k562_many_rounds.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d9d11952-37ef-41dc-9419-0fa8b53f7310",
   "metadata": {},
   "source": [
    "# Summarize the results of several rounds of active learning vs random sampling with K562 data\n",
    "Required inputs: `ModelFitting/K562/ManyRounds/init_10000_inc_3000/performance_summary.txt` and `ModelFitting/K562/ManyRounds/init_4000_inc_2000/performance_summary.txt`\n",
    "\n",
    "Outputs: TODO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "8aaa874f",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/ref/bclab/software/ryan/.conda/envs/active-learning/lib/python3.9/site-packages/tqdm/auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "import sys\n",
    "\n",
    "import numpy as np\n",
    "import scipy.io\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.patches as mpatches\n",
    "\n",
    "from src.mpra_tools import loaders, plot_utils\n",
    "\n",
    "from IPython.display import display\n",
    "\n",
    "modeling_dir = os.path.join(\"ModelFitting\", \"K562\", \"ManyRounds\")\n",
    "figures_dir = \"Figures\"\n",
    "plot_utils.set_manuscript_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e91cfa48-b996-4751-9a71-78446f09a8e0",
   "metadata": {},
   "source": [
    "Read in model performance metrics."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ab9569ea-b572-447f-9d3c-dc47b8e9b6aa",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_n_train(row, expt):\n",
    "    if row[\"Mode\"] == \"Full\":\n",
    "        filename = os.path.join(modeling_dir, expt, str(row[\"Fold\"]), row[\"Mode\"], \"train.mat\")\n",
    "    else:\n",
    "        filename = os.path.join(modeling_dir, expt, str(row[\"Fold\"]), row[\"Mode\"], row[\"Round\"], \"train.mat\")\n",
    "    train_mat = scipy.io.loadmat(filename)\n",
    "    labels = train_mat[\"activity_bin\"].flatten()\n",
    "    result = pd.Series({\n",
    "        \"n_train\": len(labels),\n",
    "        \"n_positive\": (labels == 1).sum()\n",
    "    })\n",
    "    result[\"frac_positive_in_train\"] = result[\"n_positive\"] / result[\"n_train\"]\n",
    "    return result\n",
    "\n",
    "def get_frac_pos_sampled(group):\n",
    "    return group[\"n_positive\"] / group.loc[group[\"Mode\"] == \"Full\", \"n_positive\"].squeeze()\n",
    "\n",
    "performance_dict = {}\n",
    "expts = [\"init_10000_inc_3000\", \"init_4000_inc_2000\"]\n",
    "for expt in expts:\n",
    "    df = loaders.load_data(\n",
    "        os.path.join(modeling_dir, expt, \"performance_summary.txt\"),\n",
    "        index_col=None,\n",
    "    )\n",
    "    df = df.rename(columns={\n",
    "        \"roc_auc\": \"AUROC\",\n",
    "        \"pr_auc\": \"AUPR\",\n",
    "    })\n",
    "    data_info = df.apply(get_n_train, axis=1, args=(expt,))\n",
    "    df = df.join(data_info)\n",
    "    df[\"frac_positive_sampled\"] = df.groupby(\"Fold\").apply(get_frac_pos_sampled).droplevel(level=\"Fold\")\n",
    "    performance_dict[expt] = df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "a0b1268e-62c9-4e09-8e10-2dafa22e3e4d",
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>Mode</th>\n",
       "      <th>Round</th>\n",
       "      <th colspan=\"2\" halign=\"left\">AUROC</th>\n",
       "      <th colspan=\"2\" halign=\"left\">AUPR</th>\n",
       "      <th colspan=\"2\" halign=\"left\">n_train</th>\n",
       "      <th colspan=\"2\" halign=\"left\">n_positive</th>\n",
       "      <th colspan=\"2\" halign=\"left\">frac_positive_in_train</th>\n",
       "      <th colspan=\"2\" halign=\"left\">frac_positive_sampled</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round1</td>\n",
       "      <td>0.818180</td>\n",
       "      <td>0.012692</td>\n",
       "      <td>0.692453</td>\n",
       "      <td>0.035537</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2880.2</td>\n",
       "      <td>60.228823</td>\n",
       "      <td>0.288020</td>\n",
       "      <td>0.006023</td>\n",
       "      <td>0.213494</td>\n",
       "      <td>0.003158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>Random</td>\n",
       "      <td>round1</td>\n",
       "      <td>0.818180</td>\n",
       "      <td>0.012692</td>\n",
       "      <td>0.692453</td>\n",
       "      <td>0.035537</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2880.2</td>\n",
       "      <td>60.228823</td>\n",
       "      <td>0.288020</td>\n",
       "      <td>0.006023</td>\n",
       "      <td>0.213494</td>\n",
       "      <td>0.003158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round2</td>\n",
       "      <td>0.832514</td>\n",
       "      <td>0.011626</td>\n",
       "      <td>0.718441</td>\n",
       "      <td>0.028537</td>\n",
       "      <td>13000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4179.4</td>\n",
       "      <td>184.410894</td>\n",
       "      <td>0.321492</td>\n",
       "      <td>0.014185</td>\n",
       "      <td>0.309827</td>\n",
       "      <td>0.013621</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Random</td>\n",
       "      <td>round2</td>\n",
       "      <td>0.835943</td>\n",
       "      <td>0.009468</td>\n",
       "      <td>0.724030</td>\n",
       "      <td>0.028638</td>\n",
       "      <td>13000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3741.8</td>\n",
       "      <td>79.219526</td>\n",
       "      <td>0.287831</td>\n",
       "      <td>0.006094</td>\n",
       "      <td>0.277359</td>\n",
       "      <td>0.004146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Random</td>\n",
       "      <td>round3</td>\n",
       "      <td>0.848373</td>\n",
       "      <td>0.009806</td>\n",
       "      <td>0.744334</td>\n",
       "      <td>0.029375</td>\n",
       "      <td>16000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4602.6</td>\n",
       "      <td>86.618192</td>\n",
       "      <td>0.287663</td>\n",
       "      <td>0.005414</td>\n",
       "      <td>0.341175</td>\n",
       "      <td>0.004735</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round3</td>\n",
       "      <td>0.849658</td>\n",
       "      <td>0.010917</td>\n",
       "      <td>0.744485</td>\n",
       "      <td>0.031738</td>\n",
       "      <td>16000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5388.3</td>\n",
       "      <td>414.458697</td>\n",
       "      <td>0.336769</td>\n",
       "      <td>0.025904</td>\n",
       "      <td>0.399337</td>\n",
       "      <td>0.029303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>Random</td>\n",
       "      <td>round4</td>\n",
       "      <td>0.856498</td>\n",
       "      <td>0.012672</td>\n",
       "      <td>0.756459</td>\n",
       "      <td>0.033923</td>\n",
       "      <td>19000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5456.7</td>\n",
       "      <td>103.987232</td>\n",
       "      <td>0.287195</td>\n",
       "      <td>0.005473</td>\n",
       "      <td>0.404489</td>\n",
       "      <td>0.005883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round4</td>\n",
       "      <td>0.828647</td>\n",
       "      <td>0.092253</td>\n",
       "      <td>0.714267</td>\n",
       "      <td>0.148421</td>\n",
       "      <td>19000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6388.6</td>\n",
       "      <td>697.058773</td>\n",
       "      <td>0.336242</td>\n",
       "      <td>0.036687</td>\n",
       "      <td>0.473625</td>\n",
       "      <td>0.052056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>Random</td>\n",
       "      <td>round5</td>\n",
       "      <td>0.864154</td>\n",
       "      <td>0.007370</td>\n",
       "      <td>0.768718</td>\n",
       "      <td>0.022895</td>\n",
       "      <td>22000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6317.3</td>\n",
       "      <td>117.535857</td>\n",
       "      <td>0.287150</td>\n",
       "      <td>0.005343</td>\n",
       "      <td>0.468282</td>\n",
       "      <td>0.006500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round5</td>\n",
       "      <td>0.838678</td>\n",
       "      <td>0.095505</td>\n",
       "      <td>0.732723</td>\n",
       "      <td>0.146276</td>\n",
       "      <td>22000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7408.7</td>\n",
       "      <td>704.482797</td>\n",
       "      <td>0.336759</td>\n",
       "      <td>0.032022</td>\n",
       "      <td>0.549311</td>\n",
       "      <td>0.053219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>Random</td>\n",
       "      <td>round6</td>\n",
       "      <td>0.870538</td>\n",
       "      <td>0.007876</td>\n",
       "      <td>0.779058</td>\n",
       "      <td>0.022371</td>\n",
       "      <td>25000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7183.0</td>\n",
       "      <td>139.529368</td>\n",
       "      <td>0.287320</td>\n",
       "      <td>0.005581</td>\n",
       "      <td>0.532436</td>\n",
       "      <td>0.006489</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round6</td>\n",
       "      <td>0.813028</td>\n",
       "      <td>0.121586</td>\n",
       "      <td>0.692945</td>\n",
       "      <td>0.182447</td>\n",
       "      <td>25000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8458.3</td>\n",
       "      <td>690.337124</td>\n",
       "      <td>0.338332</td>\n",
       "      <td>0.027613</td>\n",
       "      <td>0.627128</td>\n",
       "      <td>0.052597</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>Random</td>\n",
       "      <td>round7</td>\n",
       "      <td>0.874397</td>\n",
       "      <td>0.007730</td>\n",
       "      <td>0.783839</td>\n",
       "      <td>0.024882</td>\n",
       "      <td>28000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8017.8</td>\n",
       "      <td>148.487785</td>\n",
       "      <td>0.286350</td>\n",
       "      <td>0.005303</td>\n",
       "      <td>0.594326</td>\n",
       "      <td>0.007436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round7</td>\n",
       "      <td>0.877959</td>\n",
       "      <td>0.011167</td>\n",
       "      <td>0.790812</td>\n",
       "      <td>0.028749</td>\n",
       "      <td>28000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9419.2</td>\n",
       "      <td>686.508849</td>\n",
       "      <td>0.336400</td>\n",
       "      <td>0.024518</td>\n",
       "      <td>0.698280</td>\n",
       "      <td>0.051318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round8</td>\n",
       "      <td>0.885376</td>\n",
       "      <td>0.006687</td>\n",
       "      <td>0.803143</td>\n",
       "      <td>0.022791</td>\n",
       "      <td>31000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>10171.4</td>\n",
       "      <td>567.373520</td>\n",
       "      <td>0.328110</td>\n",
       "      <td>0.018302</td>\n",
       "      <td>0.754182</td>\n",
       "      <td>0.045011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>Random</td>\n",
       "      <td>round8</td>\n",
       "      <td>0.879721</td>\n",
       "      <td>0.008476</td>\n",
       "      <td>0.792363</td>\n",
       "      <td>0.024106</td>\n",
       "      <td>31000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8869.5</td>\n",
       "      <td>160.191032</td>\n",
       "      <td>0.286113</td>\n",
       "      <td>0.005167</td>\n",
       "      <td>0.657456</td>\n",
       "      <td>0.007411</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>Random</td>\n",
       "      <td>round9</td>\n",
       "      <td>0.883414</td>\n",
       "      <td>0.006410</td>\n",
       "      <td>0.799235</td>\n",
       "      <td>0.018831</td>\n",
       "      <td>34000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9713.6</td>\n",
       "      <td>164.094959</td>\n",
       "      <td>0.285694</td>\n",
       "      <td>0.004826</td>\n",
       "      <td>0.720029</td>\n",
       "      <td>0.007108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round9</td>\n",
       "      <td>0.889557</td>\n",
       "      <td>0.006631</td>\n",
       "      <td>0.808704</td>\n",
       "      <td>0.020756</td>\n",
       "      <td>34000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>10969.4</td>\n",
       "      <td>401.609318</td>\n",
       "      <td>0.322629</td>\n",
       "      <td>0.011812</td>\n",
       "      <td>0.813182</td>\n",
       "      <td>0.029695</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round10</td>\n",
       "      <td>0.892827</td>\n",
       "      <td>0.007113</td>\n",
       "      <td>0.813829</td>\n",
       "      <td>0.022172</td>\n",
       "      <td>37000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>11530.3</td>\n",
       "      <td>223.118329</td>\n",
       "      <td>0.311630</td>\n",
