000.0             NaN             NaN  \n",
       "25%         1000.0             NaN             NaN  \n",
       "50%         1000.0             NaN             NaN  \n",
       "75%         1000.0             NaN             NaN  \n",
       "max         1000.0             NaN             NaN  \n",
       "\n",
       "[8 rows x 81 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_all.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It is possible to load the specific graphs of one experiment: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# # Change this according to the experiment folder you want to load - this specific folder is not included in the repo\n",
    "# experiment_folder = 'clustercausal\\experiments\\_results\\ClusterPC_simulation_2\\erdos_renyi_15_nodes_15_edges_1_clusters_exp14-13-40-503'\n",
    "# base_est_graph, cluster_est_graph, cluster_dag = load_experiment_graphs(experiment_folder)\n",
    "# cluster_dag.true_dag.draw_pydot_graph()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data exploration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <th>base_adj_false_negative</th>\n",
       "      <th>base_adj_false_positive</th>\n",
       "      <th>base_adj_precision</th>\n",
       "      <th>base_adj_recall</th>\n",
       "      <th>base_adj_true_negative</th>\n",
       "      <th>base_adj_true_positive</th>\n",
       "      <th>base_arrow_f1_score</th>\n",
       "      <th>base_arrow_f1_score_ce</th>\n",
       "      <th>base_arrow_false_negative</th>\n",
       "      <th>base_arrow_false_negative_ce</th>\n",
       "      <th>base_arrow_false_positive</th>\n",
       "      <th>base_arrow_false_positive_ce</th>\n",
       "      <th>base_arrow_precision</th>\n",
       "      <th>base_arrow_precision_ce</th>\n",
       "      <th>base_arrow_recall</th>\n",
       "      <th>base_arrow_recall_ce</th>\n",
       "      <th>base_arrow_true_negative</th>\n",
       "      <th>base_arrow_true_negative_ce</th>\n",
       "      <th>base_arrow_true_positive</th>\n",
       "      <th>base_arrow_true_positive_ce</th>\n",
       "      <th>base_base_shd</th>\n",
       "      <th>base_sid_lower</th>\n",
       "      <th>cluster_adj_f1_score</th>\n",
       "      <th>cluster_adj_false_negative</th>\n",
       "      <th>cluster_adj_false_positive</th>\n",
       "      <th>cluster_adj_precision</th>\n",
       "      <th>cluster_adj_recall</th>\n",
       "      <th>cluster_adj_true_negative</th>\n",
       "      <th>cluster_adj_true_positive</th>\n",
       "      <th>cluster_arrow_f1_score</th>\n",
       "      <th>cluster_arrow_f1_score_ce</th>\n",
       "      <th>cluster_arrow_false_negative</th>\n",
       "      <th>cluster_arrow_false_negative_ce</th>\n",
       "      <th>cluster_arrow_false_positive</th>\n",
       "      <th>cluster_arrow_false_positive_ce</th>\n",
       "      <th>cluster_arrow_precision</th>\n",
       "      <th>cluster_arrow_precision_ce</th>\n",
       "      <th>cluster_arrow_recall</th>\n",
       "      <th>cluster_arrow_recall_ce</th>\n",
       "      <th>cluster_arrow_true_negative</th>\n",
       "      <th>cluster_arrow_true_negative_ce</th>\n",
       "      <th>cluster_arrow_true_positive</th>\n",
       "      <th>cluster_arrow_true_positive_ce</th>\n",
       "      <th>cluster_cluster_shd</th>\n",
       "      <th>cluster_sid_lower</th>\n",
       "      <th>pruned_base_adj_f1_score</th>\n",
       "      <th>pruned_base_adj_false_negative</th>\n",
       "      <th>pruned_base_adj_false_positive</th>\n",
       "      <th>pruned_base_adj_precision</th>\n",
       "      <th>pruned_base_adj_recall</th>\n",
       "      <th>pruned_base_adj_true_negative</th>\n",
       "      <th>pruned_base_adj_true_positive</th>\n",
       "      <th>pruned_base_arrow_f1_score</th>\n",
       "      <th>pruned_base_arrow_f1_score_ce</th>\n",
       "      <th>pruned_base_arrow_false_negative</th>\n",
       "      <th>pruned_base_arrow_false_negative_ce</th>\n",
       "      <th>pruned_base_arrow_false_positive</th>\n",
       "      <th>pruned_base_arrow_false_positive_ce</th>\n",
       "      <th>pruned_base_arrow_precision</th>\n",
       "      <th>pruned_base_arrow_precision_ce</th>\n",
       "      <th>pruned_base_arrow_recall</th>\n",
       "      <th>pruned_base_arrow_recall_ce</th>\n",
       "      <th>pruned_base_arrow_true_negative</th>\n",
       "      <th>pruned_base_arrow_true_negative_ce</th>\n",
       "      <th>pruned_base_arrow_true_positive</th>\n",
       "      <th>pruned_base_arrow_true_positive_ce</th>\n",
       "      <th>pruned_base_pruned_base_shd</th>\n",
       "      <th>pruned_base_sid_lower</th>\n",
       "      <th>Base indep tests</th>\n",
       "      <th>C-PC indep tests</th>\n",
       "      <th>alpha</th>\n",
       "      <th>cluster_connectivity</th>\n",
       "      <th>dag_method</th>\n",
       "      <th>distribution_type</th>\n",
       "      <th>edge_ratios</th>\n",
       "      <th>empty_graph_shd</th>\n",
       "      <th>indep_test</th>\n",
       "      <th>n_clusters</th>\n",
       "      <th>n_edges</th>\n",
       "      <th>n_nodes</th>\n",
       "      <th>noise_scale</th>\n",
       "      <th>sample_size</th>\n",
       "      <th>scm_method</th>\n",
       "      <th>seed</th>\n",
       "      <th>true_sid_lower</th>\n",
       "      <th>true_sid_upper</th>\n",
       "      <th>weight_range</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.880000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.785714</td>\n",
       "      <td>91.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.588235</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.357143</td>\n",
       "      <td>0.454545</td>\n",
       "      <td>210.0</td>\n",
       "      <td>213.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>91.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>0.782609</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.642857</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>211.0</td>\n",
       "      <td>213.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.880000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.785714</td>\n",
       "      <td>91.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.588235</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.357143</td>\n",
       "      <td>0.454545</td>\n",
       "      <td>210.0</td>\n",
       "      <td>213.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>778.0</td>\n",
       "      <td>512.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>1.0</td>\n",
       "      <td>erdos_renyi</td>\n",
       "      <td>exp</td>\n",
       "      <td>[0.15, 0.14, 0.14]</td>\n",
       "      <td>14.0</td>\n",
       "      <td>fisherz</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>linear</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>(-1, 2)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.916667</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>92.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.631579</td>\n",
       "      <td>0.705882</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.461538</td>\n",
       "      <td>0.545455</td>\n",
       "      <td>212.0</td>\n",
       "      <td>214.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.880000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.916667</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>91.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.880000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.916667</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>211.0</td>\n",
       "      <td>214.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.916667</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>92.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.631579</td>\n",
       "      <td>0.705882</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.461538</td>\n",
       "      <td>0.545455</td>\n",
       "      <td>212.0</td>\n",
       "      <td>214.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>389.0</td>\n",
       "      <td>293.0</td>\n",
       "      <td>0.05</td>\n",
       "      <td>1.0</td>\n",
       "      <td>erdos_renyi</td>\n",
       "      <td>exp</td>\n",
       "      <td>[0.12, 0.12, 0.12]</td>\n",
       "      <td>13.0</td>\n",
       "      <td>fisherz</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>linear</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>(-1, 2)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.903226</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.823529</td>\n",
       "      <td>88.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.592593</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.470588</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>206.0</td>\n",
       "      <td>209.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.903226</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.823529</td>\n",
       "      <td>88.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.642857</td>\n",
       "      <td>0.720000</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.818182</td>\n",
       "      <td>0.818182</td>\n",
       "      <td>0.529412</td>\n",
       "      <td>0.642857</td>\n",
       "      <td>206.0</td>\n",
       "      <td>209.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.903226</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.823529</td>\n",
       "      <td>88.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.592593</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.470588</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>206.0</td>\n",
       "      <td>209.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>645.0</td>\n",
       "      <td>619.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>1.0</td>\n",
       "      <td>erdos_renyi</td>\n",
       "      <td>exp</td>\n",
       "      <td>[0.33, 0.14, 0.14]</td>\n",
       "      <td>17.0</td>\n",
       "      <td>fisherz</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>linear</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>(-1, 2)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.742857</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.619048</td>\n",
       "      <td>0.928571</td>\n",
       "      <td>83.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.588235</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>0.769231</td>\n",
       "      <td>201.0</td>\n",
       "      <td>210.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.722222</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.590909</td>\n",
       "      <td>0.928571</td>\n",
       "      <td>82.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.647059</td>\n",
       "      <td>0.880000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.550000</td>\n",
       "      <td>0.916667</td>\n",
       "      <td>0.785714</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>202.0</td>\n",
       "      <td>211.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.742857</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.619048</td>\n",
       "      <td>0.928571</td>\n",
       "      <td>83.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.588235</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>0.769231</td>\n",
       "      <td>201.0</td>\n",
       "      <td>210.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1959.0</td>\n",
       "      <td>1131.0</td>\n",
       "      <td>0.25</td>\n",
       "      <td>1.0</td>\n",
       "      <td>erdos_renyi</td>\n",
       "      <td>exp</td>\n",
       "      <td>[0.22, 0.09, 0.09]</td>\n",
       "      <td>14.0</td>\n",
       "      <td>fisherz</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>linear</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>(-1, 2)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.440000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>80.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.275862</td>\n",
       "      <td>0.470588</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.222222</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.363636</td>\n",
       "      <td>0.363636</td>\n",
       "      <td>200.0</td>\n",
       "      <td>212.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.440000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>80.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.344828</td>\n",
       "      <td>0.588235</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.277778</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.454545</td>\n",
       "      <td>0.454545</td>\n",
       "      <td>201.0</td>\n",
       "      <td>213.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.440000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>80.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.275862</td>\n",
       "      <td>0.470588</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.222222</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.363636</td>\n",
       "      <td>0.363636</td>\n",
       "      <td>200.0</td>\n",
       "      <td>212.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1740.0</td>\n",
       "      <td>1423.0</td>\n",
       "      <td>0.50</td>\n",
       "      <td>1.0</td>\n",
       "      <td>erdos_renyi</td>\n",
       "      <td>exp</td>\n",
       "      <td>[0.3, 0.07, 0.07]</td>\n",
       "      <td>11.0</td>\n",
       "      <td>fisherz</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>linear</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>(-1, 2)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.933333</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>97.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.545455</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.375000</td>\n",
       "      <td>0.428571</td>\n",
       "      <td>217.0</td>\n",
       "      <td>218.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.933333</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>97.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.933333</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>217.0</td>\n",
       "      <td>218.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.933333</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>97.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.545455</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.375000</td>\n",
       "      <td>0.428571</td>\n",
       "      <td>217.0</td>\n",
       "      <td>218.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>281.0</td>\n",
       "      <td>213.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>1.0</td>\n",
       "      <td>erdos_renyi</td>\n",
       "      <td>gauss</td>\n",
       "      <td>[0.1, 0.12, 0.12]</td>\n",
       "      <td>8.0</td>\n",
       "      <td>fisherz</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>linear</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>(-1, 2)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.875000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>87.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>205.0</td>\n",
       "      <td>209.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>87.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.733333</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.785714</td>\n",
       "      <td>0.916667</td>\n",
       "      <td>0.687500</td>\n",
       "      <td>0.785714</td>\n",
       "      <td>206.0</td>\n",
       "      <td>210.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>87.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>205.0</td>\n",
       "      <td>209.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>545.0</td>\n",
       "      <td>430.0</td>\n",
       "      <td>0.05</td>\n",
       "      <td>1.0</td>\n",
       "      <td>erdos_renyi</td>\n",
       "      <td>gauss</td>\n",
       "      <td>[0.14, 0.16, 0.16]</td>\n",
       "      <td>16.0</td>\n",
       "      <td>fisherz</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>linear</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>(-1, 2)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.833333</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>91.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.545455</td>\n",
       "      <td>0.631579</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>209.0</td>\n",
       "      <td>212.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.880000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>0.916667</td>\n",
       "      <td>91.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.818182</td>\n",
       "      <td>213.0</td>\n",
       "      <td>214.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>91.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.545455</td>\n",
       "      <td>0.631579</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>209.0</td>\n",
       "      <td>212.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>543.0</td>\n",
       "      <td>428.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>1.0</td>\n",
       "      <td>erdos_renyi</td>\n",
       "      <td>gauss</td>\n",
       "      <td>[0.18, 0.18, 0.18]</td>\n",
       "      <td>12.0</td>\n",
       "      <td>fisherz</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>linear</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>(-1, 2)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.871795</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.894737</td>\n",
       "      <td>0.850000</td>\n",
       "      <td>83.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>0.424242</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.538462</td>\n",
       "      <td>0.636364</td>\n",
       "      <td>0.350000</td>\n",
       "      <td>0.411765</td>\n",
       "      <td>199.0</td>\n",
       "      <td>204.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.818182</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>81.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.882353</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.937500</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>200.0</td>\n",
       "      <td>206.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.871795</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.894737</td>\n",
       "      <td>0.850000</td>\n",
       "      <td>83.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>0.424242</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.538462</td>\n",
       "      <td>0.636364</td>\n",
       "      <td>0.350000</td>\n",
       "      <td>0.411765</td>\n",
       "      <td>199.0</td>\n",
       "      <td>204.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1701.0</td>\n",
       "      <td>1152.0</td>\n",
       "      <td>0.25</td>\n",
       "      <td>1.0</td>\n",
       "      <td>erdos_renyi</td>\n",
       "      <td>gauss</td>\n",
       "      <td>[0.42, 0.19, 0.19]</td>\n",
       "      <td>20.0</td>\n",
       "      <td>fisherz</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>linear</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>(-1, 2)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.666667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.518519</td>\n",
       "      <td>0.933333</td>\n",
       "      <td>77.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.263158</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>10.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.217391</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.357143</td>\n",
       "      <td>192.0</td>\n",
       "      <td>206.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.652174</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.483871</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>74.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>0.577778</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.433333</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>193.0</td>\n",
       "      <td>208.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.518519</td>\n",
       "      <td>0.933333</td>\n",
       "      <td>77.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.263158</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>10.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.217391</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.357143</td>\n",
       "      <td>192.0</td>\n",
       "      <td>206.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2196.0</td>\n",
       "      <td>1681.0</td>\n",
       "      <td>0.50</td>\n",
       "      <td>1.0</td>\n",
       "      <td>erdos_renyi</td>\n",
       "      <td>gauss</td>\n",
       "      <td>[0.06, 0.18, 0.18]</td>\n",
       "      <td>15.0</td>\n",
       "      <td>fisherz</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>linear</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>(-1, 2)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    base_adj_f1_score  base_adj_false_negative  base_adj_false_positive  \\\n",
       "15           0.880000                      3.0                      0.0   \n",
       "16           0.916667                      2.0                      0.0   \n",
       "17           0.903226                      3.0                      0.0   \n",
       "18           0.742857                      1.0                      8.0   \n",
       "19           0.611111                      0.0                     14.0   \n",
       "20           0.933333                      1.0                      0.0   \n",
       "21           0.875000                      2.0                      2.0   \n",
       "22           0.833333                      2.0                      2.0   \n",
       "23           0.871795                      3.0                      2.0   \n",
       "24           0.666667                      1.0                     13.0   \n",
       "\n",
       "    base_adj_precision  base_adj_recall  base_adj_true_negative  \\\n",
       "15            1.000000         0.785714                    91.0   \n",
       "16            1.000000         0.846154                    92.0   \n",
       "17            1.000000         0.823529                    88.0   \n",
       "18            0.619048         0.928571                    83.0   \n",
       "19            0.440000         1.000000                    80.0   \n",
       "20            1.000000         0.875000                    97.0   \n",
       "21            0.875000         0.875000                    87.0   \n",
       "22            0.833333         0.833333                    91.0   \n",
       "23            0.894737         0.850000                    83.0   \n",
       "24            0.518519         0.933333                    77.0   \n",
       "\n",
       "    base_adj_true_positive  base_arrow_f1_score  base_arrow_f1_score_ce  \\\n",
       "15                    11.0             0.500000                0.588235   \n",
       "16                    11.0             0.631579                0.705882   \n",
       "17                    14.0             0.592593                0.666667   \n",
       "18                    13.0             0.588235                0.800000   \n",
       "19                    11.0             0.275862                0.470588   \n",
       "20                     7.0             0.545455                0.600000   \n",
       "21                    14.0             0.571429                0.666667   \n",
       "22                    10.0             0.545455                0.631579   \n",
       "23                    17.0             0.424242                0.500000   \n",
       "24                    14.0             0.263158                0.416667   \n",
       "\n",
       "    base_arrow_false_negative  base_arrow_false_negative_ce  \\\n",
       "15                        9.0                           6.0   \n",
       "16                        7.0                           5.0   \n",
       "17                        9.0                           6.0   \n",
       "18                        4.0                           3.0   \n",
       "19                        7.0                           7.0   \n",
       "20                        5.0                           4.0   \n",
       "21                        8.0                           6.0   \n",
       "22                        6.0                           4.0   \n",
       "23                       13.0                          10.0   \n",
       "24                       10.0                           9.0   \n",
       "\n",
       "    base_arrow_false_positive  base_arrow_false_positive_ce  \\\n",
       "15                        1.0                           1.0   \n",
       "16                        0.0                           0.0   \n",
       "17                        2.0                           2.0   \n",
       "18                       10.0                           2.0   \n",
       "19                       14.0                           2.0   \n",
       "20                        0.0                           0.0   \n",
       "21                        4.0                           2.0   \n",
       "22                        4.0                           3.0   \n",
       "23                        6.0                           4.0   \n",
       "24                       18.0                           5.0   \n",
       "\n",
       "    base_arrow_precision  base_arrow_precision_ce  base_arrow_recall  \\\n",
       "15              0.833333                 0.833333           0.357143   \n",
       "16              1.000000                 1.000000           0.461538   \n",
       "17              0.800000                 0.800000           0.470588   \n",
       "18              0.500000                 0.833333           0.714286   \n",
       "19              0.222222                 0.666667           0.363636   \n",
       "20              1.000000                 1.000000           0.375000   \n",
       "21              0.666667                 0.800000           0.500000   \n",
       "22              0.600000                 0.666667           0.500000   \n",
       "23              0.538462                 0.636364           0.350000   \n",
       "24              0.217391                 0.500000           0.333333   \n",
       "\n",
       "    base_arrow_recall_ce  base_arrow_true_negative  \\\n",
       "15              0.454545                     210.0   \n",
       "16              0.545455                     212.0   \n",
       "17              0.571429                     206.0   \n",
       "18              0.769231                     201.0   \n",
       "19              0.363636                     200.0   \n",
       "20              0.428571                     217.0   \n",
       "21              0.571429                     205.0   \n",
       "22              0.600000                     209.0   \n",
       "23              0.411765                     199.0   \n",
       "24              0.357143                     192.0   \n",
       "\n",
       "    base_arrow_true_negative_ce  base_arrow_true_positive  \\\n",
       "15                        213.0                       5.0   \n",
       "16                        214.0                       6.0   \n",
       "17                        209.0                       8.0   \n",
       "18                        210.0                      10.0   \n",
       "19                        212.0                       4.0   \n",
       "20                        218.0                       3.0   \n",
       "21                        209.0                       8.0   \n",
       "22                        212.0                       6.0   \n",
       "23                        204.0                       7.0   \n",
       "24                        206.0                       5.0   \n",
       "\n",
       "    base_arrow_true_positive_ce  base_base_shd  base_sid_lower  \\\n",
       "15                          5.0            4.0             NaN   \n",
       "16                          6.0            2.0             NaN   \n",
       "17                          8.0            5.0             NaN   \n",
       "18                         10.0           11.0             NaN   \n",
       "19                          4.0           16.0             NaN   \n",
       "20                          3.0            1.0             NaN   \n",
       "21                          8.0            6.0             NaN   \n",
       "22                          6.0            7.0             NaN   \n",
       "23                          7.0            9.0             NaN   \n",
       "24                          5.0           19.0             NaN   \n",
       "\n",
       "    cluster_adj_f1_score  cluster_adj_false_negative  \\\n",
       "15              0.923077                         2.0   \n",
       "16              0.880000                         2.0   \n",
       "17              0.903226                         3.0   \n",
       "18              0.722222                         1.0   \n",
       "19              0.611111                         0.0   \n",
       "20              0.933333                         1.0   \n",
       "21              0.875000                         2.0   \n",
       "22              0.880000                         1.0   \n",
       "23              0.857143                         2.0   \n",
       "24              0.652174                         0.0   \n",
       "\n",
       "    cluster_adj_false_positive  cluster_adj_precision  cluster_adj_recall  \\\n",
       "15                         0.0               1.000000            0.857143   \n",
       "16                         1.0               0.916667            0.846154   \n",
       "17                         0.0               1.000000            0.823529   \n",
       "18                         9.0               0.590909            0.928571   \n",
       "19                        14.0               0.440000            1.000000   \n",
       "20                         0.0               1.000000            0.875000   \n",
       "21                         2.0               0.875000            0.875000   \n",
       "22                         2.0               0.846154            0.916667   \n",
       "23                         4.0               0.818182            0.900000   \n",
       "24                        16.0               0.483871            1.000000   \n",
       "\n",
       "    cluster_adj_true_negative  cluster_adj_true_positive  \\\n",
       "15                       91.0                       12.0   \n",
       "16                       91.0                       11.0   \n",
       "17                       88.0                       14.0   \n",
       "18                       82.0                       13.0   \n",
       "19                       80.0                       11.0   \n",
       "20                       97.0                        7.0   \n",
       "21                       87.0                       14.0   \n",
       "22                       91.0                       11.0   \n",
       "23                       81.0                       18.0   \n",
       "24                       74.0                       15.0   \n",
       "\n",
       "    cluster_arrow_f1_score  cluster_arrow_f1_score_ce  \\\n",
       "15                0.782609                   0.857143   \n",
       "16                0.880000                   1.000000   \n",
       "17                0.642857                   0.720000   \n",
       "18                0.647059                   0.880000   \n",
       "19                0.344828                   0.588235   \n",
       "20                0.933333                   1.000000   \n",
       "21                0.733333                   0.846154   \n",
       "22                0.857143                   0.900000   \n",
       "23                0.750000                   0.882353   \n",
       "24                0.577778                   0.866667   \n",
       "\n",
       "    cluster_arrow_false_negative  cluster_arrow_false_negative_ce  \\\n",
       "15                           5.0                              3.0   \n",
       "16                           2.0                              0.0   \n",
       "17                           8.0                              5.0   \n",
       "18                           3.0                              2.0   \n",
       "19                           6.0                              6.0   \n",
       "20                           1.0                              0.0   \n",
       "21                           5.0                              3.0   \n",
       "22                           3.0                              2.0   \n",
       "23                           5.0                              3.0   \n",
       "24                           2.0                              2.0   \n",
       "\n",
       "    cluster_arrow_false_positive  cluster_arrow_false_positive_ce  \\\n",
       "15                           0.0                              0.0   \n",
       "16                           1.0                              0.0   \n",
       "17                           2.0                              2.0   \n",
       "18                           9.0                              1.0   \n",
       "19                          13.0                              1.0   \n",
       "20                           0.0                              0.0   \n",
       "21                           3.0                              1.0   \n",
       "22                           0.0                              0.0   \n",
       "23                           5.0                              1.0   \n",
       "24                          17.0                              2.0   \n",
       "\n",
       "    cluster_arrow_precision  cluster_arrow_precision_ce  cluster_arrow_recall  \\\n",
       "15                 1.000000                    1.000000              0.642857   \n",
       "16                 0.916667                    1.000000              0.846154   \n",
       "17                 0.818182                    0.818182              0.529412   \n",
       "18                 0.550000                    0.916667              0.785714   \n",
       "19                 0.277778                    0.833333              0.454545   \n",
       "20                 1.000000                    1.000000              0.875000   \n",
       "21                 0.785714                    0.916667              0.687500   \n",
       "22                 1.000000                    1.000000              0.750000   \n",
       "23                 0.750000                    0.937500              0.750000   \n",
