results_multiclass_mnist.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "354d68e6-6e62-4a68-bf3f-8477082c7c03",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/mathieubazinet/.pyenv/versions/3.12.3/lib/python3.12/site-packages/tqdm/auto.py:21: 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 wandb\n",
    "import yaml\n",
    "from utilities.utils import correct_type_of_entry\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sn\n",
    "pd.set_option('display.max_rows', None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4ca462b4-d80c-4420-8d68-214e505a0a9b",
   "metadata": {},
   "outputs": [],
   "source": [
    "ignore_lr = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "3925cc42-39cd-4242-aa3a-162df81d23eb",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = \"mnist\"\n",
    "sweep_config = \"pretraining\"\n",
    "values_to_fetch = ['kl_bound', 'binomial_approximation_shah', 'p2l_bound', 'compression_set_size',\n",
    "                   'validation_error', 'test_error', 'CE_kl_bound', \"CE_min_val_catoni_bound\",\n",
    "                   'complement_loss', 'validation_loss', 'test_loss', '_runtime']\n",
    "\n",
    "sweep_config_name = \"./configs/sweep_configs/\" + sweep_config + \".yaml\"\n",
    "with open(sweep_config_name) as file:\n",
    "    sweep_configuration = yaml.safe_load(file)\n",
    "    \n",
    "hps = {}\n",
    "for key, item in sweep_configuration['parameters'].items():\n",
    "    if item.get('values', None) is not None:\n",
    "        hps[key] = correct_type_of_entry(item['values'])\n",
    "\n",
    "if ignore_lr:\n",
    "    hps['training_lr'] = [0.001, 0.0001]\n",
    "size_hyperparams = tuple([len(l) for l in hps.values()])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "7aa84e19-9b55-430c-b768-deba3c76252f",
   "metadata": {},
   "outputs": [],
   "source": [
    "api = wandb.Api()\n",
    "entity = \"mathieu-bazinet\"\n",
    "project = \"sweep\" + dataset\n",
    "runs = api.runs(entity + \"/\" + project)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ecedd71a-0ac1-47cd-86bb-a72647080f9b",
   "metadata": {},
   "outputs": [],
   "source": [
    "new_runs = []\n",
    "for run in runs:\n",
    "    if run.createdAt[:-4] > '2024-09-24T12:42':\n",
    "        new_runs.append(run)\n",
    "runs = new_runs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "55d18823-38b3-499f-98db-f2be574bed64",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'seed': [1, 2, 3, 4, 42],\n",
       " 'prior_size': [0.1, 0.2, 0.5],\n",
       " 'pretraining_epochs': [50, 100],\n",
       " 'pretraining_lr': [0.01, 0.001, 0.0001],\n",
       " 'dropout_probability': [0.1, 0.2],\n",
       " 'training_lr': [0.001, 0.005, 0.0001]}"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hps\n",
    "# results_matrix.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "00d981db-a8f5-428f-9abe-a0f2c7fe9939",
   "metadata": {},
   "outputs": [],
   "source": [
    "results_matrix = np.ones(((len(values_to_fetch),) + size_hyperparams))\n",
    "\n",
    "for run_idx in range(len(runs)):\n",
    "    run = runs[run_idx]\n",
    "    if run.config['training_lr'] != 0.005 or not ignore_lr:\n",
    "        for val_to_fetch_idx in range(len(values_to_fetch)):\n",
    "            matrix_idx = tuple([val_to_fetch_idx] + [hps[key].index(run.config[key]) for key in hps.keys()])\n",
    "            val_to_fetch = values_to_fetch[val_to_fetch_idx]\n",
    "            results_matrix[matrix_idx] = run.summary[val_to_fetch]\n",
    "from itertools import product\n",
    "\n",
    "hp_list = list(hps.values())[1:]\n",
    "params_product = list(product(*hp_list))\n",
    "name_list = []\n",
    "idx_list = []\n",
    "for params in params_product:\n",
    "    name = \"\"\n",
    "    for p in params:\n",
    "        name += str(p) + \" \"\n",
    "    name_list.append(name[:-1])\n",
    "    idx = ()\n",
    "    for p_idx in range(len(params)):\n",
    "        p_key = list(hps.keys())[1:][p_idx]\n",
    "        idx += (hps[p_key].index(params[p_idx]),)\n",
    "    idx_list.append(tuple(idx))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "fb54d82b-2b9f-4bd2-b747-481818a5cba5",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>kl_bound</th>\n",
       "      <th>binomial_approximation_shah</th>\n",
       "      <th>p2l_bound</th>\n",
       "      <th>compression_set_size</th>\n",
       "      <th>validation_error</th>\n",
       "      <th>test_error</th>\n",
       "      <th>CE_kl_bound</th>\n",
       "      <th>CE_min_val_catoni_bound</th>\n",
       "      <th>complement_loss</th>\n",
       "      <th>validation_loss</th>\n",
       "      <th>test_loss</th>\n",
       "      <th>_runtime</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.01 0.1 0.001</th>\n",
       "      <td>0.352127</td>\n",
       "      <td>0.352041</td>\n",
       "      <td>0.127062</td>\n",
       "      <td>5939.2</td>\n",
       "      <td>0.193691</td>\n",
       "      <td>0.193762</td>\n",
       "      <td>3.764233</td>\n",
       "      <td>3.857538</td>\n",
       "      <td>0.461629</td>\n",
       "      <td>0.520778</td>\n",
       "      <td>0.517133</td>\n",
       "      <td>2037.201954</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.01 0.1 0.005</th>\n",
       "      <td>0.458112</td>\n",
       "      <td>0.458035</td>\n",
       "      <td>0.175823</td>\n",
       "      <td>8230.4</td>\n",
       "      <td>0.193504</td>\n",
       "      <td>0.192447</td>\n",
       "      <td>4.977358</td>\n",
       "      <td>5.069166</td>\n",
       "      <td>0.460625</td>\n",
       "      <td>0.571432</td>\n",
       "      <td>0.557218</td>\n",
       "      <td>2552.396916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.01 0.1 0.0001</th>\n",
       "      <td>0.329773</td>\n",
       "      <td>0.329685</td>\n",
       "      <td>0.118833</td>\n",
       "      <td>5555.2</td>\n",
       "      <td>0.196363</td>\n",
       "      <td>0.194695</td>\n",
       "      <td>3.573397</td>\n",
       "      <td>3.673132</td>\n",
       "      <td>0.475925</td>\n",
       "      <td>0.517555</td>\n",
       "      <td>0.512739</td>\n",
       "      <td>2350.314251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.01 0.2 0.001</th>\n",
       "      <td>0.353470</td>\n",
       "      <td>0.353384</td>\n",
       "      <td>0.127741</td>\n",
       "      <td>5971.2</td>\n",
       "      <td>0.193701</td>\n",
       "      <td>0.193100</td>\n",
       "      <td>3.779415</td>\n",
       "      <td>3.872324</td>\n",
       "      <td>0.461600</td>\n",
       "      <td>0.522958</td>\n",
       "      <td>0.515490</td>\n",
       "      <td>2030.378088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.01 0.2 0.005</th>\n",
       "      <td>0.458112</td>\n",
       "      <td>0.458035</td>\n",
       "      <td>0.175823</td>\n",
       "      <td>8230.4</td>\n",
       "      <td>0.193504</td>\n",
       "      <td>0.192447</td>\n",
       "      <td>4.977358</td>\n",
       "      <td>5.069166</td>\n",
       "      <td>0.460625</td>\n",
       "      <td>0.571432</td>\n",
       "      <td>0.557218</td>\n",
       "      <td>2560.730178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.01 0.2 0.0001</th>\n",
       "      <td>0.329773</td>\n",
       "      <td>0.329685</td>\n",
       "      <td>0.118833</td>\n",
       "      <td>5555.2</td>\n",
       "      <td>0.196363</td>\n",
       "      <td>0.194695</td>\n",
       "      <td>3.573397</td>\n",
       "      <td>3.673132</td>\n",
       "      <td>0.475925</td>\n",
       "      <td>0.517555</td>\n",
       "      <td>0.512739</td>\n",
       "      <td>2341.157793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.001 0.1 0.001</th>\n",
       "      <td>0.214693</td>\n",
       "      <td>0.214588</td>\n",
       "      <td>0.062753</td>\n",
       "      <td>2880.0</td>\n",
       "      <td>0.011230</td>\n",
       "      <td>0.011348</td>\n",
       "      <td>2.472033</td>\n",
       "      <td>2.471617</td>\n",
       "      <td>0.000030</td>\n",
       "      <td>0.081253</td>\n",
       "      <td>0.075731</td>\n",
       "      <td>797.020437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.001 0.1 0.005</th>\n",
       "      <td>0.652792</td>\n",
       "      <td>0.653627</td>\n",
       "      <td>0.364699</td>\n",
       "      <td>17203.2</td>\n",
       "      <td>0.091673</td>\n",
       "      <td>0.088168</td>\n",
       "      <td>7.531131</td>\n",
       "      <td>7.565455</td>\n",
       "      <td>0.145721</td>\n",
       "      <td>1.058038</td>\n",
       "      <td>0.994486</td>\n",
       "      <td>6777.287383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.001 0.1 0.0001</th>\n",
       "      <td>0.103780</td>\n",
       "      <td>0.103664</td>\n",
       "      <td>0.023853</td>\n",
       "      <td>1068.8</td>\n",
       "      <td>0.012107</td>\n",
       "      <td>0.011648</td>\n",
       "      <td>1.209370</td>\n",
       "      <td>1.208847</td>\n",
       "      <td>0.002201</td>\n",
       "      <td>0.050163</td>\n",
       "      <td>0.042957</td>\n",
       "      <td>352.330533</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.001 0.2 0.001</th>\n",
       "      <td>0.213173</td>\n",
       "      <td>0.213068</td>\n",
       "      <td>0.062336</td>\n",
       "      <td>2860.8</td>\n",
       "      <td>0.010724</td>\n",
       "      <td>0.010422</td>\n",
       "      <td>2.454661</td>\n",
       "      <td>2.454367</td>\n",
       "      <td>0.000044</td>\n",
       "      <td>0.080642</td>\n",
       "      <td>0.075764</td>\n",
       "      <td>753.322974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.001 0.2 0.005</th>\n",
       "      <td>0.652792</td>\n",
       "      <td>0.653627</td>\n",
       "      <td>0.364699</td>\n",
       "      <td>17203.2</td>\n",
       "      <td>0.091673</td>\n",
       "      <td>0.088168</td>\n",
       "      <td>7.531131</td>\n",
       "      <td>7.565455</td>\n",
       "      <td>0.145721</td>\n",
       "      <td>1.058038</td>\n",
       "      <td>0.994486</td>\n",
       "      <td>6779.161948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.001 0.2 0.0001</th>\n",
       "      <td>0.103780</td>\n",
       "      <td>0.103664</td>\n",
       "      <td>0.023853</td>\n",
       "      <td>1068.8</td>\n",
       "      <td>0.012107</td>\n",
       "      <td>0.011648</td>\n",
       "      <td>1.209370</td>\n",
       "      <td>1.208847</td>\n",
       "      <td>0.002201</td>\n",
       "      <td>0.050163</td>\n",
       "      <td>0.042957</td>\n",
       "      <td>349.603113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.0001 0.1 0.001</th>\n",
       "      <td>0.247275</td>\n",
       "      <td>0.247172</td>\n",
       "      <td>0.076673</td>\n",
       "      <td>3532.8</td>\n",
       "      <td>0.011750</td>\n",
       "      <td>0.010770</td>\n",
       "      <td>2.847017</td>\n",
       "      <td>2.846515</td>\n",
       "      <td>0.000016</td>\n",
       "      <td>0.088303</td>\n",
       "      <td>0.079405</td>\n",
       "      <td>994.422225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.0001 0.1 0.005</th>\n",
       "      <td>0.777718</td>\n",
       "      <td>0.777676</td>\n",
       "      <td>0.382558</td>\n",
       "      <td>18054.4</td>\n",
       "      <td>0.392336</td>\n",
       "      <td>0.388250</td>\n",
       "      <td>8.953246</td>\n",
       "      <td>8.953494</td>\n",
       "      <td>3.777558</td>\n",
       "      <td>4.501556</td>\n",
       "      <td>4.439614</td>\n",
       "      <td>6550.379627</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.0001 0.1 0.0001</th>\n",
       "      <td>0.192811</td>\n",
       "      <td>0.192704</td>\n",
       "      <td>0.054000</td>\n",
       "      <td>2470.4</td>\n",
       "      <td>0.013412</td>\n",
       "      <td>0.012773</td>\n",
       "      <td>2.221881</td>\n",
       "      <td>2.223369</td>\n",
       "      <td>0.000250</td>\n",
       "      <td>0.066606</td>\n",
       "      <td>0.058348</td>\n",
       "      <td>689.702015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.0001 0.2 0.001</th>\n",
       "      <td>0.249578</td>\n",
       "      <td>0.249476</td>\n",
       "      <td>0.077762</td>\n",
       "      <td>3584.0</td>\n",
       "      <td>0.011456</td>\n",
       "      <td>0.011059</td>\n",
       "      <td>2.873543</td>\n",
       "      <td>2.873055</td>\n",
       "      <td>0.000018</td>\n",
       "      <td>0.092505</td>\n",
       "      <td>0.085001</td>\n",
       "      <td>978.228412</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.0001 0.2 0.005</th>\n",
       "      <td>0.777718</td>\n",
       "      <td>0.777676</td>\n",