       "      <td>0.006030</td>\n",
       "      <td>0.854798</td>\n",
       "      <td>0.018354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>Random</td>\n",
       "      <td>round10</td>\n",
       "      <td>0.887454</td>\n",
       "      <td>0.006690</td>\n",
       "      <td>0.806056</td>\n",
       "      <td>0.021250</td>\n",
       "      <td>37000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>10557.7</td>\n",
       "      <td>179.106455</td>\n",
       "      <td>0.285343</td>\n",
       "      <td>0.004841</td>\n",
       "      <td>0.782597</td>\n",
       "      <td>0.007628</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Random</td>\n",
       "      <td>round11</td>\n",
       "      <td>0.890131</td>\n",
       "      <td>0.006944</td>\n",
       "      <td>0.809167</td>\n",
       "      <td>0.020968</td>\n",
       "      <td>40000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>11423.3</td>\n",
       "      <td>191.145698</td>\n",
       "      <td>0.285582</td>\n",
       "      <td>0.004779</td>\n",
       "      <td>0.846760</td>\n",
       "      <td>0.007961</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round11</td>\n",
       "      <td>0.894914</td>\n",
       "      <td>0.006948</td>\n",
       "      <td>0.816381</td>\n",
       "      <td>0.022050</td>\n",
       "      <td>40000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>12010.5</td>\n",
       "      <td>233.902283</td>\n",
       "      <td>0.300262</td>\n",
       "      <td>0.005848</td>\n",
       "      <td>0.890355</td>\n",
       "      <td>0.016800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Random</td>\n",
       "      <td>round12</td>\n",
       "      <td>0.891784</td>\n",
       "      <td>0.006749</td>\n",
       "      <td>0.812859</td>\n",
       "      <td>0.019550</td>\n",
       "      <td>43000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>12279.6</td>\n",
       "      <td>197.582163</td>\n",
       "      <td>0.285572</td>\n",
       "      <td>0.004595</td>\n",
       "      <td>0.910239</td>\n",
       "      <td>0.008100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round12</td>\n",
       "      <td>0.894832</td>\n",
       "      <td>0.006386</td>\n",
       "      <td>0.816645</td>\n",
       "      <td>0.019919</td>\n",
       "      <td>43000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>12499.3</td>\n",
       "      <td>246.258694</td>\n",
       "      <td>0.290681</td>\n",
       "      <td>0.005727</td>\n",
       "      <td>0.926506</td>\n",
       "      <td>0.011865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Random</td>\n",
       "      <td>round13</td>\n",
       "      <td>0.893289</td>\n",
       "      <td>0.007558</td>\n",
       "      <td>0.813812</td>\n",
       "      <td>0.023078</td>\n",
       "      <td>46000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>13132.0</td>\n",
       "      <td>210.344691</td>\n",
       "      <td>0.285478</td>\n",
       "      <td>0.004573</td>\n",
       "      <td>0.973431</td>\n",
       "      <td>0.009316</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round13</td>\n",
       "      <td>0.894907</td>\n",
       "      <td>0.008538</td>\n",
       "      <td>0.817072</td>\n",
       "      <td>0.021641</td>\n",
       "      <td>46000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>13001.5</td>\n",
       "      <td>190.261370</td>\n",
       "      <td>0.282641</td>\n",
       "      <td>0.004136</td>\n",
       "      <td>0.963787</td>\n",
       "      <td>0.010389</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Random</td>\n",
       "      <td>round14</td>\n",
       "      <td>0.895556</td>\n",
       "      <td>0.007401</td>\n",
       "      <td>0.817689</td>\n",
       "      <td>0.023745</td>\n",
       "      <td>47274.4</td>\n",
       "      <td>463.840777</td>\n",
       "      <td>13490.4</td>\n",
       "      <td>171.697538</td>\n",
       "      <td>0.285389</td>\n",
       "      <td>0.004604</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round14</td>\n",
       "      <td>0.895556</td>\n",
       "      <td>0.007401</td>\n",
       "      <td>0.817689</td>\n",
       "      <td>0.023745</td>\n",
       "      <td>47274.4</td>\n",
       "      <td>463.840777</td>\n",
       "      <td>13490.4</td>\n",
       "      <td>171.697538</td>\n",
       "      <td>0.285389</td>\n",
       "      <td>0.004604</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>Full</td>\n",
       "      <td>full</td>\n",
       "      <td>0.895556</td>\n",
       "      <td>0.007401</td>\n",
       "      <td>0.817689</td>\n",
       "      <td>0.023745</td>\n",
       "      <td>47274.4</td>\n",
       "      <td>463.840777</td>\n",
       "      <td>13490.4</td>\n",
       "      <td>171.697538</td>\n",
       "      <td>0.285389</td>\n",
       "      <td>0.004604</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Mode    Round     AUROC                AUPR            n_train  \\\n",
       "                          mean       std      mean       std     mean   \n",
       "0   Entropy   round1  0.818180  0.012692  0.692453  0.035537  10000.0   \n",
       "15   Random   round1  0.818180  0.012692  0.692453  0.035537  10000.0   \n",
       "6   Entropy   round2  0.832514  0.011626  0.718441  0.028537  13000.0   \n",
       "21   Random   round2  0.835943  0.009468  0.724030  0.028638  13000.0   \n",
       "22   Random   round3  0.848373  0.009806  0.744334  0.029375  16000.0   \n",
       "7   Entropy   round3  0.849658  0.010917  0.744485  0.031738  16000.0   \n",
       "23   Random   round4  0.856498  0.012672  0.756459  0.033923  19000.0   \n",
       "8   Entropy   round4  0.828647  0.092253  0.714267  0.148421  19000.0   \n",
       "24   Random   round5  0.864154  0.007370  0.768718  0.022895  22000.0   \n",
       "9   Entropy   round5  0.838678  0.095505  0.732723  0.146276  22000.0   \n",
       "25   Random   round6  0.870538  0.007876  0.779058  0.022371  25000.0   \n",
       "10  Entropy   round6  0.813028  0.121586  0.692945  0.182447  25000.0   \n",
       "26   Random   round7  0.874397  0.007730  0.783839  0.024882  28000.0   \n",
       "11  Entropy   round7  0.877959  0.011167  0.790812  0.028749  28000.0   \n",
       "12  Entropy   round8  0.885376  0.006687  0.803143  0.022791  31000.0   \n",
       "27   Random   round8  0.879721  0.008476  0.792363  0.024106  31000.0   \n",
       "28   Random   round9  0.883414  0.006410  0.799235  0.018831  34000.0   \n",
       "13  Entropy   round9  0.889557  0.006631  0.808704  0.020756  34000.0   \n",
       "1   Entropy  round10  0.892827  0.007113  0.813829  0.022172  37000.0   \n",
       "16   Random  round10  0.887454  0.006690  0.806056  0.021250  37000.0   \n",
       "17   Random  round11  0.890131  0.006944  0.809167  0.020968  40000.0   \n",
       "2   Entropy  round11  0.894914  0.006948  0.816381  0.022050  40000.0   \n",
       "18   Random  round12  0.891784  0.006749  0.812859  0.019550  43000.0   \n",
       "3   Entropy  round12  0.894832  0.006386  0.816645  0.019919  43000.0   \n",
       "19   Random  round13  0.893289  0.007558  0.813812  0.023078  46000.0   \n",
       "4   Entropy  round13  0.894907  0.008538  0.817072  0.021641  46000.0   \n",
       "20   Random  round14  0.895556  0.007401  0.817689  0.023745  47274.4   \n",
       "5   Entropy  round14  0.895556  0.007401  0.817689  0.023745  47274.4   \n",
       "14     Full     full  0.895556  0.007401  0.817689  0.023745  47274.4   \n",
       "\n",
       "               n_positive             frac_positive_in_train            \\\n",
       "           std       mean         std                   mean       std   \n",
       "0     0.000000     2880.2   60.228823               0.288020  0.006023   \n",
       "15    0.000000     2880.2   60.228823               0.288020  0.006023   \n",
       "6     0.000000     4179.4  184.410894               0.321492  0.014185   \n",
       "21    0.000000     3741.8   79.219526               0.287831  0.006094   \n",
       "22    0.000000     4602.6   86.618192               0.287663  0.005414   \n",
       "7     0.000000     5388.3  414.458697               0.336769  0.025904   \n",
       "23    0.000000     5456.7  103.987232               0.287195  0.005473   \n",
       "8     0.000000     6388.6  697.058773               0.336242  0.036687   \n",
       "24    0.000000     6317.3  117.535857               0.287150  0.005343   \n",
       "9     0.000000     7408.7  704.482797               0.336759  0.032022   \n",
       "25    0.000000     7183.0  139.529368               0.287320  0.005581   \n",
       "10    0.000000     8458.3  690.337124               0.338332  0.027613   \n",
       "26    0.000000     8017.8  148.487785               0.286350  0.005303   \n",
       "11    0.000000     9419.2  686.508849               0.336400  0.024518   \n",
       "12    0.000000    10171.4  567.373520               0.328110  0.018302   \n",
       "27    0.000000     8869.5  160.191032               0.286113  0.005167   \n",
       "28    0.000000     9713.6  164.094959               0.285694  0.004826   \n",
       "13    0.000000    10969.4  401.609318               0.322629  0.011812   \n",
       "1     0.000000    11530.3  223.118329               0.311630  0.006030   \n",
       "16    0.000000    10557.7  179.106455               0.285343  0.004841   \n",
       "17    0.000000    11423.3  191.145698               0.285582  0.004779   \n",
       "2     0.000000    12010.5  233.902283               0.300262  0.005848   \n",
       "18    0.000000    12279.6  197.582163               0.285572  0.004595   \n",
       "3     0.000000    12499.3  246.258694               0.290681  0.005727   \n",
       "19    0.000000    13132.0  210.344691               0.285478  0.004573   \n",
       "4     0.000000    13001.5  190.261370               0.282641  0.004136   \n",
       "20  463.840777    13490.4  171.697538               0.285389  0.004604   \n",
       "5   463.840777    13490.4  171.697538               0.285389  0.004604   \n",
       "14  463.840777    13490.4  171.697538               0.285389  0.004604   \n",
       "\n",
       "   frac_positive_sampled            \n",
       "                    mean       std  \n",
       "0               0.213494  0.003158  \n",
       "15              0.213494  0.003158  \n",
       "6               0.309827  0.013621  \n",
       "21              0.277359  0.004146  \n",
       "22              0.341175  0.004735  \n",
       "7               0.399337  0.029303  \n",
       "23              0.404489  0.005883  \n",
       "8               0.473625  0.052056  \n",
       "24              0.468282  0.006500  \n",
       "9               0.549311  0.053219  \n",
       "25              0.532436  0.006489  \n",
       "10              0.627128  0.052597  \n",
       "26              0.594326  0.007436  \n",
       "11              0.698280  0.051318  \n",
       "12              0.754182  0.045011  \n",
       "27              0.657456  0.007411  \n",
       "28              0.720029  0.007108  \n",
       "13              0.813182  0.029695  \n",
       "1               0.854798  0.018354  \n",
       "16              0.782597  0.007628  \n",
       "17              0.846760  0.007961  \n",
       "2               0.890355  0.016800  \n",
       "18              0.910239  0.008100  \n",
       "3               0.926506  0.011865  \n",
       "19              0.973431  0.009316  \n",
       "4               0.963787  0.010389  \n",
       "20              1.000000  0.000000  \n",
       "5               1.000000  0.000000  \n",
       "14              1.000000  0.000000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>Mode</th>\n",
       "      <th>Round</th>\n",
       "      <th colspan=\"2\" halign=\"left\">AUROC</th>\n",
       "      <th colspan=\"2\" halign=\"left\">AUPR</th>\n",
       "      <th colspan=\"2\" halign=\"left\">n_train</th>\n",
       "      <th colspan=\"2\" halign=\"left\">n_positive</th>\n",
       "      <th colspan=\"2\" halign=\"left\">frac_positive_in_train</th>\n",
       "      <th colspan=\"2\" halign=\"left\">frac_positive_sampled</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round1</td>\n",
       "      <td>0.607349</td>\n",
       "      <td>0.082208</td>\n",
       "      <td>0.360280</td>\n",
       "      <td>0.106649</td>\n",
       "      <td>4000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1155.6</td>\n",
       "      <td>34.124608</td>\n",
       "      <td>0.288900</td>\n",
       "      <td>0.008531</td>\n",
       "      <td>0.085651</td>\n",
       "      <td>0.001838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>Random</td>\n",