       "24                 0.433333                    0.866667              0.866667   \n",
       "\n",
       "    cluster_arrow_recall_ce  cluster_arrow_true_negative  \\\n",
       "15                 0.750000                        211.0   \n",
       "16                 1.000000                        211.0   \n",
       "17                 0.642857                        206.0   \n",
       "18                 0.846154                        202.0   \n",
       "19                 0.454545                        201.0   \n",
       "20                 1.000000                        217.0   \n",
       "21                 0.785714                        206.0   \n",
       "22                 0.818182                        213.0   \n",
       "23                 0.833333                        200.0   \n",
       "24                 0.866667                        193.0   \n",
       "\n",
       "    cluster_arrow_true_negative_ce  cluster_arrow_true_positive  \\\n",
       "15                           213.0                          9.0   \n",
       "16                           214.0                         11.0   \n",
       "17                           209.0                          9.0   \n",
       "18                           211.0                         11.0   \n",
       "19                           213.0                          5.0   \n",
       "20                           218.0                          7.0   \n",
       "21                           210.0                         11.0   \n",
       "22                           214.0                          9.0   \n",
       "23                           206.0                         15.0   \n",
       "24                           208.0                         13.0   \n",
       "\n",
       "    cluster_arrow_true_positive_ce  cluster_cluster_shd  cluster_sid_lower  \\\n",
       "15                             9.0                  2.0                NaN   \n",
       "16                            11.0                  3.0                NaN   \n",
       "17                             9.0                  5.0                NaN   \n",
       "18                            11.0                 11.0                NaN   \n",
       "19                             5.0                 15.0                NaN   \n",
       "20                             7.0                  1.0                NaN   \n",
       "21                            11.0                  5.0                NaN   \n",
       "22                             9.0                  3.0                NaN   \n",
       "23                            15.0                  7.0                NaN   \n",
       "24                            13.0                 18.0                NaN   \n",
       "\n",
       "    pruned_base_adj_f1_score  pruned_base_adj_false_negative  \\\n",
       "15                  0.880000                             3.0   \n",
       "16                  0.916667                             2.0   \n",
       "17                  0.903226                             3.0   \n",
       "18                  0.742857                             1.0   \n",
       "19                  0.611111                             0.0   \n",
       "20                  0.933333                             1.0   \n",
       "21                  0.875000                             2.0   \n",
       "22                  0.833333                             2.0   \n",
       "23                  0.871795                             3.0   \n",
       "24                  0.666667                             1.0   \n",
       "\n",
       "    pruned_base_adj_false_positive  pruned_base_adj_precision  \\\n",
       "15                             0.0                   1.000000   \n",
       "16                             0.0                   1.000000   \n",
       "17                             0.0                   1.000000   \n",
       "18                             8.0                   0.619048   \n",
       "19                            14.0                   0.440000   \n",
       "20                             0.0                   1.000000   \n",
       "21                             2.0                   0.875000   \n",
       "22                             2.0                   0.833333   \n",
       "23                             2.0                   0.894737   \n",
       "24                            13.0                   0.518519   \n",
       "\n",
       "    pruned_base_adj_recall  pruned_base_adj_true_negative  \\\n",
       "15                0.785714                           91.0   \n",
       "16                0.846154                           92.0   \n",
       "17                0.823529                           88.0   \n",
       "18                0.928571                           83.0   \n",
       "19                1.000000                           80.0   \n",
       "20                0.875000                           97.0   \n",
       "21                0.875000                           87.0   \n",
       "22                0.833333                           91.0   \n",
       "23                0.850000                           83.0   \n",
       "24                0.933333                           77.0   \n",
       "\n",
       "    pruned_base_adj_true_positive  pruned_base_arrow_f1_score  \\\n",
       "15                           11.0                    0.500000   \n",
       "16                           11.0                    0.631579   \n",
       "17                           14.0                    0.592593   \n",
       "18                           13.0                    0.588235   \n",
       "19                           11.0                    0.275862   \n",
       "20                            7.0                    0.545455   \n",
       "21                           14.0                    0.571429   \n",
       "22                           10.0                    0.545455   \n",
       "23                           17.0                    0.424242   \n",
       "24                           14.0                    0.263158   \n",
       "\n",
       "    pruned_base_arrow_f1_score_ce  pruned_base_arrow_false_negative  \\\n",
       "15                       0.588235                               9.0   \n",
       "16                       0.705882                               7.0   \n",
       "17                       0.666667                               9.0   \n",
       "18                       0.800000                               4.0   \n",
       "19                       0.470588                               7.0   \n",
       "20                       0.600000                               5.0   \n",
       "21                       0.666667                               8.0   \n",
       "22                       0.631579                               6.0   \n",
       "23                       0.500000                              13.0   \n",
       "24                       0.416667                              10.0   \n",
       "\n",
       "    pruned_base_arrow_false_negative_ce  pruned_base_arrow_false_positive  \\\n",
       "15                                  6.0                               1.0   \n",
       "16                                  5.0                               0.0   \n",
       "17                                  6.0                               2.0   \n",
       "18                                  3.0                              10.0   \n",
       "19                                  7.0                              14.0   \n",
       "20                                  4.0                               0.0   \n",
       "21                                  6.0                               4.0   \n",
       "22                                  4.0                               4.0   \n",
       "23                                 10.0                               6.0   \n",
       "24                                  9.0                              18.0   \n",
       "\n",
       "    pruned_base_arrow_false_positive_ce  pruned_base_arrow_precision  \\\n",
       "15                                  1.0                     0.833333   \n",
       "16                                  0.0                     1.000000   \n",
       "17                                  2.0                     0.800000   \n",
       "18                                  2.0                     0.500000   \n",
       "19                                  2.0                     0.222222   \n",
       "20                                  0.0                     1.000000   \n",
       "21                                  2.0                     0.666667   \n",
       "22                                  3.0                     0.600000   \n",
       "23                                  4.0                     0.538462   \n",
       "24                                  5.0                     0.217391   \n",
       "\n",
       "    pruned_base_arrow_precision_ce  pruned_base_arrow_recall  \\\n",
       "15                        0.833333                  0.357143   \n",
       "16                        1.000000                  0.461538   \n",
       "17                        0.800000                  0.470588   \n",
       "18                        0.833333                  0.714286   \n",
       "19                        0.666667                  0.363636   \n",
       "20                        1.000000                  0.375000   \n",
       "21                        0.800000                  0.500000   \n",
       "22                        0.666667                  0.500000   \n",
       "23                        0.636364                  0.350000   \n",
       "24                        0.500000                  0.333333   \n",
       "\n",
       "    pruned_base_arrow_recall_ce  pruned_base_arrow_true_negative  \\\n",
       "15                     0.454545                            210.0   \n",
       "16                     0.545455                            212.0   \n",
       "17                     0.571429                            206.0   \n",
       "18                     0.769231                            201.0   \n",
       "19                     0.363636                            200.0   \n",
       "20                     0.428571                            217.0   \n",
       "21                     0.571429                            205.0   \n",
       "22                     0.600000                            209.0   \n",
       "23                     0.411765                            199.0   \n",
       "24                     0.357143                            192.0   \n",
       "\n",
       "    pruned_base_arrow_true_negative_ce  pruned_base_arrow_true_positive  \\\n",
       "15                               213.0                              5.0   \n",
       "16                               214.0                              6.0   \n",
       "17                               209.0                              8.0   \n",
       "18                               210.0                             10.0   \n",
       "19                               212.0                              4.0   \n",
       "20                               218.0                              3.0   \n",
       "21                               209.0                              8.0   \n",
       "22                               212.0                              6.0   \n",
       "23                               204.0                              7.0   \n",
       "24                               206.0                              5.0   \n",
       "\n",
       "    pruned_base_arrow_true_positive_ce  pruned_base_pruned_base_shd  \\\n",
       "15                                 5.0                          4.0   \n",
       "16                                 6.0                          2.0   \n",
       "17                                 8.0                          5.0   \n",
       "18                                10.0                         11.0   \n",
       "19                                 4.0                         16.0   \n",
       "20                                 3.0                          1.0   \n",
       "21                                 8.0                          6.0   \n",
       "22                                 6.0                          7.0   \n",
       "23                                 7.0                          9.0   \n",
       "24                                 5.0                         19.0   \n",
       "\n",
       "    pruned_base_sid_lower  Base indep tests  C-PC indep tests  alpha  \\\n",
       "15                    NaN             778.0             512.0   0.01   \n",
       "16                    NaN             389.0             293.0   0.05   \n",
       "17                    NaN             645.0             619.0   0.10   \n",
       "18                    NaN            1959.0            1131.0   0.25   \n",
       "19                    NaN            1740.0            1423.0   0.50   \n",
       "20                    NaN             281.0             213.0   0.01   \n",
       "21                    NaN             545.0             430.0   0.05   \n",
       "22                    NaN             543.0             428.0   0.10   \n",
       "23                    NaN            1701.0            1152.0   0.25   \n",
       "24                    NaN            2196.0            1681.0   0.50   \n",
       "\n",
       "    cluster_connectivity   dag_method distribution_type         edge_ratios  \\\n",
       "15                   1.0  erdos_renyi               exp  [0.15, 0.14, 0.14]   \n",
       "16                   1.0  erdos_renyi               exp  [0.12, 0.12, 0.12]   \n",
       "17                   1.0  erdos_renyi               exp  [0.33, 0.14, 0.14]   \n",
       "18                   1.0  erdos_renyi               exp  [0.22, 0.09, 0.09]   \n",
       "19                   1.0  erdos_renyi               exp   [0.3, 0.07, 0.07]   \n",
       "20                   1.0  erdos_renyi             gauss   [0.1, 0.12, 0.12]   \n",
       "21                   1.0  erdos_renyi             gauss  [0.14, 0.16, 0.16]   \n",
       "22                   1.0  erdos_renyi             gauss  [0.18, 0.18, 0.18]   \n",
       "23                   1.0  erdos_renyi             gauss  [0.42, 0.19, 0.19]   \n",
       "24                   1.0  erdos_renyi             gauss  [0.06, 0.18, 0.18]   \n",
       "\n",
       "    empty_graph_shd indep_test  n_clusters  n_edges  n_nodes  noise_scale  \\\n",
       "15             14.0    fisherz         2.0     15.0     15.0          1.0   \n",
       "16             13.0    fisherz         2.0     15.0     15.0          1.0   \n",
       "17             17.0    fisherz         2.0     15.0     15.0          1.0   \n",
       "18             14.0    fisherz         2.0     15.0     15.0          1.0   \n",
       "19             11.0    fisherz         2.0     15.0     15.0          1.0   \n",
       "20              8.0    fisherz         2.0     15.0     15.0          1.0   \n",
       "21             16.0    fisherz         2.0     15.0     15.0          1.0   \n",
       "22             12.0    fisherz         2.0     15.0     15.0          1.0   \n",
       "23             20.0    fisherz         2.0     15.0     15.0          1.0   \n",
       "24             15.0    fisherz         2.0     15.0     15.0          1.0   \n",
       "\n",
       "    sample_size scm_method  seed  true_sid_lower  true_sid_upper weight_range  \n",
       "15       1000.0     linear  None             NaN             NaN      (-1, 2)  \n",
       "16       1000.0     linear  None             NaN             NaN      (-1, 2)  \n",
       "17       1000.0     linear  None             NaN             NaN      (-1, 2)  \n",
       "18       1000.0     linear  None             NaN             NaN      (-1, 2)  \n",
       "19       1000.0     linear  None             NaN             NaN      (-1, 2)  \n",
       "20       1000.0     linear  None             NaN             NaN      (-1, 2)  \n",
       "21       1000.0     linear  None             NaN             NaN      (-1, 2)  \n",
       "22       1000.0     linear  None             NaN             NaN      (-1, 2)  \n",
       "23       1000.0     linear  None             NaN             NaN      (-1, 2)  \n",
       "24       1000.0     linear  None             NaN             NaN      (-1, 2)  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option('display.max_columns', None)\n",
    "data[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The data contains the following information on the settings of the simulation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# settings_results = {\n",
    "#             \"n_nodes\": simulation.n_nodes,\n",
    "#             \"n_edges\": simulation.n_edges,\n",
    "#             \"n_clusters\": simulation.n_clusters,\n",
    "#             \"edge_ratios\": edge_ratios,\n",
    "#             \"dag_method\": simulation.dag_method,\n",
    "#             \"distribution_type\": simulation.distribution_type,\n",
    "#             \"scm_method\": simulation.scm_method,\n",
    "#             \"weight_range\": simulation.weight_range,\n",
    "#             \"sample_size\": simulation.sample_size,\n",
    "#             \"seed\": simulation.seed,\n",
    "#             \"noise_scale\": simulation.noise_scale,\n",
    "#             # \"n_c_edges\": simulation.n_c_edges,\n",
    "#             \"alpha\": simulation.alpha,\n",
    "#             \"true_sid_lower\": true_sid_bounds[\"sid_lower\"],\n",
    "#             \"true_sid_upper\": true_sid_bounds[\"sid_upper\"],\n",
    "#             \"indep_test\": self.indep_test,\n",
    "#             \"empty_graph_shd\": empty_graph_shd,\n",
    "#             \"cluster_connectivity\": cluster_connectivity,\n",
    "#             \"C-PC indep tests\": clust_no_indep_tests,\n",
    "#             \"Base indep tests\": one_clust_no_indep_tests,\n",
    "#         }"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In addition the metrics for the runs are as follows (where pruned is the results of PC but edges that could be removed from C-DAG are removed):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# cluster_adjacency_confusion,\n",
    "# cluster_arrow_confusion,\n",
    "# cluster_shd,\n",
    "# cluster_sid,\n",
    "\n",
    "# base_adjacency_confusion,\n",
    "# base_arrow_confusion,\n",
    "# base_shd,\n",
    "# base_sid,\n",
    "\n",
    "# pruned_base_adjacency_confusion,\n",
    "# pruned_base_arrow_confusion,\n",
    "# pruned_base_shd,\n",
    "# pruned_base_sid,"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pruned and base results are identical, as it seems PC never failed to delete an edge between non-adjacent clusters. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <th>base_adj_false_negative</th>\n",
       "      <th>base_adj_false_positive</th>\n",
       "      <th>base_adj_precision</th>\n",
       "      <th>base_adj_recall</th>\n",
       "      <th>base_adj_true_negative</th>\n",
       "      <th>base_adj_true_positive</th>\n",
       "      <th>base_arrow_f1_score</th>\n",
       "      <th>base_arrow_f1_score_ce</th>\n",
       "      <th>base_arrow_false_negative</th>\n",
       "      <th>base_arrow_false_negative_ce</th>\n",
       "      <th>base_arrow_false_positive</th>\n",
       "      <th>base_arrow_false_positive_ce</th>\n",
       "      <th>base_arrow_precision</th>\n",
       "      <th>base_arrow_precision_ce</th>\n",
       "      <th>base_arrow_recall</th>\n",
       "      <th>base_arrow_recall_ce</th>\n",
       "      <th>base_arrow_true_negative</th>\n",
       "      <th>base_arrow_true_negative_ce</th>\n",
       "      <th>base_arrow_true_positive</th>\n",
       "      <th>base_arrow_true_positive_ce</th>\n",
       "      <th>base_base_shd</th>\n",
       "      <th>base_sid_lower</th>\n",
       "      <th>cluster_adj_f1_score</th>\n",
       "      <th>cluster_adj_false_negative</th>\n",
       "      <th>cluster_adj_false_positive</th>\n",
       "      <th>cluster_adj_precision</th>\n",
       "      <th>cluster_adj_recall</th>\n",
       "      <th>cluster_adj_true_negative</th>\n",
       "      <th>cluster_adj_true_positive</th>\n",
       "      <th>cluster_arrow_f1_score</th>\n",
       "      <th>cluster_arrow_f1_score_ce</th>\n",
       "      <th>cluster_arrow_false_negative</th>\n",
       "      <th>cluster_arrow_false_negative_ce</th>\n",
       "      <th>cluster_arrow_false_positive</th>\n",
       "      <th>cluster_arrow_false_positive_ce</th>\n",
       "      <th>cluster_arrow_precision</th>\n",
       "      <th>cluster_arrow_precision_ce</th>\n",
       "      <th>cluster_arrow_recall</th>\n",
       "      <th>cluster_arrow_recall_ce</th>\n",
       "      <th>cluster_arrow_true_negative</th>\n",
       "      <th>cluster_arrow_true_negative_ce</th>\n",
       "      <th>cluster_arrow_true_positive</th>\n",
       "      <th>cluster_arrow_true_positive_ce</th>\n",
       "      <th>cluster_cluster_shd</th>\n",
       "      <th>cluster_sid_lower</th>\n",
       "      <th>pruned_base_adj_f1_score</th>\n",
       "      <th>pruned_base_adj_false_negative</th>\n",
       "      <th>pruned_base_adj_false_positive</th>\n",
       "      <th>pruned_base_adj_precision</th>\n",
       "      <th>pruned_base_adj_recall</th>\n",
       "      <th>pruned_base_adj_true_negative</th>\n",
       "      <th>pruned_base_adj_true_positive</th>\n",
       "      <th>pruned_base_arrow_f1_score</th>\n",
       "      <th>pruned_base_arrow_f1_score_ce</th>\n",
       "      <th>pruned_base_arrow_false_negative</th>\n",
       "      <th>pruned_base_arrow_false_negative_ce</th>\n",
       "      <th>pruned_base_arrow_false_positive</th>\n",
       "      <th>pruned_base_arrow_false_positive_ce</th>\n",
       "      <th>pruned_base_arrow_precision</th>\n",
       "      <th>pruned_base_arrow_precision_ce</th>\n",
       "      <th>pruned_base_arrow_recall</th>\n",
       "      <th>pruned_base_arrow_recall_ce</th>\n",
       "      <th>pruned_base_arrow_true_negative</th>\n",
       "      <th>pruned_base_arrow_true_negative_ce</th>\n",
       "      <th>pruned_base_arrow_true_positive</th>\n",
       "      <th>pruned_base_arrow_true_positive_ce</th>\n",
       "      <th>pruned_base_pruned_base_shd</th>\n",
       "      <th>pruned_base_sid_lower</th>\n",
       "      <th>Base indep tests</th>\n",
       "      <th>C-PC indep tests</th>\n",
       "      <th>alpha</th>\n",
       "      <th>cluster_connectivity</th>\n",
       "      <th>empty_graph_shd</th>\n",
       "      <th>n_clusters</th>\n",
       "      <th>n_edges</th>\n",
       "      <th>n_nodes</th>\n",
       "      <th>noise_scale</th>\n",
       "      <th>sample_size</th>\n",
       "      <th>true_sid_lower</th>\n",
       "      <th>true_sid_upper</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1078.000000</td>\n",
       "      <td>1078.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1079.000000</td>\n",
       "      <td>1079.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1078.000000</td>\n",
       "      <td>1078.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.000000</td>\n",
       "      <td>1080.0</td>\n",
       "      <td>1080.0</td>\n",
       "      <td>1080.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.558015</td>\n",
       "      <td>33.237037</td>\n",
       "      <td>2.963889</td>\n",
       "      <td>0.863080</td>\n",
       "      <td>0.463783</td>\n",
       "      <td>51.439815</td>\n",
       "      <td>17.359259</td>\n",
       "      <td>0.343881</td>\n",
       "      <td>0.632246</td>\n",
       "      <td>40.306481</td>\n",
       "      <td>7.069444</td>\n",
       "      <td>6.856481</td>\n",
       "      <td>4.170370</td>\n",
       "      <td>0.610799</td>\n",
       "      <td>0.709428</td>\n",
       "      <td>0.264068</td>\n",
       "      <td>0.577924</td>\n",
       "      <td>167.547222</td>\n",
       "      <td>203.470370</td>\n",
       "      <td>10.289815</td>\n",
       "      <td>10.289815</td>\n",
       "      <td>40.371296</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.621049</td>\n",
       "      <td>28.457407</td>\n",
       "      <td>4.598148</td>\n",
       "      <td>0.836135</td>\n",
       "      <td>0.545386</td>\n",
       "      <td>49.805556</td>\n",
       "      <td>22.138889</td>\n",
       "      <td>0.526171</td>\n",
       "      <td>0.841170</td>\n",
       "      <td>32.137037</td>\n",
       "      <td>3.679630</td>\n",
       "      <td>6.642593</td>\n",
       "      <td>2.333333</td>\n",
       "      <td>0.731029</td>\n",
       "      <td>0.872837</td>\n",
       "      <td>0.452331</td>\n",
       "      <td>0.815458</td>\n",
       "      <td>167.761111</td>\n",
       "      <td>200.527778</td>\n",
       "      <td>18.459259</td>\n",
       "      <td>18.459259</td>\n",
       "      <td>35.388889</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.558015</td>\n",
       "      <td>33.237037</td>\n",
       "      <td>2.963889</td>\n",
       "      <td>0.863080</td>\n",
       "      <td>0.463783</td>\n",
       "      <td>51.439815</td>\n",
       "      <td>17.359259</td>\n",
       "      <td>0.343881</td>\n",
       "      <td>0.632246</td>\n",
       "      <td>40.306481</td>\n",
       "      <td>7.069444</td>\n",
       "      <td>6.856481</td>\n",
       "      <td>4.170370</td>\n",
       "      <td>0.610799</td>\n",
       "      <td>0.709428</td>\n",
       "      <td>0.264068</td>\n",
       "      <td>0.577924</td>\n",
       "      <td>167.547222</td>\n",
       "      <td>203.470370</td>\n",
       "      <td>10.289815</td>\n",
       "      <td>10.289815</td>\n",
       "      <td>40.371296</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4580.461111</td>\n",
       "      <td>2994.563889</td>\n",
       "      <td>0.182000</td>\n",
       "      <td>0.878259</td>\n",
       "      <td>50.596296</td>\n",
       "      <td>3.500000</td>\n",
       "      <td>43.750000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.197564</td>\n",
       "      <td>24.538131</td>\n",
       "      <td>2.952135</td>\n",
       "      <td>0.119324</td>\n",
       "      <td>0.253751</td>\n",
       "      <td>25.915885</td>\n",
       "      <td>3.956273</td>\n",
       "      <td>0.163435</td>\n",
       "      <td>0.171687</td>\n",
       "      <td>24.603653</td>\n",
       "      <td>2.925994</td>\n",
       "      <td>4.285324</td>\n",
       "      <td>2.853652</td>\n",
       "      <td>0.200017</td>\n",
       "      <td>0.185393</td>\n",
       "      <td>0.165490</td>\n",
       "      <td>0.177350</td>\n",
       "      <td>26.462059</td>\n",
       "      <td>5.503647</td>\n",
       "      <td>4.588163</td>\n",
       "      <td>4.588163</td>\n",
       "      <td>24.568304</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.160947</td>\n",
       "      <td>21.838272</td>\n",
       "      <td>3.999026</td>\n",
       "      <td>0.129122</td>\n",
       "      <td>0.231577</td>\n",
       "      <td>25.841951</td>\n",
       "      <td>7.121556</td>\n",
       "      <td>0.186572</td>\n",
       "      <td>0.158696</td>\n",
       "      <td>22.619177</td>\n",
       "      <td>3.083808</td>\n",
       "      <td>4.448453</td>\n",
       "      <td>2.388573</td>\n",
       "      <td>0.174157</td>\n",
       "      <td>0.146137</td>\n",
       "      <td>0.228972</td>\n",
       "      <td>0.176456</td>\n",
       "      <td>26.424677</td>\n",
       "      <td>7.520079</td>\n",
       "      <td>8.008464</td>\n",
       "      <td>8.008464</td>\n",
       "      <td>21.895714</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.197564</td>\n",
       "      <td>24.538131</td>\n",
       "      <td>2.952135</td>\n",
       "      <td>0.119324</td>\n",
       "      <td>0.253751</td>\n",
       "      <td>25.915885</td>\n",
       "      <td>3.956273</td>\n",
       "      <td>0.163435</td>\n",
       "      <td>0.171687</td>\n",
       "      <td>24.603653</td>\n",
       "      <td>2.925994</td>\n",
       "      <td>4.285324</td>\n",
       "      <td>2.853652</td>\n",
       "      <td>0.200017</td>\n",
       "      <td>0.185393</td>\n",
       "      <td>0.165490</td>\n",
       "      <td>0.177350</td>\n",
       "      <td>26.462059</td>\n",
       "      <td>5.503647</td>\n",
       "      <td>4.588163</td>\n",
       "      <td>4.588163</td>\n",
       "      <td>24.568304</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2791.169654</td>\n",
       "      <td>2140.460093</td>\n",
       "      <td>0.178677</td>\n",
       "      <td>0.172465</td>\n",
       "      <td>26.367545</td>\n",
       "      <td>1.708616</td>\n",
       "      <td>24.346177</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.231579</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.187500</td>\n",
       "      <td>0.134146</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.037037</td>\n",
       "      <td>0.125000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>125.000000</td>\n",
       "      <td>183.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.235294</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.176471</td>\n",
       "      <td>0.135135</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.037037</td>\n",
       "      <td>0.133333</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>127.000000</td>\n",
       "      <td>178.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.231579</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.187500</td>\n",
       "      <td>0.134146</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.037037</td>\n",
       "      <td>0.125000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>125.000000</td>\n",
       "      <td>183.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>251.000000</td>\n",
       "      <td>103.000000</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.166667</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.391641</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.809066</td>\n",
       "      <td>0.253679</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>0.228571</td>\n",
       "      <td>0.514286</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.469608</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>0.144697</td>\n",
       "      <td>0.454545</td>\n",
       "      <td>141.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.783490</td>\n",
       "      <td>0.360000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>0.403670</td>\n",
       "      <td>0.775510</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.635809</td>\n",
       "      <td>0.823529</td>\n",
       "      <td>0.282051</td>\n",
       "      <td>0.736842</td>\n",
       "      <td>141.000000</td>\n",
       "      <td>196.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.391641</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.809066</td>\n",
       "      <td>0.253679</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>0.228571</td>\n",
       "      <td>0.514286</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.469608</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>0.144697</td>\n",
       "      <td>0.454545</td>\n",
       "      <td>141.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2598.500000</td>\n",
       "      <td>1469.000000</td>\n",
       "      <td>0.050000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>26.250000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.507463</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.360000</td>\n",
       "      <td>53.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>0.311111</td>\n",
       "      <td>0.642857</td>\n",
       "      <td>40.500000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.718254</td>\n",
       "      <td>0.215190</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>167.000000</td>\n",
       "      <td>203.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.602740</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>0.480260</td>\n",
       "      <td>51.000000</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>0.519481</td>\n",
       "      <td>0.892857</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.760000</td>\n",
       "      <td>0.916667</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>168.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>34.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.507463</td>\n",
       "      <td>31.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.360000</td>\n",
       "      <td>53.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>0.311111</td>\n",
       "      <td>0.642857</td>\n",
       "      <td>40.500000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.718254</td>\n",
       "      <td>0.215190</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>167.000000</td>\n",
       "      <td>203.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4142.500000</td>\n",
       "      <td>2571.000000</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>3.500000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.723588</td>\n",
       "      <td>58.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.947368</td>\n",
       "      <td>0.653846</td>\n",
       "      <td>76.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>0.434783</td>\n",
       "      <td>0.756757</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.850000</td>\n",
       "      <td>0.346154</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>191.000000</td>\n",
       "      <td>207.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>65.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.736842</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.928571</td>\n",
       "      <td>0.730769</td>\n",
       "      <td>74.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>0.636700</td>\n",
       "      <td>0.956039</td>\n",
       "      <td>53.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>0.941597</td>\n",
       "      <td>192.000000</td>\n",
       "      <td>205.000000</td>\n",
       "      <td>23.250000</td>\n",
       "      <td>23.250000</td>\n",
       "      <td>56.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.723588</td>\n",
       "      <td>58.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.947368</td>\n",
       "      <td>0.653846</td>\n",
       "      <td>76.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>0.434783</td>\n",
       "      <td>0.756757</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.850000</td>\n",
       "      <td>0.346154</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>191.000000</td>\n",