       "      <td>0.382558</td>\n",
       "      <td>18054.4</td>\n",
       "      <td>0.392336</td>\n",
       "      <td>0.388250</td>\n",
       "      <td>8.953246</td>\n",
       "      <td>8.953494</td>\n",
       "      <td>3.777558</td>\n",
       "      <td>4.501556</td>\n",
       "      <td>4.439614</td>\n",
       "      <td>6687.021900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.0001 0.2 0.0001</th>\n",
       "      <td>0.192811</td>\n",
       "      <td>0.192704</td>\n",
       "      <td>0.054000</td>\n",
       "      <td>2470.4</td>\n",
       "      <td>0.013412</td>\n",
       "      <td>0.012773</td>\n",
       "      <td>2.221881</td>\n",
       "      <td>2.223369</td>\n",
       "      <td>0.000250</td>\n",
       "      <td>0.066606</td>\n",
       "      <td>0.058348</td>\n",
       "      <td>689.643832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.01 0.1 0.001</th>\n",
       "      <td>0.347797</td>\n",
       "      <td>0.347710</td>\n",
       "      <td>0.125418</td>\n",
       "      <td>5862.4</td>\n",
       "      <td>0.195992</td>\n",
       "      <td>0.195190</td>\n",
       "      <td>3.719693</td>\n",
       "      <td>3.812003</td>\n",
       "      <td>0.462505</td>\n",
       "      <td>0.523461</td>\n",
       "      <td>0.518910</td>\n",
       "      <td>2071.210029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.01 0.1 0.005</th>\n",
       "      <td>0.460731</td>\n",
       "      <td>0.460655</td>\n",
       "      <td>0.177181</td>\n",
       "      <td>8294.4</td>\n",
       "      <td>0.194235</td>\n",
       "      <td>0.193971</td>\n",
       "      <td>5.007673</td>\n",
       "      <td>5.099593</td>\n",
       "      <td>0.460646</td>\n",
       "      <td>0.560861</td>\n",
       "      <td>0.554136</td>\n",
       "      <td>2645.547310</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.01 0.1 0.0001</th>\n",
       "      <td>0.228247</td>\n",
       "      <td>0.228142</td>\n",
       "      <td>0.074192</td>\n",
       "      <td>3424.0</td>\n",
       "      <td>0.024104</td>\n",
       "      <td>0.022260</td>\n",
       "      <td>2.720026</td>\n",
       "      <td>2.730640</td>\n",
       "      <td>0.019953</td>\n",
       "      <td>0.086938</td>\n",
       "      <td>0.078881</td>\n",
       "      <td>1662.687191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.01 0.2 0.001</th>\n",
       "      <td>0.350628</td>\n",
       "      <td>0.350541</td>\n",
       "      <td>0.126514</td>\n",
       "      <td>5913.6</td>\n",
       "      <td>0.197150</td>\n",
       "      <td>0.195500</td>\n",
       "      <td>3.752947</td>\n",
       "      <td>3.845634</td>\n",
       "      <td>0.462691</td>\n",
       "      <td>0.528850</td>\n",
       "      <td>0.521348</td>\n",
       "      <td>2013.965046</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.01 0.2 0.005</th>\n",
       "      <td>0.460731</td>\n",
       "      <td>0.460655</td>\n",
       "      <td>0.177181</td>\n",
       "      <td>8294.4</td>\n",
       "      <td>0.194235</td>\n",
       "      <td>0.193971</td>\n",
       "      <td>5.007673</td>\n",
       "      <td>5.099593</td>\n",
       "      <td>0.460646</td>\n",
       "      <td>0.560861</td>\n",
       "      <td>0.554136</td>\n",
       "      <td>2640.737753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.01 0.2 0.0001</th>\n",
       "      <td>0.228247</td>\n",
       "      <td>0.228142</td>\n",
       "      <td>0.074192</td>\n",
       "      <td>3424.0</td>\n",
       "      <td>0.024104</td>\n",
       "      <td>0.022260</td>\n",
       "      <td>2.720026</td>\n",
       "      <td>2.730640</td>\n",
       "      <td>0.019953</td>\n",
       "      <td>0.086938</td>\n",
       "      <td>0.078881</td>\n",
       "      <td>1667.070111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.001 0.1 0.001</th>\n",
       "      <td>0.199226</td>\n",
       "      <td>0.199120</td>\n",
       "      <td>0.056597</td>\n",
       "      <td>2592.0</td>\n",
       "      <td>0.011022</td>\n",
       "      <td>0.010262</td>\n",
       "      <td>2.294107</td>\n",
       "      <td>2.293820</td>\n",
       "      <td>0.000045</td>\n",
       "      <td>0.079231</td>\n",
       "      <td>0.071050</td>\n",
       "      <td>767.203005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.001 0.1 0.005</th>\n",
       "      <td>0.655194</td>\n",
       "      <td>0.655597</td>\n",
       "      <td>0.375578</td>\n",
       "      <td>17721.6</td>\n",
       "      <td>0.093036</td>\n",
       "      <td>0.088932</td>\n",
       "      <td>7.534741</td>\n",
       "      <td>7.545830</td>\n",
       "      <td>0.058387</td>\n",
       "      <td>1.050746</td>\n",
       "      <td>1.000831</td>\n",
       "      <td>7036.163196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.001 0.1 0.0001</th>\n",
       "      <td>0.099972</td>\n",
       "      <td>0.099856</td>\n",
       "      <td>0.022743</td>\n",
       "      <td>1017.6</td>\n",
       "      <td>0.013769</td>\n",
       "      <td>0.011966</td>\n",
       "      <td>1.165102</td>\n",
       "      <td>1.164514</td>\n",
       "      <td>0.002122</td>\n",
       "      <td>0.049962</td>\n",
       "      <td>0.043052</td>\n",
       "      <td>351.872121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.001 0.2 0.001</th>\n",
       "      <td>0.178733</td>\n",
       "      <td>0.178625</td>\n",
       "      <td>0.048657</td>\n",
       "      <td>2220.8</td>\n",
       "      <td>0.010614</td>\n",
       "      <td>0.010162</td>\n",
       "      <td>2.058184</td>\n",
       "      <td>2.057878</td>\n",
       "      <td>0.000045</td>\n",
       "      <td>0.072113</td>\n",
       "      <td>0.065899</td>\n",
       "      <td>575.423601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.001 0.2 0.005</th>\n",
       "      <td>0.655194</td>\n",
       "      <td>0.655597</td>\n",
       "      <td>0.375578</td>\n",
       "      <td>17721.6</td>\n",
       "      <td>0.093036</td>\n",
       "      <td>0.088932</td>\n",
       "      <td>7.534741</td>\n",
       "      <td>7.545830</td>\n",
       "      <td>0.058387</td>\n",
       "      <td>1.050746</td>\n",
       "      <td>1.000831</td>\n",
       "      <td>7044.671999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.001 0.2 0.0001</th>\n",
       "      <td>0.099972</td>\n",
       "      <td>0.099856</td>\n",
       "      <td>0.022743</td>\n",
       "      <td>1017.6</td>\n",
       "      <td>0.013769</td>\n",
       "      <td>0.011966</td>\n",
       "      <td>1.165102</td>\n",
       "      <td>1.164514</td>\n",
       "      <td>0.002122</td>\n",
       "      <td>0.049962</td>\n",
       "      <td>0.043052</td>\n",
       "      <td>351.787471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.0001 0.1 0.001</th>\n",
       "      <td>0.238983</td>\n",
       "      <td>0.238880</td>\n",
       "      <td>0.072993</td>\n",
       "      <td>3360.0</td>\n",
       "      <td>0.010983</td>\n",
       "      <td>0.010733</td>\n",
       "      <td>2.751667</td>\n",
       "      <td>2.751245</td>\n",
       "      <td>0.000029</td>\n",
       "      <td>0.084314</td>\n",
       "      <td>0.080571</td>\n",
       "      <td>990.609190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.0001 0.1 0.005</th>\n",
       "      <td>0.737257</td>\n",
       "      <td>0.737861</td>\n",
       "      <td>0.361197</td>\n",
       "      <td>17036.8</td>\n",
       "      <td>0.275070</td>\n",
       "      <td>0.274120</td>\n",
       "      <td>8.477993</td>\n",
       "      <td>8.564228</td>\n",
       "      <td>2.627745</td>\n",
       "      <td>3.188697</td>\n",
       "      <td>3.153391</td>\n",
       "      <td>6445.522120</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.0001 0.1 0.0001</th>\n",
       "      <td>0.175904</td>\n",
       "      <td>0.175795</td>\n",
       "      <td>0.047694</td>\n",
       "      <td>2176.0</td>\n",
       "      <td>0.013826</td>\n",
       "      <td>0.012800</td>\n",
       "      <td>2.028423</td>\n",
       "      <td>2.030316</td>\n",
       "      <td>0.000411</td>\n",
       "      <td>0.063092</td>\n",
       "      <td>0.057025</td>\n",
       "      <td>584.755171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.0001 0.2 0.001</th>\n",
       "      <td>0.253584</td>\n",
       "      <td>0.253482</td>\n",
       "      <td>0.079667</td>\n",
       "      <td>3673.6</td>\n",
       "      <td>0.012260</td>\n",
       "      <td>0.011468</td>\n",
       "      <td>2.919797</td>\n",
       "      <td>2.919418</td>\n",
       "      <td>0.000033</td>\n",
       "      <td>0.091494</td>\n",
       "      <td>0.083414</td>\n",
       "      <td>996.880270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.0001 0.2 0.005</th>\n",
       "      <td>0.737257</td>\n",
       "      <td>0.737861</td>\n",
       "      <td>0.361197</td>\n",
       "      <td>17036.8</td>\n",
       "      <td>0.275070</td>\n",
       "      <td>0.274120</td>\n",
       "      <td>8.477993</td>\n",
       "      <td>8.564228</td>\n",
       "      <td>2.627745</td>\n",
       "      <td>3.188697</td>\n",
       "      <td>3.153391</td>\n",
       "      <td>6468.377538</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.0001 0.2 0.0001</th>\n",
       "      <td>0.175904</td>\n",
       "      <td>0.175795</td>\n",
       "      <td>0.047694</td>\n",
       "      <td>2176.0</td>\n",
       "      <td>0.013826</td>\n",
       "      <td>0.012800</td>\n",
       "      <td>2.028423</td>\n",
       "      <td>2.030316</td>\n",
       "      <td>0.000411</td>\n",
       "      <td>0.063092</td>\n",
       "      <td>0.057025</td>\n",
       "      <td>586.708536</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.01 0.1 0.001</th>\n",
       "      <td>0.175166</td>\n",
       "      <td>0.175042</td>\n",
       "      <td>0.047595</td>\n",
       "      <td>1894.4</td>\n",
       "      <td>0.022016</td>\n",
       "      <td>0.019077</td>\n",
       "      <td>2.040576</td>\n",
       "      <td>2.040253</td>\n",
       "      <td>0.003934</td>\n",
       "      <td>0.086088</td>\n",
       "      <td>0.072829</td>\n",
       "      <td>635.130060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.01 0.1 0.005</th>\n",
       "      <td>0.315312</td>\n",
       "      <td>0.315204</td>\n",
       "      <td>0.110153</td>\n",
       "      <td>4460.8</td>\n",
       "      <td>0.019357</td>\n",
       "      <td>0.018195</td>\n",
       "      <td>3.630956</td>\n",
       "      <td>3.630956</td>\n",
       "      <td>0.000102</td>\n",
       "      <td>0.130083</td>\n",
       "      <td>0.118285</td>\n",
       "      <td>1207.450799</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.01 0.1 0.0001</th>\n",
       "      <td>0.167029</td>\n",
       "      <td>0.166905</td>\n",
       "      <td>0.044611</td>\n",
       "      <td>1772.8</td>\n",
       "      <td>0.022925</td>\n",
       "      <td>0.019880</td>\n",
       "      <td>2.045735</td>\n",
       "      <td>2.063732</td>\n",
       "      <td>0.027012</td>\n",
       "      <td>0.084964</td>\n",
       "      <td>0.077047</td>\n",
       "      <td>890.482353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.01 0.2 0.001</th>\n",
       "      <td>0.180424</td>\n",
       "      <td>0.180301</td>\n",
       "      <td>0.049632</td>\n",
       "      <td>1977.6</td>\n",
       "      <td>0.021006</td>\n",
       "      <td>0.017978</td>\n",
       "      <td>2.096352</td>\n",
       "      <td>2.096439</td>\n",
       "      <td>0.003073</td>\n",
       "      <td>0.081954</td>\n",
       "      <td>0.071872</td>\n",
       "      <td>581.673505</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.01 0.2 0.005</th>\n",
       "      <td>0.315312</td>\n",
       "      <td>0.315204</td>\n",
       "      <td>0.110153</td>\n",
       "      <td>4460.8</td>\n",
       "      <td>0.019357</td>\n",
       "      <td>0.018195</td>\n",
       "      <td>3.630956</td>\n",
       "      <td>3.630956</td>\n",
       "      <td>0.000102</td>\n",
       "      <td>0.130083</td>\n",
       "      <td>0.118285</td>\n",
       "      <td>1205.217252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.01 0.2 0.0001</th>\n",
       "      <td>0.167029</td>\n",
       "      <td>0.166905</td>\n",
       "      <td>0.044611</td>\n",
       "      <td>1772.8</td>\n",
       "      <td>0.022925</td>\n",
       "      <td>0.019880</td>\n",
       "      <td>2.045735</td>\n",
       "      <td>2.063732</td>\n",
       "      <td>0.027012</td>\n",
       "      <td>0.084964</td>\n",
       "      <td>0.077047</td>\n",