       "      <td>round1</td>\n",
       "      <td>0.607349</td>\n",
       "      <td>0.082208</td>\n",
       "      <td>0.360280</td>\n",
       "      <td>0.106649</td>\n",
       "      <td>4000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1155.6</td>\n",
       "      <td>34.124608</td>\n",
       "      <td>0.288900</td>\n",
       "      <td>0.008531</td>\n",
       "      <td>0.085651</td>\n",
       "      <td>0.001838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round2</td>\n",
       "      <td>0.751011</td>\n",
       "      <td>0.072391</td>\n",
       "      <td>0.562989</td>\n",
       "      <td>0.102551</td>\n",
       "      <td>6000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2128.8</td>\n",
       "      <td>141.105792</td>\n",
       "      <td>0.354800</td>\n",
       "      <td>0.023518</td>\n",
       "      <td>0.157743</td>\n",
       "      <td>0.009223</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>Random</td>\n",
       "      <td>round2</td>\n",
       "      <td>0.734637</td>\n",
       "      <td>0.072556</td>\n",
       "      <td>0.508147</td>\n",
       "      <td>0.109664</td>\n",
       "      <td>6000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1725.0</td>\n",
       "      <td>41.387062</td>\n",
       "      <td>0.287500</td>\n",
       "      <td>0.006898</td>\n",
       "      <td>0.127856</td>\n",
       "      <td>0.001790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>Random</td>\n",
       "      <td>round3</td>\n",
       "      <td>0.696811</td>\n",
       "      <td>0.098799</td>\n",
       "      <td>0.454964</td>\n",
       "      <td>0.124715</td>\n",
       "      <td>8000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2300.1</td>\n",
       "      <td>66.865454</td>\n",
       "      <td>0.287513</td>\n",
       "      <td>0.008358</td>\n",
       "      <td>0.170473</td>\n",
       "      <td>0.003172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round3</td>\n",
       "      <td>0.791075</td>\n",
       "      <td>0.012570</td>\n",
       "      <td>0.639630</td>\n",
       "      <td>0.061464</td>\n",
       "      <td>8000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3009.9</td>\n",
       "      <td>183.760986</td>\n",
       "      <td>0.376238</td>\n",
       "      <td>0.022970</td>\n",
       "      <td>0.223017</td>\n",
       "      <td>0.011291</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round4</td>\n",
       "      <td>0.801295</td>\n",
       "      <td>0.017320</td>\n",
       "      <td>0.658504</td>\n",
       "      <td>0.038997</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3917.4</td>\n",
       "      <td>214.633641</td>\n",
       "      <td>0.391740</td>\n",
       "      <td>0.021463</td>\n",
       "      <td>0.290391</td>\n",
       "      <td>0.015636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>Random</td>\n",
       "      <td>round4</td>\n",
       "      <td>0.724926</td>\n",
       "      <td>0.060389</td>\n",
       "      <td>0.458783</td>\n",
       "      <td>0.090679</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2882.1</td>\n",
       "      <td>64.972558</td>\n",
       "      <td>0.288210</td>\n",
       "      <td>0.006497</td>\n",
       "      <td>0.213625</td>\n",
       "      <td>0.002888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>Random</td>\n",
       "      <td>round5</td>\n",
       "      <td>0.771923</td>\n",
       "      <td>0.074249</td>\n",
       "      <td>0.571432</td>\n",
       "      <td>0.119884</td>\n",
       "      <td>12000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3444.3</td>\n",
       "      <td>70.669890</td>\n",
       "      <td>0.287025</td>\n",
       "      <td>0.005889</td>\n",
       "      <td>0.255300</td>\n",
       "      <td>0.002875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round5</td>\n",
       "      <td>0.808689</td>\n",
       "      <td>0.015066</td>\n",
       "      <td>0.660321</td>\n",
       "      <td>0.055366</td>\n",
       "      <td>12000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4770.7</td>\n",
       "      <td>279.975416</td>\n",
       "      <td>0.397558</td>\n",
       "      <td>0.023331</td>\n",
       "      <td>0.353653</td>\n",
       "      <td>0.020640</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>Random</td>\n",
       "      <td>round6</td>\n",
       "      <td>0.784658</td>\n",
       "      <td>0.037743</td>\n",
       "      <td>0.556221</td>\n",
       "      <td>0.075897</td>\n",
       "      <td>14000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4011.3</td>\n",
       "      <td>83.272044</td>\n",
       "      <td>0.286521</td>\n",
       "      <td>0.005948</td>\n",
       "      <td>0.297330</td>\n",
       "      <td>0.003782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round6</td>\n",
       "      <td>0.817682</td>\n",
       "      <td>0.025080</td>\n",
       "      <td>0.682271</td>\n",
       "      <td>0.081054</td>\n",
       "      <td>14000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5598.5</td>\n",
       "      <td>341.928274</td>\n",
       "      <td>0.399893</td>\n",
       "      <td>0.024423</td>\n",
       "      <td>0.414926</td>\n",
       "      <td>0.023495</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>Random</td>\n",
       "      <td>round7</td>\n",
       "      <td>0.813147</td>\n",
       "      <td>0.026865</td>\n",
       "      <td>0.637022</td>\n",
       "      <td>0.096055</td>\n",
       "      <td>16000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>4584.8</td>\n",
       "      <td>109.645692</td>\n",
       "      <td>0.286550</td>\n",
       "      <td>0.006853</td>\n",
       "      <td>0.339834</td>\n",
       "      <td>0.005541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round7</td>\n",
       "      <td>0.819427</td>\n",
       "      <td>0.023545</td>\n",
       "      <td>0.679116</td>\n",
       "      <td>0.064703</td>\n",
       "      <td>16000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6508.5</td>\n",
       "      <td>507.066345</td>\n",
       "      <td>0.406781</td>\n",
       "      <td>0.031692</td>\n",
       "      <td>0.482342</td>\n",
       "      <td>0.035319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round8</td>\n",
       "      <td>0.810785</td>\n",
       "      <td>0.034767</td>\n",
       "      <td>0.662223</td>\n",
       "      <td>0.085232</td>\n",
       "      <td>18000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7548.8</td>\n",
       "      <td>558.538728</td>\n",
       "      <td>0.419378</td>\n",
       "      <td>0.031030</td>\n",
       "      <td>0.559383</td>\n",
       "      <td>0.037783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>Random</td>\n",
       "      <td>round8</td>\n",
       "      <td>0.799025</td>\n",
       "      <td>0.066430</td>\n",
       "      <td>0.620397</td>\n",
       "      <td>0.113105</td>\n",
       "      <td>18000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5152.8</td>\n",
       "      <td>120.494168</td>\n",
       "      <td>0.286267</td>\n",
       "      <td>0.006694</td>\n",
       "      <td>0.381941</td>\n",
       "      <td>0.006375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round9</td>\n",
       "      <td>0.840185</td>\n",
       "      <td>0.008781</td>\n",
       "      <td>0.726469</td>\n",
       "      <td>0.045156</td>\n",
       "      <td>20000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8382.5</td>\n",
       "      <td>585.581809</td>\n",
       "      <td>0.419125</td>\n",
       "      <td>0.029279</td>\n",
       "      <td>0.621265</td>\n",
       "      <td>0.041099</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>Random</td>\n",
       "      <td>round9</td>\n",
       "      <td>0.818706</td>\n",
       "      <td>0.021615</td>\n",
       "      <td>0.650075</td>\n",
       "      <td>0.075893</td>\n",
       "      <td>20000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5720.9</td>\n",
       "      <td>134.639808</td>\n",
       "      <td>0.286045</td>\n",
       "      <td>0.006732</td>\n",
       "      <td>0.424047</td>\n",
       "      <td>0.006908</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round10</td>\n",
       "      <td>0.843760</td>\n",
       "      <td>0.014879</td>\n",
       "      <td>0.723512</td>\n",
       "      <td>0.049724</td>\n",
       "      <td>22000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9032.5</td>\n",
       "      <td>691.476078</td>\n",
       "      <td>0.410568</td>\n",
       "      <td>0.031431</td>\n",
       "      <td>0.669392</td>\n",
       "      <td>0.048173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>Random</td>\n",
       "      <td>round10</td>\n",
       "      <td>0.839983</td>\n",
       "      <td>0.020217</td>\n",
       "      <td>0.709592</td>\n",
       "      <td>0.058935</td>\n",
       "      <td>22000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6288.5</td>\n",
       "      <td>142.602205</td>\n",
       "      <td>0.285841</td>\n",
       "      <td>0.006482</td>\n",
       "      <td>0.466118</td>\n",
       "      <td>0.006946</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round11</td>\n",
       "      <td>0.844047</td>\n",
       "      <td>0.021200</td>\n",
       "      <td>0.731557</td>\n",
       "      <td>0.055287</td>\n",
       "      <td>24000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9689.7</td>\n",
       "      <td>607.467064</td>\n",
       "      <td>0.403737</td>\n",
       "      <td>0.025311</td>\n",
       "      <td>0.718079</td>\n",
       "      <td>0.040583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>Random</td>\n",
       "      <td>round11</td>\n",
       "      <td>0.844123</td>\n",
       "      <td>0.016124</td>\n",
       "      <td>0.708666</td>\n",
       "      <td>0.056873</td>\n",
       "      <td>24000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6858.8</td>\n",
       "      <td>152.321152</td>\n",
       "      <td>0.285783</td>\n",
       "      <td>0.006347</td>\n",
       "      <td>0.508385</td>\n",
       "      <td>0.006750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round12</td>\n",
       "      <td>0.838876</td>\n",
       "      <td>0.018603</td>\n",
       "      <td>0.722446</td>\n",
       "      <td>0.035737</td>\n",
       "      <td>26000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>10163.3</td>\n",
       "      <td>572.547339</td>\n",
       "      <td>0.390896</td>\n",
       "      <td>0.022021</td>\n",
       "      <td>0.753215</td>\n",
       "      <td>0.038090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>Random</td>\n",
       "      <td>round12</td>\n",
       "      <td>0.850861</td>\n",
       "      <td>0.016126</td>\n",
       "      <td>0.733666</td>\n",
       "      <td>0.061500</td>\n",
       "      <td>26000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7426.3</td>\n",
       "      <td>157.555669</td>\n",
       "      <td>0.285627</td>\n",
       "      <td>0.006060</td>\n",
       "      <td>0.550459</td>\n",
       "      <td>0.007278</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round13</td>\n",
       "      <td>0.830397</td>\n",
       "      <td>0.026867</td>\n",
       "      <td>0.679576</td>\n",
       "      <td>0.088138</td>\n",
       "      <td>28000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>10785.4</td>\n",
       "      <td>749.561976</td>\n",
       "      <td>0.385193</td>\n",
       "      <td>0.026770</td>\n",
       "      <td>0.799386</td>\n",
       "      <td>0.053293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>Random</td>\n",
       "      <td>round13</td>\n",
       "      <td>0.840382</td>\n",
       "      <td>0.019362</td>\n",
       "      <td>0.712696</td>\n",
       "      <td>0.068100</td>\n",
       "      <td>28000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7999.7</td>\n",
       "      <td>169.568112</td>\n",
       "      <td>0.285704</td>\n",
       "      <td>0.006056</td>\n",
       "      <td>0.592959</td>\n",
       "      <td>0.007635</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round14</td>\n",
       "      <td>0.862656</td>\n",
       "      <td>0.014235</td>\n",
       "      <td>0.760514</td>\n",
       "      <td>0.034631</td>\n",
       "      <td>30000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>11195.1</td>\n",
       "      <td>809.118649</td>\n",
       "      <td>0.373170</td>\n",
       "      <td>0.026971</td>\n",
       "      <td>0.829859</td>\n",
       "      <td>0.059562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>Random</td>\n",