       "      <td>207.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>65.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6045.000000</td>\n",
       "      <td>3933.750000</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>79.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>57.500000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>73.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>83.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>219.000000</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>78.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>16417.000000</td>\n",
       "      <td>16417.000000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>89.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       base_adj_f1_score  base_adj_false_negative  base_adj_false_positive  \\\n",
       "count        1080.000000              1080.000000              1080.000000   \n",
       "mean            0.558015                33.237037                 2.963889   \n",
       "std             0.197564                24.538131                 2.952135   \n",
       "min             0.231579                 0.000000                 0.000000   \n",
       "25%             0.391641                 9.000000                 1.000000   \n",
       "50%             0.507463                31.000000                 2.000000   \n",
       "75%             0.723588                58.000000                 4.000000   \n",
       "max             1.000000                75.000000                26.000000   \n",
       "\n",
       "       base_adj_precision  base_adj_recall  base_adj_true_negative  \\\n",
       "count         1080.000000      1080.000000             1080.000000   \n",
       "mean             0.863080         0.463783               51.439815   \n",
       "std              0.119324         0.253751               25.915885   \n",
       "min              0.187500         0.134146               12.000000   \n",
       "25%              0.809066         0.253679               25.000000   \n",
       "50%              0.888889         0.360000               53.000000   \n",
       "75%              0.947368         0.653846               76.000000   \n",
       "max              1.000000         1.000000               98.000000   \n",
       "\n",
       "       base_adj_true_positive  base_arrow_f1_score  base_arrow_f1_score_ce  \\\n",
       "count             1080.000000          1078.000000             1078.000000   \n",
       "mean                17.359259             0.343881                0.632246   \n",
       "std                  3.956273             0.163435                0.171687   \n",
       "min                  6.000000             0.037037                0.125000   \n",
       "25%                 15.000000             0.228571                0.514286   \n",
       "50%                 17.000000             0.311111                0.642857   \n",
       "75%                 20.000000             0.434783                0.756757   \n",
       "max                 29.000000             0.923077                1.000000   \n",
       "\n",
       "       base_arrow_false_negative  base_arrow_false_negative_ce  \\\n",
       "count                1080.000000                   1080.000000   \n",
       "mean                   40.306481                      7.069444   \n",
       "std                    24.603653                      2.925994   \n",
       "min                     1.000000                      0.000000   \n",
       "25%                    17.000000                      5.000000   \n",
       "50%                    40.500000                      7.000000   \n",
       "75%                    64.000000                      9.000000   \n",
       "max                    84.000000                     18.000000   \n",
       "\n",
       "       base_arrow_false_positive  base_arrow_false_positive_ce  \\\n",
       "count                1080.000000                   1080.000000   \n",
       "mean                    6.856481                      4.170370   \n",
       "std                     4.285324                      2.853652   \n",
       "min                     0.000000                      0.000000   \n",
       "25%                     4.000000                      2.000000   \n",
       "50%                     6.000000                      4.000000   \n",
       "75%                     9.000000                      6.000000   \n",
       "max                    28.000000                     17.000000   \n",
       "\n",
       "       base_arrow_precision  base_arrow_precision_ce  base_arrow_recall  \\\n",
       "count           1080.000000              1080.000000        1080.000000   \n",
       "mean               0.610799                 0.709428           0.264068   \n",
       "std                0.200017                 0.185393           0.165490   \n",
       "min                0.000000                 0.000000           0.000000   \n",
       "25%                0.469608                 0.583333           0.144697   \n",
       "50%                0.611111                 0.718254           0.215190   \n",
       "75%                0.750000                 0.850000           0.346154   \n",
       "max                1.000000                 1.000000           0.888889   \n",
       "\n",
       "       base_arrow_recall_ce  base_arrow_true_negative  \\\n",
       "count           1080.000000               1080.000000   \n",
       "mean               0.577924                167.547222   \n",
       "std                0.177350                 26.462059   \n",
       "min                0.000000                125.000000   \n",
       "25%                0.454545                141.000000   \n",
       "50%                0.583333                167.000000   \n",
       "75%                0.700000                191.000000   \n",
       "max                1.000000                218.000000   \n",
       "\n",
       "       base_arrow_true_negative_ce  base_arrow_true_positive  \\\n",
       "count                  1080.000000               1080.000000   \n",
       "mean                    203.470370                 10.289815   \n",
       "std                       5.503647                  4.588163   \n",
       "min                     183.000000                  0.000000   \n",
       "25%                     200.000000                  7.000000   \n",
       "50%                     203.000000                 10.000000   \n",
       "75%                     207.000000                 13.000000   \n",
       "max                     218.000000                 28.000000   \n",
       "\n",
       "       base_arrow_true_positive_ce  base_base_shd  base_sid_lower  \\\n",
       "count                  1080.000000    1080.000000             0.0   \n",
       "mean                     10.289815      40.371296             NaN   \n",
       "std                       4.588163      24.568304             NaN   \n",
       "min                       0.000000       0.000000             NaN   \n",
       "25%                       7.000000      17.000000             NaN   \n",
       "50%                      10.000000      41.000000             NaN   \n",
       "75%                      13.000000      65.000000             NaN   \n",
       "max                      28.000000      84.000000             NaN   \n",
       "\n",
       "       cluster_adj_f1_score  cluster_adj_false_negative  \\\n",
       "count           1080.000000                 1080.000000   \n",
       "mean               0.621049                   28.457407   \n",
       "std                0.160947                   21.838272   \n",
       "min                0.235294                    0.000000   \n",
       "25%                0.500000                    7.000000   \n",
       "50%                0.602740                   26.000000   \n",
       "75%                0.736842                   50.000000   \n",
       "max                1.000000                   73.000000   \n",
       "\n",
       "       cluster_adj_false_positive  cluster_adj_precision  cluster_adj_recall  \\\n",
       "count                 1080.000000            1080.000000         1080.000000   \n",
       "mean                     4.598148               0.836135            0.545386   \n",
       "std                      3.999026               0.129122            0.231577   \n",
       "min                      0.000000               0.176471            0.135135   \n",
       "25%                      2.000000               0.783490            0.360000   \n",
       "50%                      4.000000               0.857143            0.480260   \n",
       "75%                      6.000000               0.928571            0.730769   \n",
       "max                     28.000000               1.000000            1.000000   \n",
       "\n",
       "       cluster_adj_true_negative  cluster_adj_true_positive  \\\n",
       "count                1080.000000                1080.000000   \n",
       "mean                   49.805556                  22.138889   \n",
       "std                    25.841951                   7.121556   \n",
       "min                    11.000000                   6.000000   \n",
       "25%                    23.000000                  17.000000   \n",
       "50%                    51.000000                  22.000000   \n",
       "75%                    74.000000                  26.000000   \n",
       "max                    98.000000                  45.000000   \n",
       "\n",
       "       cluster_arrow_f1_score  cluster_arrow_f1_score_ce  \\\n",
       "count             1079.000000                1079.000000   \n",
       "mean                 0.526171                   0.841170   \n",
       "std                  0.186572                   0.158696   \n",
       "min                  0.037037                   0.133333   \n",
       "25%                  0.403670                   0.775510   \n",
       "50%                  0.519481                   0.892857   \n",
       "75%                  0.636700                   0.956039   \n",
       "max                  1.000000                   1.000000   \n",
       "\n",
       "       cluster_arrow_false_negative  cluster_arrow_false_negative_ce  \\\n",
       "count                   1080.000000                      1080.000000   \n",
       "mean                      32.137037                         3.679630   \n",
       "std                       22.619177                         3.083808   \n",
       "min                        0.000000                         0.000000   \n",
       "25%                       10.000000                         1.000000   \n",
       "50%                       30.000000                         3.000000   \n",
       "75%                       53.000000                         6.000000   \n",
       "max                       83.000000                        15.000000   \n",
       "\n",
       "       cluster_arrow_false_positive  cluster_arrow_false_positive_ce  \\\n",
       "count                   1080.000000                      1080.000000   \n",
       "mean                       6.642593                         2.333333   \n",
       "std                        4.448453                         2.388573   \n",
       "min                        0.000000                         0.000000   \n",
       "25%                        3.000000                         0.000000   \n",
       "50%                        6.000000                         2.000000   \n",
       "75%                        9.000000                         3.000000   \n",
       "max                       28.000000                        15.000000   \n",
       "\n",
       "       cluster_arrow_precision  cluster_arrow_precision_ce  \\\n",
       "count              1080.000000                 1080.000000   \n",
       "mean                  0.731029                    0.872837   \n",
       "std                   0.174157                    0.146137   \n",
       "min                   0.000000                    0.000000   \n",
       "25%                   0.635809                    0.823529   \n",
       "50%                   0.760000                    0.916667   \n",
       "75%                   0.857143                    1.000000   \n",
       "max                   1.000000                    1.000000   \n",
       "\n",
       "       cluster_arrow_recall  cluster_arrow_recall_ce  \\\n",
       "count           1080.000000              1080.000000   \n",
       "mean               0.452331                 0.815458   \n",
       "std                0.228972                 0.176456   \n",
       "min                0.000000                 0.000000   \n",
       "25%                0.282051                 0.736842   \n",
       "50%                0.400000                 0.875000   \n",
       "75%                0.583333                 0.941597   \n",
       "max                1.000000                 1.000000   \n",
       "\n",
       "       cluster_arrow_true_negative  cluster_arrow_true_negative_ce  \\\n",
       "count                  1080.000000                     1080.000000   \n",
       "mean                    167.761111                      200.527778   \n",
       "std                      26.424677                        7.520079   \n",
       "min                     127.000000                      178.000000   \n",
       "25%                     141.000000                      196.000000   \n",
       "50%                     168.000000                      200.000000   \n",
       "75%                     192.000000                      205.000000   \n",
       "max                     218.000000                      219.000000   \n",
       "\n",
       "       cluster_arrow_true_positive  cluster_arrow_true_positive_ce  \\\n",
       "count                  1080.000000                     1080.000000   \n",
       "mean                     18.459259                       18.459259   \n",
       "std                       8.008464                        8.008464   \n",
       "min                       0.000000                        0.000000   \n",
       "25%                      13.000000                       13.000000   \n",
       "50%                      18.000000                       18.000000   \n",
       "75%                      23.250000                       23.250000   \n",
       "max                      44.000000                       44.000000   \n",
       "\n",
       "       cluster_cluster_shd  cluster_sid_lower  pruned_base_adj_f1_score  \\\n",
       "count          1080.000000                0.0               1080.000000   \n",
       "mean             35.388889                NaN                  0.558015   \n",
       "std              21.895714                NaN                  0.197564   \n",
       "min               0.000000                NaN                  0.231579   \n",
       "25%              16.000000                NaN                  0.391641   \n",
       "50%              34.000000                NaN                  0.507463   \n",
       "75%              56.000000                NaN                  0.723588   \n",
       "max              78.000000                NaN                  1.000000   \n",
       "\n",
       "       pruned_base_adj_false_negative  pruned_base_adj_false_positive  \\\n",
       "count                     1080.000000                     1080.000000   \n",
       "mean                        33.237037                        2.963889   \n",
       "std                         24.538131                        2.952135   \n",
       "min                          0.000000                        0.000000   \n",
       "25%                          9.000000                        1.000000   \n",
       "50%                         31.000000                        2.000000   \n",
       "75%                         58.000000                        4.000000   \n",
       "max                         75.000000                       26.000000   \n",
       "\n",
       "       pruned_base_adj_precision  pruned_base_adj_recall  \\\n",
       "count                1080.000000             1080.000000   \n",
       "mean                    0.863080                0.463783   \n",
       "std                     0.119324                0.253751   \n",
       "min                     0.187500                0.134146   \n",
       "25%                     0.809066                0.253679   \n",
       "50%                     0.888889                0.360000   \n",
       "75%                     0.947368                0.653846   \n",
       "max                     1.000000                1.000000   \n",
       "\n",
       "       pruned_base_adj_true_negative  pruned_base_adj_true_positive  \\\n",
       "count                    1080.000000                    1080.000000   \n",
       "mean                       51.439815                      17.359259   \n",
       "std                        25.915885                       3.956273   \n",
       "min                        12.000000                       6.000000   \n",
       "25%                        25.000000                      15.000000   \n",
       "50%                        53.000000                      17.000000   \n",
       "75%                        76.000000                      20.000000   \n",
       "max                        98.000000                      29.000000   \n",
       "\n",
       "       pruned_base_arrow_f1_score  pruned_base_arrow_f1_score_ce  \\\n",
       "count                 1078.000000                    1078.000000   \n",
       "mean                     0.343881                       0.632246   \n",
       "std                      0.163435                       0.171687   \n",
       "min                      0.037037                       0.125000   \n",
       "25%                      0.228571                       0.514286   \n",
       "50%                      0.311111                       0.642857   \n",
       "75%                      0.434783                       0.756757   \n",
       "max                      0.923077                       1.000000   \n",
       "\n",
       "       pruned_base_arrow_false_negative  pruned_base_arrow_false_negative_ce  \\\n",
       "count                       1080.000000                          1080.000000   \n",
       "mean                          40.306481                             7.069444   \n",
       "std                           24.603653                             2.925994   \n",
       "min                            1.000000                             0.000000   \n",
       "25%                           17.000000                             5.000000   \n",
       "50%                           40.500000                             7.000000   \n",
       "75%                           64.000000                             9.000000   \n",
       "max                           84.000000                            18.000000   \n",
       "\n",
       "       pruned_base_arrow_false_positive  pruned_base_arrow_false_positive_ce  \\\n",
       "count                       1080.000000                          1080.000000   \n",
       "mean                           6.856481                             4.170370   \n",
       "std                            4.285324                             2.853652   \n",
       "min                            0.000000                             0.000000   \n",
       "25%                            4.000000                             2.000000   \n",
       "50%                            6.000000                             4.000000   \n",
       "75%                            9.000000                             6.000000   \n",
       "max                           28.000000                            17.000000   \n",
       "\n",
       "       pruned_base_arrow_precision  pruned_base_arrow_precision_ce  \\\n",
       "count                  1080.000000                     1080.000000   \n",
       "mean                      0.610799                        0.709428   \n",
       "std                       0.200017                        0.185393   \n",
       "min                       0.000000                        0.000000   \n",
       "25%                       0.469608                        0.583333   \n",
       "50%                       0.611111                        0.718254   \n",
       "75%                       0.750000                        0.850000   \n",
       "max                       1.000000                        1.000000   \n",
       "\n",
       "       pruned_base_arrow_recall  pruned_base_arrow_recall_ce  \\\n",
       "count               1080.000000                  1080.000000   \n",
       "mean                   0.264068                     0.577924   \n",
       "std                    0.165490                     0.177350   \n",
       "min                    0.000000                     0.000000   \n",
       "25%                    0.144697                     0.454545   \n",
       "50%                    0.215190                     0.583333   \n",
       "75%                    0.346154                     0.700000   \n",
       "max                    0.888889                     1.000000   \n",
       "\n",
       "       pruned_base_arrow_true_negative  pruned_base_arrow_true_negative_ce  \\\n",
       "count                      1080.000000                         1080.000000   \n",
       "mean                        167.547222                          203.470370   \n",
       "std                          26.462059                            5.503647   \n",
       "min                         125.000000                          183.000000   \n",
       "25%                         141.000000                          200.000000   \n",
       "50%                         167.000000                          203.000000   \n",
       "75%                         191.000000                          207.000000   \n",
       "max                         218.000000                          218.000000   \n",
       "\n",
       "       pruned_base_arrow_true_positive  pruned_base_arrow_true_positive_ce  \\\n",
       "count                      1080.000000                         1080.000000   \n",
       "mean                         10.289815                           10.289815   \n",
       "std                           4.588163                            4.588163   \n",
       "min                           0.000000                            0.000000   \n",
       "25%                           7.000000                            7.000000   \n",
       "50%                          10.000000                           10.000000   \n",
       "75%                          13.000000                           13.000000   \n",
       "max                          28.000000                           28.000000   \n",
       "\n",
       "       pruned_base_pruned_base_shd  pruned_base_sid_lower  Base indep tests  \\\n",
       "count                  1080.000000                    0.0       1080.000000   \n",
       "mean                     40.371296                    NaN       4580.461111   \n",
       "std                      24.568304                    NaN       2791.169654   \n",
       "min                       0.000000                    NaN        251.000000   \n",
       "25%                      17.000000                    NaN       2598.500000   \n",
       "50%                      41.000000                    NaN       4142.500000   \n",
       "75%                      65.000000                    NaN       6045.000000   \n",
       "max                      84.000000                    NaN      16417.000000   \n",
       "\n",
       "       C-PC indep tests        alpha  cluster_connectivity  empty_graph_shd  \\\n",
       "count       1080.000000  1080.000000            900.000000      1080.000000   \n",
       "mean        2994.563889     0.182000              0.878259        50.596296   \n",
       "std         2140.460093     0.178677              0.172465        26.367545   \n",
       "min          103.000000     0.010000              0.166667         7.000000   \n",
       "25%         1469.000000     0.050000              0.800000        26.000000   \n",
       "50%         2571.000000     0.100000              1.000000        50.000000   \n",
       "75%         3933.750000     0.250000              1.000000        79.000000   \n",
       "max        16417.000000     0.500000              1.000000        89.000000   \n",
       "\n",
       "        n_clusters      n_edges  n_nodes  noise_scale  sample_size  \\\n",
       "count  1080.000000  1080.000000   1080.0       1080.0       1080.0   \n",
       "mean      3.500000    43.750000     15.0          1.0       1000.0   \n",
       "std       1.708616    24.346177      0.0          0.0          0.0   \n",
       "min       1.000000    15.000000     15.0          1.0       1000.0   \n",
       "25%       2.000000    26.250000     15.0          1.0       1000.0   \n",
       "50%       3.500000    40.000000     15.0          1.0       1000.0   \n",
       "75%       5.000000    57.500000     15.0          1.0       1000.0   \n",
       "max       6.000000    80.000000     15.0          1.0       1000.0   \n",
       "\n",
       "       true_sid_lower  true_sid_upper  \n",
       "count             0.0             0.0  \n",
       "mean              NaN             NaN  \n",
       "std               NaN             NaN  \n",
       "min               NaN             NaN  \n",
       "25%               NaN             NaN  \n",
       "50%               NaN             NaN  \n",
       "75%               NaN             NaN  \n",
       "max               NaN             NaN  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_all.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <th>base_adj_false_negative</th>\n",
       "      <th>base_adj_false_positive</th>\n",
       "      <th>base_adj_precision</th>\n",
       "      <th>base_adj_recall</th>\n",
       "      <th>base_adj_true_negative</th>\n",
       "      <th>base_adj_true_positive</th>\n",
       "      <th>base_arrow_f1_score</th>\n",
       "      <th>base_arrow_f1_score_ce</th>\n",
       "      <th>base_arrow_false_negative</th>\n",
       "      <th>base_arrow_false_negative_ce</th>\n",
       "      <th>base_arrow_false_positive</th>\n",
       "      <th>base_arrow_false_positive_ce</th>\n",
       "      <th>base_arrow_precision</th>\n",
       "      <th>base_arrow_precision_ce</th>\n",
       "      <th>base_arrow_recall</th>\n",
       "      <th>base_arrow_recall_ce</th>\n",
       "      <th>base_arrow_true_negative</th>\n",
       "      <th>base_arrow_true_negative_ce</th>\n",
       "      <th>base_arrow_true_positive</th>\n",
       "      <th>base_arrow_true_positive_ce</th>\n",
       "      <th>base_base_shd</th>\n",
       "      <th>base_sid_lower</th>\n",
       "      <th>cluster_adj_f1_score</th>\n",
       "      <th>cluster_adj_false_negative</th>\n",
       "      <th>cluster_adj_false_positive</th>\n",
       "      <th>cluster_adj_precision</th>\n",
       "      <th>cluster_adj_recall</th>\n",
       "      <th>cluster_adj_true_negative</th>\n",
       "      <th>cluster_adj_true_positive</th>\n",
       "      <th>cluster_arrow_f1_score</th>\n",
       "      <th>cluster_arrow_f1_score_ce</th>\n",
       "      <th>cluster_arrow_false_negative</th>\n",
       "      <th>cluster_arrow_false_negative_ce</th>\n",
       "      <th>cluster_arrow_false_positive</th>\n",
       "      <th>cluster_arrow_false_positive_ce</th>\n",
       "      <th>cluster_arrow_precision</th>\n",
       "      <th>cluster_arrow_precision_ce</th>\n",
       "      <th>cluster_arrow_recall</th>\n",
       "      <th>cluster_arrow_recall_ce</th>\n",
       "      <th>cluster_arrow_true_negative</th>\n",
       "      <th>cluster_arrow_true_negative_ce</th>\n",
       "      <th>cluster_arrow_true_positive</th>\n",
       "      <th>cluster_arrow_true_positive_ce</th>\n",
       "      <th>cluster_cluster_shd</th>\n",
       "      <th>cluster_sid_lower</th>\n",
       "      <th>pruned_base_adj_f1_score</th>\n",
       "      <th>pruned_base_adj_false_negative</th>\n",
       "      <th>pruned_base_adj_false_positive</th>\n",
       "      <th>pruned_base_adj_precision</th>\n",
       "      <th>pruned_base_adj_recall</th>\n",
       "      <th>pruned_base_adj_true_negative</th>\n",
       "      <th>pruned_base_adj_true_positive</th>\n",
       "      <th>pruned_base_arrow_f1_score</th>\n",
       "      <th>pruned_base_arrow_f1_score_ce</th>\n",
       "      <th>pruned_base_arrow_false_negative</th>\n",
       "      <th>pruned_base_arrow_false_negative_ce</th>\n",
       "      <th>pruned_base_arrow_false_positive</th>\n",
       "      <th>pruned_base_arrow_false_positive_ce</th>\n",
       "      <th>pruned_base_arrow_precision</th>\n",
       "      <th>pruned_base_arrow_precision_ce</th>\n",
       "      <th>pruned_base_arrow_recall</th>\n",
       "      <th>pruned_base_arrow_recall_ce</th>\n",
       "      <th>pruned_base_arrow_true_negative</th>\n",
       "      <th>pruned_base_arrow_true_negative_ce</th>\n",
       "      <th>pruned_base_arrow_true_positive</th>\n",
       "      <th>pruned_base_arrow_true_positive_ce</th>\n",
       "      <th>pruned_base_pruned_base_shd</th>\n",
       "      <th>pruned_base_sid_lower</th>\n",
       "      <th>Base indep tests</th>\n",
       "      <th>C-PC indep tests</th>\n",
       "      <th>alpha</th>\n",
       "      <th>cluster_connectivity</th>\n",
       "      <th>empty_graph_shd</th>\n",
       "      <th>n_clusters</th>\n",
       "      <th>n_edges</th>\n",
       "      <th>n_nodes</th>\n",
       "      <th>noise_scale</th>\n",
       "      <th>sample_size</th>\n",
       "      <th>true_sid_lower</th>\n",
       "      <th>true_sid_upper</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>899.000000</td>\n",
       "      <td>899.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>899.000000</td>\n",
       "      <td>899.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.0</td>\n",
       "      <td>900.0</td>\n",
       "      <td>900.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.559182</td>\n",
       "      <td>33.222222</td>\n",
       "      <td>2.924444</td>\n",
       "      <td>0.864585</td>\n",
       "      <td>0.464137</td>\n",
       "      <td>51.490000</td>\n",
       "      <td>17.363333</td>\n",
       "      <td>0.346783</td>\n",
       "      <td>0.635757</td>\n",
       "      <td>40.238889</td>\n",
       "      <td>7.016667</td>\n",
       "      <td>6.777778</td>\n",
       "      <td>4.125556</td>\n",
       "      <td>0.615787</td>\n",
       "      <td>0.713506</td>\n",
       "      <td>0.266378</td>\n",
       "      <td>0.581592</td>\n",
       "      <td>167.636667</td>\n",
       "      <td>203.511111</td>\n",
       "      <td>10.346667</td>\n",
       "      <td>10.346667</td>\n",
       "      <td>40.272222</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.634823</td>\n",
       "      <td>27.486667</td>\n",
       "      <td>4.885556</td>\n",
       "      <td>0.832252</td>\n",
       "      <td>0.562061</td>\n",
       "      <td>49.528889</td>\n",
       "      <td>23.098889</td>\n",
       "      <td>0.565326</td>\n",
       "      <td>0.886229</td>\n",
       "      <td>30.435556</td>\n",
       "      <td>2.948889</td>\n",
       "      <td>6.521111</td>\n",
       "      <td>1.921111</td>\n",
       "      <td>0.760063</td>\n",
       "      <td>0.909596</td>\n",
       "      <td>0.492294</td>\n",
       "      <td>0.866634</td>\n",
       "      <td>167.893333</td>\n",
       "      <td>199.980000</td>\n",
       "      <td>20.150000</td>\n",
       "      <td>20.150000</td>\n",
       "      <td>34.293333</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.559182</td>\n",
       "      <td>33.222222</td>\n",
       "      <td>2.924444</td>\n",
       "      <td>0.864585</td>\n",
       "      <td>0.464137</td>\n",
       "      <td>51.490000</td>\n",
       "      <td>17.363333</td>\n",
       "      <td>0.346783</td>\n",
       "      <td>0.635757</td>\n",
       "      <td>40.238889</td>\n",
       "      <td>7.016667</td>\n",
       "      <td>6.777778</td>\n",
       "      <td>4.125556</td>\n",
       "      <td>0.615787</td>\n",
       "      <td>0.713506</td>\n",
       "      <td>0.266378</td>\n",
       "      <td>0.581592</td>\n",
       "      <td>167.636667</td>\n",
       "      <td>203.511111</td>\n",
       "      <td>10.346667</td>\n",
       "      <td>10.346667</td>\n",
       "      <td>40.272222</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4572.567778</td>\n",
       "      <td>2669.491111</td>\n",
       "      <td>0.182000</td>\n",
       "      <td>0.878259</td>\n",
       "      <td>50.585556</td>\n",
       "      <td>4.000</td>\n",
       "      <td>43.750000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.198208</td>\n",
       "      <td>24.527018</td>\n",
       "      <td>2.907713</td>\n",
       "      <td>0.117717</td>\n",
       "      <td>0.254046</td>\n",
       "      <td>25.958144</td>\n",
       "      <td>3.944863</td>\n",
       "      <td>0.165063</td>\n",
       "      <td>0.170022</td>\n",
       "      <td>24.626095</td>\n",
       "      <td>2.897728</td>\n",
       "      <td>4.265853</td>\n",
       "      <td>2.824652</td>\n",
       "      <td>0.198433</td>\n",
       "      <td>0.183026</td>\n",
       "      <td>0.167875</td>\n",
       "      <td>0.174870</td>\n",
       "      <td>26.545870</td>\n",
       "      <td>5.472846</td>\n",
       "      <td>4.553213</td>\n",
       "      <td>4.553213</td>\n",
       "      <td>24.622225</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.149663</td>\n",
       "      <td>21.110094</td>\n",
       "      <td>4.086824</td>\n",
       "      <td>0.129235</td>\n",
       "      <td>0.223503</td>\n",
       "      <td>25.860932</td>\n",
       "      <td>7.218826</td>\n",
       "      <td>0.166598</td>\n",
       "      <td>0.107627</td>\n",
       "      <td>21.831320</td>\n",
       "      <td>2.518817</td>\n",
       "      <td>4.456014</td>\n",
       "      <td>2.010926</td>\n",
       "      <td>0.151157</td>\n",
       "      <td>0.099307</td>\n",
       "      <td>0.220628</td>\n",
       "      <td>0.120622</td>\n",
       "      <td>26.500063</td>\n",
       "      <td>7.725234</td>\n",
       "      <td>7.435802</td>\n",
       "      <td>7.435802</td>\n",
       "      <td>21.214634</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.198208</td>\n",
       "      <td>24.527018</td>\n",
       "      <td>2.907713</td>\n",
       "      <td>0.117717</td>\n",
       "      <td>0.254046</td>\n",
       "      <td>25.958144</td>\n",