       "      <td>892.006992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.001 0.1 0.001</th>\n",
       "      <td>0.176772</td>\n",
       "      <td>0.176649</td>\n",
       "      <td>0.048070</td>\n",
       "      <td>1913.6</td>\n",
       "      <td>0.011210</td>\n",
       "      <td>0.010218</td>\n",
       "      <td>2.035603</td>\n",
       "      <td>2.035253</td>\n",
       "      <td>0.000046</td>\n",
       "      <td>0.073488</td>\n",
       "      <td>0.065919</td>\n",
       "      <td>544.123728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.001 0.1 0.005</th>\n",
       "      <td>0.683026</td>\n",
       "      <td>0.683174</td>\n",
       "      <td>0.413789</td>\n",
       "      <td>17088.0</td>\n",
       "      <td>0.073311</td>\n",
       "      <td>0.073065</td>\n",
       "      <td>7.864293</td>\n",
       "      <td>7.867262</td>\n",
       "      <td>0.040295</td>\n",
       "      <td>0.840060</td>\n",
       "      <td>0.831924</td>\n",
       "      <td>6443.108022</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.001 0.1 0.0001</th>\n",
       "      <td>0.090929</td>\n",
       "      <td>0.090796</td>\n",
       "      <td>0.020264</td>\n",
       "      <td>787.2</td>\n",
       "      <td>0.011994</td>\n",
       "      <td>0.010924</td>\n",
       "      <td>1.063091</td>\n",
       "      <td>1.062450</td>\n",
       "      <td>0.002520</td>\n",
       "      <td>0.044112</td>\n",
       "      <td>0.036578</td>\n",
       "      <td>293.753623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.001 0.2 0.001</th>\n",
       "      <td>0.190151</td>\n",
       "      <td>0.190029</td>\n",
       "      <td>0.053092</td>\n",
       "      <td>2118.4</td>\n",
       "      <td>0.010828</td>\n",
       "      <td>0.010323</td>\n",
       "      <td>2.189390</td>\n",
       "      <td>2.188811</td>\n",
       "      <td>0.000019</td>\n",
       "      <td>0.076633</td>\n",
       "      <td>0.069752</td>\n",
       "      <td>525.601341</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.001 0.2 0.005</th>\n",
       "      <td>0.683026</td>\n",
       "      <td>0.683174</td>\n",
       "      <td>0.413789</td>\n",
       "      <td>17088.0</td>\n",
       "      <td>0.073311</td>\n",
       "      <td>0.073065</td>\n",
       "      <td>7.864293</td>\n",
       "      <td>7.867262</td>\n",
       "      <td>0.040295</td>\n",
       "      <td>0.840060</td>\n",
       "      <td>0.831924</td>\n",
       "      <td>6515.031295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.001 0.2 0.0001</th>\n",
       "      <td>0.090929</td>\n",
       "      <td>0.090796</td>\n",
       "      <td>0.020264</td>\n",
       "      <td>787.2</td>\n",
       "      <td>0.011994</td>\n",
       "      <td>0.010924</td>\n",
       "      <td>1.063091</td>\n",
       "      <td>1.062450</td>\n",
       "      <td>0.002520</td>\n",
       "      <td>0.044112</td>\n",
       "      <td>0.036578</td>\n",
       "      <td>311.640808</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.0001 0.1 0.001</th>\n",
       "      <td>0.248064</td>\n",
       "      <td>0.247948</td>\n",
       "      <td>0.077168</td>\n",
       "      <td>3104.0</td>\n",
       "      <td>0.011980</td>\n",
       "      <td>0.011772</td>\n",
       "      <td>2.856198</td>\n",
       "      <td>2.855705</td>\n",
       "      <td>0.000028</td>\n",
       "      <td>0.093192</td>\n",
       "      <td>0.081732</td>\n",
       "      <td>925.962744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.0001 0.1 0.005</th>\n",
       "      <td>0.611837</td>\n",
       "      <td>0.611756</td>\n",
       "      <td>0.337068</td>\n",
       "      <td>13881.6</td>\n",
       "      <td>0.051728</td>\n",
       "      <td>0.051720</td>\n",
       "      <td>7.044030</td>\n",
       "      <td>7.043582</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.602510</td>\n",
       "      <td>0.583475</td>\n",
       "      <td>5298.622682</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.0001 0.1 0.0001</th>\n",
       "      <td>0.174903</td>\n",
       "      <td>0.174779</td>\n",
       "      <td>0.047440</td>\n",
       "      <td>1888.0</td>\n",
       "      <td>0.012957</td>\n",
       "      <td>0.012351</td>\n",
       "      <td>2.016593</td>\n",
       "      <td>2.018199</td>\n",
       "      <td>0.000370</td>\n",
       "      <td>0.059193</td>\n",
       "      <td>0.053868</td>\n",
       "      <td>531.955199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.0001 0.2 0.001</th>\n",
       "      <td>0.239228</td>\n",
       "      <td>0.239111</td>\n",
       "      <td>0.073267</td>\n",
       "      <td>2944.0</td>\n",
       "      <td>0.011885</td>\n",
       "      <td>0.012280</td>\n",
       "      <td>2.754513</td>\n",
       "      <td>2.754057</td>\n",
       "      <td>0.000032</td>\n",
       "      <td>0.085326</td>\n",
       "      <td>0.085965</td>\n",
       "      <td>807.869276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.0001 0.2 0.005</th>\n",
       "      <td>0.628229</td>\n",
       "      <td>0.628150</td>\n",
       "      <td>0.354551</td>\n",
       "      <td>14611.2</td>\n",
       "      <td>0.054473</td>\n",
       "      <td>0.053175</td>\n",
       "      <td>7.232753</td>\n",
       "      <td>7.232315</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.635898</td>\n",
       "      <td>0.599823</td>\n",
       "      <td>5434.996939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.0001 0.2 0.0001</th>\n",
       "      <td>0.170652</td>\n",
       "      <td>0.170528</td>\n",
       "      <td>0.045870</td>\n",
       "      <td>1824.0</td>\n",
       "      <td>0.012998</td>\n",
       "      <td>0.012320</td>\n",
       "      <td>1.968043</td>\n",
       "      <td>1.969959</td>\n",
       "      <td>0.000422</td>\n",
       "      <td>0.058964</td>\n",
       "      <td>0.051933</td>\n",
       "      <td>473.910653</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.01 0.1 0.001</th>\n",
       "      <td>0.174960</td>\n",
       "      <td>0.174836</td>\n",
       "      <td>0.047441</td>\n",
       "      <td>1888.0</td>\n",
       "      <td>0.021376</td>\n",
       "      <td>0.019821</td>\n",
       "      <td>2.046025</td>\n",
       "      <td>2.045538</td>\n",
       "      <td>0.005467</td>\n",
       "      <td>0.083465</td>\n",
       "      <td>0.074830</td>\n",
       "      <td>736.974611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.01 0.1 0.005</th>\n",
       "      <td>0.287308</td>\n",
       "      <td>0.287197</td>\n",
       "      <td>0.095853</td>\n",
       "      <td>3872.0</td>\n",
       "      <td>0.019792</td>\n",
       "      <td>0.019237</td>\n",
       "      <td>3.309117</td>\n",
       "      <td>3.309760</td>\n",
       "      <td>0.000178</td>\n",
       "      <td>0.119976</td>\n",
       "      <td>0.118816</td>\n",
       "      <td>1089.092600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.01 0.1 0.0001</th>\n",
       "      <td>0.184633</td>\n",
       "      <td>0.184511</td>\n",
       "      <td>0.051051</td>\n",
       "      <td>2035.2</td>\n",
       "      <td>0.023892</td>\n",
       "      <td>0.021719</td>\n",
       "      <td>2.246248</td>\n",
       "      <td>2.261798</td>\n",
       "      <td>0.026591</td>\n",
       "      <td>0.092755</td>\n",
       "      <td>0.084042</td>\n",
       "      <td>1123.453475</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.01 0.2 0.001</th>\n",
       "      <td>0.177131</td>\n",
       "      <td>0.177008</td>\n",
       "      <td>0.048227</td>\n",
       "      <td>1920.0</td>\n",
       "      <td>0.021007</td>\n",
       "      <td>0.020107</td>\n",
       "      <td>2.067079</td>\n",
       "      <td>2.066850</td>\n",
       "      <td>0.004741</td>\n",
       "      <td>0.085834</td>\n",
       "      <td>0.076625</td>\n",
       "      <td>613.055591</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.01 0.2 0.005</th>\n",
       "      <td>0.286005</td>\n",
       "      <td>0.285893</td>\n",
       "      <td>0.095230</td>\n",
       "      <td>3846.4</td>\n",
       "      <td>0.019706</td>\n",
       "      <td>0.019408</td>\n",
       "      <td>3.294218</td>\n",
       "      <td>3.294958</td>\n",
       "      <td>0.000191</td>\n",
       "      <td>0.118448</td>\n",
       "      <td>0.118631</td>\n",
       "      <td>1086.484207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.01 0.2 0.0001</th>\n",
       "      <td>0.185052</td>\n",
       "      <td>0.184929</td>\n",
       "      <td>0.051208</td>\n",
       "      <td>2041.6</td>\n",
       "      <td>0.023772</td>\n",
       "      <td>0.021654</td>\n",
       "      <td>2.250532</td>\n",
       "      <td>2.265845</td>\n",
       "      <td>0.026450</td>\n",
       "      <td>0.092671</td>\n",
       "      <td>0.083746</td>\n",
       "      <td>1050.114938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.001 0.1 0.001</th>\n",
       "      <td>0.153840</td>\n",
       "      <td>0.153715</td>\n",
       "      <td>0.039894</td>\n",
       "      <td>1580.8</td>\n",
       "      <td>0.011115</td>\n",
       "      <td>0.010137</td>\n",
       "      <td>1.771596</td>\n",
       "      <td>1.771216</td>\n",
       "      <td>0.000045</td>\n",
       "      <td>0.069024</td>\n",
       "      <td>0.062030</td>\n",
       "      <td>574.430000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.001 0.1 0.005</th>\n",
       "      <td>0.678513</td>\n",
       "      <td>0.679029</td>\n",
       "      <td>0.394654</td>\n",
       "      <td>16288.0</td>\n",
       "      <td>0.094489</td>\n",
       "      <td>0.090146</td>\n",
       "      <td>7.859176</td>\n",
       "      <td>7.909655</td>\n",
       "      <td>0.239364</td>\n",
       "      <td>1.093728</td>\n",
       "      <td>1.038696</td>\n",
       "      <td>6192.622718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.001 0.1 0.0001</th>\n",
       "      <td>0.088063</td>\n",
       "      <td>0.087930</td>\n",
       "      <td>0.019466</td>\n",
       "      <td>755.2</td>\n",
       "      <td>0.012447</td>\n",
       "      <td>0.011536</td>\n",
       "      <td>1.032238</td>\n",
       "      <td>1.031551</td>\n",
       "      <td>0.002900</td>\n",
       "      <td>0.047118</td>\n",
       "      <td>0.038755</td>\n",
       "      <td>310.489978</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.001 0.2 0.001</th>\n",
       "      <td>0.145956</td>\n",
       "      <td>0.145830</td>\n",
       "      <td>0.037215</td>\n",
       "      <td>1472.0</td>\n",
       "      <td>0.011174</td>\n",
       "      <td>0.010471</td>\n",
       "      <td>1.681211</td>\n",
       "      <td>1.681212</td>\n",
       "      <td>0.000088</td>\n",
       "      <td>0.071231</td>\n",
       "      <td>0.060359</td>\n",
       "      <td>348.458644</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.001 0.2 0.005</th>\n",
       "      <td>0.668580</td>\n",
       "      <td>0.669094</td>\n",
       "      <td>0.382864</td>\n",
       "      <td>15795.2</td>\n",
       "      <td>0.086826</td>\n",
       "      <td>0.083468</td>\n",
       "      <td>7.744813</td>\n",
       "      <td>7.795287</td>\n",
       "      <td>0.239364</td>\n",
       "      <td>1.012370</td>\n",
       "      <td>0.962250</td>\n",
       "      <td>5847.276190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.001 0.2 0.0001</th>\n",
       "      <td>0.087475</td>\n",
       "      <td>0.087343</td>\n",
       "      <td>0.019306</td>\n",
       "      <td>748.8</td>\n",
       "      <td>0.012671</td>\n",
       "      <td>0.011529</td>\n",
       "      <td>1.026074</td>\n",
       "      <td>1.025390</td>\n",
       "      <td>0.003007</td>\n",
       "      <td>0.047150</td>\n",
       "      <td>0.038754</td>\n",
       "      <td>309.193037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.0001 0.1 0.001</th>\n",
       "      <td>0.246333</td>\n",
       "      <td>0.246217</td>\n",
       "      <td>0.076388</td>\n",
       "      <td>3072.0</td>\n",
       "      <td>0.012379</td>\n",
       "      <td>0.012303</td>\n",
       "      <td>2.836318</td>\n",
       "      <td>2.835864</td>\n",
       "      <td>0.000033</td>\n",
       "      <td>0.095430</td>\n",
       "      <td>0.090301</td>\n",
       "      <td>939.690883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.0001 0.1 0.005</th>\n",
       "      <td>0.679993</td>\n",
       "      <td>0.680571</td>\n",
       "      <td>0.386690</td>\n",
       "      <td>15955.2</td>\n",
       "      <td>0.096004</td>\n",
       "      <td>0.089157</td>\n",
       "      <td>7.827576</td>\n",
       "      <td>7.887833</td>\n",
       "      <td>0.311268</td>\n",