       "      <td>round14</td>\n",
       "      <td>0.851125</td>\n",
       "      <td>0.018547</td>\n",
       "      <td>0.725901</td>\n",
       "      <td>0.058836</td>\n",
       "      <td>30000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8562.9</td>\n",
       "      <td>164.746573</td>\n",
       "      <td>0.285430</td>\n",
       "      <td>0.005492</td>\n",
       "      <td>0.634716</td>\n",
       "      <td>0.007094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>Random</td>\n",
       "      <td>round15</td>\n",
       "      <td>0.855758</td>\n",
       "      <td>0.018662</td>\n",
       "      <td>0.746981</td>\n",
       "      <td>0.033667</td>\n",
       "      <td>32000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9139.3</td>\n",
       "      <td>173.271175</td>\n",
       "      <td>0.285603</td>\n",
       "      <td>0.005415</td>\n",
       "      <td>0.677443</td>\n",
       "      <td>0.007458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round15</td>\n",
       "      <td>0.853565</td>\n",
       "      <td>0.020646</td>\n",
       "      <td>0.736733</td>\n",
       "      <td>0.057999</td>\n",
       "      <td>32000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>11523.3</td>\n",
       "      <td>732.317486</td>\n",
       "      <td>0.360103</td>\n",
       "      <td>0.022885</td>\n",
       "      <td>0.854197</td>\n",
       "      <td>0.053852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>Random</td>\n",
       "      <td>round16</td>\n",
       "      <td>0.858045</td>\n",
       "      <td>0.021257</td>\n",
       "      <td>0.758950</td>\n",
       "      <td>0.037691</td>\n",
       "      <td>34000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9710.0</td>\n",
       "      <td>174.094483</td>\n",
       "      <td>0.285588</td>\n",
       "      <td>0.005120</td>\n",
       "      <td>0.719752</td>\n",
       "      <td>0.007293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round16</td>\n",
       "      <td>0.856546</td>\n",
       "      <td>0.019944</td>\n",
       "      <td>0.738329</td>\n",
       "      <td>0.049340</td>\n",
       "      <td>34000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>11730.1</td>\n",
       "      <td>645.019629</td>\n",
       "      <td>0.345003</td>\n",
       "      <td>0.018971</td>\n",
       "      <td>0.869556</td>\n",
       "      <td>0.047749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round17</td>\n",
       "      <td>0.840414</td>\n",
       "      <td>0.051824</td>\n",
       "      <td>0.720832</td>\n",
       "      <td>0.102430</td>\n",
       "      <td>36000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>11976.6</td>\n",
       "      <td>624.401438</td>\n",
       "      <td>0.332683</td>\n",
       "      <td>0.017344</td>\n",
       "      <td>0.887820</td>\n",
       "      <td>0.045896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>Random</td>\n",
       "      <td>round17</td>\n",
       "      <td>0.864072</td>\n",
       "      <td>0.012789</td>\n",
       "      <td>0.766862</td>\n",
       "      <td>0.033886</td>\n",
       "      <td>36000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>10272.6</td>\n",
       "      <td>181.575329</td>\n",
       "      <td>0.285350</td>\n",
       "      <td>0.005044</td>\n",
       "      <td>0.761458</td>\n",
       "      <td>0.007710</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round18</td>\n",
       "      <td>0.857864</td>\n",
       "      <td>0.026402</td>\n",
       "      <td>0.746541</td>\n",
       "      <td>0.056345</td>\n",
       "      <td>38000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>12325.6</td>\n",
       "      <td>409.222623</td>\n",
       "      <td>0.324358</td>\n",
       "      <td>0.010769</td>\n",
       "      <td>0.913685</td>\n",
       "      <td>0.029119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>Random</td>\n",
       "      <td>round18</td>\n",
       "      <td>0.861156</td>\n",
       "      <td>0.015207</td>\n",
       "      <td>0.758109</td>\n",
       "      <td>0.036705</td>\n",
       "      <td>38000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>10846.5</td>\n",
       "      <td>188.490053</td>\n",
       "      <td>0.285434</td>\n",
       "      <td>0.004960</td>\n",
       "      <td>0.804001</td>\n",
       "      <td>0.008088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round19</td>\n",
       "      <td>0.862374</td>\n",
       "      <td>0.019715</td>\n",
       "      <td>0.761665</td>\n",
       "      <td>0.040171</td>\n",
       "      <td>40000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>12586.7</td>\n",
       "      <td>285.605808</td>\n",
       "      <td>0.314667</td>\n",
       "      <td>0.007140</td>\n",
       "      <td>0.933040</td>\n",
       "      <td>0.019289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>Random</td>\n",
       "      <td>round19</td>\n",
       "      <td>0.861718</td>\n",
       "      <td>0.014527</td>\n",
       "      <td>0.750534</td>\n",
       "      <td>0.048973</td>\n",
       "      <td>40000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>11429.2</td>\n",
       "      <td>184.399205</td>\n",
       "      <td>0.285730</td>\n",
       "      <td>0.004610</td>\n",
       "      <td>0.847207</td>\n",
       "      <td>0.008150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>Random</td>\n",
       "      <td>round20</td>\n",
       "      <td>0.859182</td>\n",
       "      <td>0.018326</td>\n",
       "      <td>0.751793</td>\n",
       "      <td>0.047360</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>11985.3</td>\n",
       "      <td>187.374284</td>\n",
       "      <td>0.285364</td>\n",
       "      <td>0.004461</td>\n",
       "      <td>0.888434</td>\n",
       "      <td>0.008392</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round20</td>\n",
       "      <td>0.852305</td>\n",
       "      <td>0.013715</td>\n",
       "      <td>0.750159</td>\n",
       "      <td>0.024893</td>\n",
       "      <td>42000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>12729.3</td>\n",
       "      <td>274.472848</td>\n",
       "      <td>0.303079</td>\n",
       "      <td>0.006535</td>\n",
       "      <td>0.943600</td>\n",
       "      <td>0.017612</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>Random</td>\n",
       "      <td>round21</td>\n",
       "      <td>0.861966</td>\n",
       "      <td>0.012957</td>\n",
       "      <td>0.763601</td>\n",
       "      <td>0.026245</td>\n",
       "      <td>44000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>12560.2</td>\n",
       "      <td>208.806556</td>\n",
       "      <td>0.285459</td>\n",
       "      <td>0.004746</td>\n",
       "      <td>0.931040</td>\n",
       "      <td>0.009256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round21</td>\n",
       "      <td>0.863931</td>\n",
       "      <td>0.014728</td>\n",
       "      <td>0.762414</td>\n",
       "      <td>0.034367</td>\n",
       "      <td>44000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>12919.2</td>\n",
       "      <td>251.005887</td>\n",
       "      <td>0.293618</td>\n",
       "      <td>0.005705</td>\n",
       "      <td>0.957697</td>\n",
       "      <td>0.016768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round22</td>\n",
       "      <td>0.854380</td>\n",
       "      <td>0.026779</td>\n",
       "      <td>0.742682</td>\n",
       "      <td>0.059978</td>\n",
       "      <td>46000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>13097.8</td>\n",
       "      <td>193.728791</td>\n",
       "      <td>0.284735</td>\n",
       "      <td>0.004211</td>\n",
       "      <td>0.970929</td>\n",
       "      <td>0.011098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>Random</td>\n",
       "      <td>round22</td>\n",
       "      <td>0.860083</td>\n",
       "      <td>0.022169</td>\n",
       "      <td>0.760218</td>\n",
       "      <td>0.039313</td>\n",
       "      <td>46000.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>13132.7</td>\n",
       "      <td>212.225797</td>\n",
       "      <td>0.285493</td>\n",
       "      <td>0.004614</td>\n",
       "      <td>0.973484</td>\n",
       "      <td>0.009685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>Entropy</td>\n",
       "      <td>round23</td>\n",
       "      <td>0.862177</td>\n",
       "      <td>0.017081</td>\n",
       "      <td>0.758480</td>\n",
       "      <td>0.043383</td>\n",
       "      <td>47274.4</td>\n",
       "      <td>463.840777</td>\n",
       "      <td>13490.4</td>\n",
       "      <td>171.697538</td>\n",
       "      <td>0.285389</td>\n",
       "      <td>0.004604</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>Random</td>\n",
       "      <td>round23</td>\n",
       "      <td>0.862177</td>\n",
       "      <td>0.017081</td>\n",
       "      <td>0.758480</td>\n",
       "      <td>0.043383</td>\n",
       "      <td>47274.4</td>\n",
       "      <td>463.840777</td>\n",
       "      <td>13490.4</td>\n",
       "      <td>171.697538</td>\n",
       "      <td>0.285389</td>\n",
       "      <td>0.004604</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>Full</td>\n",
       "      <td>full</td>\n",
       "      <td>0.862177</td>\n",
       "      <td>0.017081</td>\n",
       "      <td>0.758480</td>\n",
       "      <td>0.043383</td>\n",
       "      <td>47274.4</td>\n",
       "      <td>463.840777</td>\n",
       "      <td>13490.4</td>\n",
       "      <td>171.697538</td>\n",
       "      <td>0.285389</td>\n",
       "      <td>0.004604</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Mode    Round     AUROC                AUPR            n_train  \\\n",
       "                          mean       std      mean       std     mean   \n",
       "0   Entropy   round1  0.607349  0.082208  0.360280  0.106649   4000.0   \n",
       "24   Random   round1  0.607349  0.082208  0.360280  0.106649   4000.0   \n",
       "11  Entropy   round2  0.751011  0.072391  0.562989  0.102551   6000.0   \n",
       "35   Random   round2  0.734637  0.072556  0.508147  0.109664   6000.0   \n",
       "40   Random   round3  0.696811  0.098799  0.454964  0.124715   8000.0   \n",
       "16  Entropy   round3  0.791075  0.012570  0.639630  0.061464   8000.0   \n",
       "17  Entropy   round4  0.801295  0.017320  0.658504  0.038997  10000.0   \n",
       "41   Random   round4  0.724926  0.060389  0.458783  0.090679  10000.0   \n",
       "42   Random   round5  0.771923  0.074249  0.571432  0.119884  12000.0   \n",
       "18  Entropy   round5  0.808689  0.015066  0.660321  0.055366  12000.0   \n",
       "43   Random   round6  0.784658  0.037743  0.556221  0.075897  14000.0   \n",
       "19  Entropy   round6  0.817682  0.025080  0.682271  0.081054  14000.0   \n",
       "44   Random   round7  0.813147  0.026865  0.637022  0.096055  16000.0   \n",
       "20  Entropy   round7  0.819427  0.023545  0.679116  0.064703  16000.0   \n",
       "21  Entropy   round8  0.810785  0.034767  0.662223  0.085232  18000.0   \n",
       "45   Random   round8  0.799025  0.066430  0.620397  0.113105  18000.0   \n",
       "22  Entropy   round9  0.840185  0.008781  0.726469  0.045156  20000.0   \n",
       "46   Random   round9  0.818706  0.021615  0.650075  0.075893  20000.0   \n",
       "1   Entropy  round10  0.843760  0.014879  0.723512  0.049724  22000.0   \n",
       "25   Random  round10  0.839983  0.020217  0.709592  0.058935  22000.0   \n",
       "2   Entropy  round11  0.844047  0.021200  0.731557  0.055287  24000.0   \n",
       "26   Random  round11  0.844123  0.016124  0.708666  0.056873  24000.0   \n",
       "3   Entropy  round12  0.838876  0.018603  0.722446  0.035737  26000.0   \n",
       "27   Random  round12  0.850861  0.016126  0.733666  0.061500  26000.0   \n",
       "4   Entropy  round13  0.830397  0.026867  0.679576  0.088138  28000.0   \n",
       "28   Random  round13  0.840382  0.019362  0.712696  0.068100  28000.0   \n",
       "5   Entropy  round14  0.862656  0.014235  0.760514  0.034631  30000.0   \n",
       "29   Random  round14  0.851125  0.018547  0.725901  0.058836  30000.0   \n",
       "30   Random  round15  0.855758  0.018662  0.746981  0.033667  32000.0   \n",
       "6   Entropy  round15  0.853565  0.020646  0.736733  0.057999  32000.0   \n",
       "31   Random  round16  0.858045  0.021257  0.758950  0.037691  34000.0   \n",
       "7   Entropy  round16  0.856546  0.019944  0.738329  0.049340  34000.0   \n",
       "8   Entropy  round17  0.840414  0.051824  0.720832  0.102430  36000.0   \n",
       "32   Random  round17  0.864072  0.012789  0.766862  0.033886  36000.0   \n",