       "      <td>3.944863</td>\n",
       "      <td>0.165063</td>\n",
       "      <td>0.170022</td>\n",
       "      <td>24.626095</td>\n",
       "      <td>2.897728</td>\n",
       "      <td>4.265853</td>\n",
       "      <td>2.824652</td>\n",
       "      <td>0.198433</td>\n",
       "      <td>0.183026</td>\n",
       "      <td>0.167875</td>\n",
       "      <td>0.174870</td>\n",
       "      <td>26.545870</td>\n",
       "      <td>5.472846</td>\n",
       "      <td>4.553213</td>\n",
       "      <td>4.553213</td>\n",
       "      <td>24.622225</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2796.341374</td>\n",
       "      <td>1825.808292</td>\n",
       "      <td>0.178694</td>\n",
       "      <td>0.172465</td>\n",
       "      <td>26.385404</td>\n",
       "      <td>1.415</td>\n",
       "      <td>24.348434</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.231579</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.187500</td>\n",
       "      <td>0.134146</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.045977</td>\n",
       "      <td>0.125000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>125.000000</td>\n",
       "      <td>183.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.292683</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.176471</td>\n",
       "      <td>0.194444</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.087912</td>\n",
       "      <td>0.296296</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.176471</td>\n",
       "      <td>0.307692</td>\n",
       "      <td>0.055556</td>\n",
       "      <td>0.285714</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>178.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.231579</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.187500</td>\n",
       "      <td>0.134146</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.045977</td>\n",
       "      <td>0.125000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>125.000000</td>\n",
       "      <td>183.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>251.000000</td>\n",
       "      <td>103.000000</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.166667</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>2.000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.393213</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.809524</td>\n",
       "      <td>0.256332</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>0.228869</td>\n",
       "      <td>0.516129</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.473684</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>0.145607</td>\n",
       "      <td>0.466667</td>\n",
       "      <td>141.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.520548</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.379638</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.448276</td>\n",
       "      <td>0.836364</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>0.320000</td>\n",
       "      <td>0.807692</td>\n",
       "      <td>141.000000</td>\n",
       "      <td>195.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.393213</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.809524</td>\n",
       "      <td>0.256332</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>0.228869</td>\n",
       "      <td>0.516129</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.473684</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>0.145607</td>\n",
       "      <td>0.466667</td>\n",
       "      <td>141.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2568.250000</td>\n",
       "      <td>1406.750000</td>\n",
       "      <td>0.050000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>3.000</td>\n",
       "      <td>26.250000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.507252</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.360000</td>\n",
       "      <td>53.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>0.311927</td>\n",
       "      <td>0.645161</td>\n",
       "      <td>40.500000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.615385</td>\n",
       "      <td>0.727273</td>\n",
       "      <td>0.215190</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>167.000000</td>\n",
       "      <td>203.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.611341</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.850926</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>51.000000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>0.548703</td>\n",
       "      <td>0.916667</td>\n",
       "      <td>28.500000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.937500</td>\n",
       "      <td>0.433849</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>168.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>33.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.507252</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.360000</td>\n",
       "      <td>53.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>0.311927</td>\n",
       "      <td>0.645161</td>\n",
       "      <td>40.500000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.615385</td>\n",
       "      <td>0.727273</td>\n",
       "      <td>0.215190</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>167.000000</td>\n",
       "      <td>203.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4117.000000</td>\n",
       "      <td>2370.500000</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>4.000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.720328</td>\n",
       "      <td>58.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.947368</td>\n",
       "      <td>0.653846</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>0.439024</td>\n",
       "      <td>0.761905</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.850000</td>\n",
       "      <td>0.346154</td>\n",
       "      <td>0.705882</td>\n",
       "      <td>191.000000</td>\n",
       "      <td>207.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>65.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.736842</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>0.742349</td>\n",
       "      <td>73.000000</td>\n",
       "      <td>27.000000</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.962963</td>\n",
       "      <td>51.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.869565</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.642857</td>\n",
       "      <td>0.954545</td>\n",
       "      <td>192.000000</td>\n",
       "      <td>205.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>54.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.720328</td>\n",
       "      <td>58.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.947368</td>\n",
       "      <td>0.653846</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>0.439024</td>\n",
       "      <td>0.761905</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.850000</td>\n",
       "      <td>0.346154</td>\n",
       "      <td>0.705882</td>\n",
       "      <td>191.000000</td>\n",
       "      <td>207.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>65.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6066.750000</td>\n",
       "      <td>3459.000000</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>79.000000</td>\n",
       "      <td>5.000</td>\n",
       "      <td>57.500000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>67.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>72.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>219.000000</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>74.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>13911.000000</td>\n",
       "      <td>11843.000000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>89.000000</td>\n",
       "      <td>6.000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       base_adj_f1_score  base_adj_false_negative  base_adj_false_positive  \\\n",
       "count         900.000000               900.000000               900.000000   \n",
       "mean            0.559182                33.222222                 2.924444   \n",
       "std             0.198208                24.527018                 2.907713   \n",
       "min             0.231579                 0.000000                 0.000000   \n",
       "25%             0.393213                 9.000000                 1.000000   \n",
       "50%             0.507252                32.000000                 2.000000   \n",
       "75%             0.720328                58.000000                 4.000000   \n",
       "max             1.000000                75.000000                26.000000   \n",
       "\n",
       "       base_adj_precision  base_adj_recall  base_adj_true_negative  \\\n",
       "count          900.000000       900.000000              900.000000   \n",
       "mean             0.864585         0.464137               51.490000   \n",
       "std              0.117717         0.254046               25.958144   \n",
       "min              0.187500         0.134146               12.000000   \n",
       "25%              0.809524         0.256332               25.000000   \n",
       "50%              0.888889         0.360000               53.000000   \n",
       "75%              0.947368         0.653846               77.000000   \n",
       "max              1.000000         1.000000               98.000000   \n",
       "\n",
       "       base_adj_true_positive  base_arrow_f1_score  base_arrow_f1_score_ce  \\\n",
       "count              900.000000           899.000000              899.000000   \n",
       "mean                17.363333             0.346783                0.635757   \n",
       "std                  3.944863             0.165063                0.170022   \n",
       "min                  6.000000             0.045977                0.125000   \n",
       "25%                 15.000000             0.228869                0.516129   \n",
       "50%                 17.000000             0.311927                0.645161   \n",
       "75%                 20.000000             0.439024                0.761905   \n",
       "max                 29.000000             0.923077                1.000000   \n",
       "\n",
       "       base_arrow_false_negative  base_arrow_false_negative_ce  \\\n",
       "count                 900.000000                    900.000000   \n",
       "mean                   40.238889                      7.016667   \n",
       "std                    24.626095                      2.897728   \n",
       "min                     1.000000                      0.000000   \n",
       "25%                    17.000000                      5.000000   \n",
       "50%                    40.500000                      7.000000   \n",
       "75%                    64.000000                      9.000000   \n",
       "max                    84.000000                     18.000000   \n",
       "\n",
       "       base_arrow_false_positive  base_arrow_false_positive_ce  \\\n",
       "count                 900.000000                    900.000000   \n",
       "mean                    6.777778                      4.125556   \n",
       "std                     4.265853                      2.824652   \n",
       "min                     0.000000                      0.000000   \n",
       "25%                     4.000000                      2.000000   \n",
       "50%                     6.000000                      4.000000   \n",
       "75%                     9.000000                      6.000000   \n",
       "max                    28.000000                     17.000000   \n",
       "\n",
       "       base_arrow_precision  base_arrow_precision_ce  base_arrow_recall  \\\n",
       "count            900.000000               900.000000         900.000000   \n",
       "mean               0.615787                 0.713506           0.266378   \n",
       "std                0.198433                 0.183026           0.167875   \n",
       "min                0.000000                 0.000000           0.000000   \n",
       "25%                0.473684                 0.583333           0.145607   \n",
       "50%                0.615385                 0.727273           0.215190   \n",
       "75%                0.750000                 0.850000           0.346154   \n",
       "max                1.000000                 1.000000           0.888889   \n",
       "\n",
       "       base_arrow_recall_ce  base_arrow_true_negative  \\\n",
       "count            900.000000                900.000000   \n",
       "mean               0.581592                167.636667   \n",
       "std                0.174870                 26.545870   \n",
       "min                0.000000                125.000000   \n",
       "25%                0.466667                141.000000   \n",
       "50%                0.583333                167.000000   \n",
       "75%                0.705882                191.000000   \n",
       "max                1.000000                218.000000   \n",
       "\n",
       "       base_arrow_true_negative_ce  base_arrow_true_positive  \\\n",
       "count                   900.000000                900.000000   \n",
       "mean                    203.511111                 10.346667   \n",
       "std                       5.472846                  4.553213   \n",
       "min                     183.000000                  0.000000   \n",
       "25%                     200.000000                  7.000000   \n",
       "50%                     203.000000                 10.000000   \n",
       "75%                     207.000000                 13.000000   \n",
       "max                     218.000000                 28.000000   \n",
       "\n",
       "       base_arrow_true_positive_ce  base_base_shd  base_sid_lower  \\\n",
       "count                   900.000000     900.000000             0.0   \n",
       "mean                     10.346667      40.272222             NaN   \n",
       "std                       4.553213      24.622225             NaN   \n",
       "min                       0.000000       0.000000             NaN   \n",
       "25%                       7.000000      17.000000             NaN   \n",
       "50%                      10.000000      41.000000             NaN   \n",
       "75%                      13.000000      65.000000             NaN   \n",
       "max                      28.000000      84.000000             NaN   \n",
       "\n",
       "       cluster_adj_f1_score  cluster_adj_false_negative  \\\n",
       "count            900.000000                  900.000000   \n",
       "mean               0.634823                   27.486667   \n",
       "std                0.149663                   21.110094   \n",
       "min                0.292683                    0.000000   \n",
       "25%                0.520548                    7.000000   \n",
       "50%                0.611341                   25.000000   \n",
       "75%                0.736842                   48.000000   \n",
       "max                1.000000                   67.000000   \n",
       "\n",
       "       cluster_adj_false_positive  cluster_adj_precision  cluster_adj_recall  \\\n",
       "count                  900.000000             900.000000          900.000000   \n",
       "mean                     4.885556               0.832252            0.562061   \n",
       "std                      4.086824               0.129235            0.223503   \n",
       "min                      0.000000               0.176471            0.194444   \n",
       "25%                      2.000000               0.777778            0.379638   \n",
       "50%                      4.000000               0.850926            0.500000   \n",
       "75%                      7.000000               0.923077            0.742349   \n",
       "max                     28.000000               1.000000            1.000000   \n",
       "\n",
       "       cluster_adj_true_negative  cluster_adj_true_positive  \\\n",
       "count                 900.000000                 900.000000   \n",
       "mean                   49.528889                  23.098889   \n",
       "std                    25.860932                   7.218826   \n",
       "min                    11.000000                   6.000000   \n",
       "25%                    23.000000                  19.000000   \n",
       "50%                    51.000000                  23.000000   \n",
       "75%                    73.000000                  27.000000   \n",
       "max                    98.000000                  45.000000   \n",
       "\n",
       "       cluster_arrow_f1_score  cluster_arrow_f1_score_ce  \\\n",
       "count              900.000000                 900.000000   \n",
       "mean                 0.565326                   0.886229   \n",
       "std                  0.166598                   0.107627   \n",
       "min                  0.087912                   0.296296   \n",
       "25%                  0.448276                   0.836364   \n",
       "50%                  0.548703                   0.916667   \n",
       "75%                  0.666667                   0.962963   \n",
       "max                  1.000000                   1.000000   \n",
       "\n",
       "       cluster_arrow_false_negative  cluster_arrow_false_negative_ce  \\\n",
       "count                    900.000000                       900.000000   \n",
       "mean                      30.435556                         2.948889   \n",
       "std                       21.831320                         2.518817   \n",
       "min                        0.000000                         0.000000   \n",
       "25%                        9.000000                         1.000000   \n",
       "50%                       28.500000                         2.000000   \n",
       "75%                       51.000000                         4.000000   \n",
       "max                       72.000000                        15.000000   \n",
       "\n",
       "       cluster_arrow_false_positive  cluster_arrow_false_positive_ce  \\\n",
       "count                    900.000000                       900.000000   \n",
       "mean                       6.521111                         1.921111   \n",
       "std                        4.456014                         2.010926   \n",
       "min                        0.000000                         0.000000   \n",
       "25%                        3.000000                         0.000000   \n",
       "50%                        6.000000                         1.000000   \n",
       "75%                        9.000000                         3.000000   \n",
       "max                       28.000000                        12.000000   \n",
       "\n",
       "       cluster_arrow_precision  cluster_arrow_precision_ce  \\\n",
       "count               900.000000                  900.000000   \n",
       "mean                  0.760063                    0.909596   \n",
       "std                   0.151157                    0.099307   \n",
       "min                   0.176471                    0.307692   \n",
       "25%                   0.666667                    0.866667   \n",
       "50%                   0.777778                    0.937500   \n",
       "75%                   0.869565                    1.000000   \n",
       "max                   1.000000                    1.000000   \n",
       "\n",
       "       cluster_arrow_recall  cluster_arrow_recall_ce  \\\n",
       "count            900.000000               900.000000   \n",
       "mean               0.492294                 0.866634   \n",
       "std                0.220628                 0.120622   \n",
       "min                0.055556                 0.285714   \n",
       "25%                0.320000                 0.807692   \n",
       "50%                0.433849                 0.900000   \n",
       "75%                0.642857                 0.954545   \n",
       "max                1.000000                 1.000000   \n",
       "\n",
       "       cluster_arrow_true_negative  cluster_arrow_true_negative_ce  \\\n",
       "count                   900.000000                      900.000000   \n",
       "mean                    167.893333                      199.980000   \n",
       "std                      26.500063                        7.725234   \n",
       "min                     128.000000                      178.000000   \n",
       "25%                     141.000000                      195.000000   \n",
       "50%                     168.000000                      200.000000   \n",
       "75%                     192.000000                      205.000000   \n",
       "max                     218.000000                      219.000000   \n",
       "\n",
       "       cluster_arrow_true_positive  cluster_arrow_true_positive_ce  \\\n",
       "count                   900.000000                      900.000000   \n",
       "mean                     20.150000                       20.150000   \n",
       "std                       7.435802                        7.435802   \n",
       "min                       4.000000                        4.000000   \n",
       "25%                      14.000000                       14.000000   \n",
       "50%                      19.000000                       19.000000   \n",
       "75%                      25.000000                       25.000000   \n",
       "max                      44.000000                       44.000000   \n",
       "\n",
       "       cluster_cluster_shd  cluster_sid_lower  pruned_base_adj_f1_score  \\\n",
       "count           900.000000                0.0                900.000000   \n",
       "mean             34.293333                NaN                  0.559182   \n",
       "std              21.214634                NaN                  0.198208   \n",
       "min               0.000000                NaN                  0.231579   \n",
       "25%              15.000000                NaN                  0.393213   \n",
       "50%              33.000000                NaN                  0.507252   \n",
       "75%              54.000000                NaN                  0.720328   \n",
       "max              74.000000                NaN                  1.000000   \n",
       "\n",
       "       pruned_base_adj_false_negative  pruned_base_adj_false_positive  \\\n",
       "count                      900.000000                      900.000000   \n",
       "mean                        33.222222                        2.924444   \n",
       "std                         24.527018                        2.907713   \n",
       "min                          0.000000                        0.000000   \n",
       "25%                          9.000000                        1.000000   \n",
       "50%                         32.000000                        2.000000   \n",
       "75%                         58.000000                        4.000000   \n",
       "max                         75.000000                       26.000000   \n",
       "\n",
       "       pruned_base_adj_precision  pruned_base_adj_recall  \\\n",
       "count                 900.000000              900.000000   \n",
       "mean                    0.864585                0.464137   \n",
       "std                     0.117717                0.254046   \n",
       "min                     0.187500                0.134146   \n",
       "25%                     0.809524                0.256332   \n",
       "50%                     0.888889                0.360000   \n",
       "75%                     0.947368                0.653846   \n",
       "max                     1.000000                1.000000   \n",
       "\n",
       "       pruned_base_adj_true_negative  pruned_base_adj_true_positive  \\\n",
       "count                     900.000000                     900.000000   \n",
       "mean                       51.490000                      17.363333   \n",
       "std                        25.958144                       3.944863   \n",
       "min                        12.000000                       6.000000   \n",
       "25%                        25.000000                      15.000000   \n",
       "50%                        53.000000                      17.000000   \n",
       "75%                        77.000000                      20.000000   \n",
       "max                        98.000000                      29.000000   \n",
       "\n",
       "       pruned_base_arrow_f1_score  pruned_base_arrow_f1_score_ce  \\\n",
       "count                  899.000000                     899.000000   \n",
       "mean                     0.346783                       0.635757   \n",
       "std                      0.165063                       0.170022   \n",
       "min                      0.045977                       0.125000   \n",
       "25%                      0.228869                       0.516129   \n",
       "50%                      0.311927                       0.645161   \n",
       "75%                      0.439024                       0.761905   \n",
       "max                      0.923077                       1.000000   \n",
       "\n",
       "       pruned_base_arrow_false_negative  pruned_base_arrow_false_negative_ce  \\\n",
       "count                        900.000000                           900.000000   \n",
       "mean                          40.238889                             7.016667   \n",
       "std                           24.626095                             2.897728   \n",
       "min                            1.000000                             0.000000   \n",
       "25%                           17.000000                             5.000000   \n",
       "50%                           40.500000                             7.000000   \n",
       "75%                           64.000000                             9.000000   \n",
       "max                           84.000000                            18.000000   \n",
       "\n",
       "       pruned_base_arrow_false_positive  pruned_base_arrow_false_positive_ce  \\\n",
       "count                        900.000000                           900.000000   \n",
       "mean                           6.777778                             4.125556   \n",
       "std                            4.265853                             2.824652   \n",
       "min                            0.000000                             0.000000   \n",
       "25%                            4.000000                             2.000000   \n",
       "50%                            6.000000                             4.000000   \n",
       "75%                            9.000000                             6.000000   \n",
       "max                           28.000000                            17.000000   \n",
       "\n",
       "       pruned_base_arrow_precision  pruned_base_arrow_precision_ce  \\\n",
       "count                   900.000000                      900.000000   \n",
       "mean                      0.615787                        0.713506   \n",
       "std                       0.198433                        0.183026   \n",
       "min                       0.000000                        0.000000   \n",
       "25%                       0.473684                        0.583333   \n",
       "50%                       0.615385                        0.727273   \n",
       "75%                       0.750000                        0.850000   \n",
       "max                       1.000000                        1.000000   \n",
       "\n",
       "       pruned_base_arrow_recall  pruned_base_arrow_recall_ce  \\\n",
       "count                900.000000                   900.000000   \n",
       "mean                   0.266378                     0.581592   \n",
       "std                    0.167875                     0.174870   \n",
       "min                    0.000000                     0.000000   \n",
       "25%                    0.145607                     0.466667   \n",
       "50%                    0.215190                     0.583333   \n",
       "75%                    0.346154                     0.705882   \n",
       "max                    0.888889                     1.000000   \n",
       "\n",
       "       pruned_base_arrow_true_negative  pruned_base_arrow_true_negative_ce  \\\n",
       "count                       900.000000                          900.000000   \n",
       "mean                        167.636667                          203.511111   \n",
       "std                          26.545870                            5.472846   \n",
       "min                         125.000000                          183.000000   \n",
       "25%                         141.000000                          200.000000   \n",
       "50%                         167.000000                          203.000000   \n",
       "75%                         191.000000                          207.000000   \n",
       "max                         218.000000                          218.000000   \n",
       "\n",
       "       pruned_base_arrow_true_positive  pruned_base_arrow_true_positive_ce  \\\n",
       "count                       900.000000                          900.000000   \n",
       "mean                         10.346667                           10.346667   \n",
       "std                           4.553213                            4.553213   \n",
       "min                           0.000000                            0.000000   \n",
       "25%                           7.000000                            7.000000   \n",
       "50%                          10.000000                           10.000000   \n",
       "75%                          13.000000                           13.000000   \n",
       "max                          28.000000                           28.000000   \n",
       "\n",
       "       pruned_base_pruned_base_shd  pruned_base_sid_lower  Base indep tests  \\\n",
       "count                   900.000000                    0.0        900.000000   \n",
       "mean                     40.272222                    NaN       4572.567778   \n",
       "std                      24.622225                    NaN       2796.341374   \n",
       "min                       0.000000                    NaN        251.000000   \n",
       "25%                      17.000000                    NaN       2568.250000   \n",
       "50%                      41.000000                    NaN       4117.000000   \n",
       "75%                      65.000000                    NaN       6066.750000   \n",
       "max                      84.000000                    NaN      13911.000000   \n",
       "\n",
       "       C-PC indep tests       alpha  cluster_connectivity  empty_graph_shd  \\\n",
       "count        900.000000  900.000000            900.000000       900.000000   \n",
       "mean        2669.491111    0.182000              0.878259        50.585556   \n",
       "std         1825.808292    0.178694              0.172465        26.385404   \n",
       "min          103.000000    0.010000              0.166667         7.000000   \n",
       "25%         1406.750000    0.050000              0.800000        26.000000   \n",
       "50%         2370.500000    0.100000              1.000000        50.000000   \n",
       "75%         3459.000000    0.250000              1.000000        79.000000   \n",
       "max        11843.000000    0.500000              1.000000        89.000000   \n",
       "\n",
       "       n_clusters     n_edges  n_nodes  noise_scale  sample_size  \\\n",
       "count     900.000  900.000000    900.0        900.0        900.0   \n",
       "mean        4.000   43.750000     15.0          1.0       1000.0   \n",
       "std         1.415   24.348434      0.0          0.0          0.0   \n",
       "min         2.000   15.000000     15.0          1.0       1000.0   \n",
       "25%         3.000   26.250000     15.0          1.0       1000.0   \n",
       "50%         4.000   40.000000     15.0          1.0       1000.0   \n",
       "75%         5.000   57.500000     15.0          1.0       1000.0   \n",
       "max         6.000   80.000000     15.0          1.0       1000.0   \n",
       "\n",
       "       true_sid_lower  true_sid_upper  \n",
       "count             0.0             0.0  \n",
       "mean              NaN             NaN  \n",
       "std               NaN             NaN  \n",
       "min               NaN             NaN  \n",
       "25%               NaN             NaN  \n",
       "50%               NaN             NaN  \n",
       "75%               NaN             NaN  \n",
       "max               NaN             NaN  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The precision of C-PC is slightly lower, but the recall is higher. In total C-PC has a higher F1 score."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <th>base_adj_false_negative</th>\n",
       "      <th>base_adj_false_positive</th>\n",
       "      <th>base_adj_precision</th>\n",
       "      <th>base_adj_recall</th>\n",
       "      <th>base_adj_true_negative</th>\n",
       "      <th>base_adj_true_positive</th>\n",
       "      <th>base_arrow_f1_score</th>\n",
       "      <th>base_arrow_f1_score_ce</th>\n",
       "      <th>base_arrow_false_negative</th>\n",
       "      <th>base_arrow_false_negative_ce</th>\n",
       "      <th>base_arrow_false_positive</th>\n",
       "      <th>base_arrow_false_positive_ce</th>\n",
       "      <th>base_arrow_precision</th>\n",
       "      <th>base_arrow_precision_ce</th>\n",
       "      <th>base_arrow_recall</th>\n",
       "      <th>base_arrow_recall_ce</th>\n",
       "      <th>base_arrow_true_negative</th>\n",
       "      <th>base_arrow_true_negative_ce</th>\n",
       "      <th>base_arrow_true_positive</th>\n",
       "      <th>base_arrow_true_positive_ce</th>\n",
       "      <th>base_base_shd</th>\n",
       "      <th>base_sid_lower</th>\n",
       "      <th>cluster_adj_f1_score</th>\n",
       "      <th>cluster_adj_false_negative</th>\n",
       "      <th>cluster_adj_false_positive</th>\n",
       "      <th>cluster_adj_precision</th>\n",
       "      <th>cluster_adj_recall</th>\n",
       "      <th>cluster_adj_true_negative</th>\n",
       "      <th>cluster_adj_true_positive</th>\n",
       "      <th>cluster_arrow_f1_score</th>\n",
       "      <th>cluster_arrow_f1_score_ce</th>\n",
       "      <th>cluster_arrow_false_negative</th>\n",
       "      <th>cluster_arrow_false_negative_ce</th>\n",
       "      <th>cluster_arrow_false_positive</th>\n",
       "      <th>cluster_arrow_false_positive_ce</th>\n",