       "      <td>1.129418</td>\n",
       "      <td>1.039157</td>\n",
       "      <td>6024.818873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.0001 0.1 0.0001</th>\n",
       "      <td>0.151629</td>\n",
       "      <td>0.151503</td>\n",
       "      <td>0.039108</td>\n",
       "      <td>1548.8</td>\n",
       "      <td>0.011918</td>\n",
       "      <td>0.012037</td>\n",
       "      <td>1.750118</td>\n",
       "      <td>1.751659</td>\n",
       "      <td>0.000571</td>\n",
       "      <td>0.058036</td>\n",
       "      <td>0.051576</td>\n",
       "      <td>385.924237</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.0001 0.2 0.001</th>\n",
       "      <td>0.235507</td>\n",
       "      <td>0.235390</td>\n",
       "      <td>0.071704</td>\n",
       "      <td>2880.0</td>\n",
       "      <td>0.011880</td>\n",
       "      <td>0.012021</td>\n",
       "      <td>2.711724</td>\n",
       "      <td>2.711295</td>\n",
       "      <td>0.000037</td>\n",
       "      <td>0.085972</td>\n",
       "      <td>0.083469</td>\n",
       "      <td>745.608170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.0001 0.2 0.005</th>\n",
       "      <td>0.673286</td>\n",
       "      <td>0.673864</td>\n",
       "      <td>0.379332</td>\n",
       "      <td>15648.0</td>\n",
       "      <td>0.089753</td>\n",
       "      <td>0.084308</td>\n",
       "      <td>7.750360</td>\n",
       "      <td>7.810613</td>\n",
       "      <td>0.311268</td>\n",
       "      <td>1.053817</td>\n",
       "      <td>0.985046</td>\n",
       "      <td>5855.812801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.0001 0.2 0.0001</th>\n",
       "      <td>0.150685</td>\n",
       "      <td>0.150559</td>\n",
       "      <td>0.038792</td>\n",
       "      <td>1536.0</td>\n",
       "      <td>0.011848</td>\n",
       "      <td>0.011837</td>\n",
       "      <td>1.739139</td>\n",
       "      <td>1.740683</td>\n",
       "      <td>0.000555</td>\n",
       "      <td>0.058257</td>\n",
       "      <td>0.051452</td>\n",
       "      <td>381.302760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.01 0.1 0.001</th>\n",
       "      <td>0.155428</td>\n",
       "      <td>0.155219</td>\n",
       "      <td>0.041008</td>\n",
       "      <td>915.2</td>\n",
       "      <td>0.020031</td>\n",
       "      <td>0.018739</td>\n",
       "      <td>1.834185</td>\n",
       "      <td>1.833995</td>\n",
       "      <td>0.008073</td>\n",
       "      <td>0.072760</td>\n",
       "      <td>0.065821</td>\n",
       "      <td>435.852056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.01 0.1 0.005</th>\n",
       "      <td>0.241293</td>\n",
       "      <td>0.241099</td>\n",
       "      <td>0.075251</td>\n",
       "      <td>1708.8</td>\n",
       "      <td>0.020038</td>\n",
       "      <td>0.017584</td>\n",
       "      <td>2.781068</td>\n",
       "      <td>2.782873</td>\n",
       "      <td>0.000405</td>\n",
       "      <td>0.095380</td>\n",
       "      <td>0.086412</td>\n",
       "      <td>364.866074</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.01 0.1 0.0001</th>\n",
       "      <td>0.164055</td>\n",
       "      <td>0.163848</td>\n",
       "      <td>0.044064</td>\n",
       "      <td>985.6</td>\n",
       "      <td>0.022602</td>\n",
       "      <td>0.019793</td>\n",
       "      <td>1.992703</td>\n",
       "      <td>2.005482</td>\n",
       "      <td>0.022332</td>\n",
       "      <td>0.084117</td>\n",
       "      <td>0.075092</td>\n",
       "      <td>478.792652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.01 0.2 0.001</th>\n",
       "      <td>0.156245</td>\n",
       "      <td>0.156036</td>\n",
       "      <td>0.041287</td>\n",
       "      <td>921.6</td>\n",
       "      <td>0.020152</td>\n",
       "      <td>0.017806</td>\n",
       "      <td>1.839748</td>\n",
       "      <td>1.839333</td>\n",
       "      <td>0.007298</td>\n",
       "      <td>0.072484</td>\n",
       "      <td>0.063873</td>\n",
       "      <td>248.319818</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.01 0.2 0.005</th>\n",
       "      <td>0.251599</td>\n",
       "      <td>0.251407</td>\n",
       "      <td>0.080716</td>\n",
       "      <td>1836.8</td>\n",
       "      <td>0.019514</td>\n",
       "      <td>0.017430</td>\n",
       "      <td>2.899517</td>\n",
       "      <td>2.901102</td>\n",
       "      <td>0.000378</td>\n",
       "      <td>0.096028</td>\n",
       "      <td>0.087905</td>\n",
       "      <td>391.323444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.01 0.2 0.0001</th>\n",
       "      <td>0.164055</td>\n",
       "      <td>0.163848</td>\n",
       "      <td>0.044064</td>\n",
       "      <td>985.6</td>\n",
       "      <td>0.022737</td>\n",
       "      <td>0.019401</td>\n",
       "      <td>1.991907</td>\n",
       "      <td>2.004390</td>\n",
       "      <td>0.022111</td>\n",
       "      <td>0.083979</td>\n",
       "      <td>0.074855</td>\n",
       "      <td>479.760224</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.001 0.1 0.001</th>\n",
       "      <td>0.111081</td>\n",
       "      <td>0.110864</td>\n",
       "      <td>0.026492</td>\n",
       "      <td>582.4</td>\n",
       "      <td>0.011276</td>\n",
       "      <td>0.009708</td>\n",
       "      <td>1.281422</td>\n",
       "      <td>1.281823</td>\n",
       "      <td>0.000307</td>\n",
       "      <td>0.054745</td>\n",
       "      <td>0.047779</td>\n",
       "      <td>298.228348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.001 0.1 0.005</th>\n",
       "      <td>0.692967</td>\n",
       "      <td>0.692846</td>\n",
       "      <td>0.432730</td>\n",
       "      <td>10169.6</td>\n",
       "      <td>0.067597</td>\n",
       "      <td>0.065122</td>\n",
       "      <td>7.978082</td>\n",
       "      <td>7.977453</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.777890</td>\n",
       "      <td>0.736431</td>\n",
       "      <td>2964.001945</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.001 0.1 0.0001</th>\n",
       "      <td>0.073244</td>\n",
       "      <td>0.073020</td>\n",
       "      <td>0.015762</td>\n",
       "      <td>339.2</td>\n",
       "      <td>0.010797</td>\n",
       "      <td>0.010995</td>\n",
       "      <td>0.861935</td>\n",
       "      <td>0.861014</td>\n",
       "      <td>0.003012</td>\n",
       "      <td>0.043493</td>\n",
       "      <td>0.037847</td>\n",
       "      <td>217.928213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.001 0.2 0.001</th>\n",
       "      <td>0.106329</td>\n",
       "      <td>0.106111</td>\n",
       "      <td>0.025086</td>\n",
       "      <td>550.4</td>\n",
       "      <td>0.010388</td>\n",
       "      <td>0.009556</td>\n",
       "      <td>1.227287</td>\n",
       "      <td>1.227418</td>\n",
       "      <td>0.000387</td>\n",
       "      <td>0.052262</td>\n",
       "      <td>0.046735</td>\n",
       "      <td>100.960557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.001 0.2 0.005</th>\n",
       "      <td>0.680507</td>\n",
       "      <td>0.680384</td>\n",
       "      <td>0.416134</td>\n",
       "      <td>9772.8</td>\n",
       "      <td>0.069274</td>\n",
       "      <td>0.065777</td>\n",
       "      <td>7.834628</td>\n",
       "      <td>7.833987</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.794435</td>\n",
       "      <td>0.744721</td>\n",
       "      <td>2813.286191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.001 0.2 0.0001</th>\n",
       "      <td>0.075341</td>\n",
       "      <td>0.075117</td>\n",
       "      <td>0.016330</td>\n",
       "      <td>352.0</td>\n",
       "      <td>0.010621</td>\n",
       "      <td>0.010788</td>\n",
       "      <td>0.883966</td>\n",
       "      <td>0.882902</td>\n",
       "      <td>0.002612</td>\n",
       "      <td>0.043230</td>\n",
       "      <td>0.037535</td>\n",
       "      <td>219.898248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.0001 0.1 0.001</th>\n",
       "      <td>0.270231</td>\n",
       "      <td>0.270042</td>\n",
       "      <td>0.088168</td>\n",
       "      <td>2009.6</td>\n",
       "      <td>0.014800</td>\n",
       "      <td>0.013748</td>\n",
       "      <td>3.111644</td>\n",
       "      <td>3.110966</td>\n",
       "      <td>0.000056</td>\n",
       "      <td>0.101851</td>\n",
       "      <td>0.088920</td>\n",
       "      <td>620.621142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.0001 0.1 0.005</th>\n",
       "      <td>0.674288</td>\n",
       "      <td>0.674165</td>\n",
       "      <td>0.409149</td>\n",
       "      <td>9606.4</td>\n",
       "      <td>0.064930</td>\n",
       "      <td>0.060006</td>\n",
       "      <td>7.763033</td>\n",
       "      <td>7.762386</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.732222</td>\n",
       "      <td>0.678990</td>\n",
       "      <td>2719.386001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.0001 0.1 0.0001</th>\n",
       "      <td>0.112941</td>\n",
       "      <td>0.112724</td>\n",
       "      <td>0.027054</td>\n",
       "      <td>595.2</td>\n",
       "      <td>0.013359</td>\n",
       "      <td>0.012187</td>\n",
       "      <td>1.313943</td>\n",
       "      <td>1.313113</td>\n",
       "      <td>0.002055</td>\n",
       "      <td>0.048798</td>\n",
       "      <td>0.043397</td>\n",
       "      <td>114.213661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.0001 0.2 0.001</th>\n",
       "      <td>0.269329</td>\n",
       "      <td>0.269139</td>\n",
       "      <td>0.087889</td>\n",
       "      <td>2003.2</td>\n",
       "      <td>0.013879</td>\n",
       "      <td>0.013165</td>\n",
       "      <td>3.101053</td>\n",
       "      <td>3.100209</td>\n",
       "      <td>0.000032</td>\n",
       "      <td>0.092935</td>\n",
       "      <td>0.088020</td>\n",
       "      <td>429.423048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.0001 0.2 0.005</th>\n",
       "      <td>0.679594</td>\n",
       "      <td>0.679471</td>\n",
       "      <td>0.415586</td>\n",
       "      <td>9760.0</td>\n",
       "      <td>0.065904</td>\n",
       "      <td>0.059904</td>\n",
       "      <td>7.824111</td>\n",
       "      <td>7.823469</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.738746</td>\n",
       "      <td>0.675537</td>\n",
       "      <td>2788.610445</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.0001 0.2 0.0001</th>\n",
       "      <td>0.112941</td>\n",
       "      <td>0.112724</td>\n",
       "      <td>0.027054</td>\n",
       "      <td>595.2</td>\n",
       "      <td>0.013359</td>\n",
       "      <td>0.012187</td>\n",
       "      <td>1.313942</td>\n",
       "      <td>1.313112</td>\n",
       "      <td>0.002054</td>\n",
       "      <td>0.048791</td>\n",
       "      <td>0.043397</td>\n",
       "      <td>112.210161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.01 0.1 0.001</th>\n",
       "      <td>0.155684</td>\n",
       "      <td>0.155475</td>\n",
       "      <td>0.041013</td>\n",
       "      <td>915.2</td>\n",
       "      <td>0.020507</td>\n",
       "      <td>0.017959</td>\n",
       "      <td>1.839851</td>\n",
       "      <td>1.840354</td>\n",
       "      <td>0.008723</td>\n",
       "      <td>0.077228</td>\n",
       "      <td>0.066065</td>\n",
       "      <td>626.477681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.01 0.1 0.005</th>\n",
       "      <td>0.227441</td>\n",
       "      <td>0.227245</td>\n",
       "      <td>0.068935</td>\n",
       "      <td>1561.6</td>\n",
       "      <td>0.019946</td>\n",
       "      <td>0.017181</td>\n",
       "      <td>2.623046</td>\n",
       "      <td>2.623897</td>\n",
       "      <td>0.000647</td>\n",
       "      <td>0.098908</td>\n",
       "      <td>0.084937</td>\n",
       "      <td>338.123212</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.01 0.1 0.0001</th>\n",
       "      <td>0.166577</td>\n",
       "      <td>0.166370</td>\n",
       "      <td>0.044900</td>\n",
       "      <td>1004.8</td>\n",
       "      <td>0.023480</td>\n",
       "      <td>0.021593</td>\n",
       "      <td>2.023296</td>\n",
       "      <td>2.035532</td>\n",
       "      <td>0.022469</td>\n",
       "      <td>0.092476</td>\n",
       "      <td>0.084056</td>\n",
       "      <td>448.308909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.01 0.2 0.001</th>\n",
       "      <td>0.157229</td>\n",
       "      <td>0.157020</td>\n",
       "      <td>0.041568</td>\n",
       "      <td>928.0</td>\n",
       "      <td>0.020434</td>\n",
       "      <td>0.017697</td>\n",
       "      <td>1.853642</td>\n",
       "      <td>1.853575</td>\n",