       "9   Entropy  round18  0.857864  0.026402  0.746541  0.056345  38000.0   \n",
       "33   Random  round18  0.861156  0.015207  0.758109  0.036705  38000.0   \n",
       "10  Entropy  round19  0.862374  0.019715  0.761665  0.040171  40000.0   \n",
       "34   Random  round19  0.861718  0.014527  0.750534  0.048973  40000.0   \n",
       "36   Random  round20  0.859182  0.018326  0.751793  0.047360  42000.0   \n",
       "12  Entropy  round20  0.852305  0.013715  0.750159  0.024893  42000.0   \n",
       "37   Random  round21  0.861966  0.012957  0.763601  0.026245  44000.0   \n",
       "13  Entropy  round21  0.863931  0.014728  0.762414  0.034367  44000.0   \n",
       "14  Entropy  round22  0.854380  0.026779  0.742682  0.059978  46000.0   \n",
       "38   Random  round22  0.860083  0.022169  0.760218  0.039313  46000.0   \n",
       "15  Entropy  round23  0.862177  0.017081  0.758480  0.043383  47274.4   \n",
       "39   Random  round23  0.862177  0.017081  0.758480  0.043383  47274.4   \n",
       "23     Full     full  0.862177  0.017081  0.758480  0.043383  47274.4   \n",
       "\n",
       "               n_positive             frac_positive_in_train            \\\n",
       "           std       mean         std                   mean       std   \n",
       "0     0.000000     1155.6   34.124608               0.288900  0.008531   \n",
       "24    0.000000     1155.6   34.124608               0.288900  0.008531   \n",
       "11    0.000000     2128.8  141.105792               0.354800  0.023518   \n",
       "35    0.000000     1725.0   41.387062               0.287500  0.006898   \n",
       "40    0.000000     2300.1   66.865454               0.287513  0.008358   \n",
       "16    0.000000     3009.9  183.760986               0.376238  0.022970   \n",
       "17    0.000000     3917.4  214.633641               0.391740  0.021463   \n",
       "41    0.000000     2882.1   64.972558               0.288210  0.006497   \n",
       "42    0.000000     3444.3   70.669890               0.287025  0.005889   \n",
       "18    0.000000     4770.7  279.975416               0.397558  0.023331   \n",
       "43    0.000000     4011.3   83.272044               0.286521  0.005948   \n",
       "19    0.000000     5598.5  341.928274               0.399893  0.024423   \n",
       "44    0.000000     4584.8  109.645692               0.286550  0.006853   \n",
       "20    0.000000     6508.5  507.066345               0.406781  0.031692   \n",
       "21    0.000000     7548.8  558.538728               0.419378  0.031030   \n",
       "45    0.000000     5152.8  120.494168               0.286267  0.006694   \n",
       "22    0.000000     8382.5  585.581809               0.419125  0.029279   \n",
       "46    0.000000     5720.9  134.639808               0.286045  0.006732   \n",
       "1     0.000000     9032.5  691.476078               0.410568  0.031431   \n",
       "25    0.000000     6288.5  142.602205               0.285841  0.006482   \n",
       "2     0.000000     9689.7  607.467064               0.403737  0.025311   \n",
       "26    0.000000     6858.8  152.321152               0.285783  0.006347   \n",
       "3     0.000000    10163.3  572.547339               0.390896  0.022021   \n",
       "27    0.000000     7426.3  157.555669               0.285627  0.006060   \n",
       "4     0.000000    10785.4  749.561976               0.385193  0.026770   \n",
       "28    0.000000     7999.7  169.568112               0.285704  0.006056   \n",
       "5     0.000000    11195.1  809.118649               0.373170  0.026971   \n",
       "29    0.000000     8562.9  164.746573               0.285430  0.005492   \n",
       "30    0.000000     9139.3  173.271175               0.285603  0.005415   \n",
       "6     0.000000    11523.3  732.317486               0.360103  0.022885   \n",
       "31    0.000000     9710.0  174.094483               0.285588  0.005120   \n",
       "7     0.000000    11730.1  645.019629               0.345003  0.018971   \n",
       "8     0.000000    11976.6  624.401438               0.332683  0.017344   \n",
       "32    0.000000    10272.6  181.575329               0.285350  0.005044   \n",
       "9     0.000000    12325.6  409.222623               0.324358  0.010769   \n",
       "33    0.000000    10846.5  188.490053               0.285434  0.004960   \n",
       "10    0.000000    12586.7  285.605808               0.314667  0.007140   \n",
       "34    0.000000    11429.2  184.399205               0.285730  0.004610   \n",
       "36    0.000000    11985.3  187.374284               0.285364  0.004461   \n",
       "12    0.000000    12729.3  274.472848               0.303079  0.006535   \n",
       "37    0.000000    12560.2  208.806556               0.285459  0.004746   \n",
       "13    0.000000    12919.2  251.005887               0.293618  0.005705   \n",
       "14    0.000000    13097.8  193.728791               0.284735  0.004211   \n",
       "38    0.000000    13132.7  212.225797               0.285493  0.004614   \n",
       "15  463.840777    13490.4  171.697538               0.285389  0.004604   \n",
       "39  463.840777    13490.4  171.697538               0.285389  0.004604   \n",
       "23  463.840777    13490.4  171.697538               0.285389  0.004604   \n",
       "\n",
       "   frac_positive_sampled            \n",
       "                    mean       std  \n",
       "0               0.085651  0.001838  \n",
       "24              0.085651  0.001838  \n",
       "11              0.157743  0.009223  \n",
       "35              0.127856  0.001790  \n",
       "40              0.170473  0.003172  \n",
       "16              0.223017  0.011291  \n",
       "17              0.290391  0.015636  \n",
       "41              0.213625  0.002888  \n",
       "42              0.255300  0.002875  \n",
       "18              0.353653  0.020640  \n",
       "43              0.297330  0.003782  \n",
       "19              0.414926  0.023495  \n",
       "44              0.339834  0.005541  \n",
       "20              0.482342  0.035319  \n",
       "21              0.559383  0.037783  \n",
       "45              0.381941  0.006375  \n",
       "22              0.621265  0.041099  \n",
       "46              0.424047  0.006908  \n",
       "1               0.669392  0.048173  \n",
       "25              0.466118  0.006946  \n",
       "2               0.718079  0.040583  \n",
       "26              0.508385  0.006750  \n",
       "3               0.753215  0.038090  \n",
       "27              0.550459  0.007278  \n",
       "4               0.799386  0.053293  \n",
       "28              0.592959  0.007635  \n",
       "5               0.829859  0.059562  \n",
       "29              0.634716  0.007094  \n",
       "30              0.677443  0.007458  \n",
       "6               0.854197  0.053852  \n",
       "31              0.719752  0.007293  \n",
       "7               0.869556  0.047749  \n",
       "8               0.887820  0.045896  \n",
       "32              0.761458  0.007710  \n",
       "9               0.913685  0.029119  \n",
       "33              0.804001  0.008088  \n",
       "10              0.933040  0.019289  \n",
       "34              0.847207  0.008150  \n",
       "36              0.888434  0.008392  \n",
       "12              0.943600  0.017612  \n",
       "37              0.931040  0.009256  \n",
       "13              0.957697  0.016768  \n",
       "14              0.970929  0.011098  \n",
       "38              0.973484  0.009685  \n",
       "15              1.000000  0.000000  \n",
       "39              1.000000  0.000000  \n",
       "23              1.000000  0.000000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 171.26x171.26 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 171.26x171.26 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 171.26x171.26 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 171.26x171.26 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 171.26x171.26 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAMwAAADDCAYAAADOdP4nAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAABCq0lEQVR4nO29d3xUVfr4/753enolEEoo0sFAIPQSmrCCimDBBnFRLLvqWnbBj6vi7voVXeW3H11cy/oBxRXsyiqW0IugVKkCQgKhBNLrZNo9vz8uM2QyM8kMJBDCfb9e80rm3nPPOffOee45zznPeR5JCCHQ0NAICvlSV0BD43JCExgNjRDQBEZDIwQ0gdHQCIGLIjB//etfGTNmDOPHj+e6664jOzsbgC+//JL8/Pyg8tizZw+bNm0KusydO3cyb948AN5++20A5s6dy6JFi+q99vTp08TGxnq+f/bZZ6SnpzNo0CC+/vprT5oxY8YwfPhwHnjgARRFAeCpp55i4MCBZGRkcPjw4TrLCfRcGpL27dsHle7XX38lIyODkSNHcs0113DmzBnA//1s3LiRAQMGMHjwYM+zraqqYvLkyQwfPpxbb70Vq9Xa4Pcyd+5cunXrRkZGBhkZGaSmpvLee+81eDlQx3MTjczevXvFqFGjPN+XLVsmpkyZIoQQYuTIkSI7OzuofJ599lmxcOHC86pDSkpKSHnMmDFDhIWFCSGEsNvtolu3bqK8vFyUlpaK1NRUYbfbxe9//3vx4YcfCiGEuP/++8WyZcvE1q1bxbXXXiuEEGLz5s2e+/RHXc+lIXHfe31MmzZNZGVlCSGEeOONN8ScOXMC3k9aWprIzc0VNptNpKeni/z8fPHyyy+LF198UQghxLx588Srr77a4PdS+/crKSkR7dq1a/ByhAj83Bq9h0lISCA7O5vFixdTWFjIpEmTWLp0Kd9++y07d+7kjjvuwOl0kpmZyTXXXEO/fv14/fXXAbj66qu57rrruO+++1i0aBHz5s1j9+7dnrxvuOEG9u7di6IoJCQksHXrVoQQpKamsmrVKjIzM5k3bx55eXk8/vjjAHz88ceMHTuW/v37s2vXLp/6rlq1ioSEBBITEwHYv38/Xbt2JSIigqioKDp06MC+ffvYuHEj11xzDQATJkxg1apVbNy4kXHjxgEwcOBAduzYEfJzAVi6dCmjRo1ixIgRTJw4EZvNxty5c7nrrrsYP348w4cP5/XXX2fChAn07duX/Px8Fi1axNSpUxkzZgxpaWmsXbvWq7yff/6ZjIwMRo0axd13343D4fA6/+qrrzJq1CgAXC4XJpPJ7/2UlZWhKApt2rTBaDQydOhQNm7c6Pd51CRQ/ZYuXcrgwYMZNmwY//jHPwDIzMxk4sSJjBw5EpvNFvAZ5uXlERYWBkBOTg5jxoxh5MiRjB8/nqNHj5KTk0NGRoYnvbvXyMzMZNasWYwbN47+/fuTnZ2Ny+VixowZDBkyhHvuuSdgmY0uMC1atODTTz8lKyuL1NRUBg4cyI8//siECRPo06cP//nPfzh27BijRo3i+++/5/vvv/c8uLKyMl566SXefPNNMjMzmTNnDr179/bkPXnyZL799lu2bdtGcnIyq1atYsuWLQwcOBBZVm9tzpw5tGzZkldeeQVQH9qKFSv44x//yJtvvulVV7vdzvPPP8/cuXM9x8rKyoiOjvZ8j4yMpLy83Ou4v2MAoo4lrkDPBeDIkSN89913rFu3Dr1ez9atWwGIjY3lu+++Iy0tjWPHjvHtt98yYsQIVq5cCYDD4WDFihV8/PHH/OlPf/Iq77777uPf//43q1evpnPnzixcuNDrfGJiIjqdji1btvDPf/6Thx56yO/9lJWVERUVFdTzqE3t+hUWFvLCCy+wZs0a1q9fz3fffceBAwcAmDRpEmvXrsVkMnnlMW/ePEaMGEFKSgqPPPIIS5YsAeCJJ55gzpw5rF27lscee8zzggxEjx49yMrKYvLkyXz44Yd88803GI1GfvjhBx566CEkSfJ7nb7OXBuAX3/9lbi4OM9Yc8WKFdx+++0cO3bMkyYxMZEffviBVatWERkZidPp9Jzr2rVrwLwnTZrEXXfd5XkDv/POO1RWVnLjjTcGvKZfv34AJCUlUVlZ6XVu3rx53H///URERHiORUVFef345eXlREdHe467/9Y85sYttKE+l6SkJKZPn054eDg5OTme3qBv376AKjjut2VsbCzV1dUAjBgxAkmS6NSpk49u+Msvv3jenFar1dNz1OS///0v//M//8OXX35JfHy83/up73nUPFab2vU7fPgweXl5jB8/HoDi4mIOHToEBP7d58yZQ2ZmJitWrOAPf/iD5zns27eP4cOHe8p55JFHfK6t+QJLTU0FoE2bNuTk5LB371769+/vORceHu63/EbvYXbs2MHvfvc77HY7oEq2xWJRC5dlFEVh4cKFxMbG8u677zJ9+nRcLte5Cp5tdO60NUlMTMTpdJKVlcXEiROprKxkxYoVjBkzxitdzQdVVyNesWIFCxYsICMjg7y8PCZNmkS3bt345ZdfPG/RAwcO0K1bNwYPHsz3338PwDfffMOQIUMYPHgwWVlZAGzevJkePXqE/FxKS0t55pln+OCDD1iwYAF6vd5T/0BvPTfbtm0DVGGMj4/3OtetWzeWLl3KmjVrmDt3rtdQBeC7777jueeeY8WKFXTu3BnA7/24e5fjx49jt9tZv3496enpfp9HffXr0KEDHTt2ZOXKlaxZs4Z77rmHXr16AXX/TgBjx47llltu4cEHH/Tc38aNGwFYu3YtV111FWazmby8PIQQnDx5ktOnT3uur/0sO3fu7JlUOnDggM/L1E2j9zA333wzx44dIz09nYiICHQ6He+88w6g/iDTpk3jnXfeYdq0aaxfv564uDhMJpOnIblJS0tj9uzZ9OrViwEDBniOT5gwgaysLEwmE0OGDOHIkSMYjUavazt27Mj9999Py5Yt66zrunXrPP+3b9+er776CoDnn3+eMWPG4HQ6+dvf/obBYODpp5/mrrvu4pVXXqFbt27ceOONyLLM1VdfzaBBg5AkyTMjN3/+fHr06MGECRPqfS5RUVEMGjSIfv36YbFYiI6O5tSpU0E966NHjzJ69GgqKytZsGCB17nXXnuNadOm4XA4CA8PZ/HixV7nH330UWRZ5tZbbwVg3LhxPPXUU37v57XXXmPq1Kk4nU5mzpxJy5YteeCBB5g+fTpDhw4lPj7eo4/VVb/ExETuu+8+RowYgd1uJy0tzSMAwTBnzhz69u1LVlYWL7/8MrNmzeK5555DkiT+/e9/07JlS4YPH86AAQPo0qULnTp1CpjXlClTWLduHYMHD/borP6QRF0DbY0GYfny5RiNRsaOHdtoZSxatIicnBwv/asp0dTrFyyN3sNoqLN9bdq0udTV0GgAtB5GQyMENNMYDY0Q0ARGQyMENIHR0AgBTWA0NELgspslUxSFkydPEhkZWe9CnsaVgRCC8vJykpOT613wvFAuO4E5efIkbdu2vdTV0GiC5ObmNvr0/WUnMJGRkYD6cGoaAWpcuZSVldG2bVtP22hMLjuBcQ/DoqKiNIHR8OJiDNE1pV9DIwQ0gdHQCAFNYDQ0QkATGI2QKC+H/fvhSrVAvOyUfo1Lh6LAjh3gcEBJCaSlQa0dxM0erYfRCJpdu8DpBJ0ObDbYuBEKCy91rS4ujdLD2Gw2brnlFgoKCsjMzOTee+8F1L3kN954IyUlJcyaNYvf/va3DVamEIKqqqoGy0/Dm+PH4dgxGYPB+/jGjZCSonDVVRenHmFhYZfUwqNR9sMsXryYiooK7rvvPiZMmMCyZcswm818/vnn7Nu3jyeffJLhw4d79mDXhc1m83K1416kKi0t9VqHce/nr709WePCqaqS+eWXWPR6/03F5ZIIC3MQGelAkkCWBbIs0OsVwsNdGI2K3+vqw+mUKC01YrfLyLIApYLx4wbRqm2iVzq3x5rabaIxaJQeZtu2bcycORNZlunduzf79++nb9++dO/enZ07d+JwODDUflUF4IUXXuC5554LKq3RaPQ42NC4cCThQHaWcuaUnnZRpRgkKzrJhgTYlTDsIgybiKZaxCAwUl4OMg6MUjlGUYHASoVko11KFZYwBZA8H5dkoVqXhJDPKUFVVTJFRUasVh1VVTqEvZIYYx7hugp02HA4qjm+20yrtr4eby4WjSIw5eXlHicC4eHhVFRUAGAwGFiyZAnvvfeeZ5hWH08++SSPPfaY57u7h9FQ0bkqsDhzserb4tL5d9zgxmaTOHQoEiFAHdUI3KMbk0khIsJBQoIDvc5JmOMYJqWAE6fCsGDinE2j+o9RrsZINZGcQULBKcxIuNBLNhT0iBpN62SunnZtq7BYzvU0eqUKsysPpxyOQ47hyOkUTpyMwKS3EaU7TitdEfowGwruEYMeIRlBvrTzVI1SekREhMdNTWVlpaeb/Oc//8kLL7zAjTfeyJQpUzh69CgpKSl15mUymXycuV2pnGvoKpJSTaTzICAT6diP0xVFlT4FRTb7XGu3S+zbFxVw/O9w6CgrM1B0/Awtw7OxWlxIcgTl5QZ0ujrqhB4ByJLag7jwLVsnQ+7xMNq2rcJiPis0koTAiORyUJBbiqFyF+0sFgySFQUDINUQlqZDo8ySpaWlsW7dOoQQ7Ny50+OULTw8nMjISI9DOHfPo1E3LhccPWph27YY8vPVRiQJF1H2X/D8hJIBvbASbd9DmP0wkuLwpFOqKzm2v4I43RHidQeJkbMJl/LQUwWoDVhHFa0Mu0myHEFRdFRWGikrq1tYQkGWIDc3DGv1OYF1OCAnJ4wqqxFJZ0Qnuc4KSdPdttEoPcytt97KbbfdxsKFC7n77rs5dOgQWVlZPPTQQ9x555389a9/pX///vTs2bMxim82uFxw8qSFM2dMSBLo9XDsWBh6nZOU8N1ICGo3LiHpMSplGO0/IyQZ4XSRczSMaJ2xVu+iIOMAZFzCiCzZERi8hlJyA7dbVWjCadu2EqFIHD9hQZakBi+nMWkUgXHPiNXE7RPZ7UlRwxshQFEkXC51dqioyMSZM+pQtOZbXqcTVOYeo7qti7DwAC1NkgA9igJHj0XgUmR8R2IyCqazyQWC4CZhLhS30AiFBuu9LibaSv8lpLpaVcLtdhlFkWqYmwjMhip0stPrjQ8QrzuERa4k90QE7dpVYjEHmOpV4NjRMFwuf8ISOuHyaVoZdtHeuAGLXMJhWwb7qidzPsMnWQJqCUtH4yq6mFdwoPoasu0ZF17hRkITmAtAUeB8d8RWVsocOBCJLEvodGrP4SZKziVWdxSBjIIOlzDgwoAkJIxyBQK9qkjnhtE+pZKaS0+VlTpKSgxUVOqRJe9Jgs0/xnHgQARpaSX07lVWZ/06GlfRzfwNZa5WhOmKCJcLyHd0xSKXYpKr6G5eThvjNo7b+3HC0Y9SV1tqCo/VKpOV1YL8fDP9+hfTJ7XUbzlGqYJExwZSoz5Hp1PoY1lCO+NPVCoJVCkJWOR8Wur3k20fxh77KBy2S2vEdtk58gu0SFVZWcnKlSvR1ejn9Xo9JpMJm83mFRHAYDBgNBqxWq1eDs6NRiMGg8HnuNlsRqfT+TioPnYsEZBo3Trfq8GHhYWhKIrHq76b8PBwXC4X+flODh+OQpZVp9smkwmn04nD4UCHjbamnxGyAaPBiNPp9HbOrpMx6A04nA4UlwIIWrWqorLKgq3ajLXagSydq4ter0en02F32Hnn311xKTrCw+xMn5GNTtb5xF8xG510Mv9Ab8sn6CQFu2JiU+l0iuiNXTHSQb+abuErOVg1kiolhvZhP9NavwMkQb69I3srxrFu10A2bWmDooDTqQMEvXsX0rtXCXFx4HQ6CeMkbcQGesZuZN+J7uQWtSO9w0+sPziMhJQIEiLyidQV0da8Db3k8MwQVlSHsV28woiZ99fbJhqDZiUws2bN4oMPPvAcu/baa/njH//I3//+d5YvX+45PmPGDDIzM/njH//oib0CaoyRiRMnkpmZydGjRz3HX3zxRQYMGMC1115bIxRdNI89tpT4+GieeupJ4GegBICvvvqKM2fOeJn+WCwWli9fzsqVu/nb374AVCFISmrJnDlz2LRpEx999CH9OpQRYVZo26YtN9xwAz/99BM/bfnJk0+P7j0YPXo0q1atYt/+fWePGhiQ3pcBAwbw5Zdfkns815N+VMYoevbsyeL3v6K09CHACphI77+PgQPDefPNN3E4HSRE2pmSXsDUgeWUu5LYtKeMvu0r+GhzC5ZtT+Dee+6loqKCJUuXePI26A3cd999jAn7E7GmYiqqDZRZY0iIKOGEtRuFJTId43LYlt2WQ3kdsRij6dtxN12T9mPQOVi5dyQf/9gGOUqmVfK1bPrBhN1xBmgDfMCAdAu3DquitfMLPtg4kVsHraN1XB45hVfR/qFD9baJxqBZCczF7GH27YtBls3Iskx1dTUul0SLFlaSk6sID/ffw1RWxpKTY8LlOheZoGYPY3SdItFwGAUDkiwF2cOo6HQ69Ho9docdofj2MNu2R3DsaATXXnuUnKORrF7dmqFDC7k+dSm9I5ajkxyctPXkkG08ha7O2O3eEcqMRiOKUHA6nF7HTSYT8qmfGNDyaz756WbOGAcwqPdB2oXvp5f5c/SyA4di4IStN9UOC22MO4kwV5JXmsR3Fc+g1wmfezqSHcn69a3p0rmUgQNLWP5NCxSXxC3DPmFQqywOm/98yXqYZqXDGAwGv6YxgRY/A5nRBDruDrJTUGBElsM8szzu9OXlJo4di6JTp0pcLh2VlRaqq3XY7TJ2u0xVlR6jUQC++Rv0EsnmUyB5r9br9Xr0et+fyaA3+P31jAb/i33Z2dF07VqGyWSiaxc7kZGn2P9TJX2GfolOVqh0xfGj9WFAHfr4e146SYfO5K2tl5To+eDzaSjidqIi7Uyffgw7bfjV1gaXMNDF/D0Hayjyh3I30y/xG7bl/4bwBO+6uu+pezc7ya1y+e77JP5vYQw6nWDQoCJsCemc7DSdERNG+b3Hi0GzEpiLgRBw/LjF75SoTidht+vYuVONvmUwCK9JgUDGiwCxcnajrddVVOg4c9rEb35zrocc2G4t93d9j58OD6Bd/DFWHLqBnQVx2Gw67DbVpquiUs+AAUUBFfaiYj1fftGaVslWKioM9OlT4nU+257hM+NlSxjED2IQJNRd5+hoJ1OnnODd91KoqjKwe3c0PXoEF3G7MdEEJkROnjTjckl1riGYTKGNcg2UEy7n17kW8tNPsezbF0W/fsX07q3OcMnYGR4xnwg5n33V1wWcjj1yJIJWrWyEhymAi96WT+hg3MBPVffx4vJJlJcbMehddOtWjsXiIibGzokTFux2HT9sjCcuzk67tt5hxAsLDXzxZWt6dC9j0KCiBpi6FkjCBTXWhHQ6SE8vZufOGB9hvFRoAhMCLpdEXp653gW3RP1+9NhQxUYgn/3PKYzYRRTlSiIuLKxencjatQncO3k5if0NXnZKLgVOnzaRnR1OTnY4xSWqyci69QkUlxjp0fUMUzr+L3H6HHSSi96Wz8i2D8dngQM4fDicTp0q6Gz6lp6WZTgUC6vKn6JCaUlaWomnQdacajYYBDt3xpCYWM03y1vSuUsFw4acxmiEvPwIvl8eQ+9epaSnFzfIOo8QEiec6YRL+cToj+G2au7dq8xTr1pq1SVBE5gQOHrUUu+6i4wDi1TkWUWviVGqxoiVKN1xnMJM7u402kUVsG1LFNu2RBMe4SIiwondrg6JdDqFq66qZOCgIioq9OzeHU1KShVGqYJRka9w/GQE3+X9lmu6fYnF7GBExHw2V96HXdSYDKmSKSt2MqXPf+ht+RJZgmqiqVDU8IU1G2RNah6vKBd8m5XC3/53LKa4GHbujGVwv6Ok9t+ELDnOLq4GLzUuRTX7AdUwU693UCT1RCBTIVpS6UggTvcr4XLxWUPMpoMmMEFSXS1TVGSqUw8BCJMK6vmRVSvcaitEW4qIiHDSp08J7VOqqKjQU1GhZ8OGBISQCA93MXbMGc+VfVJLMckljIj4/yhzJvH1/of5elUb3sm6naT4Uv4+82+MjfwbmyofoNjVAYNURTvnOpb+/luq5Nb8Wj2WVsafOWi9Jog7FsjYsSoxVFuSGTgpFrFZz0cfxQISh462JN/VHZ3LRrScS5iuENUQ1FtwnE7QGxQMeoHBKDDoFYxGF2azgt4g0ElOKg0didcpVFaWs39/FHq9nkJXNyqVYmJ1R9R1mCbSVJtGLRqI2oFkG5JDh6JwuaqpMcPrlyhDAXbhrDsRsGdPPMmtqpg46Vz4dUsYJLYAa7Vg9644evcuwu44d0/hciEZka+Sb+/Ej+V30DK5kn5pBfy0pQUtWjnZUHI33SwryYiYhwsjMi4OWrvynz2PEdG2JYoCSS1+Q36VGfB9VhIuJAHVIoJqJZpyV4sawm8jLe0UNpuLDRtaMmxYnmfRs4q2SLSiheEQZrkMxd2sBLRLqcAQaKem3UW5nIhVWAArsgxhYRJlZUYkCWyEUUpPIuR8wnWF6EUZEvX8AI1Ms1mHaaw9/U4n7N4tU1ysWgvXR6xj/Vkr4sAoAl560cCkSS569aq1fVcIullnE6EcpEpuh1XugCKZMLuOE6XspFzuyS+WF0E6NzbcuFHmhx90/OFRBwY99K24AyOFVBPPtc9/zOw5DmKiBQ4HpKYqlJSo+/Pd9yMJBzY5AYfUApuc5JV3qFhc2YQpR3ApBjp3VggPtKdNCNBHosSmeR12OmHDBtn/0Fe4aJVop096nNfhy34dJpATjPnz57Ns2TJAjVO/evVq0tLS6soqaCRJ8qyTNBRnzsDever/wfi5lhUrFoTXtlt/7N2r2qH16Sujq9UwIu0/EaXsQULBouRRaRyIhI1Ix4Gz9sWFmMze6zgjM2D7Dtiw3sSECXBaTCfR+iE/5d1Km7YSSS3U9Q6jUb2PyEgoLVUbJ4CTOFymdGT8rRCFSg8qbcn0bruThLhaO968kCFxCMi+kxQ9e8Lhw77WzHY7dLnEO0IaZQPZRx99xIQJE1i/fj0ff/yxZ8X7scceY82aNSxYsICJEycGJSw2m42ysjKvT2PjcsHPP6sfSarjN6+FUTmlbqOth40bYfBgfIQFILH6I8oMA6iWkzkVPpO88N9yKvwBTobPolpOJt9yq881OhluuRlWroT8fCg0X88vsf9hybrr6ZN6Ll3N9dh27dQNXAgHVYaGa4VOJ7S7KoaEbsPBEAmuanDZ4OyGNmQT6CMhvr9fYQFISQF/vkySkiAsrMGqel5cVCcYbp5//vmgHVuE4gQjFHJy4OBBMJvVj8WiOqUzGODoUbUHCNJPhwedUlyvdBUWwoGDcKtvu8fsPEK4Yxf7Yj/CJcd4X2e+nkLz9QHzbdcOBg6Cjz+GBx4Aq1W9v6k31ci/xu7hqCiIilQotbVGkRumFbpc0KIFdOoEoFOFwlEJOiPIwT9MSVJ7ma1bzwmOwwHdujVINS+IRulhAjnBAKioqKCwsJDOnTsHldeTTz5JaWmp55Obm1v/RfVQWgqHDqk/hqJAVZXakE+ehOxsNc35mO3rRf293w+boFdPiI72PZdo/Zhi0zU+whIsE6+FU3nq8GzPXmjZChLi1XMuBR99IqU9lNL1vMqqjRDq2//sPsFzGMJDEhY3cXEQH38u71atmoaXzUYRmEBOMACWL1/OlClTgs7LZDJ5YsE0REwYlwt27gzce/hXNhXCHHvrzFd2lSOLumfpHA7YvAmGDvM9p1eKiLWvJN98k+/JILFY4MYb4Ysv4Kcf8RqOuZwQU1NIFRfm+I4kt9ahKLVzUodWoUw6CgHp6cEPX4OhVy/1mblc0LVh5PqCuahOMADWrl3L8OHDG6PYoNi5E78NpC50SgVmZw7S2eliIeCZZ2DMGLj5ZnjySVi68BQrVplYshT+8lfY+INvPj//rL7lO/vxEplg/YJyQ19s+vZB1UkItTFZao2m+vaBMAv8ehivKXCDodZLQmeG8BSvYY4QqgvYiAi1sfbqdW5ioD6SkoKbRQwFk0nVZ1q3Dn143Fg0isDceuutZGVlMWDAAKZOncqhQ4eYP38+ANnZ2Y0ehzAQOTlQXL+a4YNBFKJIFszOI9jt8OKLquKuKOpnwACIDish9zhs2wZFRfDtt74e7jduhKFDarlKksCotxFf/SWnjbfUWxenU70mMRH69oWuXbyHd5J0VplH1QHc1NRfUBwQ3Q0k1SauY0e1Z23dGkaPhv79oWVLdRgUzNDUZlMbdmPQuXPT6V3gEjjBqLmR62Li1lvO502lExUgydhKTvLYc51xOiVmzoSvv1aV9+uvE8TYykDSs/EHVVgcdnjjDbj5FlWPOHFC/cyadS5fIdSG35rvERXxxHZMQy6HinK10et06ltbbwC9Tq17XDxE1xqVxsSoOpj7DT9mLKxeBaNGn0vjNRNtjAVTvOdr+/bqpzZuwazP4XhkpNorNQahzFJeDJrVSn8g6tNb6kMWlZw4AQv/z0nPDsf57e/bYjLB7ber5/WuYtz+vYYOUT9WK/z3K3jpJZgwXl3TSUvzntp1OCEuToGdHyOl3E5iC4nEFuq5oP0FCBfRsWYQNtwDBncdPPevQIR7iUqxQ3SvoO89JUWdDAmkcCuKOjN2pXBFCMy+farQBNMAhYDqaigrU4MHlZfD8a2VbNuho1dPPXffc5Ryk7erWqNyGmqtv1gs6tpIv36waJGaz9gx3mUZ9GCp/AkcFdBytNe5eusqXCAbIX4gss6EJfJHHFar31V6lxOiY1CHYhGdQe/rnTIQUVEQHh5Yl3E6/fdOzZUrQmDKy+tvgFYr/OlPsGfPuWPh4ZAQa6NnogPQkZcHOmFF78rHqTvnQV4n/G+wAujUEUxGKEft5SZN8s6fYx9D2xvVxh8sihPMLSCml0dAjK0GUH1kMzph9xnD6PVg1DkgugeEtQ6+nLMkJUFurv9nGBfX8Mp+U+aKCKhUa2u9D7/8Avfcoy70garsrlgBX30FHyzKZ+J1BhLiz+oEkgGLM+fcxcKFTql7/WXUaM5d775MQMuqd6B4u0/vVCeKE6K6QuzVXr1J6zY6ChiEIul9ZhssFgXi+p+XsIDag/jrYZxOSE4+rywvW5r9u8HlUtcTzH5GIS4XLFkCixfDjBnqwtvHH6uKvNuOSa+UMWSIjiE1dAKdUoSsVKLI4RiUfOp779TWKQCc9moiKz9Uv5xYBu1rLP0Lt8NuvSoUku7c38iuYPQ1bAsLA7NFR7kymCjbD0goIAQCHUrsQDCdv5WYwQCxsVDLyxSyrL5criSavcBUVvqfZTl9Gv7f/1OngF999dzU5eTJ3ulk4ccCWjZhcf5KpTEVg1KgNuxQEC662J5DMrdQe4OUWnYyilPVaUIM7RATA8XFesqMg4iyb0KRwijV9+Oq2Av3ydq6taoL1hx+JSY2rRmsi0GzF5iSEt/ZMZsNMjPVN+Tdd9c9z68TlX6PG5QzSMKBXvHWX1pW/ps4exanLXf4t/0SgjaV/x9m8iD9DTD4mY81xpxXHJT4eCgoAJ3OSKlpGCBjt0kNMovVqpU6dHVjt6v2a1cazV6HsVp9ldVNm9R1jqoqqLVc5IUknMgikAKkx+I8oK7R1KBF9YcYlTO0qnwbSdh8rkqyvkuU/UdK2r/kX1gUF1ha1nNX/mnZssbqvqQDScJg8D8cDRX3mowbi8W/PVxzp9kLjD+Ff9Uq1e4pOdm/1bAbnVJKwL3qkoTJleu1ddboOgYo2KUEXFI43Up+S4T93HJ7XPXXJFo/4YDlJWKSEv1kCiDAcn7KuV6vTgPXpCEXFFNS1N5ZiCtr7aUmzX5IVst1MJWVsHkz/Pvf9Q8pDEpRnftbhOStSMfY1lFuGER21PMgXCTYltGh/BlKjUMpN/SndeU/ORI1D5exQ2DLW2NswH0iwRATA3l557435J4695pMZeWVtfZSk2bfw3hcIZ9lwwZVUIIZf8v4118CEWNfS4lxpPpF0lFgvpH9se9idmbTrnIexaYxVBpSz62610ZxguXCpp1atjxnZawowe0UDYWkJFUom4Kp/aWgWfcwQpydUpaLceliAXU4Nnp04GtkWdVvZBl0SvACY3SdwOzKpszoPX/slBORsSIBEY4dKEodb31JOm/9xU109LmZLIfDW+9oCNq3b3ghvJxo1j2MzaYqwRHOHRhcpygtVa2JAwmMw6GasowcCS0SBYo9eKca0bZ1lBv6Yxe+SkO+5VbP9mKnE+IDuUk1xl2QAwpQZS4mRv1fp/O2XWsIDAa1F7tSaRSBsdls3HDDDQwdOpS3337bc9xqtTJt2jSGDRvG7373u8Yo2ovyctDJAlm4CHfuZ/06O127+v/BXS717Rkerr6he3atoHdvF2Fh58zlAyEERNvW4ogZQWoq9OihGjy6ce+xLzRfj8kERn9GoIrjvJX92sTGqsOxiIgrb52ksbmoTjDefvttJk6cyIYNG+jWrVuNWCuBuRAnGGVlYNJVAgoIiZN7fmbMGP9pjUa4qubGLlsBJrOJLl2gew91GGIyAWf3m9jtqpC5XJAYmUe4coiE7kNVuy0jtGuLXx9mAd0OSXowN8z4qXVrtX6NZXJ/JdMoArNt2zaGDRvm5QQDYP369Rw5coRRo0YRGRkZMKxETV544QWio6M9n7Zt29Z7jZvqatCLQgQGSkolCk6VMGHEUZ90Doe6F93rbew8ZyIQEa46dujeXd32278/pKaqPVKfPtDGsA4pti8Yzi1MJCaqs0o1zbrcb32/mOIbrDswGtWesoG9TmkQQGDKysp8vEj+3//9X9CZBnKCUVxcTHJyMllZWSxevJiCgoJ687oQJxiqwKibv3buhPYdDCRHHEJWzukmiqKuYrvH/R6cgRV+NX6KurKu0wFn1kLSSJ90HTt5y4DTBfFxPsnU4VhYw+5CjYi4ctdKGhMfgXnppZdIT0+nR48erF27lmPHjjF8+HCWLl0adKaBnGDExMSQkZGBXq8nPT2dbLeLljq4ECcYNts5W7DtO6BvGiDpiXDs8Lz6ZVnVOXyoQ2C8qM6Hsl8g0dezhU5Wt/+6LX3NpgCb2GQDmPxJ0vnTvful9+HVHPERmKVLl7J37142btzI008/zbhx45g1axbff/990JkGcoLRr18/1q5dC8DPP/9Mly5dGug2/FNdrdqC5RfAieNw9VkXQLKoxuI8hMOhCovPPg+XTd2ZGAz561RTe2Os39NRUeqb3lXXdLKxnuhC54HZrCn8jYGPwERHR6PX60lKSuLUqVN88skn3HXXXSFlGsgJxgMPPMAXX3zBgAEDGDFiBNGNaIzkdILD7kIWNnbugK7dajRYSYfReZQWscX+hy22ApCC3M98ei0k+g7HatK2rbq70q/+otghLHi9TOPS4rNwKdV4LbVp08bjvCIU6nKC8dVXX4Wc3/lQVQU6UQaSxLbtMG6s93mHy0DvhB3gHAT6WmMXe5m3ecqev0DxLugwHdrUsEC2FULpXuj9bJ11kSTVItrvzkSdGUxXoBXjZYrPT5idnc1f/vIXhBDk5OTwl7/8xXPumWeeuaiVuxBKSiBMV8TJk0YKC1QfWzXR68BikaFwCyQO9Tand9XQXxQnnF6t/n/odYjtC+Fne4Qz6yG6p5cHlkD4NSURAsxJId2XxqXFR2Dmzp3r9//LDasVDFIl//1KXTzfus1716NHIRYKFPwIiYPPrbLXVPjLD6kOtI2xYGoBP82E1jdAhxmQ7392LGiEHcLbn//1GhcdH4G5+eabWbx4MVFRUUybNs1riHY5UV2tzpD9+qu6iLd6VQCBkSRQbFC0A+L7qftRXNWgO9sllOxWe5U+L6jfK7Lh4D9hw82gVKt75c8XYzzor1ArxssUH6X/rrvuIjs7m7Vr1/LnP//5UtSpQbDZQCeqCA9TV+lrOqBwOGrtG5FksJdAyV71b82XRMluiKmhx0V0gL4vg+6sxJ04T8eEjbD2otH4+PQwBQUFfPrppwghGDdu3KWoU4NQXWXHpNix2c3cf79qqlKTyNrLObIOrKfUGTK3yyMhoHQ3pNRy4SpJ0HEGHP3Qdz9+sMgGTX+5DPERGP3ZqRxJkrjMovl5EAKU6kKqqg1UVZ0L+eDGbPEfzAhZryr57hmyqlxwVqlujWrT5nrvGbNQK2hpqS2UXIb4CIzVaiU3NxdFUTz/uwWn3WXi9cBmA9lZRlGxDovZd8U7vK4V8JrTyaW7IapbaE72AFx2NYhQIBQ7hHcILU+NJoGPwBiNRqZPnw6oZinu/yVJYtWqVRe3dudJeTkY5CoKC1Xn3TVRFAgL1iixtv4SDIoDLC1Uk5lAgYSMsSG5a9VoOvgIzJo1ay5BNRoW1ay/ioLCc1Gs3DidEOffisWX4t3Q9aHQCteZICYVSveBNc93f77ihChtZf9yxUdgRo0a5ZlKliQJs9lMWloas2fP9lggN3WqrQJZWCkq0vtYBxsM/gOO+mArguqTIXm6R7ggvJ2qm0T3AGc5OK21AsLowHwFb1m8zPERmEWLFnl9r6io4JtvvmHWrFl88MEHF6teF4StqgodTooK9T6WyEFb8JbshvCO/n2H1UXY2chCkgSxaZC/8dw5IVQn4pqyf9niIzApfkJJ9ezZk2HD/ARmbKK4qgqRMapDslqGwEELTOl56C/mJO8hmM4IsX2gaKuqzyh2iOgYWp4aTYqgd1zaajv4asK47BUoQqaoCNLiltC9+Dbiq5fhcPjZKBaIUBV+lx0i/ASvNMVCZBd1MsAYC/oG9kqhcVHx6WGOHTvm9d1ms/H555/TsWPwb0abzcYtt9xCQUEBmZmZ3HvvvYC6k7N79+507twZnU7HypUrL7D6vjidoNirKK9S/++kex+9UkWi9UNOR14f3LZdZxWU/xqawJjiAs98RaSAvRhMDezzSOOi4yMwM2bM8PoeFhZGnz59mFzbrX0duJ1g3HfffUyYMIG77roLs9nMvn37uP/++3n66aeDzstms3n1bvU5waiqAr2oJK9Q9dHlkiLRiyrKjEOwWIIMg1e2X3VIYQ5yj69iV01m6iI2te7zGpcFPs1n9erVns/SpUsZOnQoS5YsCWlPfyAnGHv37uXbb79lxIgRLFy4MKi8QnWCUVLswqCzUVgEifEu9KKEYsMoLM5DwesvxbsgOoTeRRdev4l/U4tuqnFe+H3frl69mltuuYXhw4ezaNEi1qxZQ1ZWVtCZBnKC0b59e1566SVWrFjBu+++S35+fr15heoEo7qiHElWI/9e1eYUEgrHIx7B4vyVGHYGdwOhKPyKC8I1I8orBR+B6dWrF4sXL+bBBx/kwIEDdOrUKWSTmEBOMAYNGsTgwYMxGo0MGjSoUZxgOCuLAQNFhdA1+TDVuhRccjSnDTcRXfhu/ZVXnOqiY7ACI8kQdnmYDGlcOD4CM3HiRLZv386SJUvYtGnTee2HCeQE45lnniErKwtFUdi2bRudOnW68DuohctWAZJEYSG0iztCtU4toyhyKnLFQSj+ue4MKn5Vp4DDfafXfdCMKK84fATmxRdfZPv27Vx33XW88sorbN68meeee47Dhw8HnWkgJxiPPfYY8+bNY/jw4dx0003E17ZbaQDs1apbpYJCaBlxGKtend0zhUdC25sgu55epmS3urofjI9j4YRIP1PJGs0Wv977ZVlm0qRJTJo0iYKCAhYvXszUqVPZuXNnUJnW5QRj9erVF1bjOhBCnVJWhERpKcQZj3BCfyNCnF2wTLoJNn6iKvWxV/vPpHhXcMMxxakuSgYysNRoltT7Gk1ISODRRx8NWlguJTarA1w2ioshMqwKC6ew6jricKoOujFEQtupgXsZ94ax+gRGORvzvoF8IWtcPjSrcBdVJWVIkkxhEaR2zMYhxeKSY9HJNUxi2t0EpftVS2SfDI4H3jDmRnFCZGcIu8IC1GsAzSygkkAgnVX4u7c9p79YLDX0ckMUtJsK+19UPVzGpqrOKKzHoWSPmvDkt/53UyoO1ctLRPuLdUsaTYxm1cO4KSqEjknZVOvOKvy1HbO0u1ndv28vgPwNqteY2L5ntyjb1b36tVGcYEmGqM6NfwMaTZZm1cO4KSyEtqmHsequBcBQe/+LIQq6PnLOiYW7N5GN/h1bCAHGGIjp2eh112jaNE+BKRK0sBwhW98JRQGTvw1j/pxYBHJsodghqnuj1FXj8qJZDslk+xn0spVqXQpOVx1Rv4JFbwGDFp1IoxkKjNUKraOPUC23RUhGEGC5UH8TxoZfYNW4PGl2AlNUDF1bH6baoCr8Ov3ZKGHni2IDi2ZcqaHS7ASmsBC6tj5nQ+Y3YnEoyKYrLhzFokWL6NatGxkZGWRkZDBlyhSfNEVFRaxYseIS1O7S0uyU/uIS6NDxMJV61c2t3xB5oXCFDsfmzJlDZmZmwPO7du1iw4YNjB07NmCa5kiz62HKiuwkhh/3rMEE5VIpEC47WFo1TMUucwYPHszkyZPp27cvBw8e5K233mLRokXs3r2btLQ0hgwZQm5uLtdeey3Dhw/nscceA2Ds2LFMnTqVfv36sXnzZl555RU+/fRTAB5//HF27dp1KW8rZJpdD2N0HsMpLDhkdXvxBfUwsj6oYEkXEyFUz56BeOcd+Oc/4fe/h5kz684rMjLwzoR58+Z5XG6NHDmS48ePs379ej7++GM++eQTZs2aRZcuXYiMjCQsLIwNGzbw8ssvc8cdd3DHHXcwa9YstmzZwsmTJ1m6dCnV1dU88sgjzJ8/nz//+c9MnTqVvXv3cvXVAYxgmyiNIjCBnGC4mTt3Lu3bt6+zyz9fYvRHKHV1BEnCpaiOx88bY1yT2+tSXq76KqiPxx5TP3VRWlor7EcNag/J1qxZg16vp1WrVhw4cMArrdtByuHDh7nuuusANQDwr7/+Svv27UlIUH1dlZWVkZKSwunTp9m2bRt9+vSp/0aaGI0yJHM7wXC/kaqrqz3nCgoKePvtt4POy2azUVZW5vUJhKJAi/Bsqg2qwu9yQsT5Lp8oziYZjiIyUm3ogT7z56uhzufPrztdaamaV7DU3khYM7qDfNazSIcOHdi6dSsAW7duJSUlhdzcXEpKSjh+/DixsaqP3jFjxvDMM88wderUBngiF5dGEZhATjAAXnrpJY+D82AIxQlG3mnokHAYl+WsSygJjCZUo8lQkQBL0xMYSVJ7hUCfRx+Fw4fVv3Wli4qqu/OcN2+eZ5YsIyMDq9Xqdb5jx4589tlnXn4Z3N5Rhw4dSlRUFEOGDMFkMnHnnXcydepUnnzySUCNcrdnzx7S09Mb5Rk1Jo0yJAvkBCM3N5fy8nIGDhwYdF5PPvmkR4EEtVsPJDQnjsPAltnkG9UeRq87GwfG5QShC24XpRtjXGjpmxGZmZkBh8tuAQLYvVvdIuHWdWJiYvj666+90kdERPhEznY4HNx0000NWueLRaO0iEBOMF544QVmz54dUl6hOMHIP36G2LBSqnWqjzCDEVVLjrhK9e4SLMKlOd1rJH766SemTZvGww8/fKmrcl40Sg/jdoLRs2dPdu7cyfPPPw/Ali1byMzMJC8vD4Bhw4Zx1VUNtyfeVbyPfGMrlHhV0zcaUQ0nzUngKFO96QeD4lJN+TUumNrhUwYMGMCOHTsuTWUagEYRmFtvvZXbbruNhQsXcvfdd3Po0CGysrLYsmULcK4Lb0hhAbBU76XI1t7z3WhAHagbIiCsLRTvrDsymBtTrG9cFw0NGklg6nKCATTKdDJAjLyfSjp6xplGI2q0Y0lWo4KVBSEsQgFTQv3pNK5ImpVW2yZyHzajqr+4FDCbAX0N/7DmJFWnqQuhhQPXCEyzEZjVby2ge/JuXKVHgbNrMBGofo/dhHdQdZpACKHajmmukzQC0GwE5irHy8gSdApXLWglCQw6p7q12I3erMZoCYiAmMvLVKMxqGmtPGzYMMaPHx9yfKC5c+c2i3iptWk2ApNtnk1OYXsOWG8GQK8HWXL4CoiltbqKXxvFAdE9g5sUuAKYM2cOa9asYcOGDQwePJi1a9de6io1CZqN8eWImfdTePImIrerU5ZGIyCbfQUgLBnKD3ofE4qq3zTBlX0fhKh7evzwO3Dwn9Dl99CpHutLfR3WlzU4ceIER44cYfz48VitVvr06cOrr77K+PHjiYuLY+/evbz++uukp6dz0003UVVVhd1uJyMjg59//pkHH3wQl8vFH/7wBwYNGsTMmTPR6/VERUXRqVMnvv76a/7+978zYcKEEB/GxafZCExtDAZA78eQTJLURUl7QY1jOogJIVrypcRZDh8HYX25/TH1Uxc3l6oedPwwb9483nzzTY4dO8aDDz5ImzZt+Ne//kWHDh08lhpHjx7lww8/ZP/+/SxcuJDTp08zcuRInnjiCW69VfW88+yzz7JkyRKSk5MZPnw46enpVFZWsnnzZq677jrGjx/P7bffzoIFCzSBuZQYjHjPkNUkoj3knwCdCVwOiE+/fMxg9JFqQw9EqD1MANzWyg8//DAREREkJyfz1FNPERYW5rEfS0pKIiYmhlatWmG32zly5AipqWqktbS0NABKS0s94VLat2/PmTNnPNEc4uPj6dBBndW8XGKoNluBMRmE/x4G1IVMQ5TqFjYiBUwxF7VuF4QkBewVAOj2qPppIF588UX69evH8uXLefPNN0lMTKRr164IIXwsmDt37sz27du55ppr2L17NwMHDiQyMpJjx46RnJzMr7/+Snx8/HmFUGkqNEuBcTohzGyrewHS0gqqclU/yRoBsVgsPPnkk8yaNYuJEycSHx9Py5YtOX36tE/a66+/niVLljBq1CiPyf9zzz3Hbbfdht1u54knnsB4QVtgLz2SEPWt5DUtysrKiI6OprS01McQs/BkATnbd+ASBlKvdmBqNzawUisUdTimr+1HVuNyo6420dBcJgP30JAkMJotdc8ASbImLBoh0ywFxmAASfNUqdEINEuBMZoIrPBraFwAjSIwNpuNG264gaFDh3rt3y8vL2f8+PEMGjSIDz74oDGKBsCod4DhynK+p3FxuKhOMD744ANuu+02Nm3aFLQjjFCcYLgx6l1giruge9DQ8MdFdYJx3333cdddd1FeXo6iKEHlFYoTDFAtR/Rmk2ZxfJ7UdhP797//3W86975+9183mZmZTJo0yfM9Ly8PWZbJycmps9y6jDVrl3EpuahOMABOnz7N4MGDg3aCEIoTDACnC8IiNf3lQqjPTWx95OTkUF1djdls5uuvv6Z169YNV7lLzEV1ggGQnJxMTk4O2dnZPg7h/BGKEwwABIRFBTCJaSbUHqa6XSBZrVav425zk8rKSq/jdnsde4L8UPPtH8zbfvTo0Z7w8uvXr2fEiBEArFy5kkGDBjF48GBWr16NzWbjuuuuY8yYMaxcuRKAvXv3kpGRwZAhQ/jss89CqufFoFEExu0EQwjBzp07PbZDr732GllZWUiShNlsbhQTCUkSGC0XGkGpaVN7mPrQQw8B8NBDD3kdf+GFFwCYMmWK1/H33nuvzvxr+iTLzs4OuX6TJk1i+fLl2Gw2dDodhrP+ep977jmysrL4/vvvefbZZ/nqq68YOXIkK1euJDlZdTry9NNP8/7777Nu3ToWLFhAU1tXv6hOMG6//XbuvPNO/vKXvzBq1Ci6dOnS4GWbDPZm7yKp9jDV3SBfe+015s+f7zluOhsN97PPPsPlOudmymyuO8JUXUOyYBpwly5dOHToEGvWrGHUqFGesBg6nY7Is+42JUnya6x5+PBh7rzzTgDy8/MpKCjwU8Kl46I7wWjsmCIGgxzYSrmZYDKZPMJQE4vFgsXi60w6PPzCdDqTyUR+fj42m40jR44EdU2PHj343//9XxYvXuz5zZ1OJ+VnPanbbDa/xpqdO3fmrbfeIiYmhueff564uKY129nsjC9lU1iTcyB+uTNlyhSmTZvG4sWL652ldDNp0iS2bt1KfPy56AdPP/0048aNw+Vy8corrzBs2DAfY023z+XS0lJuvvlmdBcUPq7haXbGl8W52Vw18PLz2atx/mjGlxeAJUqbUtZoPJqVwEhIRMQ3b4Vf49LSrHSY8Ohw9BccBVZDIzDNSmBM4XVPl2poXCjNakimodHYaAKjoRECmsBoaITAZafDuJeNgtkXo3Fl4G4LF2NJ8bITGLdpRbArzhpXDuXl5UQHE5P9ArjsVvoVReHkyZNERkY2uLWze69Nbm5uo68Ya+U1HEIIysvLSU5O9pjYNBaXXQ8jyzJt2jRuwKOg9t1o5TWp8hq7Z3GjKf0aGiGgCYyGRghoAlMDk8nEs88+63eviVZe0y/vYnDZKf0aGpcSrYfR0AgBTWA0NEJAExgNjRDQBEZDIwQ0gdHQCAFNYDQ0QkATGA2NENAERkMjBK4ogZk6dSo5OTl+Az7VPhZMmkDk5uYyatQohgwZwrx58xq9vNqBqhq7PFAdlC9atOiilNWkEFcANptNTJ48WbRu3VpkZ2eL9957T7z++uvC5XKJcePGCavV6nMsmDRWq9VveY8//rhYs2aNEEKI0aNHN3p5b7zxhli4cKFQFEVkZGQ0enn5+fkiOTlZLFy4sNHLampcET2MzWbj0UcfZezYsYD/gE+1j33yySf1pnEHiqrN//zP/zBs2DAAXC4XGzZsaNTyageqauz7e+mll5g+ffpFeZZNjStCYCIjIz0xSsB/wCd/x4JJ44+4uDh0Oh1vvfUWaWlpOJ3ORi0P1EBVvXv3pn///o16f7m5uZSXl3tCmDT2s2xqXBECUxt/AZ9qH1MUpd40dW2Kev/99/niiy+YN2/eRSmvZqCqkydPNlp5L7zwArNnz76oz7IpcUUKjL+AT7WPZWZm1pvG/ZatzY8//si7777Lp59+itFobPTyageqGjRoUKOVt2XLFjIzM5k3bx7z5s0jLi6uUe+tyXHx1aZLx4wZM0R2drawWq1i8uTJon///mLBggVCCOFzLJg0gZg0aZLo1auXGDlypBg5cqQ4fvx4o5Z36tQpMWbMGDFs2DDx9NNPN/r9CSHEwoULxcKFCy9KWU0JbT+MhkYIXJFDMg2N80UTGA2NENAERkMjBDSB0dAIAU1gNDRCQBMYDY0QuGIFZvbs2fTs2fOieHwHKC4u5tNPPw0qbUlJCX379uXGG2/0HPvXv/5FRkYGffr0oWXLlmRkZDBp0qSg8ps2bVrAc4sWLeLbb78NKh9/tG/fnoyMDDIyMkhPT+fJJ59EUZSA6ffs2cOmTZvOu7xLziVeB7okOJ1OcfXVV4sZM2aIFStWXJQyV69eLWbMmBFU2rVr14rbb7/9gvO5GKSkpHh9f/jhh8U//vGPgOmfffZZsXDhwsatVCNyRQrMsmXLxD333CNWrlwpbrrpJiGEEGlpaWLWrFli8ODBYvbs2eLBBx8UaWlp4g9/+IMQQojt27eLYcOGiREjRojJkyeLoqIin8brbjwjR44UjzzyiMjIyBDDhw8XRUVFYvz48SIpKUm89957XnV5//33RXp6uhgyZIh44oknhBBCpKamiqSkJPHSSy/51L12mTNmzBDXXnutGDFihMjLyxNTp04V48aNE3379hWfffaZV7369OkjHnjgATF06FAxadIk4XA4PA144cKF4qabbhK/+c1vRPfu3cXXX38thBDiqaeeEv369fNYL9SmtsAcPXpUDBs2TAghxPz588WoUaPE4MGDxfTp08WpU6dESkqK6Nq1q9i1a5fP+cuBK1JgbrzxRrF+/XqhKIq46qqrPD/ktm3bhNPpFHFxcWL//v3C4XCI5ORkIYQQAwYMEPv37xdCqPtPHn/88ToF5vPPPxdCCDFz5kyxZMkSvz1DQUGB6NGjh6iqqhJCCDFt2jTx3//+t85exJ/AvP7660IIIbZs2SKWLVsmhBBi27ZtYuzYsV71SklJETt27BBCCDFmzBixadMmL4G54YYbhBBCrF+/XowfP17s3btXjB49WrhcLlFQUCBiYmJ86lNbYKqrq0XXrl2F0+kUf/3rX4WiKMLlconU1FRx/PhxT3mBzjd1LrtwFxdKfn4+q1evpqqqClBji7zzzjsA9O3bF0mSiIyMpFu3bgAYDAbPde5jI0aM4PPPP/fSIUQtXSg1NRWANm3aUF1d7bcuhw8fpnfv3lgsFk++e/bsYdCgQSHdk9twMTk5mddee41PP/0USZJwOBw+aeuqV+1ze/fuZcCAAciyTHx8fFAGkseOHaNNmzbodDpkWebOO+8kLCyM0tJSr/rUd76pcsUJzOLFi3n44Yd57rnnADh69CgZGRkoilJngKb4+HgOHjxIly5dWLt2LVdddRVms5mTJ08CsH37dq/0tfOSZdlHGe7YsSN79uyhuroas9nMunXrmDp1asj35A4i9PLLLzN48GDuv/9+PvroIxYsWOCTtq57rH2ua9euvPnmmwghKCkp4eDBg3XWQwjByy+/zLRp09i1axfLli1j8+bNFBUVkZqaihDC8xwCnW/qXHECs2jRIq/ZqpSUFDp27MiuXbvqvO6NN95g5syZCCGIiopi0aJFxMXFYTKZGDp0KFdffTUJCQkBr+/UqRNbt27lrbfeYtasWQAkJCQwe/ZsRo4ciRCCAQMGMGXKFNatW3de9zZx4kQefPBB3n//fdq0aUNRUdF55ePm6quvZtiwYQwePJiWLVsGDFqUkZGBJEk4nU4yMjKYOXMmVqsVs9nMgAEDMJlMJCcnc+rUKdLS0pg9ezb/+te//J7v0KHDBdW5sdGslTUCcubMGb788kvuvfdeiouLGT16NDt27LjU1bqkXHE9jEbwxMfHs3btWt5++210Oh1z58691FW65Gg9jIZGCFyxK/0aGueDJjAaGiGgCYyGRghoAqOhEQL/PyJmielpSkyEAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 171.26x171.26 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 171.26x171.26 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 171.26x171.26 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for expt, df in performance_dict.items():\n",
    "    summary_df = df.groupby([\"Mode\", \"Round\"])[\n",
    "        [\"AUROC\", \"AUPR\", \"n_train\", \"n_positive\", \"frac_positive_in_train\", \"frac_positive_sampled\"]\n",
    "    ].agg([\"mean\", \"std\"])\n",
    "    summary_df = summary_df.reset_index()\n",
    "    display(summary_df.sort_values(by=(\"n_train\", \"mean\")))\n",
    "\n",
    "    # Pull out the full model since this will be plotted differently\n",
    "    full_model = summary_df[summary_df[\"Mode\"] == \"Full\"].squeeze()\n",
    "    summary_df = summary_df[summary_df[\"Mode\"] != \"Full\"]\n",
    "    summary_df = summary_df.sort_values(by=(\"n_train\", \"mean\"))\n",
    "\n",
    "    expt_splitter = expt.split(\"_\")\n",
    "    title = \"Start with \" + expt_splitter[1] + \", Sample \" + expt_splitter[3] + \" per Round\"\n",
    "\n",
    "    colors = {\n",
    "        \"Entropy\": \"blue\",\n",
    "        \"Random\": \"orange\"\n",
    "    }\n",
    "\n",
    "    for metric in [\"AUROC\", \"AUPR\", \"frac_positive_in_train\", \"n_positive\"]:\n",
    "        alpha = 0.25\n",
    "        fig, ax = plt.subplots(figsize=plot_utils.get_figsize(frac=0.25, aspect=1))\n",
    "        for mode, data in summary_df.groupby(\"Mode\"):\n",
    "            x = data[\"n_train\", \"mean\"]\n",
    "            y = data[metric]\n",
    "            c = colors[mode]\n",
    "\n",
    "            ax.plot(\n",
    "                x,\n",
    "                y[\"mean\"],\n",
    "                label=mode,\n",
    "                marker=\"o\",\n",
    "                zorder=2,\n",
    "                color=c\n",
    "            )\n",
    "            ax.fill_between(\n",
    "                x,\n",
    "                y[\"mean\"] - y[\"std\"],\n",
    "                y[\"mean\"] + y[\"std\"],\n",
    "                alpha=alpha,\n",
    "                zorder=2,\n",
    "                color=c\n",
    "            )\n",
    "\n",
    "        if metric == \"AUROC\" or metric == \"AUPR\":\n",
    "            # I like having a bit of whitespace on the xaxis, so show the full model \n",
    "            # with plot instead of axhline\n",
    "            n = len(x)\n",
    "            ax.plot(\n",
    "                x,\n",
    "                [full_model[metric, \"mean\"]] * n,\n",
    "                color=\"k\",\n",
    "                label=\"Full Model\",\n",
    "                linestyle=\"--\",\n",
    "                zorder=1\n",
    "            )\n",
    "            ax.fill_between(\n",
    "                x,\n",
    "                [full_model[metric, \"mean\"] - full_model[metric, \"std\"]] * n,\n",
    "                [full_model[metric, \"mean\"] + full_model[metric, \"std\"]] * n,\n",
    "                color=\"k\",\n",
    "                alpha=alpha,\n",
    "                zorder=1\n",
    "            )\n",
    "\n",
    "        elif metric == \"n_positive\":\n",
    "            metric = \"Number of Active Sequences\"\n",
    "        elif metric == \"frac_positive_in_train\":\n",
    "            metric = \"Fraction that is Active\"\n",
    "\n",
    "        ax.set_xlabel(\"Amount of Training Data\")\n",
    "        ax.set_ylabel(metric)\n",
    "        ax.set_title(title)\n",
    "        ax.legend(\n",
    "            frameon=False\n",
    "        )\n",
    "        plot_utils.save_fig(fig, os.path.join(figures_dir, f\"k562_manyRounds_{expt}_{metric}\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "9109b0a5-bb0c-44d6-9aec-2aef21902b1b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUAAAAFHCAYAAADQlZUSAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAAArcElEQVR4nO3dfXRU9YH/8XcmDzPkYYJSHkQSQMQEMavGipEkNDSWRmmLGlnwbLba7rbVde1ZOd3Tw2519VROWNuy29atZ40oLLG0UFqqXcSGKBBgURuN7YGoLAbJ/jA8WMwkIXMDmfv7Y5Jx8mhy70xmyP28zpkzZHJz+d4w8+F7v48JpmmaiIg4kCvWBRARiRUFoIg4lgJQRBxLASgijqUAFBHHUgDKiO3cuZMNGzYM+r2VK1cCsG/fPt55551hzxMIBCgvL2f37t2hr7/+9a9TWFjIrbfeypkzZwD49a9/zY033khBQQH//d//DcDJkycpLS2luLiY+++/n0AgAMA///M/c9NNN1FSUsLRo0eH/Ls3bNjAzp07R3PZg2poaGDhwoUUFBSwZs2aiF2HjDFTJILuuece89VXXx3y+8ePHzcXLVpkZmdnh4771a9+Zd5///2maZrmL37xC3PVqlVmV1eXmZuba7a1tZmtra3mtddea3Z1dZl///d/b/7yl780TdM077vvPvOFF14w//CHP5i33XabaZqmefDgQfPOO++M6jWapml+8YtfNN966y3TNE2zpKTEPHr0qO3rkLGnGqCM2IYNGygpKeG2227jK1/5CvPnz+eZZ54BYNasWTQ0NLBz507+4R/+gdOnTw96jvb2dn72s5+xePHi0Gv79+9nyZIlAJSVlfHKK6/Q2NhITk4O6enpeL1eZs+ezeHDhwc9dv/+/XzhC18A4KabbuKtt94a8hoeffRRNmzYwIYNG1i+fDm33XYbV199NTt27ADgP//zP/nsZz/L9ddfz1NPPTXkeX73u99x3XXX0dHRgc/nIzU11fZ1yNhTAMqotbS0sH37dl5++WV++MMfhl6/7rrrKCsr49///d+ZPHnyoD87b9485s+f3+c1n89HZmYmABkZGbS1tfV5bajXhzrWHOHY/vPnz7Njxw6efvppfvKTn3Dy5El+/OMfs3//fl577TXee++9Ic+VlJTEW2+9xTXXXMO0adOYOHGi7euQsacAlFHLy8vD5XIxY8YM/H6/7fN5vd5QALS1tZGZmdnntaFeH+pYl2tkb+trr70WIHQdTU1NXHvttbjdblJSUvi3f/s3EhIShvz566+/nqamJm6++WbWrl1r+zpk7CkAZdSGCwWXyzXqBv2bb76Z3//+9wC89NJLLFy4kNzcXN55551Qbendd98lNzd30GNvvvlmampqADh48CBXX321peu44oorOHz4MOfPn6e7u5uysrIhA/7WW2/l//7v/wBC4WX3OmTsJcW6ADK+LFiwgFWrVvHb3/6WmTNnjuhn7rzzTl566SUKCwtJSUnhF7/4BSkpKaxZs4bS0lIuXLjA448/TnJyMg8//DB//dd/zY9+9CNyc3O54447cLlc/MVf/AUFBQUkJCQM2VP9aaZMmcKDDz5IcXExpmnyt3/7t3g8nkGP/cd//EeWL1+O2+1mypQprF+/ntTUVFvXIWMvwRxpg4mIyDijGqBExfLlywf0BOfl5fHTn/50TP7+NWvWhG6Lw9XW1pKYmDji88T6OiS6VAMUEcdSJ4iIOJYCUEQcSwEoIo41LjpBAoEAJ06cICMjY9gxaiLiTKZp0tbWxvTp0/sMlB8XAXjixAmysrJiXQwRiXPNzc3MmDEj9PW4CMCMjAwgeHFerzfGpRGReOPz+cjKygplRa9xEYC9t71er1cBKCJD6t9Epk4QEXEsBaCIOJYCUEQcSwEoIo6lABQRx1IAiohjKQBFxLEUgCJy8eg2Qn80jGGOGyEFoEgcisSHO650R+CCDlTAb2fBgQoqKmDWLKiosHdKBaBInInUhzvSLIdyWHABDLFl9PC6DWipBX8LgQ9rqdtt0NICtbX2/rNQAIrEEcMIfqgj8eEG8PkiUy7LoRwWXLTUckmmwZQpkJY2yvMkumFaKXim4bqslOISN9OmQWkpuN2jPFeYqASgYRgsW7aMwsJCqqqqQq/7fD7KysooLCzk2WefHfZYESdyu4Mf6kh8uGfPhksuCT5b1uWzF8qJbnB5ABcX8PCxL3hB585ZqAkurIYvvQsLq6muhnffherqUZ6jn6gE4JYtWygrK6Ouro6tW7eG9lZ9+umnWb58Ofv37+fFF1/k3LlzQx47HMMw8Pl8fR4i8cBWjS2snczWTj1dPnw+OH4cAoHgs6WPyPbZ8KtLcL80O/TzPt8oQ7nbgIAfCOAy/aQkfXKNo1635EAFvDg31AY4d26ctgHW19dTVFSEy+UiLy+PxsZGAI4ePUphYSEAOTk5NDY2DnnscCorK8nMzAw9tBagxANbbXc97WQX6iqorYWTJy3eAveElveV2fQufJKQYCFsunxw7jgQwDx3HFcgmIDnzo0yTHtvXd1TcV1WyvIVbqZMgb/6KwtBeux5ME5hHnuerb80OHUKnn8+DtsA29raSE9PByAtLY329nYAcnNzeeWVV+ju7mb//v34/f4hjx3O6tWraW1tDT2am5ujcRkiI2brNjGsnSzpdC3eNAOXCzyeUYZEv9CakBxMqu5uCzXAFC+kZgMuzNRsznUFE9TlsnBbfmo/GKfh1H6qq+HIEQu3rl19LyDd88nXdm4AoxKA6enpdHR0ANDR0RFao+8b3/gG+/bt44477uCKK67gkksuGfLY4bjd7tDaf1oDUCLBbmeD2x0MLEvBFdbAf2FyKb4ON4EA+P2jLFeKFxI9AJiJHtr9kfhcmLiAu++Gz3wm+Dz6UD4GBODcMf72Hh9XXWWhlpzS91rCr83Oxz8qAZifn8/evXsxTZOGhgZycnIAeP3113nooYfYvn07Pp+Pq666ashjRT5NpNrbbA876TYwjL5tbpbKFgiQlGgjSLsNSO6pqSV7megNFiI11c4tsAnnjlPzko8zZ+A3vxnleRLd9MaMiYutv3Zz8iRs22ahMyUxFYCExFSSUoK/mNTUOOwFXrFiBTU1NSxYsIDy8nKOHDnCunXrmDNnDt/+9rcpLCzk7rvvJikpacCxHo8nGkWScSYS7W0cqLA/7KTnXAkHKwgEgi8FAhZugY9vg65TBI5vI3DesFYDTHTD+eD9YOC8D09qMBm8XgvXFVabDLg8nPpzMEFH3Xub6O65lQYzNZuuCxbTqssH3cEOUrPbTxLB6/T77d0CR2VJfI/Hw2/6/VeRl5cHwP/8z/986rEiwxkstEZcC+g3Ls2dZODxuK3XuHrb7s7U4kkx8He5rbWT9XABp88E/3z6tIXbzZ6QcHX7WbrEx+9e9lobTtNtgBnoKVOACW6DTsNNYiJMnjzK8/hbgufxt5B9ucH/NgXH8I2qTL1tkueOk5CazaVTvLQfh+zsOLwFFhlOJNrbLI+VC2tvY1opxgU3LS3BWltLi/Uxbq5ED+XL3UydaqGdLNEN2eXgmYp/SjmdRvCHu7tHWdsKdVwkQGo2/m4vkdglNgAkJwf/nJxs/d/PBFpOWvxdA9zeBHedhdubaGqCs2ehqclaWXopAGVMRXKal62xcrZ+uEfYGDcCfqo3Grz3no3BuWawFpoabOoiNXWUtS2AKYXgmcKFSYX2bu0T3cH/JAA806zfuoaFe/fl5dbP0yslMp0fvRSAMmYiMs2rp8PB8li53vY24yQc34Y7yaC8HKZOhfJyi7VJ91SYVkrFPW5rPZz9yrS8PDhd7I47LJynpRb8J+0Np+k9V9it64rlBtOmWfgd9TIhKTH487bOE2HjYltMuQh0G7jdbkpLg6FlqV3qQEWw3W5aKaWl1dbPM1jxuu38tMmF7mDPpt8ffF6/fpTlCnQFnwJdvPoKnDploX0z7JY84PIMGE5j5/e04TkwLlgM0pbaYLi31FK90cDX4Y5I7S0SVAOUEYvEaiDV1XDsmIXbxH6dF9UbDf74Rwvn6b0lS5kC2eUYF9w8/zycOWNhVkG3Acc2g3GKxObNEAj+cFfXKMsE4EoJPrlS7A2D6Z12FvDzxS8Y1ucUh26BXcHnRLe1AB2klpyTY2/IUS+tByhjJlKrgQRrghYKEFa7weVh9pXBnkTLE/1dwbd+/06GUXU6dBsEuwgAAqSnBj+RKSmjLEuiGwLng2cJnLc+ELrf72jDf7mt/WcDA9o3ba/nl5DAhe4IDDnaPjM0F3jmzDidCyzji+3VQMJ6XYMDYy0I+0B2X/DTcsKwNtG/XyDPuMwgMbGnqIkwY8YozhU2yDcBF5dOCg6BGfUQj87TQPAe3EU3y7542lrNLdRuFwg+2wmtQf7drLbZhk/zs1wrDWsnDRzfxou/NawNqO5HbYDyqXqneSUkWGxQX1gdfANbDT/oU7tJTPIwbbqb41bGgfV+sD/cFfpgr1wJv/89LFlioUyJHug+RyDRRpvbhMnBWQ7d5yAxlfYLkyPSSX3v1+DlmmDYWKoFLqwOji1M8VJR8Unb7ajONUit1Oez14MbgZE9IaoBytB6ahCGEaz9mabF8VtgL/x6yxJ2S/a5YoNJk6BncaHR6xkg11u7PX3aQu22yxcMLcDVfY5JXp/1XtcVHXDHKYzbO6z3cIfa7RIIeKbxco3bXo/7gQr4XU5ohRrLCz2E/bvd+1XDWhtgaDjNNBKyy/nyMndEepMVgDK4fsuY22a38bq35pYyhQuTS9n0vJvTp2HzZusrr/TOBLE1qDpsnmub1ba7XhMm2xvkHQob034nSKRuXfst9GArlBdWw7JjoQVRLbdvhtEtsAw0SEiUl7vtDV/5cBdcdgsVP6u2disFwWWVus6QcGb/gHm3VoaK4PJAopvqaovDRBLdMOtuaNlFwrRbKC6x8TuCUDOBrfJMKw3+200rZcN/ua0Pf4nkuXqaQJISbQ6D6i1Xj0gMf0owzUi0NsSWz+cjMzOT1tZWLY0F9tvbIDTmjmmlwTcwFj+U3QZsnQgBPwGXh9nf+Zjj/y94+3Ls2CjO1+WDX11CsNfVxTWPnaXxf71kZ49yOlS3EazZ+luCt4vLjkXm9rznHJZDYpDfdyTKY1skz9XD7phEK4bKCN0CxylbY+56hgrYOk/Y/gu9LL9pAz0T9AN+Fn8+OOti1DWAsAU6Sc3muhu9TJ5soQ0wbMUUzvsi8+G2WysZZKhQpMpjW4TDD+JjBkgvBWAcsjXmLmxK1b1fNayPleppAA9vA7Q8DCLMH14zOH0a9u+3cC4ATALYmAoX1nFB97nQSsORWlvQkkgNFZJRUwDGmUhti2gSXLzS0lipQWoklkM5xRtayDKQmMqh97wEAsHb31GN3wtboNN17jhLl/hsd1yACxLdEVtb0JawBn4ZOwrAOGN7qafscnDbXHljkCWjbA2D6FmlmCQvnpTgD496zbywBTpJ9PDMRq+1bRF7Oy7cU2HW3favLV5vXWVE1AscT+z2AvZKSCApMRiix49bmJkAwZpI5+ng0Ayw3ns3yJp5NTXwhS9YGOLRs0AnZoB7v2rwco3bWm9y2ABf29cW1lOqALv4KADjRb9eQLuN6YEPa3uWVndbWw1k+2w49wGkzoTbm6iuDg4WHvX6dP0Gwia5DFwue0FhAq++YnFFaBjwu7b1H04kZrlIzOgWOB4Mcitla7CwZxquy0opLnFbu5UO7eRlBp+7fMyejbXFB8JWA7E1ELbfTADL1zbEbautnkmF30VLNcB40O9WquIet/XBwhBa7bi6GmvzLvt9oH0dwVVF4JPOi1Gfs+e23NZA2LDaVvXCyAwWVng5mwIwXvR8uAdrlB/1hj89i0/aaifDRe+gY7c72GkRCFjovBhkHb9ILYhpudam21bpoVvgOBOpDX9sz7vsWaATVwpud3CjnylTLG74M60UPBFYEDOS85MVfoJqgPEjrGG+uro6IvMuLS9h1buqyLkPQqsB2+ooOLUf/KcInNofse0sVYOTSFAN0C67Y796zxGJVZN79SxeaXkJq7BVRcJXA7ZUpkgOYNZsCYkw1QDtiNQE9kg3zHcbgM3FRwcpj6XOj7ANrUnN5pmNXusLYqrtTiJMAWhVNG7JIrEwT8/SU+7LbqG83MbOaWEDoSE4/KV3BeZRb0Z9e5P9lYV7hf2OY7GqiIwvugW2aohbMlv7JvT03loeB9hvMYTqjYb1RSO3z4bfTIPts/H5guFnaQ+OXineiM1zhshusC7OpQC0o98EdssfykG2DrT84e7ZXzb0bEWo3S4A547j9fjIzg4OgRn1HhxhbPVwh4lkkIqzRSUADcNg2bJlFBYWUlVVFXq9s7OTsrIyCgoKePbZZ4HgQoWXX345JSUllJaWRqM40RVW87P9oezZOnDbtuB5LO14FTZ85d6v2VjBJWztPVK8NDXB2bMWbn/7icRS5pEKUpGoBOCWLVsoKyujrq6OrVu34vcHF8TcuXMnxcXFHDhwgPXr1wNw+PBh7rvvPnbv3k1tbe2Izm8YBj6fr88j1mzv5dDTnug6VUtKksUqTSQ3xbm9Ce46G3zuEanFtiMRWJHaE0KcLSoBWF9fT1FRES6Xi7y8PBobGwGYN28eXV1dnD9/nuTkZAAOHTrEzp07WbRoEc8999yIzl9ZWUlmZmbokZWVFY3LGDXLH8qwwcKuy0q59DPBhBj1Ki6R3BQHgjXBOKaan9gVlQBsa2sjPT0dgLS0NNrb2wFITk5m8+bN5ObmsqRnE9ZZs2bxxBNPsGvXLjZu3Mjp06c/9fyrV6+mtbU19Ghubh59Ie3uUjYEyx/KsMHCLS3Bl0Y9fm+QPVhVSxIZWlQCMD09nY6ODgA6OjpCm5A8+eSTVFZWcvToUV5//XU++OADCgoKuPnmm0lJSaGgoICmETQyud1uvF5vn8eohE2piovexLDBwgnnjpPusXhL32/pKdsDqkXGuagEYH5+Pnv37sU0TRoaGsjJyQGCtcGMjAxcLhder5f29nYeeeQRampqCAQC1NfXM2fOnGgU6RP91syr223Y67iIxEyQsNWOExI9fHGplylTLGz6rNkSIqMSlQBcsWIFNTU1LFiwgPLyco4cOcK6det48MEH+cEPfkBxcTGTJk1i/vz5rFq1irVr11JcXMxdd93FpEmTolGkT0RqzTwYdHK+5RDtXTY+2csbB21sHKS9JURGzLn7Atvdy3WQ/WUr7nGzaxfccouFdreeaXVdk0qZ8Pnq0NJTZ8/a633VbAkR7Qs8kN29XHtrkilTQhsHbdtmcRc2CNbYbv0jKZ+rjsigY9BsCZFPo7nAdpzaD11ngs92he/BQZPtacGDDcxWTVCkL+fWAO3qN13MneDD1fPbHPWqyWF7cJjnjvHnUz5M08a8WzRbQmQkVAO0qrfntvscJHrw+b2cOxf81rlzo1w6KrRZd4AEXEy7zE17k/1bYMt7gog4hGqAVvXbp9adZFivAfZu1u0Jbtb97v+67c277RmaU1GB9eXnRRxANcAI6d03o7cX2M6OZ2Cj1tbTm3xhcim1tdVqAxQZhgLQqkQ3JPRU+RJc9vfN6D2nHWGDvJNO1/LFL3yyK5zCT2QgBaBVXT7oDq5yQ7c/tOJxTPVbyn7Df7lV8xMZhgLQqn6dIBFZ7j0S+t1KK/xEhubcThC7q8H0m75mnDPiZ5VizQEWGRFnBmAkVoPpt/CAO9WtcXciFxnnzQXuNmDrRAj4MV0eJn7jY3ztwU3EP/7YQnD1a/tTm5tI/NFc4HA9GwaZgS66evYO6rKyh9CBCvhdTp/VYETk4uHMAOzZOCjBlUJKzx5Cvc8jNsi+wFp8QOTi4rwATHRDdjl4ppGQXc6XlwXXA7S7+KhxwR0/nSAiMiLOHAazsDrUdle90MZ82bAhJ26CnR+7dqkTRORi4bwaIPRpu7M9X1ZDTkQuWuoFttsL3MMwYOJE8PuxfS4RiSz1Ag/iok9+EbHFeQEYtoiBK8HFhUCwmjbqJaz6cbuDHSlTp1roUBGRmHBeJ0jYIgZmt58kfIAXv9/+4qG2V4MRkTHlvBpgihdSswEXCanZXDrFG5ENiHop/EQuHs6rAQLc3hQaBtN0u5aNF3Eq59UAe4XN31X4iTiTcwNQRBzPuQFodz1AEbnoOTMAD1TA9pmhmSAzZ2oBAxEnikoAGobBsmXLKCwspKqqKvR6Z2cnZWVlFBQU8Oyzzw57bNR0G3B8GxgnCRzfxou/NTh5ErZtU01QxGmiEoBbtmyhrKyMuro6tm7dit8fHHe3c+dOiouLOXDgAOvXrx/22OEYhoHP5+vzsCLB0k+JyHgRlQCsr6+nqKgIl8tFXl4ejY2NAMybN4+uri7Onz9PcnLysMcOp7KykszMzNAjKytr5IWL1HJYInLRi0oAtrW1kZ6eDkBaWhrt7e0AJCcns3nzZnJzc1myZMmwxw5n9erVtLa2hh7Nzc2jK+DCalh2DBZWU10Nx47FcBc3EYmZqARgeno6HR0dAHR0dIRWX3jyySeprKzk6NGjvP7663zwwQdDHjsct9uN1+vt8xi1sGWsVPMTcaaoBGB+fj579+7FNE0aGhrIyckBgjW8jIwMXC4XXq+X9vb2IY8VEYm2qATgihUrqKmpYcGCBZSXl3PkyBHWrVvHgw8+yA9+8AOKi4uZNGkS8+fPH3Csx+OJRpFERAZw3oKoIuI4WhBVRKQfBaCIOJYCUEQcSwEoIo6lABQRx1IAiohjKQBFxLEUgCLiWApAEXEsBaCIOJYCUEQcSwEoIo6lABQRx1IAiohjKQBFxLEUgCLiWApAEXEsBaCIOJYCUEQcSwEoIo6lABQRx1IAiohjKQBFxLEGDcDq6mouXLgAwGuvvcamTZvGtFAiImNhQAB+//vf59e//jVdXV0ATJ48mW3btrF27doxL5yISDQlmKZphr9w0003cfDgQRISEkKvnT9/noULF/LGG2+MeQFHYqhd30VEYOiMGFADTE1N7RN+AMnJyWRkZES/lCIiY2hAAF566aUcOnSoz2uHDh0iJSVlzAolIjIWkvq/sG7dOsrLy5k7dy4zZ87k/fff5/3332fz5s0jPqlhGPzlX/4lZ86c4d577+Ub3/hG6NwvvPACAG+99RavvvoqV155JfPmzWPu3LkkJiZSW1sboUsTERnegDZAgO7ubvbt28eHH35IVlYWBQUFJCYmjvikmzZtor29nW9961uUlZXxwgsv4PF4Qt8/dOgQa9as4ec//zkHDx6kpqaGhx9+eMTnNwwDwzBCX/t8PrKystQGKCKDGnEbIEBLSwsHDhxgz5497N27lxMnTozqL6uvr6eoqAiXy0VeXh6NjY19vr9mzRoee+wxIBiGO3fuZNGiRTz33HMjOn9lZSWZmZmhR1ZW1qjKJyICgwTg66+/zpe+9CUmTZrE7bffTmZmJkuXLuXgwYMjPmlbWxvp6ekApKWl0d7eHvpee3s7H330EXPnzgVg1qxZPPHEE+zatYuNGzdy+vTpTz3/6tWraW1tDT2am5tHXDYRkV4D2gC/973v8eKLLzJjxozQa0uXLuVv/uZv2LVr14hOmp6eTkdHBwAdHR19qpw7duzgzjvvDH1dUFDAhAkTcLlcFBQU0NTUxOTJk4c9v9vtxu12j6gsIiJDGVADvHDhQp/wA5g5c2ZoYPRI5Ofns3fvXkzTpKGhgZycnND39uzZQ3FxcejrRx55hJqaGgKBAPX19cyZM8fKdYiIjNqAAOzu7qa7u3vAa+GdDp9mxYoV1NTUsGDBAsrLyzly5Ajr1q0DoKmpqU/Arlq1irVr11JcXMxdd93FpEmTrF6LiMioDOgF/ulPf8rhw4f54Q9/SFpaGp2dnTz00ENceeWVfOc734lVOYelmSAiMpwR9wI/+OCD5OXlcfPNNzN79mzy8/OZN29e3IafiIhVg44DHMyZM2f4zGc+E+3yWKIaoIgMZ8Q1wIaGBu666y6+9a1v8ec//xmAqqoqrr/++rErrYjIGBgwDOab3/wma9eu5dixY6Hxdh999BGvvvpqLMonIhI1A2qAEyZM4POf/zxf//rX2bFjB9dccw01NTVceeWVsSifiEjUDAjA8Dm/M2bM4Hvf+96YFkhEZKwMuAX2+/00NzcTCARISEigubmZ3n6S7OzsMS+giEi0DOgFLikpGbAgKkBCQgKvvPLKmBVsNNQLLCLDGSojBtQAd+/ePZblEhGJmQEBuHjx4lANMCEhAY/HQ35+Pt/97ndDK7yIiIwHAwJww4YNfb5ub2/npZde4pvf/CY///nPx6pcIiJRNyAAZ86cOeCg+fPnU1RUNCYFEhEZK4OuCD2Y0awGIyJyMRhQAzx+/Hifrw3D4De/+Q1XXHHFmBVKRGQsDAjAe+65p8/XqampXHfdddx+++1jVSYRkTExIADD5/yePHmS9evX88wzzzBnzhzuvvvuMS2ciEg0DQhACIbgU089RUNDAxAcG6hZICIy3gzoBLnmmmvYtGkTf/d3f8e7777LnDlzFH4iMi4NqAEuXbqUl19+meTkZFJSUgadFiciMh4MqAH+67/+K2+++SZf/vKX+dGPfsTBgwd57LHHOHr0aCzKJyISNYOOA3S5XHzpS19i27ZtvPfee3i9XsrLy8e6bCIiUTXiPUHimVaDEZHhjHhPEBERp1AAiohjKQBFxLEUgCLiWApAEXGsqASgYRgsW7aMwsJCqqqqQq+vW7eOkpISSkpKyMzM5M033xzyWBGRaItKAG7ZsoWysjLq6urYunUrfr8fgFWrVrF7927+4z/+g6VLl5Kfnz/kscMxDAOfz9fnISIyWlEJwPr6eoqKinC5XOTl5dHY2Njn+2vWrOGxxx4b0bGDqaysJDMzM/TIysqKxmWIyDgXlQBsa2sLbaCUlpZGe3t76Hvt7e189NFHzJ0791OPHcrq1atpbW0NPZqbm6NwFSIy3g26HJZd6enpdHR0ANDR0dFn5PWOHTu48847R3TsUNxuN263O8KlFhGniUoNMD8/n71792KaJg0NDeTk5IS+t2fPHoqLi0d0rIhINEUlAFesWEFNTQ0LFiygvLycI0eOsG7dOgCampqYMWPGkMd6PJ5oFElEZAAthiAi454WQxAR6UcBKCKOpQAUEcdSAIqIYykARcSxFIAi4lgKQBFxLAWgiDiWAlBEHEsBKCKOpQAUEcdSAIqIYykARcSxFIAi4lgKQBFxLAWgiDiWAlBEHEsBKCKOpQAUEcdSAIqIYykARcSxFIAi4lgKQBFxLAWgiDiWAlBEHEsBKCKOpQAUEcdSAIqIY0UlAA3DYNmyZRQWFlJVVRV6vbOzk5UrV1JUVMQDDzwAgM/n4/LLL6ekpITS0tJoFEdEZFBRCcAtW7ZQVlZGXV0dW7duxe/3A1BVVcXSpUvZt28fubm5dHZ2cvjwYe677z52795NbW3tiM5vGAY+n6/PQ0RktKISgPX19RQVFeFyucjLy6OxsRGAuro63n//fRYvXkxGRgYTJkzg0KFD7Ny5k0WLFvHcc8+N6PyVlZVkZmaGHllZWdG4DBEZ56ISgG1tbaSnpwOQlpZGe3s7AGfPnmX69OnU1NSwadMmzpw5w6xZs3jiiSfYtWsXGzdu5PTp0596/tWrV9Pa2hp6NDc3R+MyRGScS4rGSdPT0+no6ACgo6MDr9cLwMSJEykpKSEpKYkbb7yRpqYmCgoKmDBhAi6Xi4KCApqampg8efKw53e73bjd7mgUXUQcJCo1wPz8fPbu3YtpmjQ0NJCTkwPADTfcwJ49ewB4++23ueqqq3jkkUeoqakhEAhQX1/PnDlzolEkEZEBohKAK1asoKamhgULFlBeXs6RI0dYt24d999/P9u3b2fBggUsWrSIzMxMVq1axdq1aykuLuauu+5i0qRJ0SiSiMgACaZpmrEuhF0+n4/MzExaW1tDt9siIr2GyggNhBYRx1IAiohjKQBFxLEUgCLiWApAEXEsBaCIOJYCUEQcSwEoIo6lABQRx1IAiohjKQBFxLEUgCLiWApAEXEsBaCIOJYCUEQcSwEoIo6lABQRx1IAiohjKQBFxLEUgCLiWApAEXEsBaCIOJYCUEQcSwEoIo6lABQRx1IAiohjKQBFxLGiEoCGYbBs2TIKCwupqqoKvd7Z2cnKlSspKirigQceGPZYEZFoi0oAbtmyhbKyMurq6ti6dSt+vx+Aqqoqli5dyr59+8jNzaWzs3PIY4djGAY+n6/PQ0RktKISgPX19RQVFeFyucjLy6OxsRGAuro63n//fRYvXkxGRgYTJkwY8tjhVFZWkpmZGXpkZWVF4zJEZJyLSgC2tbWRnp4OQFpaGu3t7QCcPXuW6dOnU1NTw6ZNmzhz5syQxw5n9erVtLa2hh7Nzc3RuAwRGeeSonHS9PR0Ojo6AOjo6MDr9QIwceJESkpKSEpK4sYbb6SpqWnIY4fjdrtxu93RKLqIOEhUaoD5+fns3bsX0zRpaGggJycHgBtuuIE9e/YA8Pbbb3PVVVcNeayISLRFJQBXrFhBTU0NCxYsoLy8nCNHjrBu3Truv/9+tm/fzoIFC1i0aBGZmZkDjvV4PNEoksQRw4h1CUSCEkzTNGNdCLt8Ph+ZmZm0traO6BZaYqeiAmprobQUqqtjXRpxiqEyQgOhZcwYRjD8WlqCz6oJSqwpAGXMuN3Bmt+0acFn9WNJrEWlF1hkKNXVwZqfwk/igWqAMuYUfhIvFIAi4lgKQBFxLAWgiDiWAlBEHEsBKCKOpQAUEcdSAIqIYykAZcxpCpzECwWgjKmKCpg1K/gsEmsKQBkzWgxB4o0CUMaMFkOQeKPFEMa5SC48EIlzaTEEiSeqAY5jkWxvi+S5FH4SLxSA41Qk29vUdifjlQJwnIpke5va7mS80p4g41y8tQGKxIL2BHGoSAaWwk/GGwWgiDiWAlBEHEsBKCKOpQAUEcdSAIqIYykARcSxFIAi4ljjYjGE3rHcPp8vxiURkXjUmw39532MiwBsa2sDICsrK8YlEZF41tbWRmZmZujrcTEVLhAIcOLECTIyMkhISBjzv9/n85GVlUVzc3PcTMWLtzLFW3lAZboYywPWymSaJm1tbUyfPh2X65OWv3FRA3S5XMyYMSPWxcDr9cbNm6RXvJUp3soDKtNIxFt5YPRlCq/59VIniIg4lgJQRBxLARgBbrebf/mXf8EdR8ulxFuZ4q08oDKNRLyVByJbpnHRCSIiYoVqgCLiWApAEXEsBaCIOJYCUEQcSwEoIo6lABQRx1IAiohjKQBFxLEUgDY1NzezePFiFi5cyNq1a2NdnD4effRRNmzYEOti0NnZycqVKykqKuKBBx6IdXGAYJnKysooKCjg2WefjXVxKC8v59ixYxiGwbJlyygsLKSqqiouyhQv7/He8vSKxPtbAWjTj3/8Yx599FEOHDhATU1NaG3CWDtz5kzMP0C9qqqqWLp0Kfv27SM3N5fOzs5YF4mdO3dSXFzMgQMHWL9+fczK0dXVxR133MFrr70GwJYtWygrK6Ouro6tW7fi9/tjXqZYv8f7lwci9/5WANr0T//0TxQVFQHQ3d1NUlJ8rDD2xBNP8NWvfjXWxQCgrq6O999/n8WLF5ORkcGECRNiXSTmzZtHV1cX58+fJzk5OWblMAyDhx56iFtuuQWA+vp6ioqKcLlc5OXl0djYGPMyxfo93r88ELn3twLQpksvvZTExESefvpp8vPz4+LD3dzcTFtbGzk5ObEuCgBnz55l+vTp1NTUsGnTJs6cORPrIpGcnMzmzZvJzc1lyZIlMStHRkYGixYtCn3d1tZGeno6AGlpabS3t8e8TLF+j/cvTyTf3wrACKiurmb79u1x0wZYWVnJd7/73VgXI2TixImUlJSQlJTEjTfeSFNTU6yLxJNPPkllZSVHjx7l9ddf54MPPoh1kQBIT0+no6MDgI6OjrhZhDSe3uORfH/Hx/3aRey1115j48aNvPDCC6SkpMS6OAC88cYb3HvvvbS0tABQVFTElVdeGbPy3HDDDezZs4e5c+fy9ttvs3r16piVpVdaWhoZGRm4XC68Xm9MalqDyc/PZ+/evcyfP5+GhgbWrFkT6yLF3Xs8ku9vBaBNjz/+OC0tLdx6660APP/881x++eUxLdMbb7wBEOohi2X4Adx///1UVFTw9NNPc8cddwy6NPlYe/DBB6moqOD73/8+n/3sZ5k/f36siwTAihUruPvuu3nuuef42te+hsfjiXWR4u49Hsn3t9YDFBHHUhugiDiWAlBEHEsBKCKOpQAUEcdSAIqIYykARcSxFIAS93bv3s20adMoKSlh8eLF3HDDDXz729+O+N9TUlLSZ7URGf80EFouCmVlZaGBr6Zp8rnPfY4//elP5OXlxbZgclFTDVAuOu3t7Xz88cd4vV4qKiooLi5m4cKFvPLKKwDMmjUrdOy9997L7t272bBhA8uXL+e2227j6quvZseOHQD85Cc/IT8/n6985SuhqVXiHKoBykVh586dlJSU8OGHH5KRkcHDDz/Miy++yMyZM6murubkyZMUFRXx7rvvDnmO8+fPs2PHDvbt28fjjz8eWnT0zTffpKurK25Wz5GxoxqgXBTKysrYvXs3NTU1dHR0MHv2bA4fPkxxcTEAU6dO5dJLL+XUqVN9fi58pue1114LwIwZM/D7/bzzzjtcc801JCcnk5aWxnXXXTdm1yPxQQEoF5Xs7GyeeuopVq5cSU5ODvv37wfg5MmTnDx5kkmTJuHxeDhx4gQXLlzgT3/6U+hnExIS+pxr1qxZ/PGPf6SrqwvDMDh06NCYXovEnm6B5aJTUlJCSUkJLS0tHD9+nOLiYjo7O3nyySdJTk7mO9/5DkuWLGHmzJnMmDFjyPNMnTqV1atXU1BQwGWXXcakSZPG8CokHmg1GBFxLN0Ci4hjKQBFxLEUgCLiWApAEXEsBaCIOJYCUEQc6/8D7B1saCV0RvQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 342.519x342.519 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 342.519x342.519 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 342.519x342.519 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 342.519x342.519 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for expt, df in performance_dict.items():\n",
    "    df = df[df[\"Mode\"] != \"Full\"]\n",
    "    expt_splitter = expt.split(\"_\")\n",
    "    batch_size = int(expt_splitter[3])\n",
    "\n",
    "    for metric in [\"AUROC\", \"AUPR\"]:\n",
    "        fig, ax = plt.subplots(figsize=plot_utils.get_figsize(frac=0.5, aspect=1))\n",
    "        offset = 0.09\n",
    "        for mode, data in df.groupby(\"Mode\"):\n",
    "            # x = data[\"n_train\"]\n",
    "            x = data[\"Round\"].str[5:].astype(int)\n",
    "            y = data[metric]\n",
    "            c = colors[mode]\n",
    "\n",
    "            if mode == \"Entropy\":\n",
    "                x += offset\n",
    "            elif mode == \"Random\":\n",
    "                x -= offset\n",
    "            else:\n",
    "                raise ValueError(f\"Mode should only be Entropy or Random, saw {mode}\")\n",
    "\n",
    "            ax.scatter(x, y, label=mode, color=c)\n",
    "        ax.set_xlabel(\"Round\")\n",
    "        ax.set_ylabel(metric)\n",
    "        ax.set_title(expt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f0b433a9",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}