       "      <th>cluster_arrow_precision</th>\n",
       "      <th>cluster_arrow_precision_ce</th>\n",
       "      <th>cluster_arrow_recall</th>\n",
       "      <th>cluster_arrow_recall_ce</th>\n",
       "      <th>cluster_arrow_true_negative</th>\n",
       "      <th>cluster_arrow_true_negative_ce</th>\n",
       "      <th>cluster_arrow_true_positive</th>\n",
       "      <th>cluster_arrow_true_positive_ce</th>\n",
       "      <th>cluster_cluster_shd</th>\n",
       "      <th>cluster_sid_lower</th>\n",
       "      <th>pruned_base_adj_f1_score</th>\n",
       "      <th>pruned_base_adj_false_negative</th>\n",
       "      <th>pruned_base_adj_false_positive</th>\n",
       "      <th>pruned_base_adj_precision</th>\n",
       "      <th>pruned_base_adj_recall</th>\n",
       "      <th>pruned_base_adj_true_negative</th>\n",
       "      <th>pruned_base_adj_true_positive</th>\n",
       "      <th>pruned_base_arrow_f1_score</th>\n",
       "      <th>pruned_base_arrow_f1_score_ce</th>\n",
       "      <th>pruned_base_arrow_false_negative</th>\n",
       "      <th>pruned_base_arrow_false_negative_ce</th>\n",
       "      <th>pruned_base_arrow_false_positive</th>\n",
       "      <th>pruned_base_arrow_false_positive_ce</th>\n",
       "      <th>pruned_base_arrow_precision</th>\n",
       "      <th>pruned_base_arrow_precision_ce</th>\n",
       "      <th>pruned_base_arrow_recall</th>\n",
       "      <th>pruned_base_arrow_recall_ce</th>\n",
       "      <th>pruned_base_arrow_true_negative</th>\n",
       "      <th>pruned_base_arrow_true_negative_ce</th>\n",
       "      <th>pruned_base_arrow_true_positive</th>\n",
       "      <th>pruned_base_arrow_true_positive_ce</th>\n",
       "      <th>pruned_base_pruned_base_shd</th>\n",
       "      <th>pruned_base_sid_lower</th>\n",
       "      <th>Base indep tests</th>\n",
       "      <th>C-PC indep tests</th>\n",
       "      <th>alpha</th>\n",
       "      <th>cluster_connectivity</th>\n",
       "      <th>empty_graph_shd</th>\n",
       "      <th>n_clusters</th>\n",
       "      <th>n_edges</th>\n",
       "      <th>n_nodes</th>\n",
       "      <th>noise_scale</th>\n",
       "      <th>sample_size</th>\n",
       "      <th>true_sid_lower</th>\n",
       "      <th>true_sid_upper</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.0</td>\n",
       "      <td>417.0</td>\n",
       "      <td>417.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.613622</td>\n",
       "      <td>26.865707</td>\n",
       "      <td>2.923261</td>\n",
       "      <td>0.863344</td>\n",
       "      <td>0.540058</td>\n",
       "      <td>58.378897</td>\n",
       "      <td>16.832134</td>\n",
       "      <td>0.383271</td>\n",
       "      <td>0.640144</td>\n",
       "      <td>33.616307</td>\n",
       "      <td>6.750600</td>\n",
       "      <td>6.513189</td>\n",
       "      <td>3.858513</td>\n",
       "      <td>0.626084</td>\n",
       "      <td>0.725695</td>\n",
       "      <td>0.308170</td>\n",
       "      <td>0.584133</td>\n",
       "      <td>174.788969</td>\n",
       "      <td>204.309353</td>\n",
       "      <td>10.081535</td>\n",
       "      <td>10.081535</td>\n",
       "      <td>33.647482</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.683436</td>\n",
       "      <td>21.177458</td>\n",
       "      <td>5.158273</td>\n",
       "      <td>0.820270</td>\n",
       "      <td>0.644706</td>\n",
       "      <td>56.143885</td>\n",
       "      <td>22.520384</td>\n",
       "      <td>0.638167</td>\n",
       "      <td>0.927616</td>\n",
       "      <td>23.079137</td>\n",
       "      <td>1.901679</td>\n",
       "      <td>6.122302</td>\n",
       "      <td>1.304556</td>\n",
       "      <td>0.778635</td>\n",
       "      <td>0.940886</td>\n",
       "      <td>0.593473</td>\n",
       "      <td>0.915803</td>\n",
       "      <td>175.179856</td>\n",
       "      <td>201.175060</td>\n",
       "      <td>20.618705</td>\n",
       "      <td>20.618705</td>\n",
       "      <td>27.640288</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.613622</td>\n",
       "      <td>26.865707</td>\n",
       "      <td>2.923261</td>\n",
       "      <td>0.863344</td>\n",
       "      <td>0.540058</td>\n",
       "      <td>58.378897</td>\n",
       "      <td>16.832134</td>\n",
       "      <td>0.383271</td>\n",
       "      <td>0.640144</td>\n",
       "      <td>33.616307</td>\n",
       "      <td>6.750600</td>\n",
       "      <td>6.513189</td>\n",
       "      <td>3.858513</td>\n",
       "      <td>0.626084</td>\n",
       "      <td>0.725695</td>\n",
       "      <td>0.308170</td>\n",
       "      <td>0.584133</td>\n",
       "      <td>174.788969</td>\n",
       "      <td>204.309353</td>\n",
       "      <td>10.081535</td>\n",
       "      <td>10.081535</td>\n",
       "      <td>33.647482</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4035.211031</td>\n",
       "      <td>2029.055156</td>\n",
       "      <td>0.189808</td>\n",
       "      <td>0.737250</td>\n",
       "      <td>43.697842</td>\n",
       "      <td>5.026379</td>\n",
       "      <td>38.297362</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.205361</td>\n",
       "      <td>24.679118</td>\n",
       "      <td>3.272681</td>\n",
       "      <td>0.132053</td>\n",
       "      <td>0.272973</td>\n",
       "      <td>26.523762</td>\n",
       "      <td>4.181648</td>\n",
       "      <td>0.174715</td>\n",
       "      <td>0.173975</td>\n",
       "      <td>24.792243</td>\n",
       "      <td>2.949928</td>\n",
       "      <td>4.536756</td>\n",
       "      <td>2.837190</td>\n",
       "      <td>0.208914</td>\n",
       "      <td>0.185748</td>\n",
       "      <td>0.183161</td>\n",
       "      <td>0.179640</td>\n",
       "      <td>27.214578</td>\n",
       "      <td>5.810942</td>\n",
       "      <td>4.634535</td>\n",
       "      <td>4.634535</td>\n",
       "      <td>24.720514</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.148385</td>\n",
       "      <td>20.352829</td>\n",
       "      <td>4.550347</td>\n",
       "      <td>0.147972</td>\n",
       "      <td>0.228748</td>\n",
       "      <td>26.530070</td>\n",
       "      <td>8.209996</td>\n",
       "      <td>0.158778</td>\n",
       "      <td>0.072930</td>\n",
       "      <td>20.916523</td>\n",
       "      <td>1.833823</td>\n",
       "      <td>4.641152</td>\n",
       "      <td>1.513021</td>\n",
       "      <td>0.154644</td>\n",
       "      <td>0.069868</td>\n",
       "      <td>0.226315</td>\n",
       "      <td>0.080915</td>\n",
       "      <td>27.063650</td>\n",
       "      <td>8.692147</td>\n",
       "      <td>7.895102</td>\n",
       "      <td>7.895102</td>\n",
       "      <td>20.361401</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.205361</td>\n",
       "      <td>24.679118</td>\n",
       "      <td>3.272681</td>\n",
       "      <td>0.132053</td>\n",
       "      <td>0.272973</td>\n",
       "      <td>26.523762</td>\n",
       "      <td>4.181648</td>\n",
       "      <td>0.174715</td>\n",
       "      <td>0.173975</td>\n",
       "      <td>24.792243</td>\n",
       "      <td>2.949928</td>\n",
       "      <td>4.536756</td>\n",
       "      <td>2.837190</td>\n",
       "      <td>0.208914</td>\n",
       "      <td>0.185748</td>\n",
       "      <td>0.183161</td>\n",
       "      <td>0.179640</td>\n",
       "      <td>27.214578</td>\n",
       "      <td>5.810942</td>\n",
       "      <td>4.634535</td>\n",
       "      <td>4.634535</td>\n",
       "      <td>24.720514</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2800.471409</td>\n",
       "      <td>1520.365094</td>\n",
       "      <td>0.184236</td>\n",
       "      <td>0.164742</td>\n",
       "      <td>27.083823</td>\n",
       "      <td>0.954128</td>\n",
       "      <td>23.279300</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.255814</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.187500</td>\n",
       "      <td>0.159420</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.071429</td>\n",
       "      <td>0.142857</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.047619</td>\n",
       "      <td>0.166667</td>\n",
       "      <td>0.056818</td>\n",
       "      <td>0.090909</td>\n",
       "      <td>126.000000</td>\n",
       "      <td>190.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.292683</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.176471</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.285714</td>\n",
       "      <td>0.588235</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.176471</td>\n",
       "      <td>0.588235</td>\n",
       "      <td>0.181818</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>178.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.255814</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.187500</td>\n",
       "      <td>0.159420</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.071429</td>\n",
       "      <td>0.142857</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.047619</td>\n",
       "      <td>0.166667</td>\n",
       "      <td>0.056818</td>\n",
       "      <td>0.090909</td>\n",
       "      <td>126.000000</td>\n",
       "      <td>190.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>251.000000</td>\n",
       "      <td>103.000000</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.166667</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.422018</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.807692</td>\n",
       "      <td>0.280000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>0.255319</td>\n",
       "      <td>0.514286</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.466667</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.169014</td>\n",
       "      <td>0.461538</td>\n",
       "      <td>145.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.566667</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.761905</td>\n",
       "      <td>0.433735</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>0.521739</td>\n",
       "      <td>0.894737</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.689655</td>\n",
       "      <td>0.914286</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>144.000000</td>\n",
       "      <td>195.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.422018</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.807692</td>\n",
       "      <td>0.280000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>0.255319</td>\n",
       "      <td>0.514286</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.466667</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.169014</td>\n",
       "      <td>0.461538</td>\n",
       "      <td>145.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2028.000000</td>\n",
       "      <td>877.000000</td>\n",
       "      <td>0.050000</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.603774</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.892857</td>\n",
       "      <td>0.476190</td>\n",
       "      <td>66.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>0.347826</td>\n",
       "      <td>0.648649</td>\n",
       "      <td>27.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.619048</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.253012</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>204.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>27.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.844444</td>\n",
       "      <td>0.615385</td>\n",
       "      <td>63.000000</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>0.615385</td>\n",
       "      <td>0.943396</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.804348</td>\n",
       "      <td>0.954545</td>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.928571</td>\n",
       "      <td>182.000000</td>\n",
       "      <td>201.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.603774</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.892857</td>\n",
       "      <td>0.476190</td>\n",
       "      <td>66.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>0.347826</td>\n",
       "      <td>0.648649</td>\n",
       "      <td>27.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.619048</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>0.253012</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>204.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>27.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3526.000000</td>\n",
       "      <td>1787.000000</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>36.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.790698</td>\n",
       "      <td>55.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.952381</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.480000</td>\n",
       "      <td>0.772727</td>\n",
       "      <td>61.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.785714</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>0.423077</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>197.000000</td>\n",
       "      <td>208.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>60.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>42.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.925926</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>78.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>0.756757</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.894737</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.785714</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>198.000000</td>\n",
       "      <td>208.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>47.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.790698</td>\n",
       "      <td>55.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.952381</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>0.480000</td>\n",
       "      <td>0.772727</td>\n",
       "      <td>61.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.785714</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>0.423077</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>197.000000</td>\n",
       "      <td>208.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>60.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5360.000000</td>\n",
       "      <td>2729.000000</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.866667</td>\n",
       "      <td>74.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>73.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>83.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>82.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>66.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>72.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>219.000000</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>74.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>73.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>83.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.888889</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>218.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>82.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>13911.000000</td>\n",
       "      <td>9185.000000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.933333</td>\n",
       "      <td>89.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       base_adj_f1_score  base_adj_false_negative  base_adj_false_positive  \\\n",
       "count         417.000000               417.000000               417.000000   \n",
       "mean            0.613622                26.865707                 2.923261   \n",
       "std             0.205361                24.679118                 3.272681   \n",
       "min             0.255814                 0.000000                 0.000000   \n",
       "25%             0.422018                 4.000000                 1.000000   \n",
       "50%             0.603774                18.000000                 2.000000   \n",
       "75%             0.790698                55.000000                 4.000000   \n",
       "max             1.000000                73.000000                26.000000   \n",
       "\n",
       "       base_adj_precision  base_adj_recall  base_adj_true_negative  \\\n",
       "count          417.000000       417.000000              417.000000   \n",
       "mean             0.863344         0.540058               58.378897   \n",
       "std              0.132053         0.272973               26.523762   \n",
       "min              0.187500         0.159420               12.000000   \n",
       "25%              0.807692         0.280000               28.000000   \n",
       "50%              0.892857         0.476190               66.000000   \n",
       "75%              0.952381         0.791667               80.000000   \n",
       "max              1.000000         1.000000               98.000000   \n",
       "\n",
       "       base_adj_true_positive  base_arrow_f1_score  base_arrow_f1_score_ce  \\\n",
       "count              417.000000           417.000000              417.000000   \n",
       "mean                16.832134             0.383271                0.640144   \n",
       "std                  4.181648             0.174715                0.173975   \n",
       "min                  6.000000             0.071429                0.142857   \n",
       "25%                 14.000000             0.255319                0.514286   \n",
       "50%                 17.000000             0.347826                0.648649   \n",
       "75%                 19.000000             0.480000                0.772727   \n",
       "max                 28.000000             0.923077                1.000000   \n",
       "\n",
       "       base_arrow_false_negative  base_arrow_false_negative_ce  \\\n",
       "count                 417.000000                    417.000000   \n",
       "mean                   33.616307                      6.750600   \n",
       "std                    24.792243                      2.949928   \n",
       "min                     1.000000                      0.000000   \n",
       "25%                    10.000000                      5.000000   \n",
       "50%                    27.000000                      7.000000   \n",
       "75%                    61.000000                      9.000000   \n",
       "max                    83.000000                     15.000000   \n",
       "\n",
       "       base_arrow_false_positive  base_arrow_false_positive_ce  \\\n",
       "count                 417.000000                    417.000000   \n",
       "mean                    6.513189                      3.858513   \n",
       "std                     4.536756                      2.837190   \n",
       "min                     0.000000                      0.000000   \n",
       "25%                     3.000000                      2.000000   \n",
       "50%                     6.000000                      3.000000   \n",
       "75%                     9.000000                      6.000000   \n",
       "max                    28.000000                     14.000000   \n",
       "\n",
       "       base_arrow_precision  base_arrow_precision_ce  base_arrow_recall  \\\n",
       "count            417.000000               417.000000         417.000000   \n",
       "mean               0.626084                 0.725695           0.308170   \n",
       "std                0.208914                 0.185748           0.183161   \n",
       "min                0.047619                 0.166667           0.056818   \n",
       "25%                0.466667                 0.600000           0.169014   \n",
       "50%                0.619048                 0.750000           0.253012   \n",
       "75%                0.785714                 0.875000           0.423077   \n",
       "max                1.000000                 1.000000           0.888889   \n",
       "\n",
       "       base_arrow_recall_ce  base_arrow_true_negative  \\\n",
       "count            417.000000                417.000000   \n",
       "mean               0.584133                174.788969   \n",
       "std                0.179640                 27.214578   \n",
       "min                0.090909                126.000000   \n",
       "25%                0.461538                145.000000   \n",
       "50%                0.583333                180.000000   \n",
       "75%                0.714286                197.000000   \n",
       "max                1.000000                218.000000   \n",
       "\n",
       "       base_arrow_true_negative_ce  base_arrow_true_positive  \\\n",
       "count                   417.000000                417.000000   \n",
       "mean                    204.309353                 10.081535   \n",
       "std                       5.810942                  4.634535   \n",
       "min                     190.000000                  1.000000   \n",
       "25%                     200.000000                  7.000000   \n",
       "50%                     204.000000                  9.000000   \n",
       "75%                     208.000000                 13.000000   \n",
       "max                     218.000000                 25.000000   \n",
       "\n",
       "       base_arrow_true_positive_ce  base_base_shd  base_sid_lower  \\\n",
       "count                   417.000000     417.000000             0.0   \n",
       "mean                     10.081535      33.647482             NaN   \n",
       "std                       4.634535      24.720514             NaN   \n",
       "min                       1.000000       0.000000             NaN   \n",
       "25%                       7.000000      13.000000             NaN   \n",
       "50%                       9.000000      27.000000             NaN   \n",
       "75%                      13.000000      60.000000             NaN   \n",
       "max                      25.000000      82.000000             NaN   \n",
       "\n",
       "       cluster_adj_f1_score  cluster_adj_false_negative  \\\n",
       "count            417.000000                  417.000000   \n",
       "mean               0.683436                   21.177458   \n",
       "std                0.148385                   20.352829   \n",
       "min                0.292683                    0.000000   \n",
       "25%                0.566667                    2.000000   \n",
       "50%                0.666667                   14.000000   \n",
       "75%                0.800000                   42.000000   \n",
       "max                1.000000                   66.000000   \n",
       "\n",
       "       cluster_adj_false_positive  cluster_adj_precision  cluster_adj_recall  \\\n",
       "count                  417.000000             417.000000          417.000000   \n",
       "mean                     5.158273               0.820270            0.644706   \n",
       "std                      4.550347               0.147972            0.228748   \n",
       "min                      0.000000               0.176471            0.250000   \n",
       "25%                      2.000000               0.761905            0.433735   \n",
       "50%                      4.000000               0.844444            0.615385   \n",
       "75%                      7.000000               0.925926            0.857143   \n",
       "max                     28.000000               1.000000            1.000000   \n",
       "\n",
       "       cluster_adj_true_negative  cluster_adj_true_positive  \\\n",
       "count                 417.000000                 417.000000   \n",
       "mean                   56.143885                  22.520384   \n",
       "std                    26.530070                   8.209996   \n",
       "min                    12.000000                   6.000000   \n",
       "25%                    25.000000                  16.000000   \n",
       "50%                    63.000000                  22.000000   \n",
       "75%                    78.000000                  28.000000   \n",
       "max                    98.000000                  45.000000   \n",
       "\n",
       "       cluster_arrow_f1_score  cluster_arrow_f1_score_ce  \\\n",
       "count              417.000000                 417.000000   \n",
       "mean                 0.638167                   0.927616   \n",
       "std                  0.158778                   0.072930   \n",
       "min                  0.285714                   0.588235   \n",
       "25%                  0.521739                   0.894737   \n",
       "50%                  0.615385                   0.943396   \n",
       "75%                  0.756757                   1.000000   \n",
       "max                  1.000000                   1.000000   \n",
       "\n",
       "       cluster_arrow_false_negative  cluster_arrow_false_negative_ce  \\\n",
       "count                    417.000000                       417.000000   \n",
       "mean                      23.079137                         1.901679   \n",
       "std                       20.916523                         1.833823   \n",
       "min                        0.000000                         0.000000   \n",
       "25%                        3.000000                         0.000000   \n",
       "50%                       16.000000                         2.000000   \n",
       "75%                       44.000000                         3.000000   \n",
       "max                       72.000000                        10.000000   \n",
       "\n",
       "       cluster_arrow_false_positive  cluster_arrow_false_positive_ce  \\\n",
       "count                    417.000000                       417.000000   \n",
       "mean                       6.122302                         1.304556   \n",
       "std                        4.641152                         1.513021   \n",
       "min                        0.000000                         0.000000   \n",
       "25%                        3.000000                         0.000000   \n",
       "50%                        5.000000                         1.000000   \n",
       "75%                        9.000000                         2.000000   \n",
       "max                       28.000000                         7.000000   \n",
       "\n",
       "       cluster_arrow_precision  cluster_arrow_precision_ce  \\\n",
       "count               417.000000                  417.000000   \n",
       "mean                  0.778635                    0.940886   \n",
       "std                   0.154644                    0.069868   \n",
       "min                   0.176471                    0.588235   \n",
       "25%                   0.689655                    0.914286   \n",
       "50%                   0.804348                    0.954545   \n",
       "75%                   0.894737                    1.000000   \n",
       "max                   1.000000                    1.000000   \n",
       "\n",
       "       cluster_arrow_recall  cluster_arrow_recall_ce  \\\n",
       "count            417.000000               417.000000   \n",
       "mean               0.593473                 0.915803   \n",
       "std                0.226315                 0.080915   \n",
       "min                0.181818                 0.583333   \n",
       "25%                0.400000                 0.875000   \n",
       "50%                0.555556                 0.928571   \n",
       "75%                0.785714                 1.000000   \n",
       "max                1.000000                 1.000000   \n",
       "\n",
       "       cluster_arrow_true_negative  cluster_arrow_true_negative_ce  \\\n",
       "count                   417.000000                      417.000000   \n",
       "mean                    175.179856                      201.175060   \n",
       "std                      27.063650                        8.692147   \n",
       "min                     128.000000                      178.000000   \n",
       "25%                     144.000000                      195.000000   \n",
       "50%                     182.000000                      201.000000   \n",
       "75%                     198.000000                      208.000000   \n",
       "max                     218.000000                      219.000000   \n",
       "\n",
       "       cluster_arrow_true_positive  cluster_arrow_true_positive_ce  \\\n",
       "count                   417.000000                      417.000000   \n",
       "mean                     20.618705                       20.618705   \n",
       "std                       7.895102                        7.895102   \n",
       "min                       6.000000                        6.000000   \n",
       "25%                      14.000000                       14.000000   \n",
       "50%                      20.000000                       20.000000   \n",
       "75%                      25.000000                       25.000000   \n",
       "max                      44.000000                       44.000000   \n",
       "\n",
       "       cluster_cluster_shd  cluster_sid_lower  pruned_base_adj_f1_score  \\\n",
       "count           417.000000                0.0                417.000000   \n",
       "mean             27.640288                NaN                  0.613622   \n",
       "std              20.361401                NaN                  0.205361   \n",
       "min               0.000000                NaN                  0.255814   \n",
       "25%               9.000000                NaN                  0.422018   \n",
       "50%              23.000000                NaN                  0.603774   \n",
       "75%              47.000000                NaN                  0.790698   \n",
       "max              74.000000                NaN                  1.000000   \n",
       "\n",
       "       pruned_base_adj_false_negative  pruned_base_adj_false_positive  \\\n",
       "count                      417.000000                      417.000000   \n",
       "mean                        26.865707                        2.923261   \n",
       "std                         24.679118                        3.272681   \n",
       "min                          0.000000                        0.000000   \n",
       "25%                          4.000000                        1.000000   \n",
       "50%                         18.000000                        2.000000   \n",
       "75%                         55.000000                        4.000000   \n",
       "max                         73.000000                       26.000000   \n",
       "\n",
       "       pruned_base_adj_precision  pruned_base_adj_recall  \\\n",
       "count                 417.000000              417.000000   \n",
       "mean                    0.863344                0.540058   \n",
       "std                     0.132053                0.272973   \n",
       "min                     0.187500                0.159420   \n",
       "25%                     0.807692                0.280000   \n",
       "50%                     0.892857                0.476190   \n",
       "75%                     0.952381                0.791667   \n",
       "max                     1.000000                1.000000   \n",
       "\n",
       "       pruned_base_adj_true_negative  pruned_base_adj_true_positive  \\\n",
       "count                     417.000000                     417.000000   \n",
       "mean                       58.378897                      16.832134   \n",
       "std                        26.523762                       4.181648   \n",
       "min                        12.000000                       6.000000   \n",
       "25%                        28.000000                      14.000000   \n",
       "50%                        66.000000                      17.000000   \n",
       "75%                        80.000000                      19.000000   \n",
       "max                        98.000000                      28.000000   \n",
       "\n",
       "       pruned_base_arrow_f1_score  pruned_base_arrow_f1_score_ce  \\\n",
       "count                  417.000000                     417.000000   \n",
       "mean                     0.383271                       0.640144   \n",
       "std                      0.174715                       0.173975   \n",
       "min                      0.071429                       0.142857   \n",
       "25%                      0.255319                       0.514286   \n",
       "50%                      0.347826                       0.648649   \n",
       "75%                      0.480000                       0.772727   \n",
       "max                      0.923077                       1.000000   \n",
       "\n",
       "       pruned_base_arrow_false_negative  pruned_base_arrow_false_negative_ce  \\\n",
       "count                        417.000000                           417.000000   \n",
       "mean                          33.616307                             6.750600   \n",
       "std                           24.792243                             2.949928   \n",
       "min                            1.000000                             0.000000   \n",
       "25%                           10.000000                             5.000000   \n",
       "50%                           27.000000                             7.000000   \n",
       "75%                           61.000000                             9.000000   \n",
       "max                           83.000000                            15.000000   \n",
       "\n",
       "       pruned_base_arrow_false_positive  pruned_base_arrow_false_positive_ce  \\\n",
       "count                        417.000000                           417.000000   \n",
       "mean                           6.513189                             3.858513   \n",
       "std                            4.536756                             2.837190   \n",
       "min                            0.000000                             0.000000   \n",
       "25%                            3.000000                             2.000000   \n",
       "50%                            6.000000                             3.000000   \n",
       "75%                            9.000000                             6.000000   \n",
       "max                           28.000000                            14.000000   \n",
       "\n",
       "       pruned_base_arrow_precision  pruned_base_arrow_precision_ce  \\\n",
       "count                   417.000000                      417.000000   \n",
       "mean                      0.626084                        0.725695   \n",
       "std                       0.208914                        0.185748   \n",