       "      <td>0.007857</td>\n",
       "      <td>0.076744</td>\n",
       "      <td>0.064709</td>\n",
       "      <td>252.080682</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.01 0.2 0.005</th>\n",
       "      <td>0.228695</td>\n",
       "      <td>0.228498</td>\n",
       "      <td>0.069485</td>\n",
       "      <td>1574.4</td>\n",
       "      <td>0.020046</td>\n",
       "      <td>0.016934</td>\n",
       "      <td>2.637200</td>\n",
       "      <td>2.637741</td>\n",
       "      <td>0.000610</td>\n",
       "      <td>0.100453</td>\n",
       "      <td>0.084770</td>\n",
       "      <td>338.387936</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.01 0.2 0.0001</th>\n",
       "      <td>0.166577</td>\n",
       "      <td>0.166370</td>\n",
       "      <td>0.044900</td>\n",
       "      <td>1004.8</td>\n",
       "      <td>0.023503</td>\n",
       "      <td>0.021379</td>\n",
       "      <td>2.023296</td>\n",
       "      <td>2.035533</td>\n",
       "      <td>0.022470</td>\n",
       "      <td>0.091741</td>\n",
       "      <td>0.083627</td>\n",
       "      <td>470.652769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.001 0.1 0.001</th>\n",
       "      <td>0.111967</td>\n",
       "      <td>0.111750</td>\n",
       "      <td>0.026772</td>\n",
       "      <td>588.8</td>\n",
       "      <td>0.009288</td>\n",
       "      <td>0.010751</td>\n",
       "      <td>1.290202</td>\n",
       "      <td>1.290037</td>\n",
       "      <td>0.000124</td>\n",
       "      <td>0.059646</td>\n",
       "      <td>0.051774</td>\n",
       "      <td>524.105930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.001 0.1 0.005</th>\n",
       "      <td>0.681184</td>\n",
       "      <td>0.681061</td>\n",
       "      <td>0.417465</td>\n",
       "      <td>9804.8</td>\n",
       "      <td>0.065212</td>\n",
       "      <td>0.058716</td>\n",
       "      <td>7.842418</td>\n",
       "      <td>7.841778</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.739514</td>\n",
       "      <td>0.660612</td>\n",
       "      <td>3002.656749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.001 0.1 0.0001</th>\n",
       "      <td>0.071070</td>\n",
       "      <td>0.070845</td>\n",
       "      <td>0.015193</td>\n",
       "      <td>326.4</td>\n",
       "      <td>0.010663</td>\n",
       "      <td>0.011426</td>\n",
       "      <td>0.827947</td>\n",
       "      <td>0.826882</td>\n",
       "      <td>0.001425</td>\n",
       "      <td>0.053495</td>\n",
       "      <td>0.046665</td>\n",
       "      <td>270.894093</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.001 0.2 0.001</th>\n",
       "      <td>0.105545</td>\n",
       "      <td>0.105327</td>\n",
       "      <td>0.024808</td>\n",
       "      <td>544.0</td>\n",
       "      <td>0.010052</td>\n",
       "      <td>0.010761</td>\n",
       "      <td>1.216881</td>\n",
       "      <td>1.217111</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.059043</td>\n",
       "      <td>0.048941</td>\n",
       "      <td>118.750645</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.001 0.2 0.005</th>\n",
       "      <td>0.677711</td>\n",
       "      <td>0.677587</td>\n",
       "      <td>0.413175</td>\n",
       "      <td>9702.4</td>\n",
       "      <td>0.064972</td>\n",
       "      <td>0.058984</td>\n",
       "      <td>7.802431</td>\n",
       "      <td>7.801788</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.741049</td>\n",
       "      <td>0.668210</td>\n",
       "      <td>2989.542676</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.001 0.2 0.0001</th>\n",
       "      <td>0.072157</td>\n",
       "      <td>0.071932</td>\n",
       "      <td>0.015477</td>\n",
       "      <td>332.8</td>\n",
       "      <td>0.010631</td>\n",
       "      <td>0.011377</td>\n",
       "      <td>0.839784</td>\n",
       "      <td>0.838718</td>\n",
       "      <td>0.001303</td>\n",
       "      <td>0.053292</td>\n",
       "      <td>0.046549</td>\n",
       "      <td>272.812829</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.0001 0.1 0.001</th>\n",
       "      <td>0.272876</td>\n",
       "      <td>0.272687</td>\n",
       "      <td>0.089800</td>\n",
       "      <td>2048.0</td>\n",
       "      <td>0.013968</td>\n",
       "      <td>0.012671</td>\n",
       "      <td>3.142075</td>\n",
       "      <td>3.141397</td>\n",
       "      <td>0.000054</td>\n",
       "      <td>0.101837</td>\n",
       "      <td>0.089858</td>\n",
       "      <td>872.970864</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.0001 0.1 0.005</th>\n",
       "      <td>0.679842</td>\n",
       "      <td>0.679719</td>\n",
       "      <td>0.416117</td>\n",
       "      <td>9772.8</td>\n",
       "      <td>0.055112</td>\n",
       "      <td>0.051775</td>\n",
       "      <td>7.826966</td>\n",
       "      <td>7.826325</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.631676</td>\n",
       "      <td>0.585317</td>\n",
       "      <td>3071.242072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.0001 0.1 0.0001</th>\n",
       "      <td>0.101650</td>\n",
       "      <td>0.101431</td>\n",
       "      <td>0.023681</td>\n",
       "      <td>518.4</td>\n",
       "      <td>0.012442</td>\n",
       "      <td>0.011355</td>\n",
       "      <td>1.180872</td>\n",
       "      <td>1.180126</td>\n",
       "      <td>0.001540</td>\n",
       "      <td>0.047541</td>\n",
       "      <td>0.044647</td>\n",
       "      <td>109.585345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.0001 0.2 0.001</th>\n",
       "      <td>0.261756</td>\n",
       "      <td>0.261565</td>\n",
       "      <td>0.084323</td>\n",
       "      <td>1920.0</td>\n",
       "      <td>0.013488</td>\n",
       "      <td>0.013968</td>\n",
       "      <td>3.013983</td>\n",
       "      <td>3.013222</td>\n",
       "      <td>0.000045</td>\n",
       "      <td>0.094206</td>\n",
       "      <td>0.090477</td>\n",
       "      <td>412.693623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.0001 0.2 0.005</th>\n",
       "      <td>0.673616</td>\n",
       "      <td>0.673492</td>\n",
       "      <td>0.408344</td>\n",
       "      <td>9587.2</td>\n",
       "      <td>0.052055</td>\n",
       "      <td>0.048579</td>\n",
       "      <td>7.755290</td>\n",
       "      <td>7.754643</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.598673</td>\n",
       "      <td>0.552160</td>\n",
       "      <td>2738.859563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.0001 0.2 0.0001</th>\n",
       "      <td>0.100712</td>\n",
       "      <td>0.100493</td>\n",
       "      <td>0.023400</td>\n",
       "      <td>512.0</td>\n",
       "      <td>0.012407</td>\n",
       "      <td>0.011453</td>\n",
       "      <td>1.170631</td>\n",
       "      <td>1.169791</td>\n",
       "      <td>0.001627</td>\n",
       "      <td>0.047760</td>\n",
       "      <td>0.044375</td>\n",
       "      <td>110.212280</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           kl_bound  binomial_approximation_shah  p2l_bound  \\\n",
       "0.1 50 0.01 0.1 0.001      0.352127                     0.352041   0.127062   \n",
       "0.1 50 0.01 0.1 0.005      0.458112                     0.458035   0.175823   \n",
       "0.1 50 0.01 0.1 0.0001     0.329773                     0.329685   0.118833   \n",
       "0.1 50 0.01 0.2 0.001      0.353470                     0.353384   0.127741   \n",
       "0.1 50 0.01 0.2 0.005      0.458112                     0.458035   0.175823   \n",
       "0.1 50 0.01 0.2 0.0001     0.329773                     0.329685   0.118833   \n",
       "0.1 50 0.001 0.1 0.001     0.214693                     0.214588   0.062753   \n",
       "0.1 50 0.001 0.1 0.005     0.652792                     0.653627   0.364699   \n",
       "0.1 50 0.001 0.1 0.0001    0.103780                     0.103664   0.023853   \n",
       "0.1 50 0.001 0.2 0.001     0.213173                     0.213068   0.062336   \n",
       "0.1 50 0.001 0.2 0.005     0.652792                     0.653627   0.364699   \n",
       "0.1 50 0.001 0.2 0.0001    0.103780                     0.103664   0.023853   \n",
       "0.1 50 0.0001 0.1 0.001    0.247275                     0.247172   0.076673   \n",
       "0.1 50 0.0001 0.1 0.005    0.777718                     0.777676   0.382558   \n",
       "0.1 50 0.0001 0.1 0.0001   0.192811                     0.192704   0.054000   \n",
       "0.1 50 0.0001 0.2 0.001    0.249578                     0.249476   0.077762   \n",
       "0.1 50 0.0001 0.2 0.005    0.777718                     0.777676   0.382558   \n",
       "0.1 50 0.0001 0.2 0.0001   0.192811                     0.192704   0.054000   \n",
       "0.1 100 0.01 0.1 0.001     0.347797                     0.347710   0.125418   \n",
       "0.1 100 0.01 0.1 0.005     0.460731                     0.460655   0.177181   \n",
       "0.1 100 0.01 0.1 0.0001    0.228247                     0.228142   0.074192   \n",
       "0.1 100 0.01 0.2 0.001     0.350628                     0.350541   0.126514   \n",
       "0.1 100 0.01 0.2 0.005     0.460731                     0.460655   0.177181   \n",
       "0.1 100 0.01 0.2 0.0001    0.228247                     0.228142   0.074192   \n",
       "0.1 100 0.001 0.1 0.001    0.199226                     0.199120   0.056597   \n",
       "0.1 100 0.001 0.1 0.005    0.655194                     0.655597   0.375578   \n",
       "0.1 100 0.001 0.1 0.0001   0.099972                     0.099856   0.022743   \n",
       "0.1 100 0.001 0.2 0.001    0.178733                     0.178625   0.048657   \n",
       "0.1 100 0.001 0.2 0.005    0.655194                     0.655597   0.375578   \n",
       "0.1 100 0.001 0.2 0.0001   0.099972                     0.099856   0.022743   \n",
       "0.1 100 0.0001 0.1 0.001   0.238983                     0.238880   0.072993   \n",
       "0.1 100 0.0001 0.1 0.005   0.737257                     0.737861   0.361197   \n",
       "0.1 100 0.0001 0.1 0.0001  0.175904                     0.175795   0.047694   \n",
       "0.1 100 0.0001 0.2 0.001   0.253584                     0.253482   0.079667   \n",
       "0.1 100 0.0001 0.2 0.005   0.737257                     0.737861   0.361197   \n",
       "0.1 100 0.0001 0.2 0.0001  0.175904                     0.175795   0.047694   \n",
       "0.2 50 0.01 0.1 0.001      0.175166                     0.175042   0.047595   \n",
       "0.2 50 0.01 0.1 0.005      0.315312                     0.315204   0.110153   \n",
       "0.2 50 0.01 0.1 0.0001     0.167029                     0.166905   0.044611   \n",
       "0.2 50 0.01 0.2 0.001      0.180424                     0.180301   0.049632   \n",
       "0.2 50 0.01 0.2 0.005      0.315312                     0.315204   0.110153   \n",
       "0.2 50 0.01 0.2 0.0001     0.167029                     0.166905   0.044611   \n",
       "0.2 50 0.001 0.1 0.001     0.176772                     0.176649   0.048070   \n",
       "0.2 50 0.001 0.1 0.005     0.683026                     0.683174   0.413789   \n",
       "0.2 50 0.001 0.1 0.0001    0.090929                     0.090796   0.020264   \n",
       "0.2 50 0.001 0.2 0.001     0.190151                     0.190029   0.053092   \n",
       "0.2 50 0.001 0.2 0.005     0.683026                     0.683174   0.413789   \n",
       "0.2 50 0.001 0.2 0.0001    0.090929                     0.090796   0.020264   \n",
       "0.2 50 0.0001 0.1 0.001    0.248064                     0.247948   0.077168   \n",
       "0.2 50 0.0001 0.1 0.005    0.611837                     0.611756   0.337068   \n",
       "0.2 50 0.0001 0.1 0.0001   0.174903                     0.174779   0.047440   \n",
       "0.2 50 0.0001 0.2 0.001    0.239228                     0.239111   0.073267   \n",
       "0.2 50 0.0001 0.2 0.005    0.628229                     0.628150   0.354551   \n",
       "0.2 50 0.0001 0.2 0.0001   0.170652                     0.170528   0.045870   \n",
       "0.2 100 0.01 0.1 0.001     0.174960                     0.174836   0.047441   \n",
       "0.2 100 0.01 0.1 0.005     0.287308                     0.287197   0.095853   \n",
       "0.2 100 0.01 0.1 0.0001    0.184633                     0.184511   0.051051   \n",
       "0.2 100 0.01 0.2 0.001     0.177131                     0.177008   0.048227   \n",
       "0.2 100 0.01 0.2 0.005     0.286005                     0.285893   0.095230   \n",
       "0.2 100 0.01 0.2 0.0001    0.185052                     0.184929   0.051208   \n",
       "0.2 100 0.001 0.1 0.001    0.153840                     0.153715   0.039894   \n",
       "0.2 100 0.001 0.1 0.005    0.678513                     0.679029   0.394654   \n",