       "min                       0.047619                        0.166667   \n",
       "25%                       0.466667                        0.600000   \n",
       "50%                       0.619048                        0.750000   \n",
       "75%                       0.785714                        0.875000   \n",
       "max                       1.000000                        1.000000   \n",
       "\n",
       "       pruned_base_arrow_recall  pruned_base_arrow_recall_ce  \\\n",
       "count                417.000000                   417.000000   \n",
       "mean                   0.308170                     0.584133   \n",
       "std                    0.183161                     0.179640   \n",
       "min                    0.056818                     0.090909   \n",
       "25%                    0.169014                     0.461538   \n",
       "50%                    0.253012                     0.583333   \n",
       "75%                    0.423077                     0.714286   \n",
       "max                    0.888889                     1.000000   \n",
       "\n",
       "       pruned_base_arrow_true_negative  pruned_base_arrow_true_negative_ce  \\\n",
       "count                       417.000000                          417.000000   \n",
       "mean                        174.788969                          204.309353   \n",
       "std                          27.214578                            5.810942   \n",
       "min                         126.000000                          190.000000   \n",
       "25%                         145.000000                          200.000000   \n",
       "50%                         180.000000                          204.000000   \n",
       "75%                         197.000000                          208.000000   \n",
       "max                         218.000000                          218.000000   \n",
       "\n",
       "       pruned_base_arrow_true_positive  pruned_base_arrow_true_positive_ce  \\\n",
       "count                       417.000000                          417.000000   \n",
       "mean                         10.081535                           10.081535   \n",
       "std                           4.634535                            4.634535   \n",
       "min                           1.000000                            1.000000   \n",
       "25%                           7.000000                            7.000000   \n",
       "50%                           9.000000                            9.000000   \n",
       "75%                          13.000000                           13.000000   \n",
       "max                          25.000000                           25.000000   \n",
       "\n",
       "       pruned_base_pruned_base_shd  pruned_base_sid_lower  Base indep tests  \\\n",
       "count                   417.000000                    0.0        417.000000   \n",
       "mean                     33.647482                    NaN       4035.211031   \n",
       "std                      24.720514                    NaN       2800.471409   \n",
       "min                       0.000000                    NaN        251.000000   \n",
       "25%                      13.000000                    NaN       2028.000000   \n",
       "50%                      27.000000                    NaN       3526.000000   \n",
       "75%                      60.000000                    NaN       5360.000000   \n",
       "max                      82.000000                    NaN      13911.000000   \n",
       "\n",
       "       C-PC indep tests       alpha  cluster_connectivity  empty_graph_shd  \\\n",
       "count        417.000000  417.000000            417.000000       417.000000   \n",
       "mean        2029.055156    0.189808              0.737250        43.697842   \n",
       "std         1520.365094    0.184236              0.164742        27.083823   \n",
       "min          103.000000    0.010000              0.166667         7.000000   \n",
       "25%          877.000000    0.050000              0.666667        18.000000   \n",
       "50%         1787.000000    0.100000              0.800000        36.000000   \n",
       "75%         2729.000000    0.250000              0.866667        74.000000   \n",
       "max         9185.000000    0.500000              0.933333        89.000000   \n",
       "\n",
       "       n_clusters     n_edges  n_nodes  noise_scale  sample_size  \\\n",
       "count  417.000000  417.000000    417.0        417.0        417.0   \n",
       "mean     5.026379   38.297362     15.0          1.0       1000.0   \n",
       "std      0.954128   23.279300      0.0          0.0          0.0   \n",
       "min      3.000000   15.000000     15.0          1.0       1000.0   \n",
       "25%      4.000000   15.000000     15.0          1.0       1000.0   \n",
       "50%      5.000000   30.000000     15.0          1.0       1000.0   \n",
       "75%      6.000000   50.000000     15.0          1.0       1000.0   \n",
       "max      6.000000   80.000000     15.0          1.0       1000.0   \n",
       "\n",
       "       true_sid_lower  true_sid_upper  \n",
       "count             0.0             0.0  \n",
       "mean              NaN             NaN  \n",
       "std               NaN             NaN  \n",
       "min               NaN             NaN  \n",
       "25%               NaN             NaN  \n",
       "50%               NaN             NaN  \n",
       "75%               NaN             NaN  \n",
       "max               NaN             NaN  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_nonadj_clusters.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_precision</th>\n",
       "      <th>cluster_adj_precision</th>\n",
       "      <th>base_adj_recall</th>\n",
       "      <th>cluster_adj_recall</th>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <th>cluster_adj_f1_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.864585</td>\n",
       "      <td>0.832252</td>\n",
       "      <td>0.464137</td>\n",
       "      <td>0.562061</td>\n",
       "      <td>0.559182</td>\n",
       "      <td>0.634823</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.117717</td>\n",
       "      <td>0.129235</td>\n",
       "      <td>0.254046</td>\n",
       "      <td>0.223503</td>\n",
       "      <td>0.198208</td>\n",
       "      <td>0.149663</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.187500</td>\n",
       "      <td>0.176471</td>\n",
       "      <td>0.134146</td>\n",
       "      <td>0.194444</td>\n",
       "      <td>0.231579</td>\n",
       "      <td>0.292683</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.809524</td>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.256332</td>\n",
       "      <td>0.379638</td>\n",
       "      <td>0.393213</td>\n",
       "      <td>0.520548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.850926</td>\n",
       "      <td>0.360000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.507252</td>\n",
       "      <td>0.611341</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.947368</td>\n",
       "      <td>0.923077</td>\n",
       "      <td>0.653846</td>\n",
       "      <td>0.742349</td>\n",
       "      <td>0.720328</td>\n",
       "      <td>0.736842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       base_adj_precision  cluster_adj_precision  base_adj_recall  \\\n",
       "count          900.000000             900.000000       900.000000   \n",
       "mean             0.864585               0.832252         0.464137   \n",
       "std              0.117717               0.129235         0.254046   \n",
       "min              0.187500               0.176471         0.134146   \n",
       "25%              0.809524               0.777778         0.256332   \n",
       "50%              0.888889               0.850926         0.360000   \n",
       "75%              0.947368               0.923077         0.653846   \n",
       "max              1.000000               1.000000         1.000000   \n",
       "\n",
       "       cluster_adj_recall  base_adj_f1_score  cluster_adj_f1_score  \n",
       "count          900.000000         900.000000            900.000000  \n",
       "mean             0.562061           0.559182              0.634823  \n",
       "std              0.223503           0.198208              0.149663  \n",
       "min              0.194444           0.231579              0.292683  \n",
       "25%              0.379638           0.393213              0.520548  \n",
       "50%              0.500000           0.507252              0.611341  \n",
       "75%              0.742349           0.720328              0.736842  \n",
       "max              1.000000           1.000000              1.000000  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prec_recall_metrics_with_pruned = ['base_adj_precision', 'pruned_base_adj_precision', 'cluster_adj_precision', \\\n",
    "           'base_adj_recall', 'pruned_base_adj_recall','cluster_adj_recall', \\\n",
    "           'base_adj_f1_score', 'pruned_base_adj_f1_score', 'cluster_adj_f1_score']\n",
    "prec_recall_metrics = ['base_adj_precision', 'cluster_adj_precision', \\\n",
    "              'base_adj_recall', 'cluster_adj_recall', \\\n",
    "                'base_adj_f1_score', 'cluster_adj_f1_score']\n",
    "data[prec_recall_metrics].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "indep_test_metrics = ['Base indep tests', 'C-PC indep tests']\n",
    "dist_metrics = ['base_base_shd', 'cluster_cluster_shd', 'empty_graph_shd']\n",
    "overall_metrics = prec_recall_metrics + dist_metrics + indep_test_metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>base_adj_precision</th>\n",
       "      <td>0.863080</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cluster_adj_precision</th>\n",
       "      <td>0.836135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>base_adj_recall</th>\n",
       "      <td>0.463783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cluster_adj_recall</th>\n",
       "      <td>0.545386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <td>0.558015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cluster_adj_f1_score</th>\n",
       "      <td>0.621049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>base_base_shd</th>\n",
       "      <td>40.371296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cluster_cluster_shd</th>\n",
       "      <td>35.388889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>empty_graph_shd</th>\n",
       "      <td>50.596296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Base indep tests</th>\n",
       "      <td>4580.461111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>C-PC indep tests</th>\n",
       "      <td>2994.563889</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 0\n",
       "base_adj_precision        0.863080\n",
       "cluster_adj_precision     0.836135\n",
       "base_adj_recall           0.463783\n",
       "cluster_adj_recall        0.545386\n",
       "base_adj_f1_score         0.558015\n",
       "cluster_adj_f1_score      0.621049\n",
       "base_base_shd            40.371296\n",
       "cluster_cluster_shd      35.388889\n",
       "empty_graph_shd          50.596296\n",
       "Base indep tests       4580.461111\n",
       "C-PC indep tests       2994.563889"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(data_all[overall_metrics].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lr}\n",
      "\\toprule\n",
      "{} &            0 \\\\\n",
      "\\midrule\n",
      "base\\_adj\\_precision    &     0.863080 \\\\\n",
      "cluster\\_adj\\_precision &     0.836135 \\\\\n",
      "base\\_adj\\_recall       &     0.463783 \\\\\n",
      "cluster\\_adj\\_recall    &     0.545386 \\\\\n",
      "base\\_adj\\_f1\\_score     &     0.558015 \\\\\n",
      "cluster\\_adj\\_f1\\_score  &     0.621049 \\\\\n",
      "base\\_base\\_shd         &    40.371296 \\\\\n",
      "cluster\\_cluster\\_shd   &    35.388889 \\\\\n",
      "empty\\_graph\\_shd       &    50.596296 \\\\\n",
      "Base indep tests      &  4580.461111 \\\\\n",
      "C-PC indep tests      &  2994.563889 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "In future versions `DataFrame.to_latex` is expected to utilise the base implementation of `Styler.to_latex` for formatting and rendering. The arguments signature may therefore change. It is recommended instead to use `DataFrame.style.to_latex` which also contains additional functionality.\n"
     ]
    }
   ],
   "source": [
    "print(pd.DataFrame(data_all[overall_metrics].mean()).to_latex())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # colors = ['#A4C2F4', '#3333FF', '#64d2a6', '#065535', '#e49645', '#b45f06']\n",
    "# # colors = ['#A4C2F4', '#3333FF', '#64d2a6', '#065535', '#e49645', '#b45f06']\n",
    "# # colors = ['#A4C2F4', '#3333FF', '#64d2a6', '#065535', '#e49645', '#b45f06']\n",
    "# colors = ['#9ABCE4', '#005293', '#C7D97D', '#7D922A', '#FAD080', '#F7811E'] #TUM COLORS\n",
    "# mean_values = data_all[prec_recall_metrics].mean()\n",
    "# ax = mean_values.plot(kind='bar', figsize=(15, 8), color=colors)\n",
    "# plt.title('Precision, Recall and F1-Score', fontsize=20)\n",
    "# plt.xlabel('Metric', fontsize=18)\n",
    "# plt.tick_params(axis='both', which='major', labelsize=14)\n",
    "# plt.tick_params(axis='x', which='major', labelsize=16, rotation=30)\n",
    "# label_names = ['PC precision', 'Cluster PC precision', 'PC recall', 'Cluster PC recall', 'PC F1-score', 'Cluster PC F1-score']\n",
    "# ax.set_xticklabels(label_names)\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "mean_values = data_all[prec_recall_metrics].mean()\n",
    "\n",
    "# Group data into a new DataFrame\n",
    "grouped_data = pd.DataFrame({\n",
    "    'Base PC': [mean_values['base_adj_precision'],mean_values['base_adj_recall'], mean_values['base_adj_f1_score']],\n",
    "    'Cluster-PC': [mean_values['cluster_adj_precision'], mean_values['cluster_adj_recall'], mean_values['cluster_adj_f1_score']],\n",
    "})\n",
    "\n",
    "# Set the index to the group names\n",
    "grouped_data.index = ['precision', 'recall', 'F1-score' ]\n",
    "\n",
    "# Plot the data\n",
    "# colors = ['#A4C2F4', '#3333FF', '#64d2a6', '#065535', '#e49645', '#b45f06']\n",
    "# colors = ['#64a0c8', '#005293', '#64d2a6', '#065535', '#e49645', '#b45f06']\n",
    "# colors = ['#A4C2F4', '#3333FF', '#4D68D5']\n",
    "colors = ['#9ABCE4', '#005293', '#5E94D4']\n",
    "ax = grouped_data.plot(kind='bar', figsize=(15, 8), color=colors[:2])\n",
    "\n",
    "# Customizing the plot\n",
    "plt.title('Base PC vs Cluster-PC: comparison of precision, recall and F1-score', fontsize=20)\n",
    "plt.xlabel('metric', fontsize=18)\n",
    "plt.xticks(rotation=0)\n",
    "plt.tick_params(axis='both', which='major', labelsize=20)\n",
    "\n",
    "# Show the plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_precision</th>\n",
       "      <th>cluster_adj_precision</th>\n",
       "      <th>base_adj_recall</th>\n",
       "      <th>cluster_adj_recall</th>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <th>cluster_adj_f1_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "      <td>417.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.863344</td>\n",
       "      <td>0.820270</td>\n",
       "      <td>0.540058</td>\n",
       "      <td>0.644706</td>\n",
       "      <td>0.613622</td>\n",
       "      <td>0.683436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.132053</td>\n",
       "      <td>0.147972</td>\n",
       "      <td>0.272973</td>\n",
       "      <td>0.228748</td>\n",
       "      <td>0.205361</td>\n",
       "      <td>0.148385</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.187500</td>\n",
       "      <td>0.176471</td>\n",
       "      <td>0.159420</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.255814</td>\n",
       "      <td>0.292683</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.807692</td>\n",
       "      <td>0.761905</td>\n",
       "      <td>0.280000</td>\n",
       "      <td>0.433735</td>\n",
       "      <td>0.422018</td>\n",
       "      <td>0.566667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.892857</td>\n",
       "      <td>0.844444</td>\n",
       "      <td>0.476190</td>\n",
       "      <td>0.615385</td>\n",
       "      <td>0.603774</td>\n",
       "      <td>0.666667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.952381</td>\n",
       "      <td>0.925926</td>\n",
       "      <td>0.791667</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>0.790698</td>\n",
       "      <td>0.800000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       base_adj_precision  cluster_adj_precision  base_adj_recall  \\\n",
       "count          417.000000             417.000000       417.000000   \n",
       "mean             0.863344               0.820270         0.540058   \n",
       "std              0.132053               0.147972         0.272973   \n",
       "min              0.187500               0.176471         0.159420   \n",
       "25%              0.807692               0.761905         0.280000   \n",
       "50%              0.892857               0.844444         0.476190   \n",
       "75%              0.952381               0.925926         0.791667   \n",
       "max              1.000000               1.000000         1.000000   \n",
       "\n",
       "       cluster_adj_recall  base_adj_f1_score  cluster_adj_f1_score  \n",
       "count          417.000000         417.000000            417.000000  \n",
       "mean             0.644706           0.613622              0.683436  \n",
       "std              0.228748           0.205361              0.148385  \n",
       "min              0.250000           0.255814              0.292683  \n",
       "25%              0.433735           0.422018              0.566667  \n",
       "50%              0.615385           0.603774              0.666667  \n",
       "75%              0.857143           0.790698              0.800000  \n",
       "max              1.000000           1.000000              1.000000  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_nonadj_clusters[prec_recall_metrics].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# latex = data_nonadj_clusters[prec_recall_metrics].describe().style.to_latex()\n",
    "# print(latex)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "C-PC has slightly lower true negative and higher true positive. Positive is defined as an edge that is present in the true graph. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_false_negative</th>\n",
       "      <th>cluster_adj_false_negative</th>\n",
       "      <th>base_adj_false_positive</th>\n",
       "      <th>cluster_adj_false_positive</th>\n",
       "      <th>base_adj_true_negative</th>\n",
       "      <th>cluster_adj_true_negative</th>\n",
       "      <th>base_adj_true_positive</th>\n",
       "      <th>cluster_adj_true_positive</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>33.222222</td>\n",
       "      <td>27.486667</td>\n",
       "      <td>2.924444</td>\n",
       "      <td>4.885556</td>\n",
       "      <td>51.490000</td>\n",
       "      <td>49.528889</td>\n",
       "      <td>17.363333</td>\n",
       "      <td>23.098889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>24.527018</td>\n",
       "      <td>21.110094</td>\n",
       "      <td>2.907713</td>\n",
       "      <td>4.086824</td>\n",
       "      <td>25.958144</td>\n",
       "      <td>25.860932</td>\n",
       "      <td>3.944863</td>\n",
       "      <td>7.218826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>9.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>19.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>32.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>53.000000</td>\n",
       "      <td>51.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>23.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>58.000000</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>73.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>27.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>75.000000</td>\n",
       "      <td>67.000000</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>45.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       base_adj_false_negative  cluster_adj_false_negative  \\\n",
       "count               900.000000                  900.000000   \n",
       "mean                 33.222222                   27.486667   \n",
       "std                  24.527018                   21.110094   \n",
       "min                   0.000000                    0.000000   \n",
       "25%                   9.000000                    7.000000   \n",
       "50%                  32.000000                   25.000000   \n",
       "75%                  58.000000                   48.000000   \n",
       "max                  75.000000                   67.000000   \n",
       "\n",
       "       base_adj_false_positive  cluster_adj_false_positive  \\\n",
       "count               900.000000                  900.000000   \n",
       "mean                  2.924444                    4.885556   \n",
       "std                   2.907713                    4.086824   \n",
       "min                   0.000000                    0.000000   \n",
       "25%                   1.000000                    2.000000   \n",
       "50%                   2.000000                    4.000000   \n",
       "75%                   4.000000                    7.000000   \n",
       "max                  26.000000                   28.000000   \n",
       "\n",
       "       base_adj_true_negative  cluster_adj_true_negative  \\\n",
       "count              900.000000                 900.000000   \n",
       "mean                51.490000                  49.528889   \n",
       "std                 25.958144                  25.860932   \n",
       "min                 12.000000                  11.000000   \n",
       "25%                 25.000000                  23.000000   \n",
       "50%                 53.000000                  51.000000   \n",
       "75%                 77.000000                  73.000000   \n",
       "max                 98.000000                  98.000000   \n",
       "\n",
       "       base_adj_true_positive  cluster_adj_true_positive  \n",
       "count              900.000000                 900.000000  \n",
       "mean                17.363333                  23.098889  \n",
       "std                  3.944863                   7.218826  \n",
       "min                  6.000000                   6.000000  \n",
       "25%                 15.000000                  19.000000  \n",
       "50%                 17.000000                  23.000000  \n",
       "75%                 20.000000                  27.000000  \n",
       "max                 29.000000                  45.000000  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "count_metrics_with_pruned = ['base_adj_false_negative', 'pruned_base_adj_false_negative', 'cluster_adj_false_negative', \\\n",
    "                 'base_adj_false_positive', 'pruned_base_adj_false_positive','cluster_adj_false_positive', \\\n",
    "                    'base_adj_true_negative', 'pruned_base_adj_true_negative', 'cluster_adj_true_negative', \\\n",
    "                        'base_adj_true_positive', 'pruned_base_adj_true_positive', 'cluster_adj_true_positive']\n",
    "count_metrics = ['base_adj_false_negative', 'cluster_adj_false_negative', \\\n",
    "                    'base_adj_false_positive','cluster_adj_false_positive', \\\n",
    "                        'base_adj_true_negative', 'cluster_adj_true_negative', \\\n",
    "                            'base_adj_true_positive', 'cluster_adj_true_positive']\n",
    "data[count_metrics].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Calculating the total of correctly identified edges ratio: $TCIE = \\frac{TP + TN}{TP + TN + FP + FN}$. \n",
    "\n",
    "In total C-PC has slightly higher TCIE. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_TCIE</th>\n",
       "      <th>cluster_TCIE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.655746</td>\n",
       "      <td>0.691693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.230079</td>\n",
       "      <td>0.196015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.266667</td>\n",
       "      <td>0.342857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.428571</td>\n",
       "      <td>0.504762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.671429</td>\n",
       "      <td>0.704762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.876190</td>\n",
       "      <td>0.866667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        base_TCIE  cluster_TCIE\n",
       "count  900.000000    900.000000\n",
       "mean     0.655746      0.691693\n",
       "std      0.230079      0.196015\n",
       "min      0.266667      0.342857\n",
       "25%      0.428571      0.504762\n",
       "50%      0.671429      0.704762\n",
       "75%      0.876190      0.866667\n",
       "max      1.000000      1.000000"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['cluster_TCIE'] = (data['cluster_adj_true_positive'] + \n",
    "                data['cluster_adj_true_negative']) / (data['cluster_adj_true_positive'] + \n",
    "                                                                                        data['cluster_adj_true_negative'] + \n",
    "                                                                                        data['cluster_adj_false_positive'] + \n",
    "                                                                                        data['cluster_adj_false_negative'])\n",
    "data['base_TCIE'] = (data['base_adj_true_positive'] +\n",
    "                data['base_adj_true_negative']) / (data['base_adj_true_positive'] + \n",
    "                                                                                        data['base_adj_true_negative'] + \n",
    "                                                                                        data['base_adj_false_positive'] + \n",
    "                                                                                        data['base_adj_false_negative'])\n",
    "data[['base_TCIE', 'cluster_TCIE']].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "All in all one can say that C-PC removes slightly less edges than PC. It doesn't remove as many edges that shouldn't be removed, but also keeps more edges that should be removed. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "C-PC has a slightly lower SHD. If one was to calculate the SID, it would show a similar trend to the SHD. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_base_shd</th>\n",
       "      <th>cluster_cluster_shd</th>\n",
       "      <th>empty_graph_shd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "      <td>900.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>40.272222</td>\n",
       "      <td>34.293333</td>\n",
       "      <td>50.585556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>24.622225</td>\n",
       "      <td>21.214634</td>\n",
       "      <td>26.385404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>17.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>26.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>41.000000</td>\n",
       "      <td>33.000000</td>\n",
       "      <td>50.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>65.000000</td>\n",
       "      <td>54.000000</td>\n",
       "      <td>79.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>84.000000</td>\n",
       "      <td>74.000000</td>\n",
       "      <td>89.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       base_base_shd  cluster_cluster_shd  empty_graph_shd\n",
       "count     900.000000           900.000000       900.000000\n",
       "mean       40.272222            34.293333        50.585556\n",
       "std        24.622225            21.214634        26.385404\n",
       "min         0.000000             0.000000         7.000000\n",
       "25%        17.000000            15.000000        26.000000\n",
       "50%        41.000000            33.000000        50.000000\n",
       "75%        65.000000            54.000000        79.000000\n",
       "max        84.000000            74.000000        89.000000"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dist_metrics_with_pruned = ['base_base_shd', 'pruned_base_pruned_base_shd', 'cluster_cluster_shd', 'empty_graph_shd']#, \\\n",
    "         #   'base_sid_lower', 'base_sid_upper', 'pruned_base_sid_lower', 'pruned_base_sid_upper', \\\n",
    "         #      'cluster_sid_lower', 'cluster_sid_upper']\n",
    "dist_metrics = ['base_base_shd', 'cluster_cluster_shd', 'empty_graph_shd']#, \\\n",
    "            #   'base_sid_lower', 'base_sid_upper \\\n",
    "            #      'cluster_sid_lower', 'cluster_sid_upper']\n",
    "data[dist_metrics].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is the edge density, intra_cluster is number_edges_within_cluster/number_possible_edges_within_cluster, \n",
    "\n",
    "inter_cluster is number_edges_between_connected_clusters/number_possible_edges_between_connected_clusters and\n",
    "\n",
    "inter_including_disconnected_cluster is number_edges_between_all_clusters/number_possible_edges_between_all_clusters.\n",
    "\n",
    "This means inter_including_disconnected_cluster also has edges between disconnected clusters in the denominator (and is therefore slightly lower)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Edge Ratios \n",
      "    intra_cluster  inter_cluster  inter_including_disconnected_cluster\n",
      "0       0.398322       0.560289                              0.507978\n"
     ]
    }
   ],
   "source": [
    "edge_ratios = data['edge_ratios'].apply(pd.Series)\n",
    "edge_ratios.columns = ['intra_cluster', 'inter_cluster', 'inter_including_disconnected_cluster']\n",
    "mean_values = edge_ratios.mean()\n",
    "mean_values = pd.DataFrame([mean_values])\n",
    "print('Edge Ratios \\n', mean_values)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Grouping by clusters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_precision</th>\n",
       "      <th>cluster_adj_precision</th>\n",
       "      <th>base_adj_recall</th>\n",
       "      <th>cluster_adj_recall</th>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <th>cluster_adj_f1_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n_clusters</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1.0</th>\n",
       "      <td>0.855552</td>\n",
       "      <td>0.855552</td>\n",
       "      <td>0.462014</td>\n",
       "      <td>0.462014</td>\n",
       "      <td>0.552180</td>\n",
       "      <td>0.552180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.0</th>\n",
       "      <td>0.858904</td>\n",
       "      <td>0.841061</td>\n",
       "      <td>0.466046</td>\n",
       "      <td>0.513265</td>\n",
       "      <td>0.560737</td>\n",
       "      <td>0.597901</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.0</th>\n",
       "      <td>0.853105</td>\n",
       "      <td>0.822118</td>\n",
       "      <td>0.459563</td>\n",
       "      <td>0.538257</td>\n",
       "      <td>0.550054</td>\n",
       "      <td>0.611988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.0</th>\n",
       "      <td>0.865708</td>\n",
       "      <td>0.831746</td>\n",
       "      <td>0.474261</td>\n",
       "      <td>0.575726</td>\n",
       "      <td>0.566705</td>\n",
       "      <td>0.644463</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.0</th>\n",
       "      <td>0.871712</td>\n",
       "      <td>0.829852</td>\n",
       "      <td>0.471323</td>\n",
       "      <td>0.594562</td>\n",
       "      <td>0.566068</td>\n",
       "      <td>0.659212</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.0</th>\n",
       "      <td>0.873498</td>\n",
       "      <td>0.836482</td>\n",
       "      <td>0.449494</td>\n",
       "      <td>0.588495</td>\n",
       "      <td>0.552346</td>\n",
       "      <td>0.660550</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            base_adj_precision  cluster_adj_precision  base_adj_recall  \\\n",
       "n_clusters                                                               \n",
       "1.0                   0.855552               0.855552         0.462014   \n",
       "2.0                   0.858904               0.841061         0.466046   \n",
       "3.0                   0.853105               0.822118         0.459563   \n",
       "4.0                   0.865708               0.831746         0.474261   \n",
       "5.0                   0.871712               0.829852         0.471323   \n",
       "6.0                   0.873498               0.836482         0.449494   \n",
       "\n",
       "            cluster_adj_recall  base_adj_f1_score  cluster_adj_f1_score  \n",
       "n_clusters                                                               \n",
       "1.0                   0.462014           0.552180              0.552180  \n",
       "2.0                   0.513265           0.560737              0.597901  \n",
       "3.0                   0.538257           0.550054              0.611988  \n",
       "4.0                   0.575726           0.566705              0.644463  \n",
       "5.0                   0.594562           0.566068              0.659212  \n",
       "6.0                   0.588495           0.552346              0.660550  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_all.groupby('n_clusters').mean()[prec_recall_metrics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b546911d20>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# colors = ['#A4C2F4', '#3333FF', '#64d2a6', '#065535', '#e49645', '#b45f06']\n",
    "# colors = ['#A4C2F4', '#3333FF', '#64d2a6', '#065535', '#e49645', '#b45f06']\n",
    "colors = ['#9ABCE4', '#005293', '#C7D97D', '#7D922A', '#FAD080', '#F7811E'] #TUM COLORS\n",
    "# colors = ['#64a0c8', '#005293', '#8F81EA', '#6955E2', '#FAD080', '#F7811E']\n",
    "data_all.groupby('n_clusters').mean()[prec_recall_metrics].plot(kind='bar', figsize=(15, 8), color = colors)\n",
    "plt.title('Precision, Recall and F1-Score for different numbers of clusters', fontsize = 20)\n",
    "plt.xlabel('Number of clusters', fontsize=18)\n",
    "plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "label_names = ['PC precision', 'Cluster PC precision', 'PC recall', 'Cluster PC recall', 'PC F1-score', 'Cluster PC F1-score']\n",
    "plt.legend(label_names,  bbox_to_anchor = (1.3, 1),  fontsize=16, loc='upper right')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # colors = ['#A4C2F4', '#3333FF', '#64d2a6', '#065535', '#e49645', '#b45f06']\n",