       "0.2 100 0.001 0.1 0.0001   0.088063                     0.087930   0.019466   \n",
       "0.2 100 0.001 0.2 0.001    0.145956                     0.145830   0.037215   \n",
       "0.2 100 0.001 0.2 0.005    0.668580                     0.669094   0.382864   \n",
       "0.2 100 0.001 0.2 0.0001   0.087475                     0.087343   0.019306   \n",
       "0.2 100 0.0001 0.1 0.001   0.246333                     0.246217   0.076388   \n",
       "0.2 100 0.0001 0.1 0.005   0.679993                     0.680571   0.386690   \n",
       "0.2 100 0.0001 0.1 0.0001  0.151629                     0.151503   0.039108   \n",
       "0.2 100 0.0001 0.2 0.001   0.235507                     0.235390   0.071704   \n",
       "0.2 100 0.0001 0.2 0.005   0.673286                     0.673864   0.379332   \n",
       "0.2 100 0.0001 0.2 0.0001  0.150685                     0.150559   0.038792   \n",
       "0.5 50 0.01 0.1 0.001      0.155428                     0.155219   0.041008   \n",
       "0.5 50 0.01 0.1 0.005      0.241293                     0.241099   0.075251   \n",
       "0.5 50 0.01 0.1 0.0001     0.164055                     0.163848   0.044064   \n",
       "0.5 50 0.01 0.2 0.001      0.156245                     0.156036   0.041287   \n",
       "0.5 50 0.01 0.2 0.005      0.251599                     0.251407   0.080716   \n",
       "0.5 50 0.01 0.2 0.0001     0.164055                     0.163848   0.044064   \n",
       "0.5 50 0.001 0.1 0.001     0.111081                     0.110864   0.026492   \n",
       "0.5 50 0.001 0.1 0.005     0.692967                     0.692846   0.432730   \n",
       "0.5 50 0.001 0.1 0.0001    0.073244                     0.073020   0.015762   \n",
       "0.5 50 0.001 0.2 0.001     0.106329                     0.106111   0.025086   \n",
       "0.5 50 0.001 0.2 0.005     0.680507                     0.680384   0.416134   \n",
       "0.5 50 0.001 0.2 0.0001    0.075341                     0.075117   0.016330   \n",
       "0.5 50 0.0001 0.1 0.001    0.270231                     0.270042   0.088168   \n",
       "0.5 50 0.0001 0.1 0.005    0.674288                     0.674165   0.409149   \n",
       "0.5 50 0.0001 0.1 0.0001   0.112941                     0.112724   0.027054   \n",
       "0.5 50 0.0001 0.2 0.001    0.269329                     0.269139   0.087889   \n",
       "0.5 50 0.0001 0.2 0.005    0.679594                     0.679471   0.415586   \n",
       "0.5 50 0.0001 0.2 0.0001   0.112941                     0.112724   0.027054   \n",
       "0.5 100 0.01 0.1 0.001     0.155684                     0.155475   0.041013   \n",
       "0.5 100 0.01 0.1 0.005     0.227441                     0.227245   0.068935   \n",
       "0.5 100 0.01 0.1 0.0001    0.166577                     0.166370   0.044900   \n",
       "0.5 100 0.01 0.2 0.001     0.157229                     0.157020   0.041568   \n",
       "0.5 100 0.01 0.2 0.005     0.228695                     0.228498   0.069485   \n",
       "0.5 100 0.01 0.2 0.0001    0.166577                     0.166370   0.044900   \n",
       "0.5 100 0.001 0.1 0.001    0.111967                     0.111750   0.026772   \n",
       "0.5 100 0.001 0.1 0.005    0.681184                     0.681061   0.417465   \n",
       "0.5 100 0.001 0.1 0.0001   0.071070                     0.070845   0.015193   \n",
       "0.5 100 0.001 0.2 0.001    0.105545                     0.105327   0.024808   \n",
       "0.5 100 0.001 0.2 0.005    0.677711                     0.677587   0.413175   \n",
       "0.5 100 0.001 0.2 0.0001   0.072157                     0.071932   0.015477   \n",
       "0.5 100 0.0001 0.1 0.001   0.272876                     0.272687   0.089800   \n",
       "0.5 100 0.0001 0.1 0.005   0.679842                     0.679719   0.416117   \n",
       "0.5 100 0.0001 0.1 0.0001  0.101650                     0.101431   0.023681   \n",
       "0.5 100 0.0001 0.2 0.001   0.261756                     0.261565   0.084323   \n",
       "0.5 100 0.0001 0.2 0.005   0.673616                     0.673492   0.408344   \n",
       "0.5 100 0.0001 0.2 0.0001  0.100712                     0.100493   0.023400   \n",
       "\n",
       "                           compression_set_size  validation_error  test_error  \\\n",
       "0.1 50 0.01 0.1 0.001                    5939.2          0.193691    0.193762   \n",
       "0.1 50 0.01 0.1 0.005                    8230.4          0.193504    0.192447   \n",
       "0.1 50 0.01 0.1 0.0001                   5555.2          0.196363    0.194695   \n",
       "0.1 50 0.01 0.2 0.001                    5971.2          0.193701    0.193100   \n",
       "0.1 50 0.01 0.2 0.005                    8230.4          0.193504    0.192447   \n",
       "0.1 50 0.01 0.2 0.0001                   5555.2          0.196363    0.194695   \n",
       "0.1 50 0.001 0.1 0.001                   2880.0          0.011230    0.011348   \n",
       "0.1 50 0.001 0.1 0.005                  17203.2          0.091673    0.088168   \n",
       "0.1 50 0.001 0.1 0.0001                  1068.8          0.012107    0.011648   \n",
       "0.1 50 0.001 0.2 0.001                   2860.8          0.010724    0.010422   \n",
       "0.1 50 0.001 0.2 0.005                  17203.2          0.091673    0.088168   \n",
       "0.1 50 0.001 0.2 0.0001                  1068.8          0.012107    0.011648   \n",
       "0.1 50 0.0001 0.1 0.001                  3532.8          0.011750    0.010770   \n",
       "0.1 50 0.0001 0.1 0.005                 18054.4          0.392336    0.388250   \n",
       "0.1 50 0.0001 0.1 0.0001                 2470.4          0.013412    0.012773   \n",
       "0.1 50 0.0001 0.2 0.001                  3584.0          0.011456    0.011059   \n",
       "0.1 50 0.0001 0.2 0.005                 18054.4          0.392336    0.388250   \n",
       "0.1 50 0.0001 0.2 0.0001                 2470.4          0.013412    0.012773   \n",
       "0.1 100 0.01 0.1 0.001                   5862.4          0.195992    0.195190   \n",
       "0.1 100 0.01 0.1 0.005                   8294.4          0.194235    0.193971   \n",
       "0.1 100 0.01 0.1 0.0001                  3424.0          0.024104    0.022260   \n",
       "0.1 100 0.01 0.2 0.001                   5913.6          0.197150    0.195500   \n",
       "0.1 100 0.01 0.2 0.005                   8294.4          0.194235    0.193971   \n",
       "0.1 100 0.01 0.2 0.0001                  3424.0          0.024104    0.022260   \n",
       "0.1 100 0.001 0.1 0.001                  2592.0          0.011022    0.010262   \n",
       "0.1 100 0.001 0.1 0.005                 17721.6          0.093036    0.088932   \n",
       "0.1 100 0.001 0.1 0.0001                 1017.6          0.013769    0.011966   \n",
       "0.1 100 0.001 0.2 0.001                  2220.8          0.010614    0.010162   \n",
       "0.1 100 0.001 0.2 0.005                 17721.6          0.093036    0.088932   \n",
       "0.1 100 0.001 0.2 0.0001                 1017.6          0.013769    0.011966   \n",
       "0.1 100 0.0001 0.1 0.001                 3360.0          0.010983    0.010733   \n",
       "0.1 100 0.0001 0.1 0.005                17036.8          0.275070    0.274120   \n",
       "0.1 100 0.0001 0.1 0.0001                2176.0          0.013826    0.012800   \n",
       "0.1 100 0.0001 0.2 0.001                 3673.6          0.012260    0.011468   \n",
       "0.1 100 0.0001 0.2 0.005                17036.8          0.275070    0.274120   \n",
       "0.1 100 0.0001 0.2 0.0001                2176.0          0.013826    0.012800   \n",
       "0.2 50 0.01 0.1 0.001                    1894.4          0.022016    0.019077   \n",
       "0.2 50 0.01 0.1 0.005                    4460.8          0.019357    0.018195   \n",
       "0.2 50 0.01 0.1 0.0001                   1772.8          0.022925    0.019880   \n",
       "0.2 50 0.01 0.2 0.001                    1977.6          0.021006    0.017978   \n",
       "0.2 50 0.01 0.2 0.005                    4460.8          0.019357    0.018195   \n",
       "0.2 50 0.01 0.2 0.0001                   1772.8          0.022925    0.019880   \n",
       "0.2 50 0.001 0.1 0.001                   1913.6          0.011210    0.010218   \n",
       "0.2 50 0.001 0.1 0.005                  17088.0          0.073311    0.073065   \n",
       "0.2 50 0.001 0.1 0.0001                   787.2          0.011994    0.010924   \n",
       "0.2 50 0.001 0.2 0.001                   2118.4          0.010828    0.010323   \n",
       "0.2 50 0.001 0.2 0.005                  17088.0          0.073311    0.073065   \n",
       "0.2 50 0.001 0.2 0.0001                   787.2          0.011994    0.010924   \n",
       "0.2 50 0.0001 0.1 0.001                  3104.0          0.011980    0.011772   \n",
       "0.2 50 0.0001 0.1 0.005                 13881.6          0.051728    0.051720   \n",
       "0.2 50 0.0001 0.1 0.0001                 1888.0          0.012957    0.012351   \n",
       "0.2 50 0.0001 0.2 0.001                  2944.0          0.011885    0.012280   \n",
       "0.2 50 0.0001 0.2 0.005                 14611.2          0.054473    0.053175   \n",
       "0.2 50 0.0001 0.2 0.0001                 1824.0          0.012998    0.012320   \n",
       "0.2 100 0.01 0.1 0.001                   1888.0          0.021376    0.019821   \n",
       "0.2 100 0.01 0.1 0.005                   3872.0          0.019792    0.019237   \n",
       "0.2 100 0.01 0.1 0.0001                  2035.2          0.023892    0.021719   \n",
       "0.2 100 0.01 0.2 0.001                   1920.0          0.021007    0.020107   \n",
       "0.2 100 0.01 0.2 0.005                   3846.4          0.019706    0.019408   \n",
       "0.2 100 0.01 0.2 0.0001                  2041.6          0.023772    0.021654   \n",
       "0.2 100 0.001 0.1 0.001                  1580.8          0.011115    0.010137   \n",
       "0.2 100 0.001 0.1 0.005                 16288.0          0.094489    0.090146   \n",
       "0.2 100 0.001 0.1 0.0001                  755.2          0.012447    0.011536   \n",
       "0.2 100 0.001 0.2 0.001                  1472.0          0.011174    0.010471   \n",
       "0.2 100 0.001 0.2 0.005                 15795.2          0.086826    0.083468   \n",
       "0.2 100 0.001 0.2 0.0001                  748.8          0.012671    0.011529   \n",
       "0.2 100 0.0001 0.1 0.001                 3072.0          0.012379    0.012303   \n",
       "0.2 100 0.0001 0.1 0.005                15955.2          0.096004    0.089157   \n",
       "0.2 100 0.0001 0.1 0.0001                1548.8          0.011918    0.012037   \n",
       "0.2 100 0.0001 0.2 0.001                 2880.0          0.011880    0.012021   \n",
       "0.2 100 0.0001 0.2 0.005                15648.0          0.089753    0.084308   \n",
       "0.2 100 0.0001 0.2 0.0001                1536.0          0.011848    0.011837   \n",
       "0.5 50 0.01 0.1 0.001                     915.2          0.020031    0.018739   \n",
       "0.5 50 0.01 0.1 0.005                    1708.8          0.020038    0.017584   \n",
       "0.5 50 0.01 0.1 0.0001                    985.6          0.022602    0.019793   \n",
       "0.5 50 0.01 0.2 0.001                     921.6          0.020152    0.017806   \n",
       "0.5 50 0.01 0.2 0.005                    1836.8          0.019514    0.017430   \n",
       "0.5 50 0.01 0.2 0.0001                    985.6          0.022737    0.019401   \n",
       "0.5 50 0.001 0.1 0.001                    582.4          0.011276    0.009708   \n",
       "0.5 50 0.001 0.1 0.005                  10169.6          0.067597    0.065122   \n",
       "0.5 50 0.001 0.1 0.0001                   339.2          0.010797    0.010995   \n",
       "0.5 50 0.001 0.2 0.001                    550.4          0.010388    0.009556   \n",
       "0.5 50 0.001 0.2 0.005                   9772.8          0.069274    0.065777   \n",
       "0.5 50 0.001 0.2 0.0001                   352.0          0.010621    0.010788   \n",
       "0.5 50 0.0001 0.1 0.001                  2009.6          0.014800    0.013748   \n",
       "0.5 50 0.0001 0.1 0.005                  9606.4          0.064930    0.060006   \n",
       "0.5 50 0.0001 0.1 0.0001                  595.2          0.013359    0.012187   \n",
       "0.5 50 0.0001 0.2 0.001                  2003.2          0.013879    0.013165   \n",
       "0.5 50 0.0001 0.2 0.005                  9760.0          0.065904    0.059904   \n",
       "0.5 50 0.0001 0.2 0.0001                  595.2          0.013359    0.012187   \n",
       "0.5 100 0.01 0.1 0.001                    915.2          0.020507    0.017959   \n",
       "0.5 100 0.01 0.1 0.005                   1561.6          0.019946    0.017181   \n",
       "0.5 100 0.01 0.1 0.0001                  1004.8          0.023480    0.021593   \n",
       "0.5 100 0.01 0.2 0.001                    928.0          0.020434    0.017697   \n",
       "0.5 100 0.01 0.2 0.005                   1574.4          0.020046    0.016934   \n",
       "0.5 100 0.01 0.2 0.0001                  1004.8          0.023503    0.021379   \n",
       "0.5 100 0.001 0.1 0.001                   588.8          0.009288    0.010751   \n",
       "0.5 100 0.001 0.1 0.005                  9804.8          0.065212    0.058716   \n",
       "0.5 100 0.001 0.1 0.0001                  326.4          0.010663    0.011426   \n",
       "0.5 100 0.001 0.2 0.001                   544.0          0.010052    0.010761   \n",
       "0.5 100 0.001 0.2 0.005                  9702.4          0.064972    0.058984   \n",
       "0.5 100 0.001 0.2 0.0001                  332.8          0.010631    0.011377   \n",
       "0.5 100 0.0001 0.1 0.001                 2048.0          0.013968    0.012671   \n",
       "0.5 100 0.0001 0.1 0.005                 9772.8          0.055112    0.051775   \n",
       "0.5 100 0.0001 0.1 0.0001                 518.4          0.012442    0.011355   \n",
       "0.5 100 0.0001 0.2 0.001                 1920.0          0.013488    0.013968   \n",
       "0.5 100 0.0001 0.2 0.005                 9587.2          0.052055    0.048579   \n",
       "0.5 100 0.0001 0.2 0.0001                 512.0          0.012407    0.011453   \n",
       "\n",
       "                           CE_kl_bound  CE_min_val_catoni_bound  \\\n",
       "0.1 50 0.01 0.1 0.001         3.764233                 3.857538   \n",
       "0.1 50 0.01 0.1 0.005         4.977358                 5.069166   \n",
       "0.1 50 0.01 0.1 0.0001        3.573397                 3.673132   \n",
       "0.1 50 0.01 0.2 0.001         3.779415                 3.872324   \n",
       "0.1 50 0.01 0.2 0.005         4.977358                 5.069166   \n",
       "0.1 50 0.01 0.2 0.0001        3.573397                 3.673132   \n",
       "0.1 50 0.001 0.1 0.001        2.472033                 2.471617   \n",
       "0.1 50 0.001 0.1 0.005        7.531131                 7.565455   \n",
       "0.1 50 0.001 0.1 0.0001       1.209370                 1.208847   \n",
       "0.1 50 0.001 0.2 0.001        2.454661                 2.454367   \n",
       "0.1 50 0.001 0.2 0.005        7.531131                 7.565455   \n",
       "0.1 50 0.001 0.2 0.0001       1.209370                 1.208847   \n",
       "0.1 50 0.0001 0.1 0.001       2.847017                 2.846515   \n",
       "0.1 50 0.0001 0.1 0.005       8.953246                 8.953494   \n",
       "0.1 50 0.0001 0.1 0.0001      2.221881                 2.223369   \n",
       "0.1 50 0.0001 0.2 0.001       2.873543                 2.873055   \n",
       "0.1 50 0.0001 0.2 0.005       8.953246                 8.953494   \n",
       "0.1 50 0.0001 0.2 0.0001      2.221881                 2.223369   \n",
       "0.1 100 0.01 0.1 0.001        3.719693                 3.812003   \n",
       "0.1 100 0.01 0.1 0.005        5.007673                 5.099593   \n",
       "0.1 100 0.01 0.1 0.0001       2.720026                 2.730640   \n",
       "0.1 100 0.01 0.2 0.001        3.752947                 3.845634   \n",
       "0.1 100 0.01 0.2 0.005        5.007673                 5.099593   \n",
       "0.1 100 0.01 0.2 0.0001       2.720026                 2.730640   \n",
       "0.1 100 0.001 0.1 0.001       2.294107                 2.293820   \n",
       "0.1 100 0.001 0.1 0.005       7.534741                 7.545830   \n",
       "0.1 100 0.001 0.1 0.0001      1.165102                 1.164514   \n",
       "0.1 100 0.001 0.2 0.001       2.058184                 2.057878   \n",
       "0.1 100 0.001 0.2 0.005       7.534741                 7.545830   \n",
       "0.1 100 0.001 0.2 0.0001      1.165102                 1.164514   \n",
       "0.1 100 0.0001 0.1 0.001      2.751667                 2.751245   \n",
       "0.1 100 0.0001 0.1 0.005      8.477993                 8.564228   \n",
       "0.1 100 0.0001 0.1 0.0001     2.028423                 2.030316   \n",
       "0.1 100 0.0001 0.2 0.001      2.919797                 2.919418   \n",
       "0.1 100 0.0001 0.2 0.005      8.477993                 8.564228   \n",
       "0.1 100 0.0001 0.2 0.0001     2.028423                 2.030316   \n",
       "0.2 50 0.01 0.1 0.001         2.040576                 2.040253   \n",
       "0.2 50 0.01 0.1 0.005         3.630956                 3.630956   \n",
       "0.2 50 0.01 0.1 0.0001        2.045735                 2.063732   \n",
       "0.2 50 0.01 0.2 0.001         2.096352                 2.096439   \n",
       "0.2 50 0.01 0.2 0.005         3.630956                 3.630956   \n",
       "0.2 50 0.01 0.2 0.0001        2.045735                 2.063732   \n",
       "0.2 50 0.001 0.1 0.001        2.035603                 2.035253   \n",
       "0.2 50 0.001 0.1 0.005        7.864293                 7.867262   \n",
       "0.2 50 0.001 0.1 0.0001       1.063091                 1.062450   \n",
       "0.2 50 0.001 0.2 0.001        2.189390                 2.188811   \n",
       "0.2 50 0.001 0.2 0.005        7.864293                 7.867262   \n",
       "0.2 50 0.001 0.2 0.0001       1.063091                 1.062450   \n",
       "0.2 50 0.0001 0.1 0.001       2.856198                 2.855705   \n",
       "0.2 50 0.0001 0.1 0.005       7.044030                 7.043582   \n",
       "0.2 50 0.0001 0.1 0.0001      2.016593                 2.018199   \n",
       "0.2 50 0.0001 0.2 0.001       2.754513                 2.754057   \n",
       "0.2 50 0.0001 0.2 0.005       7.232753                 7.232315   \n",
       "0.2 50 0.0001 0.2 0.0001      1.968043                 1.969959   \n",
       "0.2 100 0.01 0.1 0.001        2.046025                 2.045538   \n",
       "0.2 100 0.01 0.1 0.005        3.309117                 3.309760   \n",
       "0.2 100 0.01 0.1 0.0001       2.246248                 2.261798   \n",
       "0.2 100 0.01 0.2 0.001        2.067079                 2.066850   \n",
       "0.2 100 0.01 0.2 0.005        3.294218                 3.294958   \n",
       "0.2 100 0.01 0.2 0.0001       2.250532                 2.265845   \n",
       "0.2 100 0.001 0.1 0.001       1.771596                 1.771216   \n",
       "0.2 100 0.001 0.1 0.005       7.859176                 7.909655   \n",
       "0.2 100 0.001 0.1 0.0001      1.032238                 1.031551   \n",
       "0.2 100 0.001 0.2 0.001       1.681211                 1.681212   \n",
       "0.2 100 0.001 0.2 0.005       7.744813                 7.795287   \n",
       "0.2 100 0.001 0.2 0.0001      1.026074                 1.025390   \n",
       "0.2 100 0.0001 0.1 0.001      2.836318                 2.835864   \n",
       "0.2 100 0.0001 0.1 0.005      7.827576                 7.887833   \n",
       "0.2 100 0.0001 0.1 0.0001     1.750118                 1.751659   \n",
       "0.2 100 0.0001 0.2 0.001      2.711724                 2.711295   \n",
       "0.2 100 0.0001 0.2 0.005      7.750360                 7.810613   \n",
       "0.2 100 0.0001 0.2 0.0001     1.739139                 1.740683   \n",
       "0.5 50 0.01 0.1 0.001         1.834185                 1.833995   \n",
       "0.5 50 0.01 0.1 0.005         2.781068                 2.782873   \n",
       "0.5 50 0.01 0.1 0.0001        1.992703                 2.005482   \n",
       "0.5 50 0.01 0.2 0.001         1.839748                 1.839333   \n",
       "0.5 50 0.01 0.2 0.005         2.899517                 2.901102   \n",
       "0.5 50 0.01 0.2 0.0001        1.991907                 2.004390   \n",
       "0.5 50 0.001 0.1 0.001        1.281422                 1.281823   \n",
       "0.5 50 0.001 0.1 0.005        7.978082                 7.977453   \n",
       "0.5 50 0.001 0.1 0.0001       0.861935                 0.861014   \n",
       "0.5 50 0.001 0.2 0.001        1.227287                 1.227418   \n",
       "0.5 50 0.001 0.2 0.005        7.834628                 7.833987   \n",
       "0.5 50 0.001 0.2 0.0001       0.883966                 0.882902   \n",
       "0.5 50 0.0001 0.1 0.001       3.111644                 3.110966   \n",
       "0.5 50 0.0001 0.1 0.005       7.763033                 7.762386   \n",
       "0.5 50 0.0001 0.1 0.0001      1.313943                 1.313113   \n",
       "0.5 50 0.0001 0.2 0.001       3.101053                 3.100209   \n",
       "0.5 50 0.0001 0.2 0.005       7.824111                 7.823469   \n",
       "0.5 50 0.0001 0.2 0.0001      1.313942                 1.313112   \n",
       "0.5 100 0.01 0.1 0.001        1.839851                 1.840354   \n",
       "0.5 100 0.01 0.1 0.005        2.623046                 2.623897   \n",
       "0.5 100 0.01 0.1 0.0001       2.023296                 2.035532   \n",
       "0.5 100 0.01 0.2 0.001        1.853642                 1.853575   \n",
       "0.5 100 0.01 0.2 0.005        2.637200                 2.637741   \n",
       "0.5 100 0.01 0.2 0.0001       2.023296                 2.035533   \n",
       "0.5 100 0.001 0.1 0.001       1.290202                 1.290037   \n",
       "0.5 100 0.001 0.1 0.005       7.842418                 7.841778   \n",
       "0.5 100 0.001 0.1 0.0001      0.827947                 0.826882   \n",
       "0.5 100 0.001 0.2 0.001       1.216881                 1.217111   \n",
       "0.5 100 0.001 0.2 0.005       7.802431                 7.801788   \n",
       "0.5 100 0.001 0.2 0.0001      0.839784                 0.838718   \n",
       "0.5 100 0.0001 0.1 0.001      3.142075                 3.141397   \n",
       "0.5 100 0.0001 0.1 0.005      7.826966                 7.826325   \n",
       "0.5 100 0.0001 0.1 0.0001     1.180872                 1.180126   \n",
       "0.5 100 0.0001 0.2 0.001      3.013983                 3.013222   \n",
       "0.5 100 0.0001 0.2 0.005      7.755290                 7.754643   \n",
       "0.5 100 0.0001 0.2 0.0001     1.170631                 1.169791   \n",
       "\n",
       "                           complement_loss  validation_loss  test_loss  \\\n",
       "0.1 50 0.01 0.1 0.001             0.461629         0.520778   0.517133   \n",
       "0.1 50 0.01 0.1 0.005             0.460625         0.571432   0.557218   \n",
       "0.1 50 0.01 0.1 0.0001            0.475925         0.517555   0.512739   \n",
       "0.1 50 0.01 0.2 0.001             0.461600         0.522958   0.515490   \n",
       "0.1 50 0.01 0.2 0.005             0.460625         0.571432   0.557218   \n",
       "0.1 50 0.01 0.2 0.0001            0.475925         0.517555   0.512739   \n",
       "0.1 50 0.001 0.1 0.001            0.000030         0.081253   0.075731   \n",
       "0.1 50 0.001 0.1 0.005            0.145721         1.058038   0.994486   \n",
       "0.1 50 0.001 0.1 0.0001           0.002201         0.050163   0.042957   \n",
       "0.1 50 0.001 0.2 0.001            0.000044         0.080642   0.075764   \n",
       "0.1 50 0.001 0.2 0.005            0.145721         1.058038   0.994486   \n",
       "0.1 50 0.001 0.2 0.0001           0.002201         0.050163   0.042957   \n",
       "0.1 50 0.0001 0.1 0.001           0.000016         0.088303   0.079405   \n",
       "0.1 50 0.0001 0.1 0.005           3.777558         4.501556   4.439614   \n",
       "0.1 50 0.0001 0.1 0.0001          0.000250         0.066606   0.058348   \n",
       "0.1 50 0.0001 0.2 0.001           0.000018         0.092505   0.085001   \n",
       "0.1 50 0.0001 0.2 0.005           3.777558         4.501556   4.439614   \n",
       "0.1 50 0.0001 0.2 0.0001          0.000250         0.066606   0.058348   \n",
       "0.1 100 0.01 0.1 0.001            0.462505         0.523461   0.518910   \n",
       "0.1 100 0.01 0.1 0.005            0.460646         0.560861   0.554136   \n",
       "0.1 100 0.01 0.1 0.0001           0.019953         0.086938   0.078881   \n",
       "0.1 100 0.01 0.2 0.001            0.462691         0.528850   0.521348   \n",
       "0.1 100 0.01 0.2 0.005            0.460646         0.560861   0.554136   \n",
       "0.1 100 0.01 0.2 0.0001           0.019953         0.086938   0.078881   \n",
       "0.1 100 0.001 0.1 0.001           0.000045         0.079231   0.071050   \n",
       "0.1 100 0.001 0.1 0.005           0.058387         1.050746   1.000831   \n",
       "0.1 100 0.001 0.1 0.0001          0.002122         0.049962   0.043052   \n",
       "0.1 100 0.001 0.2 0.001           0.000045         0.072113   0.065899   \n",
       "0.1 100 0.001 0.2 0.005           0.058387         1.050746   1.000831   \n",
       "0.1 100 0.001 0.2 0.0001          0.002122         0.049962   0.043052   \n",
       "0.1 100 0.0001 0.1 0.001          0.000029         0.084314   0.080571   \n",
       "0.1 100 0.0001 0.1 0.005          2.627745         3.188697   3.153391   \n",