    "# colors = ['#9ABCE4', '#005293', '#C7D97D', '#7D922A', '#FAD080', '#F7811E'] #TUM COLORS\n",
    "# data_nonadj_clusters.groupby('n_clusters').mean()[prec_recall_metrics].plot(kind='bar', figsize=(15, 8), color = colors)\n",
    "# plt.title('Precision, Recall and F1-Score for different numbers of clusters for not fully connected C-DAGs', fontsize = 20)\n",
    "# plt.xlabel('Number of clusters', fontsize=18)\n",
    "# plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "# plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "# label_names = ['PC precision', 'Cluster PC precision', 'PC recall', 'Cluster PC recall', 'PC F1-score', 'Cluster PC F1-score']\n",
    "# plt.legend(label_names,  bbox_to_anchor = (1.3, 1),  fontsize=16, loc='upper right')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_base_shd</th>\n",
       "      <th>cluster_cluster_shd</th>\n",
       "      <th>empty_graph_shd</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n_clusters</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1.0</th>\n",
       "      <td>40.866667</td>\n",
       "      <td>40.866667</td>\n",
       "      <td>50.650000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.0</th>\n",
       "      <td>39.805556</td>\n",
       "      <td>37.205556</td>\n",
       "      <td>50.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.0</th>\n",
       "      <td>40.750000</td>\n",
       "      <td>35.877778</td>\n",
       "      <td>50.433333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.0</th>\n",
       "      <td>39.300000</td>\n",
       "      <td>33.405556</td>\n",
       "      <td>49.894444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.0</th>\n",
       "      <td>40.105556</td>\n",
       "      <td>32.383333</td>\n",
       "      <td>50.461111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.0</th>\n",
       "      <td>41.400000</td>\n",
       "      <td>32.594444</td>\n",
       "      <td>51.805556</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            base_base_shd  cluster_cluster_shd  empty_graph_shd\n",
       "n_clusters                                                     \n",
       "1.0             40.866667            40.866667        50.650000\n",
       "2.0             39.805556            37.205556        50.333333\n",
       "3.0             40.750000            35.877778        50.433333\n",
       "4.0             39.300000            33.405556        49.894444\n",
       "5.0             40.105556            32.383333        50.461111\n",
       "6.0             41.400000            32.594444        51.805556"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_all.groupby('n_clusters').mean()[dist_metrics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b546911300>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# colors = ['#A4C2F4', '#3333FF', '#4D68D5']\n",
    "colors = ['#9ABCE4', '#005293', '#5E94D4']\n",
    "data_all.groupby('n_clusters').mean()[dist_metrics].plot(kind='bar', figsize=(10, 6), color = colors)\n",
    "plt.title('SHD for different numbers of clusters', fontsize = 20)\n",
    "plt.xlabel('Number of clusters', fontsize=18)\n",
    "plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "label_names = ['PC SHD', 'Cluster PC SHD', 'Empty graph SHD']\n",
    "plt.legend(label_names, bbox_to_anchor = (1.42, 1),  fontsize=16, loc='upper right')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_base_shd</th>\n",
       "      <th>cluster_cluster_shd</th>\n",
       "      <th>empty_graph_shd</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n_clusters</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3.0</th>\n",
       "      <td>19.093750</td>\n",
       "      <td>19.062500</td>\n",
       "      <td>25.406250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.0</th>\n",
       "      <td>28.954545</td>\n",
       "      <td>25.284091</td>\n",
       "      <td>39.465909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.0</th>\n",
       "      <td>33.343284</td>\n",
       "      <td>27.134328</td>\n",
       "      <td>43.529851</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.0</th>\n",
       "      <td>39.288344</td>\n",
       "      <td>31.012270</td>\n",
       "      <td>49.711656</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            base_base_shd  cluster_cluster_shd  empty_graph_shd\n",
       "n_clusters                                                     \n",
       "3.0             19.093750            19.062500        25.406250\n",
       "4.0             28.954545            25.284091        39.465909\n",
       "5.0             33.343284            27.134328        43.529851\n",
       "6.0             39.288344            31.012270        49.711656"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_nonadj_clusters.groupby('n_clusters').mean()[dist_metrics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b5468a3a60>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# colors = ['#A4C2F4', '#3333FF', '#4D68D5']\n",
    "colors = ['#9ABCE4', '#005293', '#5E94D4']\n",
    "data_nonadj_clusters.groupby('n_clusters').mean()[dist_metrics].plot(kind='bar', figsize=(10, 6), color = colors)\n",
    "plt.title('SHD for different numbers of clusters for not fully connected C-DAGs', fontsize = 20)\n",
    "plt.xlabel('Number of clusters', fontsize=18)\n",
    "plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "label_names = ['PC SHD', 'Cluster PC SHD', 'Empty graph SHD']\n",
    "plt.legend(label_names, bbox_to_anchor = (1.42, 1),  fontsize=16, loc='upper right')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Goes up as less C-DAGs are fully connected at 5,6 or 7 clusters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Base indep tests</th>\n",
       "      <th>C-PC indep tests</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n_clusters</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1.0</th>\n",
       "      <td>4619.927778</td>\n",
       "      <td>4619.927778</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.0</th>\n",
       "      <td>4548.122222</td>\n",
       "      <td>3481.400000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.0</th>\n",
       "      <td>4630.138889</td>\n",
       "      <td>2949.438889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.0</th>\n",
       "      <td>4431.344444</td>\n",
       "      <td>2511.738889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.0</th>\n",
       "      <td>4598.761111</td>\n",
       "      <td>2315.611111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.0</th>\n",
       "      <td>4654.472222</td>\n",
       "      <td>2089.266667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Base indep tests  C-PC indep tests\n",
       "n_clusters                                    \n",
       "1.0              4619.927778       4619.927778\n",
       "2.0              4548.122222       3481.400000\n",
       "3.0              4630.138889       2949.438889\n",
       "4.0              4431.344444       2511.738889\n",
       "5.0              4598.761111       2315.611111\n",
       "6.0              4654.472222       2089.266667"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indep_test_metrics = ['Base indep tests', 'C-PC indep tests']\n",
    "data_all.groupby('n_clusters').mean()[indep_test_metrics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b546883bb0>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# colors = ['#A4C2F4', '#3333FF', '#4D68D5']\n",
    "colors = ['#9ABCE4', '#005293', '#5E94D4']\n",
    "data_all.groupby('n_clusters').mean()[indep_test_metrics].plot(kind='bar', figsize=(10, 6), color = colors)\n",
    "plt.title('Number of CI tests needed for different numbers of clusters', fontsize = 20)\n",
    "plt.xlabel('Number of clusters', fontsize=18)\n",
    "plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "label_names = ['PC #CI tests', 'Cluster PC #CI tests']\n",
    "plt.legend(label_names, bbox_to_anchor = (1.42, 1),  fontsize=16, loc='upper right')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Grouping by alpha"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_precision</th>\n",
       "      <th>cluster_adj_precision</th>\n",
       "      <th>base_adj_recall</th>\n",
       "      <th>cluster_adj_recall</th>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <th>cluster_adj_f1_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>alpha</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.01</th>\n",
       "      <td>0.916112</td>\n",
       "      <td>0.902991</td>\n",
       "      <td>0.429894</td>\n",
       "      <td>0.524422</td>\n",
       "      <td>0.551538</td>\n",
       "      <td>0.640852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.05</th>\n",
       "      <td>0.904338</td>\n",
       "      <td>0.885166</td>\n",
       "      <td>0.445340</td>\n",
       "      <td>0.537709</td>\n",
       "      <td>0.559914</td>\n",
       "      <td>0.642846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.10</th>\n",
       "      <td>0.894794</td>\n",
       "      <td>0.866311</td>\n",
       "      <td>0.455244</td>\n",
       "      <td>0.556900</td>\n",
       "      <td>0.565258</td>\n",
       "      <td>0.650317</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.25</th>\n",
       "      <td>0.834739</td>\n",
       "      <td>0.792300</td>\n",
       "      <td>0.480221</td>\n",
       "      <td>0.586292</td>\n",
       "      <td>0.563331</td>\n",
       "      <td>0.636353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.50</th>\n",
       "      <td>0.772943</td>\n",
       "      <td>0.714490</td>\n",
       "      <td>0.509988</td>\n",
       "      <td>0.604982</td>\n",
       "      <td>0.555869</td>\n",
       "      <td>0.603745</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       base_adj_precision  cluster_adj_precision  base_adj_recall  \\\n",
       "alpha                                                               \n",
       "0.01             0.916112               0.902991         0.429894   \n",
       "0.05             0.904338               0.885166         0.445340   \n",
       "0.10             0.894794               0.866311         0.455244   \n",
       "0.25             0.834739               0.792300         0.480221   \n",
       "0.50             0.772943               0.714490         0.509988   \n",
       "\n",
       "       cluster_adj_recall  base_adj_f1_score  cluster_adj_f1_score  \n",
       "alpha                                                               \n",
       "0.01             0.524422           0.551538              0.640852  \n",
       "0.05             0.537709           0.559914              0.642846  \n",
       "0.10             0.556900           0.565258              0.650317  \n",
       "0.25             0.586292           0.563331              0.636353  \n",
       "0.50             0.604982           0.555869              0.603745  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('alpha').mean()[prec_recall_metrics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b546881a80>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(10, 6))\n",
    "# colors = ['#9ABCE4', '#005293', '#C7D97D', '#7D922A', '#FAD080', '#F7811E'] #TUM COLORS\n",
    "ax.plot(data.groupby('alpha').mean()['base_adj_precision'], data.groupby('alpha').mean()['base_adj_recall'], color = '#F7811E')\n",
    "ax.plot(data.groupby('alpha').mean()['cluster_adj_precision'], data.groupby('alpha').mean()['cluster_adj_recall'], color = '#005293')\n",
    "ax.set_title('Precision-Recall Curve for different alpha values', fontsize = 20)\n",
    "ax.set_xlabel('Precision', fontsize=18)\n",
    "ax.set_ylabel('Recall', fontsize=18)\n",
    "ax.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "ax.legend(['PC', 'Cluster PC'], fontsize=16)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b546c83130>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# colors = ['#A4C2F4', '#3333FF', '#64d2a6', '#065535', '#e49645', '#b45f06']\n",
    "colors = ['#9ABCE4', '#005293', '#C7D97D', '#7D922A', '#FAD080', '#F7811E'] #TUM COLORS\n",
    "data.groupby('alpha').mean()[prec_recall_metrics].plot(kind='bar', figsize=(15, 8), color = colors)\n",
    "plt.title('Precision, Recall and F1-Score for different alpha', fontsize = 20)\n",
    "plt.xlabel('alpha', fontsize=18)\n",
    "plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "label_names = ['PC precision', 'Cluster PC precision', 'PC recall', 'Cluster PC recall', 'PC F1-score', 'Cluster PC F1-score']\n",
    "plt.legend(label_names, bbox_to_anchor = (1.3, 1),  fontsize=16, loc='upper right')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_base_shd</th>\n",
       "      <th>cluster_cluster_shd</th>\n",
       "      <th>empty_graph_shd</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>alpha</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.01</th>\n",
       "      <td>40.061111</td>\n",
       "      <td>33.483333</td>\n",
       "      <td>50.472222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.05</th>\n",
       "      <td>39.561111</td>\n",
       "      <td>33.772222</td>\n",
       "      <td>50.855556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.10</th>\n",
       "      <td>39.772222</td>\n",
       "      <td>33.088889</td>\n",
       "      <td>50.838889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.25</th>\n",
       "      <td>40.022222</td>\n",
       "      <td>33.855556</td>\n",
       "      <td>50.072222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.50</th>\n",
       "      <td>41.944444</td>\n",
       "      <td>37.266667</td>\n",
       "      <td>50.688889</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       base_base_shd  cluster_cluster_shd  empty_graph_shd\n",
       "alpha                                                     \n",
       "0.01       40.061111            33.483333        50.472222\n",
       "0.05       39.561111            33.772222        50.855556\n",
       "0.10       39.772222            33.088889        50.838889\n",
       "0.25       40.022222            33.855556        50.072222\n",
       "0.50       41.944444            37.266667        50.688889"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('alpha').mean()[dist_metrics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b546dc6410>"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# colors = ['#A4C2F4', '#3333FF', '#4D68D5']\n",
    "colors = ['#9ABCE4', '#005293', '#5E94D4']\n",
    "data.groupby('alpha').mean()[dist_metrics].plot(kind='bar', figsize=(10, 6), color = colors)\n",
    "plt.title('SHD for different alpha', fontsize = 20)\n",
    "plt.xlabel('alpha', fontsize=18)\n",
    "plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "label_names = ['PC SHD', 'Cluster PC SHD', 'Empty graph SHD']\n",
    "plt.legend(label_names, bbox_to_anchor = (1.42, 1),  fontsize=16, loc='upper right')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Grouping by distribution type\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_precision</th>\n",
       "      <th>cluster_adj_precision</th>\n",
       "      <th>base_adj_recall</th>\n",
       "      <th>cluster_adj_recall</th>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <th>cluster_adj_f1_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>distribution_type</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>exp</th>\n",
       "      <td>0.857808</td>\n",
       "      <td>0.828024</td>\n",
       "      <td>0.464721</td>\n",
       "      <td>0.564380</td>\n",
       "      <td>0.557131</td>\n",
       "      <td>0.635359</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gauss</th>\n",
       "      <td>0.869965</td>\n",
       "      <td>0.837394</td>\n",
       "      <td>0.465413</td>\n",
       "      <td>0.564434</td>\n",
       "      <td>0.562328</td>\n",
       "      <td>0.638352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gumbel</th>\n",
       "      <td>0.865983</td>\n",
       "      <td>0.831336</td>\n",
       "      <td>0.462278</td>\n",
       "      <td>0.557369</td>\n",
       "      <td>0.558088</td>\n",
       "      <td>0.630756</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   base_adj_precision  cluster_adj_precision  base_adj_recall  \\\n",
       "distribution_type                                                               \n",
       "exp                          0.857808               0.828024         0.464721   \n",
       "gauss                        0.869965               0.837394         0.465413   \n",
       "gumbel                       0.865983               0.831336         0.462278   \n",
       "\n",
       "                   cluster_adj_recall  base_adj_f1_score  cluster_adj_f1_score  \n",
       "distribution_type                                                               \n",
       "exp                          0.564380           0.557131              0.635359  \n",
       "gauss                        0.564434           0.562328              0.638352  \n",
       "gumbel                       0.557369           0.558088              0.630756  "
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('distribution_type').mean()[prec_recall_metrics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b546d139a0>"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABGMAAAICCAYAAABr1QDBAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAB4/ElEQVR4nO3deXxU1f3/8fcnewJJgIQ1LEEKoiCRRdFUERRxY1FoLai4VUXiThW0WEWtWKsCUqQIPxXBDb5VxIVNXHCtLVipG4gIiKxhyUr2nN8fd4YOISEJSSaAr+fjMY8w9557z7l3Fmbec8655pwTAAAAAAAAgiOkvhsAAAAAAADwS0IYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxwFHOzPqamTOzCYe5/dW+7a+u3ZahPBWdbzPbaGYb66dV1Wdms33HkVzfbcGhmVlHM1tgZtt9j1lGfbcpUEXvYWb2gZm5empWuc9xM0v2LZtdX+3ytaNezw0AAKg5whigAr4P3IG3EjPbZWbvmdll9d2+X6oKHpc9vi8nV5uZ1XcbcbCAL7YV3T4oU/73Zva0mX1uZvt8Zf5cg/qjzOxO3/4yzazQzLaZ2Sozm2ZmZ9X4II9AZhYq6XVJF0p6S9IDkv5Sn20KlqMt4Ax0NIadR/P5BgCgPoTVdwOAo8ADvr/hkjpLGiKpn5n1cs6Nqb9m7fcvSSdI2nWY2y+Q9E9J22qtRcER+Lj8StIlks6S1EvSzfXVKFRqoaQvy1m+scz9JyTFS9oraaukDodboZk1lLRCUg9J2yW96vvbUFKKpBskNfKVOda0l3SipFnOuRvquzHVdKWkmHqs/x55wdWWemxDRer73AAAgBoijAEq4ZybEHjfzM6R9I6k281sqnNuY320y885t0/Smhpsnykps/ZaFBzlPC6/lvShpDQze8I5t6FeGobKvO6cm12FcsMlfeec2+Qb0vVcDeq8XV4Qs0zSIOdcYeBKM2ssL9A8FrXy/d1ar604DM65n+q5/m06QkPq+j43AACg5himBFSTc+5deeGHSTpFksxsgq9LeV8zu8w3FCInsMu2mcWY2T1m9qWZ5frWf2ZmIyqqy8wGmNmbZrbTzArMbLOZLTSz/gFlKppv4Tgzm2lmP5hZnm8oz1dmNsPMEgLKVThnjJn1NLNXA+rfZGbTzaxlOWX3d6s3s1G+uvLNbIevHfHVOc/V5Zz7RP97XHqW077eZvYP8+bNKPSdy6fNrNVBO/PKNzGzh83sa98wmUwzW21mfzGzBgHleprZk751e3zHvM7MnvB9ya8TZnaxmb1gZt/7nk+5viE3t5rZQe/th/v4mFl/M/vIt/89Zva6mXWuq+Pyc84tcc5tqqXdpfr+/r1sEOOra69z7tOyy80s1MxuNLNPfI9/nu/19P/MrGOZsvFm9oiZrfWd171mtjTwtRpQdv9r1sxONbO3fee27PwkI8zsfTPL8O3zOzO718wiq3LQ5s0p4u/tc7/9b0jYhLps9yHa09zMnvE95/LMey+86hDlD5oXxTxXmdmnZpbua/NmX5t/F9hOSe0ktbMDh8PNDjw/vjpa+B7TLeYNe7zat/6QQ4XMrLPv9bDH9/r42MwGlFNu//8P5aw7aA4aX9v952VDQNs3Hurc+JaH+J6z/zbv/5hc379HV/C+4D8Hib73gW3mvdd/Y2bXlHfc5ezjkOfbzBqb9x663qz8YaTm/T/nzKxX2fNS1fMcsK8qv27M7Exf3T/7jnu7mf3TzO6vyrEDAFAT9IwBDo//A2XZD8N/kHSupDclvS9vmIXMrJGk9yR1l/SFpGflhaHnSXrJzLo45+49oAKzByTdJylH3pwPm+X9yp0q6QpJyytsnBeW/FtSnKRF8oZlRMkbsjBS0jRJuw95gGYDfduZpH9I2iQv5BgtaYiZnVFB75O/+o7rTXk9EfpJul7eUKKzD1VnLSoKvGNm10qaKalA0hvyzmVHSddJGmRmpwX+0mxm7eU9fu0krZL0d3mPVydJd0iaISnXV/x6eUOkVsh7TELknacxki4ws97Ouew6OMa/SCqV9Lm8YRTx8s7vk/JCwpEVbFflx8fMfiNpnqRC399tks6Q9Jmk/9bq0dQt/3O9U1U3MLMIeXOsnCvv+fKSpCxJyfIe748lrfOVbSTpE3nDgf4taYqkREmXSlpmZqOdc0+XU83p8obCfCzvPSFR3rmWmT0r6RpJP8t7HWZIOk3SQ5LOMbNznXPFlRzGA772XiXv+fmBb/kHddXuiphZoqRPJR3n2+5jSS3lvZaWVXIcgR721b1B0nx5vfpaynvO/1be83SjvGO/3bfNlIDtvyyzvybyhmnmSHpN3mtqRxXa0V7e6+ArSU/72vA7SYvN7DLn3LxqHFNZD0i6WN4QuiflPfYK+HsocyVdJu85+//k/R91iaTp8l67l5ezTSN5z4NCee/1kfLO5bNmVuqce76SOjfqEOfbObfXzF6R93zuL69n6X5m1kbSBZJWOedWltl3tc5zdV43Zna+pLflva7fkPc+2kReL7k0/W8oLAAAdcM5x40bt3Ju8j7EunKW95f3gb1UUjvfsgm+8rmSupezzWzf+rFllkdJWuLb18kBywf4yv8oKamc/bUO+HdfX9kJActu8S27rZxtG0iKDrh/ta/s1QHLGsr7Alsi6cwy24/zlV9WwTH+JKltwPIwecOHnKRT6/Bx6eNrb4GklgHLO8n7kvFD2XMp6RzfNgvKLP/UV8895dSTKCkq4H47SaHllPu9bx/jyiw/6Hz7lm+UtLEa56FDOctCJD3v23/vmjw+Ac+BIkm9yuxrsv9xkJRcxfb6639d3uul7K3RIbb1n7M/H+ZzZqBv+wJ5X0ovCnyOVLDNRN82b0iKLLMuUlLTgPtP+8o+LckClneUFxYUBJ4n/e816ySNOsTxvqaA16pv3QRV8Nqu4Dj8dU0oZ12ttruSdsz0bTe5zPJevufYQW2UFxq5Mst2y/uiHVNOHYnVeU0FHMscSWGHeM4GnoPkgO0eq+BY9kqKK+cx61tOHf79za6s7iqcmxG+bb6Q1DBgeQNJK33rLqvgHPw/BbyPyQvoiiV9W43HuMLz7Ts3TtI/ylnnPz/X1/A8X61qvG7khTVOUkplzyVu3Lhx48atLm4MUwIq4etiPsG8ISv/kBeemKQp7uBhFDOdc/8ps32CvJ4sK51zfw1c55zLlxdumLxfM/1u8f39g3PuoMkjnXM/V7H5eeVsm+ucO2h5GUPk/UI4zzn3UZl1T8j70H2umbUtZ9sHXUAvE+f9Cvmc7+6pVWx3pco8LvPk9UoxSXc6b64Hv9HyJvm9rey5dN6Qszfk9Y6J9e23p7xf/r+U9GjZep1zu3yPm//+JudcSTlNfFbeL67n1eAwK+ScW1/OslJ5v6TrEPVW9fHxPwdecgf/Wj1Bhz/P0BBJ95dza3SY+6uUc+4tSbfJez2MltfjZatvSMaLZtYnsLx5VyBK85W/0TlXUGZ/Bc65dF/ZCHmv7xx54Z0LKLdO0lRJEfImXC3rS1d+z5Pb5H0Rvrac1+pD8gKJ8no4VFkdtbuiusJ97c2W99zZz/fcerGazS+SF6IewDl3OJOYF8p7z6isl1FZmZIeLFO//1gayeuNEmzX+v7e7ZzLCWhXrrz/ZySvN2BZ+ySNCXwfc859K6+3zAnmTYBdI75zs1Jer8oW/uW+19rv5T03Xi5n0+qc58N93ZT3/+ThTogPAECVMUwJqNz9vr9OXpfnjyQ945x7oZyy/ypn2SmSQiUdNK+LT7jvb+AEoqf56ltyGO2VvIBhoqSnzOw8SUvlfbD+NvBL1yH08P19r+wK51yxmX0o75fL7vJ6WgQq+8Vd8rrMS1JtzqFyf5n7TtLvnXPPlVl+uu/vWWZ2Sjn7aSbv8ekkb0jSab7lS33hxiH5vmiOkjfh7InyhgsFBt1Jle3jcPhCvrvkXbL4OHm/fgeqqN6qPj7+58BBVxhyzmWa2Zfyrl5VXde4qk3gW2VmdrK8YR2BMpxzU/x3nHNTzez/yRt2lCrvuZsqLwS9zMwecs7d5yveWd7j+LlzrrKJb4+Xd1WbT5xze8pZ/56ke331lXXQ+4WZxcgbnrJL3iTh5dVZoJpPOFyr7a5EZ19dHzlvwvCyPtD/5kipzIvywupvzWy+vOfnZxXstyo2Oud2HsZ2X7jyhx9+IO9YusvrpRZMPeT1svygnHUr5AVY5T2e65xzWeUsD3xfyClnfXVNlxdSXyvv/yfJe/9qLW8+p/LqqNJ5PszXzYuShkr63Bfovy/v9VDVHzsAAKgRwhigEs65cj/VVWB7Ocv8k+We4rtVJPDXx0aS9lahB0u5nHcFmlPl/Qp9vrwPnJK02cwed85NrWQX8b6/FV1JxL+8UTnrMspZ5v/VObSSeqvM/7iYN5nu6ZKekTTDzDY55wJDJP/5v6uSXfrPfyPf36peznaevF9nf5R32ebt8j70S94cClWabLU6fHN9/FvefAr/kjfMYo+889xI3i/EFdWbUc6y8h4f/3Ogovkzynuu15eTdXA4t0kHzl0h5115bKHv5u8dcr283kR/MrPXnHNfqnrPgZq8Vso7h43l9fBqqoOPqTbVdrurUldtPJfukPdau0bS3b5bsZktkteT8Idqtu1wn8eVHUt8BevrUrykPa78SaqLzWyXvPC5rIwK9lfb79uvyOtZeb2Z/cUXdvsvt15RT6uqnudqv26cc6/55kb7g7yAaJQkmdkqeb3F3jnU9gAA1BTDlIDaVV6vE/8vtpOdc3aIW7+AbTIkNTaz6MNuiHPfOed+Jy+M6CXvS0uIpCfN7PeVbO5vc4sK1rcsU67e+IZdLZc0SN6XBv+vpH7+NsZXcv79PUAyfH8r7dHiu/LHJfKGSB3vnLvGOXeP8y67/aC8YR514Tp5QcwDzrnezrk059y9vnprMnFoIP95a17B+oqeG0HnnJtdzuOZXIXtCp1zT+l/wyP8Exhn+P5WpVdTTV4rh3q/+E8lz9fqhMTlqe12V6WuGj+XnHMlzrkpzrkU3/6GSVogabCkJeVdMaeyXVazvF9lxxJ43vw97Mr7AazRYdZfnkxJTXy99Q5gZmHy5rsqrwdMUPh+XJgtr1flgICJez93zq2uYLOqnufDet045952zp0tL8w5R958WF0kvWVmJx7GYQIAUGWEMUDd+5e8D+NnVmObf8r7le/8mlbunCt2zq1yzj0qb4JH6eAhHWX5573pW3aF70O9/1i+qGn7aotz7r+SZsnr8n5HwKp/+v5W9fz7y59n5VwKtoxf+f6+Uc6cE6dKOuwwrYr1vlrOusMZOlQe/2N70P7Muwz2ybVUz5HAPwzC/0VtjbxApptVcOnzAGvlzbmR4uuxVJY/ZK3Sa8U3VOMbSV3MrElVtjlMtdruSqzx1XWylX8J9b6Hs1Pn3E7n3GvOuUvlDavqIKlrQJES1WJvvDJ6+OeZKqOv72/g3GF7fX/blFO+VwX798/fUp32/0fe57o+5azr49tXXb5nV+V8/12+CaDlzRUTqop7xUhVPM81fd34Qv33nHNj5A2hipAXFAEAUGcIY4A65puP4EVJvczsT74JCw9gZh18l1P2+5vv7xNmdtCv8+UtK7O+ZwVfevy/Mu6rpNmvyxv2MsLMTiuz7nZ5vTKWu4CJYA+HmV1tZs7MZtdkPwH+LG+I0J1m5p//ZJq8CT8nm9lBlzY2swgz2x/UOOdWybua0sn636SXgeUTzCzKd3ej72/fMmWaSXqqJgdSiYrq7S7vsr+1YaG8L5GX+XoABZqg+hmGcVjM7MZynsf+dZ3lXcZX8q4qJd9EptPlhWkzyva28D1nmvrKFsp7fcfKmyQ0sFwHSbfKe/7NrUaTJ8n7MvhseUGJmTU2sx4HbVUNddTuiuoqCqhrQpm6eqmKkxGbWaSZ/bqc5eHyJpuWDnxv2y2paU16GB5CvKT7AhcEHEumvN46fv45dq7xhdn+8m3K7iOA/3Ls5U2SXpFnfX8fCewd6Pv3X3x3n6nG/qqr0vPtvMmh35V3hbMb5YWerxxin9U5z9V63ZhZn8DHI0BV/58EAKBGmDMGCI6b5V0u9kFJI83sY3lj4VvJm1DwFHm9VjZIknNumZn9Wd4Emt+Z2evyJlNsLukMeb03rj5EfSMljfLVs17el+oO8obyFKjMXBplOedyzOxaSf8naYWZ/Z+8iXp7yrvs9nb5xtfXkD8Qru6VTMrlnNtiZjPkzZkyVt64/zW+Y3lW0jdmtkTS9/ImTm4rr8dMurxJRv2ukDdB5EQzG+b7t8l7DAf4ym6UN2/LJ5KGmtmnkj6W9xhdIK/nQWWTvx6uOfLmwJliZv0krfO1baC8y7r+rqYV+J4DN8gb9vSRb4LLbfKef13lBRfl/QJfK8zsOl9d0v96Ag0ys9a+f69xzv3l4C3Ldb6kv5vZRnmP12Z5c+p0lHfVqXBJU51z/w7Y5gFJveW9Zr43s7fk9aBpI+85cJe8IReSNwTwTEk3+yaJfl/ekJBL5QUQNzvnNlSxrXLOPWveVb3SJK03s6XyXn9N5AWhfeRdAevGqu6zArXa7kr8Ud4wkNt9X6Y/ljcU6neSFskbZlSZaEkfm9kP8ibb3iQpSt6kzCfI66H2XUD5d+W9ty4xb9LxAkmrnXNv1sLxfCjpOjPrLe855T+WEHmX/d4/HMg597mv/j6S/mVm78l7nxgkb3L18nrMvCvvOTbLzF6V99zLcM5Nq6hBzrmXzGyIvMfvG9//G05eT8j28q6OV90rV1VHVc/3dEn95Z2Dv1UyN1p1znN1XzdTJSWZ2Sfy3s8L5f0fd7a859ahQiIAAGrOHQHX1+bG7Ui8yfsQ66pYdoKvfN9DlImQF8p8Ku8XvQJ5HxTfldfbJKGcbS6Ud0WlPb7ym+X9Enh2QJm+vronBCzrLa87+GrftnmSfpD3QbRrmTqu9m1/dTn1n+KrL13eB9WffPttVU7Z2b79JJez7qA2+pZP9i3vX1uPi7wP+Lm+W/OA5Sf52rjJdy73SPpaXhf5s8vZT4K8S1uvlZQv7xfcLyU9LCkmoFwTeV8uNvrKrZfXzT3Gt2xjVc53eWUrOQ8nyrtq1k7fsa6SN5dMsm//s2v6+PjWnSvvi/M+eaHeQnlhVIX7q6C9/vIHPc8qKV/R7YNqnKtO8ibpXCzvdZCr/73+XpM0sILtwuS9Zv8l72oyufKCr5mSflWmbCPf82Wdb98Zkt6RNKA657tMuYHyLsO9U97rb7uvLX+W1LmKx37Iuuqi3YdoSwt5oWi6vPekL32vh3L3Ky8EdQH3w+WFrIt9j12+b1//lPcFO6LM9g3kvV/9LC/wPeB1UdnzqLznuAJeX/ICIH8Psn3ywoLzDnGeZ/keywJ57z03qILXq2+bMZK+85V3Cnh/KHtuApaHyAsjVvratE/ee8NNkkLKKV/hOSjv+Ct5fA95vgPKhfoeNyepSwX7OqzzXJ3XjbzQ6mV5z/0cefPpfC3vPb7p4TzHuXHjxo0bt+rczLnDnbsOAGrGzL6QVOSc613fbQEA1D0zO05eKPqJc67cubzMLFleT9HnnXNXB691AAAED8OUANQL35w2KfKuhgIA+GW4U96wzwqHXAEA8EtAGAOgXjjnMlV3VzoBABwhzKytpMvkzdN0jbwhtP9Xr40CAKCeEcYAAACgLh0n6RF5c768I2m0c660fpsEAED9Ys4YAAAAAACAIAqpvAgAAAAAAABqyxE3TCkxMdElJyfXdzMAAAAAoF6sWrVql3OuaX23A0DdOeLCmOTkZK1cubK+mwEAAAAA9cLMNtV3GwDULYYpAQAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEB1xl7YGAAAAANS9VatWdQwLC/ujmaU45xqJH+uB2lBqZtuLi4sf6NGjx9KKChHGAAAAAMAvzKpVqy6IjIyc2qJFC8XFxeWGh4fvMbP6bhZw1CstLbW8vLz4jRs3Tvviiy9uriiQIfkEAAAAgF+Y8PDwscnJyUWJiYmZERERxQQxQO0ICQlxDRo0yEtOTi4MCwu7v8JywWwUAAAAAKD+OeeSGzRosK++2wEcq6Kjo/Odcy0qWk8YAwAAAAC/PEZvGKDuhISEOB0icyGMAQAAAAAACCLCGAAAAAAAgCDiakoAAAAAgP1mvbulZ33Wf/05SasOZ7upU6cm3Hbbbcn++zExMaWtW7cuuOqqq9Lvuuuu9PDw8P1lCwoKbPLkyYnz589PWLduXVRBQUFI06ZNi1JTU7Nvu+22nWecccZRMZ+O/5jXrFnz1fHHH19YlW2GDRuW/M9//jN2y5YtX9V1+1AxwhgAAAAAwDHj2Wef/bFt27aFmZmZofPnz288fvz4tjt37gyfMmXKVknKysoKOfvsszt+/fXXDS6//PL0sWPHbouNjS35/vvvo1555ZUmF154YaesrKwv6/kwquQ3v/lNZpcuXda0bdu2qKrbPPjgg9syMjJ21GW7UDnCGAAAAADAMeOUU07Z17Vr1wJJGjp0aNaGDRsin3nmmWb+MOb6669v89///rfBokWL1vbv3z/Xv91FF12Uc8cdd+yaM2dOo7pqW15enkVHR7va2l+rVq2KW7VqVVydbbp06VJQW/Xj8DFnDAAAAADgmNW9e/d9OTk5oVu2bAnbtGlT+KuvvpowfPjwXYFBTKArr7wy41D7GzZsWHLz5s27vfPOOw26du16QmRkZI+kpKSTHn744WaB5aZOnZpgZj0XL17c8IILLjguNjb25O7du58gSUVFRbrnnntatG/fvktERESPZs2adbv++utb79u374BLXGVlZYWkpaUltWnTpmtERESPxMTElPPOO6/D5s2bwwLrWLt2bYR/mxkzZjQ54YQTToyJienesGHD7p06dTrxscceSwxsf1JS0kmB9WzatCn8kksuSW7cuHFKREREj06dOp04ffr0JuUdz7vvvttg8ODB7Rs2bNi9WbNm3a6++uo2ZduNytEzBgAAAABwzNq0aVNEaGio4uLiSl988cVGJSUldvHFF2fUZJ+5ubmhI0eO7HDrrbdu69SpU8HLL7/c5N57720TGxtbcuutt+4OLHvNNdccd/HFF+8ZNWrU+uLiYpOkSy655Lh33303/qabbtp+xhln5HzzzTfRf/nLX1r99NNPkUuXLl0vSfn5+XbWWWd1WrNmTfTNN9+8PTU1NTcjIyN06dKlcbt27Qpr06bNQT1ili5d2jAtLa391VdfvfORRx7ZXFpaat9++21URkZGhd/9s7KyQs4666zjMzMzQ8ePH7+lbdu2hS+88ELCTTfd1H7fvn0hd955567A8tdee237Sy65ZM+11177wyeffNJw0qRJrRo3blwyefLkrTU5p780hDEAAAAAgGNGSUmJioqKlJGRETp79uzGy5Yta9yvX7+M2NjY0s2bN0dIUocOHWo0VCc3Nzdk0qRJG2+44Ya9kvSb3/wmKzU1Nfwvf/lLq5tvvnl3SMj/BqFcdNFFe2fMmPGz//6SJUsavv32243/9re/bbz55pt3S9LFF1+c3aRJk+K0tLT2n376aXRqamrejBkzmnz55ZcNXnjhhR8uv/zyTP/211xzzd6K2vXJJ580iI2NLXn22Wc3+5cNHTo061DHMm3atIRNmzZFvvnmm98PHDgwW5IuvfTSrNTU1PCJEycm3X777bvCwv4XHQwdOnSPP3i5+OKLs1euXNlgwYIFTQhjqocwBgB+wWa9u6XO9n39OUl1tm8AAICKnHzyyV39/w4JCdHgwYN3z5gxY/Ohtqmu0NBQXXXVVRmBy37729/uHTNmTLsNGzaEd+jQYf+Eur/5zW8OCE/efvvt+PDwcHfVVVftLSr637y7Q4YMyUpLS9N7770Xm5qamrd8+fL4xMTEosAgpjK9e/fOzcrKCh0yZEj74cOH7zn33HNzEhMTSw61zccffxzbrFmzIn8Q4zd8+PDdt912W/IXX3wRfeqpp+b5lw8ePPiA4z7xxBPzPv3007iqthEewhgAAAAAwDFjzpw569u1a1cYHx9f0rFjx8KYmJj9E+a2adOmUJLWr18fmZKScti9Y2JjY4sjIyMPmIi3RYsWRZI3LCowjGnduvUBVzpKT08PKyoqsri4uO7l7Xv37t1hkrRnz57Q5s2bV/kqSZI3CfGzzz774/Tp05uNHDmygySdeuqpOZMnT97cu3fvvPK2ycjICGvatOlB9bRq1arI197QwOVNmzY9INyJjIx0hYWFzBlTTYQxAAAAqFX0ugNQn7p3757nv5pSWeeff352aGioFi5cGF/Z8J1Dyc7ODisoKLDAQGb79u3hktSuXbvCwLJmB+YUTZo0KY6MjHTLli1bU96+/ZepTkhIKF63bl10ddt2zTXX7L3mmmv2ZmZmhixatCj2T3/6U+tBgwZ13LZt239DQ0MPKt+oUaPiH3/8Mars8q1bt4ZLB4cvqB1cTQkAAAAA8IuQnJxcNGzYsF0vv/xy0+XLlzcor8zcuXMbVbafkpISPf/88weU+7//+7/GLVu2LGzfvv0he7NceOGFWQUFBbZ3797QPn367Ct7S05OLpKk/v37Z+3atSv8pZdeiq/6Ef5PfHx86YgRIzKvvvrq9PT09PAdO3aU2xnjzDPPzN6xY0f4smXLDjgf8+bNa9KkSZPiHj16lNujBjVDzxgAAAAAwC/GzJkzN69fvz5q0KBBnS6//PL0c889NzsuLq5k3bp1ka+88kqTr7/+usHIkSO/PNQ+GjRoUHr//fe33rVrV9jxxx9f8NJLLzX57LPP4qZOnboxcPLe8gwcODB74MCBe0aOHNlh1KhRO0477bTckJAQ/fjjjxFLliyJf+KJJ37u1q1bwY033rhn9uzZTa+77rrjVq9eve3000/PzcrKCl2yZEncXXfdtbN79+75Zfd9++23t9q5c2d43759s9q0aVO0adOmiJkzZzbr3LlzXqtWrQ66+pIk3XTTTbuffvrp5iNGjPhV4NWUPv3007jHHntsU+Dkvag9nNWjBN19AQAAAATD9eckrarvNtSl+Pj40k8++eT7SZMmJc6fPz/hlVdeaVpQUGDNmjUrOuOMM7ImT578c2X7aNCgQcmcOXN+vOOOO9quW7cuOiEhoeihhx7afMstt+yubFtJev311zdMnDix2QsvvJA4derUlhEREaWtWrUq7NevX1ZSUlKx5M3F8sEHH3w/duzYVnPnzm06adKkVo0aNSru2bNnTrNmzcoNVk477bTcadOmNRs/fnybzMzMsCZNmhT36dMn869//WuFVzqKi4srXbFixdrbbrut9UMPPZSUm5sbmpycnP/UU09tSEtL21OV40H1mXOu8lJB1KtXL7dy5cr6bsYRhzAGQF3gvQVAXeC9BagZM1vlnOtVl3WsXr16Y0pKyq66rONYNWzYsOSPP/44bseOHf+t77bgyLZ69erElJSU5PLWMWcMAAAAAABAEBHGAAAAAAAABBFzxgAAAAAAUEWvvvrqxvpuA45+9IwBAAAAAAAIIsIYAAAAAACAICKMAQAAAAAACCLCGAAAAAAAgCAijAEAAAAAAAgirqYEAKgTNmxane7fvXpzne4fAAAAqCv0jAEAAAAAAAgiesYAAAAAAPazYdN61mf97tWbV9Vk++XLlzeYNGlS85UrVzbcu3dvWIMGDUpPPPHE3BEjRuwZPXr07rCwME2dOjXhtttuS16zZs1Xxx9/fGFttV2S3nrrrdj33nsv9rHHHtsaGhpam7uusL5BgwZ18t+PjIx0SUlJBZdccsmeCRMmbG/YsKHzrystLdXTTz/dZM6cOYnfffddTE5OTkhCQkLxKaeckjNq1Kj0QYMGZdd5g2uB/5jffPPN7wcOHFilNo8ZM6bV5MmTWzrnavT8qi2EMQAAADhqMAQSwKE8+OCDzSZMmNDmtNNOy77//vt/Pu644wp3794dtnTp0rixY8e2bdy4cckVV1yRUZdteO+992InT57c8tFHHw1KGOP35z//efNpp52Wm5ubG7J48eK4yZMnt1q/fn3kggULNkpScXGxBg0adNw777zTeOjQobvS0tJ2JiYmFm/cuDHi1VdfbTxkyJBO6enpXyYkJJQErdGHKTU1NXf58uVrunfvnlfVbW666ab0iy66KLMu21UdhDEAAAAAgKPe4sWLG06YMKHNlVdeuXP27NmbA9ddccUVGXffffeOnJyco3Kqjry8PIuOjnaHKtOlS5e8c845J1eSBg8enJ2enh7+6quvJuzYsWNz8+bNS/74xz+2XLJkSePnnntu/dVXX50RuO3o0aP3vPbaa3ERERGHrKMu218dTZo0KfUfa1V16NChqEOHDkW11YaaOiqfiKhdNmxand4AAAAAoK49+uijLeLj44unT5/+c3nru3TpUtC7d+8Ke1KYWc8xY8a0Cly2du3aCDPrOXXq1AT/shUrVsSkpqZ2bNSo0clRUVE9WrdufdIVV1zRVvrfUBhJioiI6GlmPc1s/7Cv7OzskNGjRyclJSWdFB4e3iMpKemkcePGtSgp+V9nlLfeeivWzHo+//zzjYYPH96ucePGKc2aNUup7vno1atXriR9++23kfn5+TZjxozmffv2zSwbxPgNHTo0KzY2trSi/Y0ZM6aVmfX817/+Fd27d+9O0dHR3Zs2bdrt9ttvb1Wd9j/++OOJxx9//ImRkZE9GjdunHLppZe227FjxwFdiIqKijR+/PgWHTp06OIvd+aZZ3b8z3/+ExVYx1tvvRXr3+bVV1+N6969e+fY2NiTY2JiuicnJ3e98847W5Ztf2A9e/bsCbnyyivbNmvWrFtERESP5OTkrg888ECz0tL/nQZ/XS+++GL8lVde2bZx48YpjRs3ThkyZEj7Xbt2HXbXJ3rGAAAAAACOasXFxfr8889j+/fvnxETE1MnvTskKTMzM2Tw4MGdunXrljt9+vQNcXFxpT/++GPEZ5991lDyhsJs2bIlfP78+YlLly5dEzhMqaioSH379u24fv366DFjxmxNSUnJ+/TTTxtMmTKl1Z49e8JmzZp1QIh01113te3Xr1/mrFmzNuTl5VW7I8WGDRsiJSkhIaHko48+isnOzg696KKLMmp2BqRhw4Z1uOyyy3aNGzdu++LFi+OefPLJliEhIZo0adLWytqflpaWNHPmzObXXHPNzkceeeTnzZs3hz/88MNJ5557bvQXX3yxJizMiygGDRp03PLlyxtfe+21OwYMGJCVl5cXsmLFioY///xzePfu3fPLtunbb7+NGDFixK/OP//8vePHj98aERHh1q5dG/njjz9GVnQcJSUlOvfcczt+++23MWPHjt2akpKS9+abb8ZPmDChTXp6evi0adO2BJYfO3Zs2/79+2c888wzG7777ruohx56qPUNN9zgXnvttY2Hcx4JYwAAAAAAR7Vt27aF5efnh7Rt27ZWJ+Mta/Xq1VFZWVmhjz/++M+BvWxuvfXW3ZI3FCYpKalIkvr165cbHh6+f9uZM2c2+eKLLxouWrRo7QUXXJAjSUOGDMmWpEmTJrWaMGHC9qSkpGJ/+ZSUlNx58+ZtqmrbSktLraioSNnZ2SELFy6Mnzt3btPOnTvndevWrWDWrFkxktS+ffsan5+RI0fumjhx4nbJ602TnZ0d+vTTTzf/4x//uCMxMXF/F5my7V+7dm3E008/3eKOO+7Y+vjjj2/zLz/hhBPyzzvvvM4vv/xyo5EjR2a88cYbsUuXLm380EMPbb733nt3BtSbUVGbPv/88wZFRUU2e/bsTU2aNPF3aznkxL7z58+P/+KLLxo++eSTG/2P39ChQ7P27dsXMnPmzObjx4/f0bJly/2PR+/evbOff/55//C3rLVr10bNmzcvsbS0dGNISPUHHTFMCQAAAACAKujSpUtBbGxsyahRo9pNnz69yQ8//BBe+VaepUuXxrdq1aqwf//+OUVFRfLfLrzwwqzi4mL74IMPGgSWHzJkSEZ12jZs2LCOERERPRMSErpfe+21x5122mnZCxcu/KE6+6iKK664Yk/g/REjRuzZt29fyKpVq6IDl5dt/1tvvRVXWlqqa665Zk/g8ffr1y+3QYMGpStWrGgoSUuWLIkzM91+++3pVW3TKaecsi8sLMxdcsklxz333HONt2zZUmnHkxUrVsSGhITohhtuOOB4Ro4cuaeoqMjef//9Ax6PCy+88IDJf0866aR9hYWF9vPPPx9WJxfCGAAAAADAUa1FixbFUVFRpT/99FNEXdaTkJBQsmTJkrXNmzcvGjt2bLuOHTt269ixY5fZs2c3qmzbXbt2hW3dujUiIiKiZ+Ctb9++J/jXB5b397CpqkceeeSnFStWfPfvf//7m8zMzP+89957P3Tq1KlQkpKTkwslacOGDTU+P61bty4OvN+qVasiSfrpp58OCKbKtn/nzp1hktS1a9euZc9Bbm5uyJ49e8Ikac+ePWHx8fHFgZfkrkzXrl0LXnvttXWlpaU2evTo9m3atElJSUnp/PbbbzesaJu9e/eGxsXFFUdFRR1Qj7/du3fvPuDxSExMPOC4IyMjnSQdzhAyiWFKAAAAAICjXHh4uE499dTsjz/+OO5wr9wTERHhCgsLLXCZP0AIlJqamrd06dL1RUVF+vDDDxtMnDixxe9///sOXbp0+eaUU045aD4TvyZNmpQkJSUVvvTSS+vLW9+xY8cDhhCZWbWOoXPnzvl9+vTZV966M888c19sbGzJokWLGt155527qrPfsn7++eewE088cX9bt27dGi5Jbdu2PSB8Kdt+/yWzX3vttXUJCQkHBBuS1KxZs2JfueLMzMywnJwcq04gM2jQoOxBgwZl5+Xl2TvvvNPwgQceaPXb3/624/r1678KHG7k17hx45KsrKyw/Px8CwxktmzZEu5vR1XrPhz0jAEAAAAAHPXuvvvu7ZmZmWFpaWmty1u/Zs2aiM8//zy6vHWS1LJly8Jvv/32gPULFy6Mr6h8eHi4zjnnnNyJEyduLS0t1VdffRUtSZGRkaWSVPYy2gMGDMjcvn17eGxsbGmfPn32lb2VFxjUlqioKHfjjTfueP/99+Mr6sWzYMGCuOzs7EozghdeeKFJ4P2XX365SUxMTGmvXr0qvFKVJF100UVZISEh2rhxY0R5x9+5c+dCSTr//POznHN68sknm1bjEPeLjo52gwcPzh4zZsz2vLy8kO+//77c3kB9+/bNLi0t1XPPPde47PGFh4e7fv36VevS2dVFzxgAAAAAwFHvggsuyJkwYcLmCRMmtFm7dm3UyJEjd7dv375w9+7docuXL4975ZVXEmfNmrWhostbX3LJJXumTp3acty4cS1SU1NzV6xYEfvaa6+VDR7iZ82a1XTw4MF7O3ToUJiTkxMybdq0Zg0aNCg966yzciSpS5cu+ZL00EMPtRg4cGBmWFiY69Onz75Ro0btmTt3buJ5553XKS0tbUf37t33FRQU2A8//BD59ttvN1qyZMn6Q11auqYmTpy47auvvoq+7rrrOixevHjXoEGDMhMTE4s3bdoUsWDBgkbLli1rnJ6e/mVl+5k7d25iaWmpevfuvW/x4sVx8+bNSxwzZsxWf8+XinTp0qVg9OjR2++55562a9eujerbt292dHR06aZNmyKWL18ed/311+/y924577zz9t5///2tN2/eHNG/f/+swsJCW7FiReygQYMyBw4ceNDEvH/961+bfvTRRw0vuOCCzHbt2hWmp6eHPf744y2bNm1a1LNnz3If79/+9reZjz32WM6dd97ZLj09Peykk07Kf+utt+LnzZuXeNNNN22vy3BMIowBAAAAAARwr968qr7bcLjuu+++naeffnrupEmTmt93332t9+7dG9agQYPSrl275j7xxBObRowYkVHRtg8//PC2jIyM0GeffbbZtGnTQvv27Zv53HPP/Xj22Wef4C9z4okn5kdHR5c+9thjrXbt2hUeExNT0q1bt9w33njj+w4dOhRJ0vDhwzOWLl2aPnv27KZTpkxp6ZyTc25VZGSkW7Fixff33ntvy+effz5x4sSJkdHR0aVt2rQpGDBgQGZUVFSdBTGSFBYWprfffvvHGTNmNJkzZ05iWlpa8r59+0ISEhKKTz311OxFixatrSxQkaTXXnvth5tvvrntlClTWjVs2LDk1ltv3fbXv/51W2XbSdK0adO2nHDCCXkzZ85s9vzzzzc1M7Vo0aLwjDPOyD7xxBP3D/F68803f7z33ntbzJs3L/GZZ55p1rBhw5Ju3brlpqWllTupb48ePfYtXbo07sEHH2ztn3OmV69eOS+++OKPFQ11Cg0N1TvvvLPu9ttvb/23v/2tRUZGRlhSUlLhhAkTNv/pT3/aWd42tcmcq9oQLDNLk3SXpJaSvpF0u3Puo0OUv0zSWEmdJGVJWi7pTufc9kPV06tXL7dy5cqqtf4XZNa7WyovdJhumL6gzvYtSe7Vm+t0/wAOH+8tAOoC7y1AzZjZKudcr7qsY/Xq1RtTUlJqNHcIflnGjBnTavLkyS0LCwtXBV6yGxVbvXp1YkpKSnJ566o0Z4yZ/U7Sk5ImSuou6VNJi82sbQXlfy1prqTnJXWRdLGkEyW9WM22AwAAAAAAHFOqOoHvGEmznXOznHPfOedukbRN0ugKyp8u6Wfn3GTn3Abn3D8l/U1S75o3GQAAAAAA4OhVaRhjZhGSekpaVmbVMkmpFWz2iaSWZjbIPImShktaVJPGAgAAAACA4Js0adJW5xxDlGpJVXrGJEoKlbSjzPIdklqUt4Fz7jN54cuLkgolpUsySVeVV97MbjCzlWa2Mj293Pl4AAAAAAAAjglVHaZULWZ2orxhSQ/J61Vzvrzg5unyyjvnZjrnejnnejVteliXEgcAAAAAADgqVOXS1rsklUhqXmZ5c0kVXRnpHkn/cs495rv/XzPLlfSRmf3ROffzYbUWAAAAAADgKFdpzxjnXKGkVZLOLbPqXHlXVSpPjLwAJ5D/fp30xgEAAAAAADgaVKVnjCRNkjTXzP4lb3LeGyW1kjRDksxsjiQ55670lX9T0iwzGy1pqaSWkqZI+sI591OttR4AAAAAAOAoU6Uwxjk3z8wSJN0rL1j5WtKFzrlNviJty5SfbWaxkm6W9ISkTEnvSRpXWw0HAAAAAAA4GlW1Z4ycc9MlTa9gXd9ylv1N3iS+AAAAAAAA8KlyGAMAAAAAOPb98/v7etZn/ad1enDV4Ww3derUhNtuuy3Zfz8mJqa0devWBVdddVX6XXfdlR4eHr6/bEFBgU2ePDlx/vz5CevWrYsqKCgIadq0aVFqamr2bbfdtvOMM87YV/MjqXunnnrq8ZL0r3/9a60kvfXWW7GDBg3q9Oabb34/cODA7PptHQ6FMAYAAAAAcMx49tlnf2zbtm1hZmZm6Pz58xuPHz++7c6dO8OnTJmyVZKysrJCzj777I5ff/11g8svvzx97Nix22JjY0u+//77qFdeeaXJhRde2CkrK+vLej4MHOMIYwAAAAAAx4xTTjllX9euXQskaejQoVkbNmyIfOaZZ5r5w5jrr7++zX//+98GixYtWtu/f/9c/3YXXXRRzh133LFrzpw5japbZ15enkVHR7taOwgc87jMNAAAAADgmNW9e/d9OTk5oVu2bAnbtGlT+KuvvpowfPjwXYFBTKArr7wy41D7GzZsWHLz5s27LV++vEH37t07R0VF9UhLS2stSVu3bg277LLL2jZr1qxbREREj/bt23d5/PHHE8vuY82aNREXX3xx+8TExJSIiIgerVu3Pumaa65p41+/YsWKmPPPP/+45s2bd4uKiuqRnJzc9eabb07KycmxGp4OHCHoGQMAAAAAOGZt2rQpIjQ0VHFxcaUvvvhio5KSErv44oszarLPnJyc0CuvvPK4m266aUdKSsqWmJiY0j179oSkpqZ2LigosHHjxm3t0KFDweLFi+PHjRvXrqCgIGT8+PE7JS+IOf3000+Ijo4uvfvuu7ccf/zxBRs3box455134vz737BhQ0S3bt3yrrrqqt1xcXElX331VfTjjz/eauPGjZFvvfXWjzU8JTgCEMYAAAAAAI4ZJSUlKioqUkZGRujs2bMbL1u2rHG/fv0yYmNjSzdv3hwhSR06dCioSR379u0LefrppzdfccUVGf5ld911V8tt27ZFrFy58puTTjqpQJIuvvji7MzMzNDHH3+85dixY3eGh4frnnvuaVVQUBDyn//859vk5OQi//a33HLLbv+/r7766gxJGZJUWlqqAQMG5MTFxZXcfPPN7bdv3x7aokWLkpq0H/WPMAYAAAAAcMw4+eSTu/r/HRISosGDB++eMWPG5tqsIywszA0fPjwjcNm7774b361bt9zOnTsXFBXtz1h03nnnZc2bNy/xiy++iO7du3feRx99FH/22WdnBgYxZe3Zsydk/PjxLd96663G27dvjyguLt4/POmbb76JatGiRblDrHD0IIwBAAAAABwz5syZs75du3aF8fHxJR07diyMiYnZP7FumzZtCiVp/fr1kSkpKYfdO6Zx48bFYWEHfp3evXt32E8//RQZERFR7qXBd+7cGSZJGRkZoa1atSo81P5HjBjR/tNPP40dO3bs1h49euyLjY0t/fTTTxvcc889bfPy8pj79RhAGAMAAAAAOGZ07949z381pbLOP//87NDQUC1cuDB+6NChWYdbh9nB8+g2atSoOCEhoXjKlCk/lbfNSSedlC95Qc62bdvCK9r3vn377N133200ZsyYrX/60592+pf/5z//iT7c9uLIQxgDAAAAAPhFSE5OLho2bNiul19+uenIkSP3lHdFpblz5zYaOXJkRnX3fc4552Q9++yzzTp06FCYlJRUXFG5M888M2vp0qWNN23aFN6uXbuDhirl5eWFlJSUKDw8/IBLZb/wwgsHXZUJRy/CGAAAAADAL8bMmTM3r1+/PmrQoEGdLr/88vRzzz03Oy4urmTdunWRr7zySpOvv/66wciRI7+s7n7Hjx+/4/XXX2/861//unNaWtqOE044IT8nJyfku+++i/rkk08avvvuu+sl6ZFHHtn6/vvvx6empnYeM2bMtk6dOhVs3rw5fOnSpfELFy7ckJCQUJKSkpL797//vXnLli2LmjZtWvzcc88l7Nixo8LeNDj6EMYAAAAAAPY7rdODq+q7DXUpPj6+9JNPPvl+0qRJifPnz0945ZVXmhYUFFizZs2KzjjjjKzJkyf/fDj7TUhIKPn888/X3H333a2efPLJFjt37gyPjY0tad++ff6QIUP2+ssdf/zxhR9//PGau+++u9Wf//znpH379oU2a9as8Lzzzsvwl5k3b96P119/fbtx48a1jYyMLB04cODeq6++evOIESN+VQunAEcAc85VXiqIevXq5VauXFnfzTjizHp3S53t+4bpC+ps35LkXr25TvcP4PDx3gKgLvDeAtSMma1yzvWqyzpWr169MSUlZVdd1gH80q1evToxJSUlubx1zMIMAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEEUVt8NAAAAAAAcOe77W5ee9Vn/g7d8s6om2y9fvrzBpEmTmq9cubLh3r17wxo0aFB64okn5o4YMWLP6NGjd4eFhWnq1KkJt912W/KaNWu+Ov744wtrq+2S9NZbb8W+9957sY899tjW0NDQ2tx1hfUNGjSok/9+ZGSkS0pKKrjkkkv2TJgwYXvDhg2df11paamefvrpJnPmzEn87rvvYnJyckISEhKKTznllJxRo0alDxo0KLvOG1wLhg0blvzPf/4zdsuWLV9J0tq1ayM6d+580pNPPrnx1ltv3V3f7asKesYAAAAAAI4JDz74YLMBAwZ0zsjICLv//vt/fuONN76fNm3axg4dOhSMHTu27SuvvNKortvw3nvvxU6ePLllaWlpXVd1gD//+c+bly9fvmb+/PnrBgwYkDF58uRWI0eObOdfX1xcrIsuuui4W265pX2bNm0K/va3v2188803v58wYcLP+fn5NmTIkE67d++u+/QIkugZAwAAAAA4BixevLjhhAkT2lx55ZU7Z8+evTlw3RVXXJFx991378jJyTkqOyTk5eVZdHS0O1SZLl265J1zzjm5kjR48ODs9PT08FdffTVhx44dm5s3b17yxz/+seWSJUsaP/fcc+uvvvrqjMBtR48evee1116Li4iIOGQdh9MulO+ofCICAAAAABDo0UcfbREfH188ffr0n8tb36VLl4LevXvnVbS9mfUcM2ZMq8Bla9eujTCznlOnTk3wL1uxYkVMampqx0aNGp0cFRXVo3Xr1iddccUVbSVpzJgxrSZPntxSkiIiInqaWU8z2z/sKzs7O2T06NFJSUlJJ4WHh/dISko6ady4cS1KSkr21/nWW2/FmlnP559/vtHw4cPbNW7cOKVZs2Yp1T0fvXr1ypWkb7/9NjI/P99mzJjRvG/fvpllgxi/oUOHZsXGxlbYnWfMmDGtzKznv//976gzzjijY0xMTPeBAwceV9XjkqStW7eGXXHFFW1btGjRLSIiokeLFi26XXzxxe3z8vJMkr7++uvIiy++uH1SUtJJ/nN7+eWXt01PTz/meuzQMwYAAAAAcFQrLi7W559/Htu/f/+MmJiYOuupkZmZGTJ48OBO3bp1y50+ffqGuLi40h9//DHis88+ayhJN910U/qWLVvC58+fn7h06dI1gXPGFBUVqW/fvh3Xr18fPWbMmK0pKSl5n376aYMpU6a02rNnT9isWbMOCJHuuuuutv369cucNWvWhry8vGp3pNiwYUOkJCUkJJR89NFHMdnZ2aEXXXRRRs3OgDR06NBfXX755bvGjh27PSQkpMrHlZ6eHnr66ad3zszMDLvjjju2nnzyyXnbt28Pf/PNNxvl5+dbdHS027x5c3jr1q0Lf/Ob32xOSEgoXrduXeSkSZNannvuuTFffvnlmpq2/UhCGAMAAAAAOKpt27YtLD8/P6Rt27a1OhlvWatXr47KysoKffzxx38O7GXjnzS2Q4cORUlJSUWS1K9fv9zw8PD9286cObPJF1980XDRokVrL7jgghxJGjJkSLYkTZo0qdWECRO2JyUlFfvLp6Sk5M6bN29TVdtWWlpqRUVFys7ODlm4cGH83Llzm3bu3DmvW7duBbNmzYqRpPbt29f4/Nxwww07//SnP+3033/qqaeqdFwPP/xw859//jnyww8//PbXv/71/nM3atSoPf5/X3DBBTn+fUhS//79c44//viC888///hPPvkkOnC7ox3DlAAAAAAAqIIuXboUxMbGlowaNard9OnTm/zwww/hlW/lWbp0aXyrVq0K+/fvn1NUVCT/7cILL8wqLi62Dz74oEFg+SFDhmRUp23Dhg3rGBER0TMhIaH7tddee9xpp52WvXDhwh+qs4+qGD58+AHtqupxvf/++3Fdu3bNPVSgkp+fb3fffXeL9u3bd4mKiuoRERHR8/zzzz9ekr755puo2j6W+kTPGAAAAADAUa1FixbFUVFRpT/99FNEXdaTkJBQsmTJkrUPPPBAq7Fjx7a76aabQn71q1/ljx8/fktFc7H47dq1K2zr1q0RERER5V46fNeuXQd8P/f3sKmqRx555KfU1NTcmJiY0k6dOhXGxcXtn/8lOTm5UJI2bNhQ4/PTtm3bA9pV1ePKyMgIO+GEE/Ydat+33HJL0uzZs5vdcccd284444yc+Pj4kk2bNkVcddVVHfLz84+pziSEMQAAAACAo1p4eLhOPfXU7I8//jjucK/wExER4QoLCy1w2c6dOw/6zpyampq3dOnS9UVFRfrwww8bTJw4scXvf//7Dl26dPnmlFNOya9o/02aNClJSkoqfOmll9aXt75jx44HDCEys2odQ+fOnfP79OlTbthx5pln7ouNjS1ZtGhRozvvvHNXdfZbVkhIyAHtqupxNW7cuHj79u2HDIMWLlzYZOjQobv/+te/bvMve/PNN4+5yXslhikBAAAAAI4Bd9999/bMzMywtLS01uWtX7NmTcTnn38eXdH2LVu2LPz2228PWL9w4cL4isqHh4frnHPOyZ04ceLW0tJSffXVV9GSFBkZWSpJZS+jPWDAgMzt27eHx8bGlvbp02df2VvLli2Ly6unNkRFRbkbb7xxx/vvvx8/e/bsRuWVWbBgQVx2dna1M4KqHle/fv2yvvrqqwafffZZhY9Bfn5+SFhY2AFhz7PPPptQUfmjGT1jAAAAAABHvQsuuCBnwoQJmydMmNBm7dq1USNHjtzdvn37wt27d4cuX7487pVXXkmcNWvWhooub33JJZfsmTp1astx48a1SE1NzV2xYkXsa6+91iSwzMsvvxw/a9aspoMHD97boUOHwpycnJBp06Y1a9CgQelZZ52VI0ldunTJl6SHHnqoxcCBAzPDwsJcnz599o0aNWrP3LlzE88777xOaWlpO7p3776voKDAfvjhh8i333670ZIlS9Yf6tLSNTVx4sRtX331VfR1113XYfHixbsGDRqUmZiYWLxp06aIBQsWNFq2bFnj9PT0L6u736oe1/jx43e89tprTS688MJOY8aM2ZaSkpK3c+fOsDfeeKPR888/v6lx48alZ511VuZrr72W8Je//CWvU6dOBf/4xz8arVq1qmEdnI56RxgDAAAAANjvwVu+WVXfbThc9913387TTz89d9KkSc3vu+++1nv37g1r0KBBadeuXXOfeOKJTSNGjMioaNuHH354W0ZGRuizzz7bbNq0aaF9+/bNfO655348++yzT/CXOfHEE/Ojo6NLH3vssVa7du0Kj4mJKenWrVvuG2+88X2HDh2KJG+C26VLl6bPnj276ZQpU1o65+ScWxUZGelWrFjx/b333tvy+eefT5w4cWJkdHR0aZs2bQoGDBiQGRUVVWdBjCSFhYXp7bff/nHGjBlN5syZk5iWlpa8b9++kISEhOJTTz01e9GiRWsTEhJKqrvfqh5XYmJiyaeffrrmzjvvTJo6dWqLjIyMsISEhOLU1NSsqKgoJ0kzZ87cfP3119vEiROTJKlv376Zc+fO/bFv374nHKoNRyNzrs4uwX5YevXq5VauXFnfzTjizHp3S53t+4bpC+ps35LkXr25TvcP4PDx3gKgLvDeAtSMma1yzvWqyzpWr169MSUlpUZzhwA4tNWrVyempKQkl7eOOWMAAAAAAACCiDAGAAAAAAAgiAhjAAAAAAAAgogwBgAAAAAAIIgIYwAAAAAAAIKIMAYAAAAAACCICGMAAAAAAACCiDAGAAAAAAAgiAhjAAAAAAAAgogwBgAAAAAAIIjC6rsBAAAAAIAjR9HqB3rWZ/3hKfevOpztpk6dmnDbbbcl++/HxMSUtm7duuCqq65Kv+uuu9LDw8P3ly0oKLDJkycnzp8/P2HdunVRBQUFIU2bNi1KTU3Nvu2223aeccYZ+yqqZ8yYMa0mT57csrx1X3311dddu3Yt2Lt3b8i4ceNaffnllw2+/fbbmNzc3JA333zz+4EDB2YfzrHh2EMYAwAAAAA4Zjz77LM/tm3btjAzMzN0/vz5jcePH992586d4VOmTNkqSVlZWSFnn312x6+//rrB5Zdfnj527NhtsbGxJd9//33UK6+80uTCCy/slJWV9WVl9SxdunRNaGjoAcs6dOhQKEk7d+4MmzdvXmKXLl32/frXv85atmxZozo4VBzFCGMAAAAAAMeMU045ZV/Xrl0LJGno0KFZGzZsiHzmmWea+cOY66+/vs1///vfBosWLVrbv3//XP92F110Uc4dd9yxa86cOY2qUk+/fv1yA3vbBOrYsWNhZmbml5L0+uuvxx5tYUxeXp5FR0e7+m7HsYw5YwAAAAAAx6zu3bvvy8nJCd2yZUvYpk2bwl999dWE4cOH7woMYgJdeeWVGTWtMySkZl+1f/rpp7ChQ4cmN2vWrFtERESPpk2bduvXr9+vtmzZsr9DRVZWVkhaWlpSmzZtukZERPRITExMOe+88zps3rx5f5n3338/JjU1tVNMTEz36Ojo7qeffnqn999/PyawrmHDhiU3b9682/Llyxt07969c1RUVI+0tLTWkrR169awyy67rK2/He3bt+/y+OOPJ9bo4CCJnjEAAAAAgGPYpk2bIkJDQxUXF1f64osvNiopKbGLL744o6b7LS4uNkn7e4+EhISo7LClwzV8+PD2W7dujXzggQd+Tk5OLty2bVv48uXLY3NyckIkKT8/384666xOa9asib755pu3p6am5mZkZIQuXbo0bteuXWFt2rQp/vzzz6MvuOCCzh06dMh76qmnNpqZe/zxx1tecMEFnd9///3vTj/99Dx/fTk5OaFXXnnlcTfddNOOlJSULTExMaV79uwJSU1N7VxQUGDjxo3b2qFDh4LFixfHjxs3rl1BQUHI+PHjd9bKwf5CEcYAAAAAAI4ZJSUlKioqUkZGRujs2bMbL1u2rHG/fv0yYmNjSzdv3hwhSR06dCioaT0xMTE9Au8PHjx4z8KFCzfUdL+S9OWXXzb84x//uGX06NF7/Muuvfbavf5/z5gxo8mXX37Z4IUXXvjh8ssvz/Qvv+aaa/aXuf/++1uGh4eXrlix4vvExMQSXxuz2rdv3+3+++9vtWzZsvX+svv27Qt5+umnN19xxRUZ/mV33XVXy23btkWsXLnym5NOOqlAki6++OLszMzM0Mcff7zl2LFjd1Y0TAuVI4wBAAAAABwzTj755K7+f4eEhGjw4MG7Z8yYsbm263n33XfXhIWF7e8Z06xZs+Lq7qO4uFjO/W9qltDQUIWEhOikk07KnTZtWgvnnM4777zsXr165QUOfVq+fHl8YmJiUWAQU9a//vWv2LPPPjvTH8RIUpMmTUr79++f8e677zYKLBsWFuaGDx+eUeb44rt165bbuXPngqKiov3LzzvvvKx58+YlfvHFF9G9e/fOEw4LYQwAAAAA4JgxZ86c9e3atSuMj48v6dixY2FMTMz+tKNNmzaFkrR+/frIlJSUGvWOOfPMMyucwLeq2rVrd9LWrVsj/PeffPLJjbfeeuvu11577ce777671d/+9rcW9913X5umTZsWXXnllemPPvrottDQUO3Zsye0efPmRYfad1ZWVliLFi0OKtO8efOirKysA8ZTNW7cuDgs7MB4YPfu3WE//fRTZERERLmXOt+5cyd5Qg1w8gAAAAAAx4zu3bvn+a+mVNb555+fHRoaqoULF8YPHTo0K9htK2vBggU/5Ofnm/9+p06dCiQpKSmpeO7cuT9J+mn16tWRs2bNSnziiSdaNW3atHjcuHHpCQkJxevWrYs+1L7j4uKKd+zYcVBatGPHjvC4uLiSwGVmVraYGjVqVJyQkFA8ZcqUn8rb/0knnZRftaNEeQhjAAAAAAC/CMnJyUXDhg3b9fLLLzcdOXLknvKuqDR37txGI0eOzAhGe0499dRKh/mkpKQUTJs2bcvcuXObfv3119GS1L9//6y33nqryUsvvRR/2WWXlTtUqXfv3tkffPBB/N69e0MaN25cKkl79+4Neffddxv17t07u7J6zznnnKxnn322WYcOHQqTkpKqPQQLh0YYAwAAAAD4xZg5c+bm9evXRw0aNKjT5Zdfnn7uuedmx8XFlaxbty7ylVdeafL11183GDly5Jc1rWf+/Plxubm5of/973+jJen9999vmJ6eHtagQYOSSy+9tMJeObt37w4966yzOl166aW7TzzxxPzw8HC3YMGCRllZWaHnnXdeliTdeOONe2bPnt30uuuuO2716tXbTj/99NysrKzQJUuWxN111107u3fvnj9hwoRtZ511VqM+ffoc/4c//GGbmemJJ55okZ+fHzJhwoStlbV//PjxO15//fXGv/71rzunpaXtOOGEE/JzcnJCvvvuu6hPPvmk4bvvvru+sn2gYoQxAAAAAID9wlPuX1XfbahL8fHxpZ988sn3kyZNSpw/f37CK6+80rSgoMCaNWtWdMYZZ2RNnjz559qo54477mgXOB/MpEmTWklSq1atCi+99NKvKtouJiamtFu3bvvmzJnTdOvWrREhISFKTk7O//vf/77Bf7WjyMhI98EHH3w/duzYVnPnzm06adKkVo0aNSru2bNnjn8i4d69e+ctWrRo7b333puUlpbW3jmnk08+OXfx4sVrAi9rXZGEhISSzz//fM3dd9/d6sknn2yxc+fO8NjY2JL27dvnDxkyZG9l2+PQLHDm5iNBr1693MqVK+u7GUecWe9uqbN93zB9QZ3tW5LcqzfX6f4BHD7eWwDUBd5bgJoxs1XOuV51Wcfq1as3pqSk7KrLOoBfutWrVyempKQkl7cupLyFAAAAAAAAqBuEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBFFbfDQAAAAAAHDnyH2zXsz7rj7pv06qabL98+fIGkyZNar5y5cqGe/fuDWvQoEHpiSeemDtixIg9o0eP3h0WFqapU6cm3Hbbbclr1qz56vjjjy+srbZL0ltvvRX73nvvxT722GNbQ0NDa3PXFdY3aNCgTv77kZGRLikpqeCSSy7ZM2HChO0NGzZ0/nWlpaV6+umnm8yZMyfxu+++i8nJyQlJSEgoPuWUU3JGjRqVPmjQoOyK6vGfs/LWLViw4PuLL744W5JuvvnmpP/85z8x33zzTYPMzMzQJ598cuOtt966uxYP+ZhAGAMAwDEi/8F2dbr/qPs21en+AQCoqQcffLDZhAkT2px22mnZ999//8/HHXdc4e7du8OWLl0aN3bs2LaNGzcuueKKKzLqsg3vvfde7OTJk1s++uijQQlj/P785z9vPu2003Jzc3NDFi9eHDd58uRW69evj1ywYMFGSSouLtagQYOOe+eddxoPHTp0V1pa2s7ExMTijRs3Rrz66quNhwwZ0ik9Pf3LhISEkkPV8+yzz/7Ytm3bAwKs7t275/n//dxzzzXr3LnzvrPPPjtjwYIFCXVysMcAwhigHvCFCQAAAKhdixcvbjhhwoQ2V1555c7Zs2dvDlx3xRVXZNx99907cnJyjsqpOvLy8iw6OtodqkyXLl3yzjnnnFxJGjx4cHZ6enr4q6++mrBjx47NzZs3L/njH//YcsmSJY2fe+659VdffXVG4LajR4/e89prr8VFREQcsg5JOuWUU/Z17dq1oKL1WVlZ/wkNDdXXX38debSFMVU5z7XlqHwiAgAAAAAQ6NFHH20RHx9fPH369J/LW9+lS5eC3r1755W3TpLMrOeYMWNaBS5bu3ZthJn1nDp16v5QYcWKFTGpqakdGzVqdHJUVFSP1q1bn3TFFVe0laQxY8a0mjx5cktJioiI6GlmPc1s/7Cv7OzskNGjRyclJSWdFB4e3iMpKemkcePGtSgp+V9nlLfeeivWzHo+//zzjYYPH96ucePGKc2aNUup7vno1atXriR9++23kfn5+TZjxozmffv2zSwbxPgNHTo0KzY2trS69ZRVk95AmZmZIVdddVWbli1bnhQREdGjSZMmKampqZ3+85//RPnLFBUVafz48S06dOjQJTIyskfjxo1TzjzzzI6BZVavXh157rnndoiNjT05KiqqR0pKSud//OMfcYF1jRkzppWZ9fz3v/8ddcYZZ3SMiYnpPnDgwOOkqj1ONUXPGAAAAADAUa24uFiff/55bP/+/TNiYmLqrGdDZmZmyODBgzt169Ytd/r06Rvi4uJKf/zxx4jPPvusoSTddNNN6Vu2bAmfP39+4tKlS9cEBhNFRUXq27dvx/Xr10ePGTNma0pKSt6nn37aYMqUKa327NkTNmvWrANCpLvuuqttv379MmfNmrUhLy+v2h0pNmzYEClJCQkJJR999FFMdnZ26EUXXZRRszMglZSUqKioaP99M1NYWO1EC6NGjWrzzjvvNLr33nu3dO7cOT89PT3s448/brhnz579J3LQoEHHLV++vPG11167Y8CAAVl5eXkhK1asaPjzzz+Hd+/ePX/jxo3hffv27dygQYPSRx999KdGjRqV/P3vf2/2u9/9ruPLL7+87tJLL80KrHPo0KG/uvzyy3eNHTt2e0hISLUfp8NFGAMAAACgQgyvxtFg27ZtYfn5+SFl5zKpbatXr47KysoKffzxx38O7GXjn6C2Q4cORUlJSUWS1K9fv9zw8PD9286cObPJF1980XDRokVrL7jgghxJGjJkSLYkTZo0qdWECRO2JyUlFfvLp6Sk5M6bN6/KL5DS0lIrKipSdnZ2yMKFC+Pnzp3btHPnznndunUrmDVrVowktW/fvsbn5+STT+4aeL9Hjx45q1atWlvT/UrSqlWrGl5yySW777jjjl3+ZVdeeWWG/99vvPFG7NKlSxs/9NBDm++9996d/uUjR47cX2bixInNs7Ozwz766KOv/cOpLr300sxf/epXXSdMmJBUNoy54YYbdv7pT3/av6+nnnqqWo/T4WKYEgAAAAAAVdClS5eC2NjYklGjRrWbPn16kx9++CG88q08S5cujW/VqlVh//79c4qKiuS/XXjhhVnFxcX2wQcfNAgsP2TIkIzqtG3YsGEdIyIieiYkJHS/9tprjzvttNOyFy5c+EN19lEVc+bMWb9ixYrv/Ldnn312Y3X3UVxcrMBz4B/+k5KSkvt///d/iXfffXeLDz/8MKa4+MDMY8mSJXFmpttvvz29on1/9tlnsSkpKTmB89qEhYVp6NChe9asWROzZ8+eA3KQ4cOHZwTer+7jdLjoGQMAAAAAOKq1aNGiOCoqqvSnn36KqMt6EhISSpYsWbL2gQceaDV27Nh2N910U8ivfvWr/PHjx2+paC4Wv127doVt3bo1IiIiotxLh+/ateuA7+f+HjZV9cgjj/yUmpqaGxMTU9qpU6fCuLi4/fO/JCcnF0rShg0banx+unfvnneoCXyrIjU19fh///vfDf3377jjjm2TJk3a+swzz/z0xz/+seill15KfPTRR5Pi4+NLhg0btnvKlClbYmNjS/fs2RMWHx9fHHi57rIyMzNDu3TpclD7WrRoUeSc065du8KaNGmyv4dQ27ZtDzjP1X2cDhdhDAAAAADgqBYeHq5TTz01++OPP4473CviREREuMLCQgtctnPnzoO+M6empuYtXbp0fVFRkT788MMGEydObPH73/++Q5cuXb455ZRT8ivaf5MmTUqSkpIKX3rppfXlre/YseMBQ4jMrFrH0Llz5/w+ffrsK2/dmWeeuS82NrZk0aJFje68885d5ZUJplmzZm3MzMzcPw+MPxCJj48vfeqpp7Y89dRTW77//vuIF198sfHDDz+cFBERUfr3v/99S0JCQnFmZmZYTk6OVRTIxMfHl+zcufOgHkvbt28PNzMlJiYe0N0mJCTkgP1U93E6XAxTAgAAAAAc9e6+++7tmZmZYWlpaa3LW79mzZqIzz//PLqi7Vu2bFn47bffHrB+4cKF8RWVDw8P1znnnJM7ceLEraWlpfrqq6+iJSkyMrJUkspeRnvAgAGZ27dvD4+NjS3t06fPvrK3li1b1ngekopERUW5G2+8ccf7778fP3v27EbllVmwYEFcdnZ2UDKClJSUgsBjT05OPqgXUKdOnQofeOCBHZ06dcr77rvvoiXp/PPPz3LO6cknn2xa0b5TU1OzV69e3WDt2rX7ewEVFxfr9ddfb3zCCSfsa9KkySGvGBWsx4meMQAAAACAo94FF1yQM2HChM0TJkxos3bt2qiRI0fubt++feHu3btDly9fHvfKK68kzpo1a0NFl7e+5JJL9kydOrXluHHjWqSmpuauWLEi9rXXXmsSWObll1+OnzVrVtPBgwfv7dChQ2FOTk7ItGnTmjVo0KD0rLPOypGkLl265EvSQw891GLgwIGZYWFhrk+fPvtGjRq1Z+7cuYnnnXdep7S0tB3du3ffV1BQYD/88EPk22+/3WjJkiXra+PS0hWZOHHitq+++ir6uuuu67B48eJdgwYNykxMTCzetGlTxIIFCxotW7ascXp6+pc1reftt99uuHPnzvBt27aFSdLKlSsbPPfcc6WSdM011+w91LYnn3xy5wsuuCCjW7duebGxsaXvv/9+w7Vr18aMGDFisyQNGjQo+7zzztt7//33t968eXNE//79swoLC23FihWxgwYNyhw4cGD2Pffcs2P+/PkJAwYM6HTPPfdsjY+PL5kxY0bTTZs2Rc2bN29dZe0P1uNEGAMAAAAA2C/qvk2r6rsNh+u+++7befrpp+dOmjSp+X333dd67969YQ0aNCjt2rVr7hNPPLFpxIgRGRVt+/DDD2/LyMgIffbZZ5tNmzYttG/fvpnPPffcj2efffYJ/jInnnhifnR0dOljjz3WateuXeExMTEl3bp1y33jjTe+79ChQ5HkTQi7dOnS9NmzZzedMmVKS+ecnHOrIiMj3YoVK76/9957Wz7//POJEydOjIyOji5t06ZNwYABAzKjoqLqLIiRvEls33777R9nzJjRZM6cOYlpaWnJ+/btC0lISCg+9dRTsxctWrQ2ISGhpKb1PPDAA0mB88HMnTu36dy5c5tK0jXXXHPI59bpp5+e/frrrzeZNm1aRHFxsbVp06bggQceOODKSW+++eaP9957b4t58+YlPvPMM80aNmxY0q1bt9y0tLR0SUpOTi764IMP1vzhD39ofdddd7UtLCwM6dy587558+at+81vfpNVce2eYD1O5lydXYL9sPTq1cutXLmyvptxxJn17pY62/cN0xfU2b4lyb16c53u/2jEJSJxpOC95djCewuOFLy3HFt4bwk+M1vlnOtVl3WsXr16Y0pKSr3PHQIcy1avXp2YkpKSXN465owBAAAAAAAIIsIYAAAAAACAIKpyGGNmaWa2wczyzWyVmZ1ZSfkIM3vQt02Bmf1kZrfWvMkAAAAAAABHrypN4Gtmv5P0pKQ0SR/7/i42sxOdcz9VsNkrklpLukHSOknNJVV4GTEAAAAAAIBfgqpeTWmMpNnOuVm++7eY2fmSRku6p2xhMxsg6RxJHZxz/kmhNtawrQAAAAAAAEe9SocpmVmEpJ6SlpVZtUxSagWbXSzp35LGmNnPZrbOzKaaWcMKygMAAAAAgscdaVfWBY4lpaWlJqnCy2BXZc6YREmhknaUWb5DUosKtjlO0hmSUiQNk3SzpPMlzS6vsJndYGYrzWxlenp6FZoEAAAAADhcZrYxNzc3pr7bARyr8vLyosxse0Xr6+pqSiGSnKTLnHOfO+eWygtkhplZ87KFnXMznXO9nHO9mjZtWkdNAgAAAABIUlFR0V83btwYtmvXrkaFhYVh9JIBakdpaanl5uZGb9y4MaK4uPiBispVZc6YXZJK5E3AG6i5pIpSnm2StjjnMgOWfef721YH97IBAAAAAARJz549F69ateqHLVu23LN169YU51wT1d2P9cAvSamZbS8uLn6gR48eSysqVGkY45wrNLNVks6V9H8Bq86V9GoFm30i6bdm1tA5l+Nb1sn3d1PlbQcAAAAA1KWePXuuk3RtfbcD+CWqavI5SdLVZnadmZ1gZk9KaiVphiSZ2RwzmxNQ/iVJuyU9Z2ZdzOzX8i6N/Q/n3M5abD8AAAAAAMBRpUqXtnbOzTOzBEn3Smop6WtJFzrn/L1c2pYpn2Nm/SX9Td5VlfZKel3S3bXUbgAAAAAAgKNSlcIYSXLOTZc0vYJ1fctZtlbSgMNuGQAAAAAAwDGICZoAAAAAAACCiDAGAAAAAAAgiAhjAAAAAAAAgogwBgAAAAAAIIgIYwAAAAAAAIKIMAYAAAAAACCICGMAAAAAAACCiDAGAAAAAAAgiAhjAAAAAAAAgogwBgAAAAAAIIjC6rsBQE3c97cudbbvB2/5ps72DeDIxnsLAAAA6hJhDAAAQVS0+oH6bgIAAADqGcOUAAAAAAAAgoieMQAAAEAQ1OUQyD/1ubTO9g0AqH2EMUAFGEoAAAAAAKgLDFMCAAAAAAAIIsIYAAAAAACAICKMAQAAAAAACCLCGAAAAAAAgCAijAEAAAAAAAgiwhgAAAAAAIAgIowBAAAAAAAIIsIYAAAAAACAICKMAQAAAAAACCLCGAAAAAAAgCAijAEAAAAAAAgiwhgAAAAAAIAgIowBAAAAAAAIIsIYAAAAAACAICKMAQAAAAAACCLCGAAAAAAAgCAijAEAAAAAAAgiwhgAAAAAAIAgIowBAAAAAAAIIsIYAAAAAACAICKMAQAAAAAACCLCGAAAAAAAgCAijAEAAAAAAAgiwhgAAAAAAIAgIowBAAAAAAAIIsIYAAAAAACAICKMAQAAAAAACCLCGAAAAAAAgCAijAEAAAAAAAgiwhgAAAAAAIAgIowBAAAAAAAIIsIYAAAAAACAICKMAQAAAAAACCLCGAAAAAAAgCAijAEAAAAAAAgiwhgAAAAAAIAgIowBAAAAAAAIIsIYAAAAAACAICKMAQAAAAAACCLCGAAAAAAAgCAijAEAAAAAAAgiwhgAAAAAAIAgCqvvBgAAcDj++f199d0EAAAA4LAQxqDO8YUJAAAAAID/IYwBAAAAfPgRCQAQDMwZAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQVTlMMbM0sxsg5nlm9kqMzuzitudYWbFZvb14TcTAAAAAADg2FClMMbMfifpSUkTJXWX9KmkxWbWtpLtGkuaI+ndGrYTAAAAAADgmFDVnjFjJM12zs1yzn3nnLtF0jZJoyvZ7hlJz0v6rAZtBAAAAAAAOGZUGsaYWYSknpKWlVm1TFLqIbZLk9Rc0p9r0kAAAAAAAIBjSVV6xiRKCpW0o8zyHZJalLeBmZ0k6X5JVzjnSiqrwMxuMLOVZrYyPT29Ck0CAAAAAAA4OtX61ZTMLFLSPEl3Ouc2VGUb59xM51wv51yvpk2b1naTAAAAAAAAjhhhVSizS1KJvCFHgZpL2l5O+ZaSTpD0nJk951sWIsnMrFjShc65skOeAAAAAAAAfhEq7RnjnCuUtErSuWVWnSvvqkplbZF0kqSTA24zJP3g+3d52wAAAAAAAPwiVKVnjCRNkjTXzP4l6RNJN0pqJS9kkZnNkSTn3JXOuSJJXwdubGY7JRU45w5YDgAAAAAA8EtTpTDGOTfPzBIk3StvGNLX8oYbbfIVaVtH7QMAAAAAADimVLVnjJxz0yVNr2Bd30q2nSBpQjXaBQAAAAAAcEyq9aspAQAAAAAAoGKEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQVTlMMbM0sxsg5nlm9kqMzvzEGWHmtkyM0s3s2wz+9zMBtdOkwEAAAAAAI5eVQpjzOx3kp6UNFFSd0mfSlpsZm0r2OQsSe9JushXfpGkBYcKcAAAAAAAAH4JwqpYboyk2c65Wb77t5jZ+ZJGS7qnbGHn3G1lFj1gZhdJuljSR4fZVgAAAAAAgKNepT1jzCxCUk9Jy8qsWiYptRp1xUraW43yAAAAAAAAx5yqDFNKlBQqaUeZ5TsktahKJWZ2k6TWkuZWq3UAAAAAAADHmDq/mpKZDZP0mKTLnHObKihzg5mtNLOV6enpdd0kAAAAAACAelOVMGaXpBJJzcssby5p+6E2NLPfyOsNc6Vz7s2KyjnnZjrnejnnejVt2rQKTQIAAAAAADg6VRrGOOcKJa2SdG6ZVefKu6pSuczsUnlBzNXOuX/UpJEAAAAAAADHiqpeTWmSpLlm9i9Jn0i6UVIrSTMkyczmSJJz7krf/eHygpg7JX1oZv65ZQqdc3tqr/kAAAAAAABHlyqFMc65eWaWIOleSS0lfS3pwoA5YNqW2eRG376n+G5+KyT1PfzmAgAAAAAAHN2q2jNGzrnpkqZXsK7voe4DAAAAAADAU+dXUwIAAAAAAMD/EMYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQEcYAAAAAAAAEEWEMAAAAAABAEBHGAAAAAAAABBFhDAAAAAAAQBARxgAAAAAAAAQRYQwAAAAAAEAQVTmMMbM0M9tgZvlmtsrMzqyk/Fm+cvlm9qOZ3Vjz5gIAAAAAABzdqhTGmNnvJD0paaKk7pI+lbTYzNpWUL69pEW+ct0lPSLpb2Y2rDYaDQAAAAAAcLSqas+YMZJmO+dmOee+c87dImmbpNEVlL9R0lbn3C2+8rMkPS/pzpo3GQAAAAAA4OhVaRhjZhGSekpaVmbVMkmpFWx2ejnll0rqZWbh1W0kAAAAAADAscKcc4cuYNZK0hZJZznnPgxYfp+ky51zx5ezzfeSXnDOPRiwrI+kFZJaOee2lSl/g6QbfHePl7T28A4Hv0CJknbVdyMAHHN4bwFQF3hvQVW1c841re9GAKg7YfXdAElyzs2UNLO+24Gjj5mtdM71qu92ADi28N4CoC7w3gIA8KvKnDG7JJVIal5meXNJ2yvYZnsF5YvFrwEAAAAAAOAXrNIwxjlXKGmVpHPLrDpX3tWSyvNZBeVXOueKqttIAAAAAACAY0VVr6Y0SdLVZnadmZ1gZk9KaiVphiSZ2RwzmxNQfoakJDOb4it/naSrJT1ei20HJIa3AagbvLcAqAu8twAAJFVhAt/9Bc3SJI2V1FLS15Lu8E/oa2YfSJJzrm9A+bMkTZbURdJWSY8652bUYtsBAAAAAACOOlUOYwAAAAAAAFBzVR2mBAAAAKAOmNnVZubM7Fe1tL++vv31rY39AQBqH2EMAAAAAABAEBHGAAAAAAAABBFhDI4oZpZiZm+Y2V4zyzOzT8zsTN+6c82s1MxuL7PNi2a2x8za+O77u/r2MbPXzSzHzHab2VNmFl0PhwWgDpnZCDNbY2b5ZvaVmQ02sw/8k8ubWZSZTTazr33vB9vN7E0z61xmPxPM7KCJ1MxstpltDLgfZmYPmdl6X527zOxjMzsjoMxlZvYfX31ZvnaNqruzAOBwVOH9w/+ZIrnMdge9X/jK/dnM/mBmm8xsn5m9bWbNfLf5ZpZpZpvNbFwFTWpV2WcXM4sxs0fNbIOZFfr+jjczPtcDwFEkrL4bAPiZWQ9JH0n6j6TrJe2TdKOk5WaW6px7x8wel/QXM3vfObfazK6WdJmk3zjnNpfZ5QuS5kuaLulUSfdJaiDvMusAjgFmdq6kFyW9IWmMpKaSpkiKkvS9r1ikpFhJf5a0TVITSWmSPjOzE5xz26tZ7ThJd0gaL+lLSXGSevn2K18o84KkqZLukvfDR2dJjap/hADqShXfP6prpLyrjqZJau7b3xx570GL5V3a+rfyPst85ZxbVGb7Q352MbMwSUslnSjpIUlfSTpN0p/kvQf94TDbDQAIMsIYHEkek/STpLOdc4WSZGZL5X2o+ZOki+V9+ekn6WUzu0zS3yTNdM69Ws7+Fjnn7vT9e5nvF6wHzWyic+5wP2QBOLI8IOlbSZc43+UBzexrSSvl+zLlnMuUdJ1/AzMLlfdlZoekEZImV7PO0yUtc849GbDszYB/nyYpwzl3e8CyZdWsA0Ddq/T94zAUSBrinCv27a+rvPD2T865P/uWfSDpEnmhTNkwprLPLiMknSHpLOfch75y75qZJN1vZo8653YeZtsBAEFEd0YcEXxdcM+S9H+SSn3DAMIkmaTlkvpIknOuSF5PmDaS/ilps7wPOeWZX+b+K/Ke86fW+gEACDpfqNJL0qv+L1KS5JxbJWlDmbKXmtnnZpYhqVhSrqSGko4/jKr/LelCM3vYzM4ws4hy1jc2sxfMbKCZNTqMOgDUoeq8f1TTO/4gxmeN7+/SgDqKJf0g77NMWZV9djlf0iZJn/o/K/k+Ly2TFC4vDAYAHAUIY3CkaCIpVF4PmKIyt5vlfbEJkSTn3DpJn8obejDTObevgn3uqOB+Uu02HUA9SZT35aO8X4H3v/7NbJCkeZK+kxfm9pZ0iqR0ecMRqmuipPslDZY3tHK3mT1nZomS5JxbIe8X7zaSFkhKN7PlZtbtMOoCUDeq9P5xGPaWuV94iOXlvf9U9tmlmaR2Oviz0r986xOq2V4AQD1hmBKOFBmSSiU9JW9s9UGcc6WSZGbXShogaZW8LrmvOed+KmeT5pK+KXNfkrbUUpsB1K9d8r6ENCtnXXN5wx4labikH5xzV/tXmlm4fHO8BMj3rYvwD5X0OeDLja+H3qOSHjWzFpIGSpokKUbS73xl/iHpH2bWUFJfX/klZtba/14GoF5V9f0j3/e3bA+4ugo9Kvvssltez51LK9h+Y900CwBQ2+gZgyOCcy5X3i/MKZK+cM6tLHuTJDPrJG9SzOmSzpYX4rzo625cVtkPKsPlBT6f19FhAAgi51yJvLkdhplvwgRJMrOektoHFI2RNzQp0Eh5vfECbfL97Rqwr0aSUg/Rhu3Ouf8nbzhl13LW5zjn3pL0tKSW4ldr4IhQjfeP8t4XwuT9KFQXKvvsskRer7uc8j4rOed21VG7AAC1jJ4xOJKMkfShpKVm9oy8q54kSuoh70vTfZJelveL0B+cc/m+SXw/lHSvvIn4Al1oZo/JG0d9qrxhBXN8w5wAHBvul/caX2BmM+W9Z0yQtF3eFxjJ+/JysZlNlvSWvHkibpEX5gZaLClT0iwzu1/eUMixknICC5nZQkmrJX0hb+hBd3nzODztW/+gvF+z35e0VVJrSbdK+tI5l147hw2gFlTl/ePfktZLesw3XLpA3pWSIuuoTZV9dnlR0jXyJu19Qt57UYSkDvKGTl58iOHbAIAjCD1jcMRwzn0hbx6H3fJ6vyyT9KSkk+QFLhPlXcpxhHMu37fNZ/JCmD+ZWdlfr6+Q1EnenA1/kDRL3gcoAMcI59w7ki6XdIK81/o4ea/37fKCFcl77T8sbwjRm5IulDQoYL1/XxnyhhyVyptE8xF5V2x7v0y1H8r7VfwZeUHPaEl/lRfcSN4v2MnyrtL0jrwhSiskXVTT4wVQe6ry/uGbbHeIvAsGzJY3nPod37/rwiE/u/iGSZ7nW36DvKsxvSjpKnnz6RUKAHBUsIAJ5IFjgpldLek5SR2dcz/Uc3MABJmZtZZ3pZKHnXMP1Xd7ABw9eP8AAAQLw5QAAEctM4uWN3nucnkTch4nr4fKPkn/rx6bBuAIx/sHAKA+EcYAAI5mJZJaSJomb3Jc/2Tgv3XObavPhgE44vH+AQCoNwxTAgAAAAAACCIm8AUAAAAAAAgiwhgAAAAAAIAgIowBAAAAAAAIIsIYAECdM7ONZvZBZct+Ke2oz3oBAABQ/whjAABHFTPra2YTzKxRfbelMr52Xlzf7QAAAMCRhUtbAwDqy/GSDueSfn0l3S9ptqSMINV5uO6X9Lyk14+AtgAAAOAIQRgDAKgXzrmCYNRjZuGSQp1z+cGqsyqOpLYAAAAguBimBACoNWbWxszmm1mmmWWZ2Ztm1qGCsuXN35JqZovNbLuZ5ZvZFjNbZGan+dbPltfbRJI2mJnz3Sb41k/w3e9iZpPM7GdJ+ZJOq6jOgLp7mNl7ZpZjZnvM7Hkza1amjH//yYc6HjNLNjN/r5erAtrpyitfZj8Xm9knZpbra8snZjakovrMrLOZvW1m2b7z/g8za1HeMQIAAODIQM8YAECt8M3h8qGkNpJmSPpW0lmS3pcUXYXtj5f0jqTtkp6UtENSc0lnSEqR9E9JT0uKk3SJpDsk7fJt/t8yu3tRUp6kJ+QNBdpWSfWtJb0r6VVJ/5DUQ9K1knqZ2SnOuX2Vtb+MdEkjJc2V9JGkmVXZyMzSJD0laY2kB32Lr5b0upmNcs6V3U+SpA8kLZB0l7zzNEreORpQzTYDAAAgSAhjAAC1ZaykZEnXOuee8y2bbmZTJN1Whe3PkxQjaYRz7l/lFXDOfWZm/5UXxrzunNtYwb4yJPV3zhVXse0dJN3hnJviX2Bm30iaJOlWSX+p4n787cyV9IKZzZX0o3Puhcq2MbPGkv4qab2k3s65LN/yv0v6j6QnzGy+cy4jYLNfSfqdc25+wH5KJaWZ2fHOubXVaTcAAACCg2FKAIDacrG83ixzyix/tIrbZ/r+DjGzqBq2ZUo1ghhJypI0vcyy6b7ll9SwLVV1rqQGkqb6gxhJ8v17qqSGkvqX2WZrYBDj857vb8e6aigAAABqhjAGAFBbjpO0zjlXErjQObdNVbvq0SuSlkv6o6Q9vvlbxplZu8Noy/fVLP+jc64wcIFvgt0f5R1XMLT3/f2mnHX+ZWXb8mM5ZXf7/ibURqMAAABQ+whjAABHBOdcgXPuXEm9JT0iqUTevClrzKy6vVOqO8dLVR3qUtT1MfS35BDrLGitAAAAQLUQxgAAasuPkjqaWWjgQjNrKalRVXfinPuXc+4hXzDzK0m5kv4cWKQW2lrWcWYWEbjAzCLl9UQJ7H2yx/e3SZmyUZJa1rAN/nq6lLPuxDJlAAAAcBQjjAEA1JaF8q5+dGWZ5eOqsrGZJZaz+Gd5VyYKDD9yfH+bHFz8sMVJSiuzLM23/PWAZf7hT2XnbrlD5f+fmqOqt/MdecHTLWYW61/o+/ctvn29U8V9AQAA4AjG1ZQAALXlr5IukzTLzHrKm+ekr6TT9b9LUB/KvWY2QNJbkjbIG2YzSFJn3779/un7+6iZvSgpX9LXzrmva9D29ZLuN7OuklZJ6inv0tZr5E2e67dc0lpJD5pZgq+dZ0g6TeUf4z8l9TezcZJ+kuScc6+U1wDnXIaZjZV3aevPzWy2b9XV8noIjXLOZZa3LQAAAI4uhDEAgFrhnNtrZmfKuxy0v3fMCkn9JL1bhV28Lm+oz6XyetjkSVon6XpJzwTU84kv3LhR0ix5/5c9IKkmYczPvnoflzRCUqGkFyXd6btMtb/uEjMbLC+gucVXbpmksyR9Us5+0+SFK+Ml+Xu7lBvG+PY/3cy2SbpL0v2+xaslXeKce/1wDw4AAABHFnOuLobeAwAAAAAAoDzMGQMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQUQYAwAAAAAAEESEMQAAAAAAAEFEGAMAAAAAABBEhDEAAAAAAABBRBgDAAAAAAAQRIQxAAAAAAAAQfT/AYXoH3iDPksnAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# colors = ['#A4C2F4', '#3333FF', '#64d2a6', '#065535', '#e49645', '#b45f06']\n",
    "colors = ['#9ABCE4', '#005293', '#C7D97D', '#7D922A', '#FAD080', '#F7811E'] #TUM COLORS\n",
    "data.groupby('distribution_type').mean()[prec_recall_metrics].plot(kind='bar', figsize=(15, 8), color = colors)\n",
    "plt.title('Precision, Recall and F1-Score for distribution types', fontsize = 20)\n",
    "plt.xlabel('distribution', fontsize=18)\n",
    "plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "label_names = ['PC precision', 'Cluster PC precision', 'PC recall', 'Cluster PC recall', 'PC F1-score', 'Cluster PC F1-score']\n",
    "plt.legend(label_names, bbox_to_anchor = (1.3, 1),  fontsize=16, loc='upper right')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_base_shd</th>\n",
       "      <th>cluster_cluster_shd</th>\n",
       "      <th>empty_graph_shd</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>distribution_type</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>exp</th>\n",
       "      <td>40.310000</td>\n",
       "      <td>34.326667</td>\n",
       "      <td>50.520000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gauss</th>\n",
       "      <td>40.123333</td>\n",
       "      <td>33.933333</td>\n",
       "      <td>50.453333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gumbel</th>\n",
       "      <td>40.383333</td>\n",
       "      <td>34.620000</td>\n",
       "      <td>50.783333</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   base_base_shd  cluster_cluster_shd  empty_graph_shd\n",
       "distribution_type                                                     \n",
       "exp                    40.310000            34.326667        50.520000\n",
       "gauss                  40.123333            33.933333        50.453333\n",
       "gumbel                 40.383333            34.620000        50.783333"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('distribution_type').mean()[dist_metrics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b5476204c0>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# colors = ['#A4C2F4', '#3333FF', '#4D68D5']\n",
    "colors = ['#9ABCE4', '#005293', '#5E94D4']\n",
    "data.groupby('distribution_type').mean()[dist_metrics].plot(kind='bar', figsize=(10, 6), color = colors)\n",
    "plt.title('SHD for different distributions', fontsize = 20)\n",
    "plt.xlabel('distribution', fontsize=18)\n",
    "plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "label_names = ['PC SHD', 'Cluster PC SHD', 'Empty graph SHD']\n",
    "plt.legend(label_names, bbox_to_anchor = (1.42, 1),  fontsize=16, loc='upper right')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Grouping by DAG method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_precision</th>\n",
       "      <th>cluster_adj_precision</th>\n",
       "      <th>base_adj_recall</th>\n",
       "      <th>cluster_adj_recall</th>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <th>cluster_adj_f1_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dag_method</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>erdos_renyi</th>\n",
       "      <td>0.848732</td>\n",
       "      <td>0.808552</td>\n",
       "      <td>0.544638</td>\n",
       "      <td>0.635218</td>\n",
       "      <td>0.618296</td>\n",
       "      <td>0.675626</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hierarchical</th>\n",
       "      <td>0.903317</td>\n",
       "      <td>0.891991</td>\n",
       "      <td>0.248063</td>\n",
       "      <td>0.367453</td>\n",
       "      <td>0.386609</td>\n",
       "      <td>0.515766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>scale_free</th>\n",
       "      <td>0.841707</td>\n",
       "      <td>0.796212</td>\n",
       "      <td>0.599711</td>\n",
       "      <td>0.683512</td>\n",
       "      <td>0.672641</td>\n",
       "      <td>0.713075</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              base_adj_precision  cluster_adj_precision  base_adj_recall  \\\n",
       "dag_method                                                                 \n",
       "erdos_renyi             0.848732               0.808552         0.544638   \n",
       "hierarchical            0.903317               0.891991         0.248063   \n",
       "scale_free              0.841707               0.796212         0.599711   \n",
       "\n",
       "              cluster_adj_recall  base_adj_f1_score  cluster_adj_f1_score  \n",
       "dag_method                                                                 \n",
       "erdos_renyi             0.635218           0.618296              0.675626  \n",
       "hierarchical            0.367453           0.386609              0.515766  \n",
       "scale_free              0.683512           0.672641              0.713075  "
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('dag_method').mean()[prec_recall_metrics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b5476d6350>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# colors = ['#A4C2F4', '#3333FF', '#64d2a6', '#065535', '#e49645', '#b45f06']\n",
    "colors = ['#9ABCE4', '#005293', '#C7D97D', '#7D922A', '#FAD080', '#F7811E'] #TUM COLORS\n",
    "data.groupby('dag_method').mean()[prec_recall_metrics].plot(kind='bar', figsize=(15, 8), color = colors)\n",
    "plt.title('Precision, Recall and F1-Score for DAG generation methods', fontsize = 20)\n",
    "plt.xlabel('DAG method', fontsize=18)\n",
    "plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "label_names = ['PC precision', 'Cluster PC precision', 'PC recall', 'Cluster PC recall', 'PC F1-score', 'Cluster PC F1-score']\n",
    "plt.legend(label_names, bbox_to_anchor = (1.3, 1),  fontsize=16, loc='upper right')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_base_shd</th>\n",
       "      <th>cluster_cluster_shd</th>\n",
       "      <th>empty_graph_shd</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dag_method</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>erdos_renyi</th>\n",
       "      <td>32.130000</td>\n",
       "      <td>27.410000</td>\n",
       "      <td>40.443333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hierarchical</th>\n",
       "      <td>65.573333</td>\n",
       "      <td>56.003333</td>\n",
       "      <td>79.813333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>scale_free</th>\n",
       "      <td>23.113333</td>\n",
       "      <td>19.466667</td>\n",
       "      <td>31.500000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              base_base_shd  cluster_cluster_shd  empty_graph_shd\n",
       "dag_method                                                       \n",
       "erdos_renyi       32.130000            27.410000        40.443333\n",
       "hierarchical      65.573333            56.003333        79.813333\n",
       "scale_free        23.113333            19.466667        31.500000"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('dag_method').mean()[dist_metrics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b547655ab0>"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# colors = ['#A4C2F4', '#3333FF', '#4D68D5']\n",
    "colors = ['#9ABCE4', '#005293', '#5E94D4']\n",
    "data.groupby('dag_method').mean()[dist_metrics].plot(kind='bar', figsize=(10, 6), color = colors)\n",
    "plt.title('SHD for different DAG generation methods', fontsize = 20)\n",
    "plt.xlabel('DAG method', fontsize=18)\n",
    "plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "label_names = ['PC SHD', 'Cluster PC SHD', 'Empty graph SHD']\n",
    "plt.legend(label_names, bbox_to_anchor = (1.42, 1),  fontsize=16, loc='upper right')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Grouping by #edges"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_adj_precision</th>\n",
       "      <th>cluster_adj_precision</th>\n",
       "      <th>base_adj_recall</th>\n",
       "      <th>cluster_adj_recall</th>\n",
       "      <th>base_adj_f1_score</th>\n",
       "      <th>cluster_adj_f1_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n_edges</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>15.0</th>\n",
       "      <td>0.843480</td>\n",
       "      <td>0.801699</td>\n",
       "      <td>0.679761</td>\n",
       "      <td>0.736713</td>\n",
       "      <td>0.684001</td>\n",
       "      <td>0.710425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30.0</th>\n",
       "      <td>0.885125</td>\n",
       "      <td>0.845446</td>\n",
       "      <td>0.521864</td>\n",
       "      <td>0.615081</td>\n",
       "      <td>0.627920</td>\n",
       "      <td>0.685672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50.0</th>\n",
       "      <td>0.872614</td>\n",
       "      <td>0.837464</td>\n",
       "      <td>0.384349</td>\n",
       "      <td>0.499568</td>\n",
       "      <td>0.518946</td>\n",
       "      <td>0.610957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80.0</th>\n",
       "      <td>0.857122</td>\n",
       "      <td>0.844398</td>\n",
       "      <td>0.270576</td>\n",
       "      <td>0.396882</td>\n",
       "      <td>0.405862</td>\n",
       "      <td>0.532237</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         base_adj_precision  cluster_adj_precision  base_adj_recall  \\\n",
       "n_edges                                                               \n",
       "15.0               0.843480               0.801699         0.679761   \n",
       "30.0               0.885125               0.845446         0.521864   \n",
       "50.0               0.872614               0.837464         0.384349   \n",
       "80.0               0.857122               0.844398         0.270576   \n",
       "\n",
       "         cluster_adj_recall  base_adj_f1_score  cluster_adj_f1_score  \n",
       "n_edges                                                               \n",
       "15.0               0.736713           0.684001              0.710425  \n",
       "30.0               0.615081           0.627920              0.685672  \n",
       "50.0               0.499568           0.518946              0.610957  \n",
       "80.0               0.396882           0.405862              0.532237  "
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('n_edges').mean()[prec_recall_metrics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b547795420>"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# colors = ['#A4C2F4', '#3333FF', '#64d2a6', '#065535', '#e49645', '#b45f06']\n",
    "colors = ['#9ABCE4', '#005293', '#C7D97D', '#7D922A', '#FAD080', '#F7811E'] #TUM COLORS\n",
    "data.groupby('n_edges').mean()[prec_recall_metrics].plot(kind='bar', figsize=(15, 8), color = colors)\n",
    "plt.title('Precision, Recall and F1-Score by approx. number of edges', fontsize = 20)\n",
    "plt.xlabel('number of edges', fontsize=18)\n",
    "plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "label_names = ['PC precision', 'Cluster PC precision', 'PC recall', 'Cluster PC recall', 'PC F1-score', 'Cluster PC F1-score']\n",
    "plt.legend(label_names, bbox_to_anchor = (1.3, 1),  fontsize=16, loc='upper right')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>base_base_shd</th>\n",
       "      <th>cluster_cluster_shd</th>\n",
       "      <th>empty_graph_shd</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n_edges</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>15.0</th>\n",
       "      <td>26.782222</td>\n",
       "      <td>23.475556</td>\n",
       "      <td>35.751111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30.0</th>\n",
       "      <td>32.364444</td>\n",
       "      <td>27.582222</td>\n",
       "      <td>44.253333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50.0</th>\n",
       "      <td>43.017778</td>\n",
       "      <td>36.262222</td>\n",
       "      <td>53.933333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80.0</th>\n",
       "      <td>58.924444</td>\n",
       "      <td>49.853333</td>\n",
       "      <td>68.404444</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         base_base_shd  cluster_cluster_shd  empty_graph_shd\n",
       "n_edges                                                     \n",
       "15.0         26.782222            23.475556        35.751111\n",
       "30.0         32.364444            27.582222        44.253333\n",
       "50.0         43.017778            36.262222        53.933333\n",
       "80.0         58.924444            49.853333        68.404444"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('n_edges').mean()[dist_metrics]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1b547de4d60>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# colors = ['#A4C2F4', '#3333FF', '#4D68D5']\n",
    "colors = ['#9ABCE4', '#005293', '#5E94D4']\n",
    "data.groupby('n_edges').mean()[dist_metrics].plot(kind='bar', figsize=(10, 6), color = colors)\n",
    "plt.title('SHD for different number of edges', fontsize = 20)\n",
    "plt.xlabel('number of edges', fontsize=18)\n",
    "plt.tick_params(axis = 'both', which = 'major', labelsize = 14)\n",
    "plt.tick_params(axis = 'x', which = 'major', labelsize = 16, rotation = 0)\n",
    "label_names = ['PC SHD', 'Cluster PC SHD', 'Empty graph SHD']\n",
    "plt.legend(label_names, bbox_to_anchor = (1.42, 1),  fontsize=16, loc='upper right')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "mthesis",
   "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": "3.10.0"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}