       "0.1 100 0.0001 0.1 0.0001         0.000411         0.063092   0.057025   \n",
       "0.1 100 0.0001 0.2 0.001          0.000033         0.091494   0.083414   \n",
       "0.1 100 0.0001 0.2 0.005          2.627745         3.188697   3.153391   \n",
       "0.1 100 0.0001 0.2 0.0001         0.000411         0.063092   0.057025   \n",
       "0.2 50 0.01 0.1 0.001             0.003934         0.086088   0.072829   \n",
       "0.2 50 0.01 0.1 0.005             0.000102         0.130083   0.118285   \n",
       "0.2 50 0.01 0.1 0.0001            0.027012         0.084964   0.077047   \n",
       "0.2 50 0.01 0.2 0.001             0.003073         0.081954   0.071872   \n",
       "0.2 50 0.01 0.2 0.005             0.000102         0.130083   0.118285   \n",
       "0.2 50 0.01 0.2 0.0001            0.027012         0.084964   0.077047   \n",
       "0.2 50 0.001 0.1 0.001            0.000046         0.073488   0.065919   \n",
       "0.2 50 0.001 0.1 0.005            0.040295         0.840060   0.831924   \n",
       "0.2 50 0.001 0.1 0.0001           0.002520         0.044112   0.036578   \n",
       "0.2 50 0.001 0.2 0.001            0.000019         0.076633   0.069752   \n",
       "0.2 50 0.001 0.2 0.005            0.040295         0.840060   0.831924   \n",
       "0.2 50 0.001 0.2 0.0001           0.002520         0.044112   0.036578   \n",
       "0.2 50 0.0001 0.1 0.001           0.000028         0.093192   0.081732   \n",
       "0.2 50 0.0001 0.1 0.005           0.000000         0.602510   0.583475   \n",
       "0.2 50 0.0001 0.1 0.0001          0.000370         0.059193   0.053868   \n",
       "0.2 50 0.0001 0.2 0.001           0.000032         0.085326   0.085965   \n",
       "0.2 50 0.0001 0.2 0.005           0.000000         0.635898   0.599823   \n",
       "0.2 50 0.0001 0.2 0.0001          0.000422         0.058964   0.051933   \n",
       "0.2 100 0.01 0.1 0.001            0.005467         0.083465   0.074830   \n",
       "0.2 100 0.01 0.1 0.005            0.000178         0.119976   0.118816   \n",
       "0.2 100 0.01 0.1 0.0001           0.026591         0.092755   0.084042   \n",
       "0.2 100 0.01 0.2 0.001            0.004741         0.085834   0.076625   \n",
       "0.2 100 0.01 0.2 0.005            0.000191         0.118448   0.118631   \n",
       "0.2 100 0.01 0.2 0.0001           0.026450         0.092671   0.083746   \n",
       "0.2 100 0.001 0.1 0.001           0.000045         0.069024   0.062030   \n",
       "0.2 100 0.001 0.1 0.005           0.239364         1.093728   1.038696   \n",
       "0.2 100 0.001 0.1 0.0001          0.002900         0.047118   0.038755   \n",
       "0.2 100 0.001 0.2 0.001           0.000088         0.071231   0.060359   \n",
       "0.2 100 0.001 0.2 0.005           0.239364         1.012370   0.962250   \n",
       "0.2 100 0.001 0.2 0.0001          0.003007         0.047150   0.038754   \n",
       "0.2 100 0.0001 0.1 0.001          0.000033         0.095430   0.090301   \n",
       "0.2 100 0.0001 0.1 0.005          0.311268         1.129418   1.039157   \n",
       "0.2 100 0.0001 0.1 0.0001         0.000571         0.058036   0.051576   \n",
       "0.2 100 0.0001 0.2 0.001          0.000037         0.085972   0.083469   \n",
       "0.2 100 0.0001 0.2 0.005          0.311268         1.053817   0.985046   \n",
       "0.2 100 0.0001 0.2 0.0001         0.000555         0.058257   0.051452   \n",
       "0.5 50 0.01 0.1 0.001             0.008073         0.072760   0.065821   \n",
       "0.5 50 0.01 0.1 0.005             0.000405         0.095380   0.086412   \n",
       "0.5 50 0.01 0.1 0.0001            0.022332         0.084117   0.075092   \n",
       "0.5 50 0.01 0.2 0.001             0.007298         0.072484   0.063873   \n",
       "0.5 50 0.01 0.2 0.005             0.000378         0.096028   0.087905   \n",
       "0.5 50 0.01 0.2 0.0001            0.022111         0.083979   0.074855   \n",
       "0.5 50 0.001 0.1 0.001            0.000307         0.054745   0.047779   \n",
       "0.5 50 0.001 0.1 0.005            0.000000         0.777890   0.736431   \n",
       "0.5 50 0.001 0.1 0.0001           0.003012         0.043493   0.037847   \n",
       "0.5 50 0.001 0.2 0.001            0.000387         0.052262   0.046735   \n",
       "0.5 50 0.001 0.2 0.005            0.000000         0.794435   0.744721   \n",
       "0.5 50 0.001 0.2 0.0001           0.002612         0.043230   0.037535   \n",
       "0.5 50 0.0001 0.1 0.001           0.000056         0.101851   0.088920   \n",
       "0.5 50 0.0001 0.1 0.005           0.000000         0.732222   0.678990   \n",
       "0.5 50 0.0001 0.1 0.0001          0.002055         0.048798   0.043397   \n",
       "0.5 50 0.0001 0.2 0.001           0.000032         0.092935   0.088020   \n",
       "0.5 50 0.0001 0.2 0.005           0.000000         0.738746   0.675537   \n",
       "0.5 50 0.0001 0.2 0.0001          0.002054         0.048791   0.043397   \n",
       "0.5 100 0.01 0.1 0.001            0.008723         0.077228   0.066065   \n",
       "0.5 100 0.01 0.1 0.005            0.000647         0.098908   0.084937   \n",
       "0.5 100 0.01 0.1 0.0001           0.022469         0.092476   0.084056   \n",
       "0.5 100 0.01 0.2 0.001            0.007857         0.076744   0.064709   \n",
       "0.5 100 0.01 0.2 0.005            0.000610         0.100453   0.084770   \n",
       "0.5 100 0.01 0.2 0.0001           0.022470         0.091741   0.083627   \n",
       "0.5 100 0.001 0.1 0.001           0.000124         0.059646   0.051774   \n",
       "0.5 100 0.001 0.1 0.005           0.000000         0.739514   0.660612   \n",
       "0.5 100 0.001 0.1 0.0001          0.001425         0.053495   0.046665   \n",
       "0.5 100 0.001 0.2 0.001           0.000200         0.059043   0.048941   \n",
       "0.5 100 0.001 0.2 0.005           0.000000         0.741049   0.668210   \n",
       "0.5 100 0.001 0.2 0.0001          0.001303         0.053292   0.046549   \n",
       "0.5 100 0.0001 0.1 0.001          0.000054         0.101837   0.089858   \n",
       "0.5 100 0.0001 0.1 0.005          0.000000         0.631676   0.585317   \n",
       "0.5 100 0.0001 0.1 0.0001         0.001540         0.047541   0.044647   \n",
       "0.5 100 0.0001 0.2 0.001          0.000045         0.094206   0.090477   \n",
       "0.5 100 0.0001 0.2 0.005          0.000000         0.598673   0.552160   \n",
       "0.5 100 0.0001 0.2 0.0001         0.001627         0.047760   0.044375   \n",
       "\n",
       "                              _runtime  \n",
       "0.1 50 0.01 0.1 0.001      2037.201954  \n",
       "0.1 50 0.01 0.1 0.005      2552.396916  \n",
       "0.1 50 0.01 0.1 0.0001     2350.314251  \n",
       "0.1 50 0.01 0.2 0.001      2030.378088  \n",
       "0.1 50 0.01 0.2 0.005      2560.730178  \n",
       "0.1 50 0.01 0.2 0.0001     2341.157793  \n",
       "0.1 50 0.001 0.1 0.001      797.020437  \n",
       "0.1 50 0.001 0.1 0.005     6777.287383  \n",
       "0.1 50 0.001 0.1 0.0001     352.330533  \n",
       "0.1 50 0.001 0.2 0.001      753.322974  \n",
       "0.1 50 0.001 0.2 0.005     6779.161948  \n",
       "0.1 50 0.001 0.2 0.0001     349.603113  \n",
       "0.1 50 0.0001 0.1 0.001     994.422225  \n",
       "0.1 50 0.0001 0.1 0.005    6550.379627  \n",
       "0.1 50 0.0001 0.1 0.0001    689.702015  \n",
       "0.1 50 0.0001 0.2 0.001     978.228412  \n",
       "0.1 50 0.0001 0.2 0.005    6687.021900  \n",
       "0.1 50 0.0001 0.2 0.0001    689.643832  \n",
       "0.1 100 0.01 0.1 0.001     2071.210029  \n",
       "0.1 100 0.01 0.1 0.005     2645.547310  \n",
       "0.1 100 0.01 0.1 0.0001    1662.687191  \n",
       "0.1 100 0.01 0.2 0.001     2013.965046  \n",
       "0.1 100 0.01 0.2 0.005     2640.737753  \n",
       "0.1 100 0.01 0.2 0.0001    1667.070111  \n",
       "0.1 100 0.001 0.1 0.001     767.203005  \n",
       "0.1 100 0.001 0.1 0.005    7036.163196  \n",
       "0.1 100 0.001 0.1 0.0001    351.872121  \n",
       "0.1 100 0.001 0.2 0.001     575.423601  \n",
       "0.1 100 0.001 0.2 0.005    7044.671999  \n",
       "0.1 100 0.001 0.2 0.0001    351.787471  \n",
       "0.1 100 0.0001 0.1 0.001    990.609190  \n",
       "0.1 100 0.0001 0.1 0.005   6445.522120  \n",
       "0.1 100 0.0001 0.1 0.0001   584.755171  \n",
       "0.1 100 0.0001 0.2 0.001    996.880270  \n",
       "0.1 100 0.0001 0.2 0.005   6468.377538  \n",
       "0.1 100 0.0001 0.2 0.0001   586.708536  \n",
       "0.2 50 0.01 0.1 0.001       635.130060  \n",
       "0.2 50 0.01 0.1 0.005      1207.450799  \n",
       "0.2 50 0.01 0.1 0.0001      890.482353  \n",
       "0.2 50 0.01 0.2 0.001       581.673505  \n",
       "0.2 50 0.01 0.2 0.005      1205.217252  \n",
       "0.2 50 0.01 0.2 0.0001      892.006992  \n",
       "0.2 50 0.001 0.1 0.001      544.123728  \n",
       "0.2 50 0.001 0.1 0.005     6443.108022  \n",
       "0.2 50 0.001 0.1 0.0001     293.753623  \n",
       "0.2 50 0.001 0.2 0.001      525.601341  \n",
       "0.2 50 0.001 0.2 0.005     6515.031295  \n",
       "0.2 50 0.001 0.2 0.0001     311.640808  \n",
       "0.2 50 0.0001 0.1 0.001     925.962744  \n",
       "0.2 50 0.0001 0.1 0.005    5298.622682  \n",
       "0.2 50 0.0001 0.1 0.0001    531.955199  \n",
       "0.2 50 0.0001 0.2 0.001     807.869276  \n",
       "0.2 50 0.0001 0.2 0.005    5434.996939  \n",
       "0.2 50 0.0001 0.2 0.0001    473.910653  \n",
       "0.2 100 0.01 0.1 0.001      736.974611  \n",
       "0.2 100 0.01 0.1 0.005     1089.092600  \n",
       "0.2 100 0.01 0.1 0.0001    1123.453475  \n",
       "0.2 100 0.01 0.2 0.001      613.055591  \n",
       "0.2 100 0.01 0.2 0.005     1086.484207  \n",
       "0.2 100 0.01 0.2 0.0001    1050.114938  \n",
       "0.2 100 0.001 0.1 0.001     574.430000  \n",
       "0.2 100 0.001 0.1 0.005    6192.622718  \n",
       "0.2 100 0.001 0.1 0.0001    310.489978  \n",
       "0.2 100 0.001 0.2 0.001     348.458644  \n",
       "0.2 100 0.001 0.2 0.005    5847.276190  \n",
       "0.2 100 0.001 0.2 0.0001    309.193037  \n",
       "0.2 100 0.0001 0.1 0.001    939.690883  \n",
       "0.2 100 0.0001 0.1 0.005   6024.818873  \n",
       "0.2 100 0.0001 0.1 0.0001   385.924237  \n",
       "0.2 100 0.0001 0.2 0.001    745.608170  \n",
       "0.2 100 0.0001 0.2 0.005   5855.812801  \n",
       "0.2 100 0.0001 0.2 0.0001   381.302760  \n",
       "0.5 50 0.01 0.1 0.001       435.852056  \n",
       "0.5 50 0.01 0.1 0.005       364.866074  \n",
       "0.5 50 0.01 0.1 0.0001      478.792652  \n",
       "0.5 50 0.01 0.2 0.001       248.319818  \n",
       "0.5 50 0.01 0.2 0.005       391.323444  \n",
       "0.5 50 0.01 0.2 0.0001      479.760224  \n",
       "0.5 50 0.001 0.1 0.001      298.228348  \n",
       "0.5 50 0.001 0.1 0.005     2964.001945  \n",
       "0.5 50 0.001 0.1 0.0001     217.928213  \n",
       "0.5 50 0.001 0.2 0.001      100.960557  \n",
       "0.5 50 0.001 0.2 0.005     2813.286191  \n",
       "0.5 50 0.001 0.2 0.0001     219.898248  \n",
       "0.5 50 0.0001 0.1 0.001     620.621142  \n",
       "0.5 50 0.0001 0.1 0.005    2719.386001  \n",
       "0.5 50 0.0001 0.1 0.0001    114.213661  \n",
       "0.5 50 0.0001 0.2 0.001     429.423048  \n",
       "0.5 50 0.0001 0.2 0.005    2788.610445  \n",
       "0.5 50 0.0001 0.2 0.0001    112.210161  \n",
       "0.5 100 0.01 0.1 0.001      626.477681  \n",
       "0.5 100 0.01 0.1 0.005      338.123212  \n",
       "0.5 100 0.01 0.1 0.0001     448.308909  \n",
       "0.5 100 0.01 0.2 0.001      252.080682  \n",
       "0.5 100 0.01 0.2 0.005      338.387936  \n",
       "0.5 100 0.01 0.2 0.0001     470.652769  \n",
       "0.5 100 0.001 0.1 0.001     524.105930  \n",
       "0.5 100 0.001 0.1 0.005    3002.656749  \n",
       "0.5 100 0.001 0.1 0.0001    270.894093  \n",
       "0.5 100 0.001 0.2 0.001     118.750645  \n",
       "0.5 100 0.001 0.2 0.005    2989.542676  \n",
       "0.5 100 0.001 0.2 0.0001    272.812829  \n",
       "0.5 100 0.0001 0.1 0.001    872.970864  \n",
       "0.5 100 0.0001 0.1 0.005   3071.242072  \n",
       "0.5 100 0.0001 0.1 0.0001   109.585345  \n",
       "0.5 100 0.0001 0.2 0.001    412.693623  \n",
       "0.5 100 0.0001 0.2 0.005   2738.859563  \n",
       "0.5 100 0.0001 0.2 0.0001   110.212280  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reshaped_matrix = results_matrix.mean(1).reshape(results_matrix.shape[0],np.prod(results_matrix.shape[2:])).T\n",
    "\n",
    "mean_df = pd.DataFrame(reshaped_matrix, index=name_list, columns=values_to_fetch)\n",
    "mean_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "386e3cf4-ff16-4dfc-95cf-ffed84a675aa",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "std_matrix = results_matrix.std(1).reshape(results_matrix.shape[0],np.prod(results_matrix.shape[2:])).T\n",
    "std_df = pd.DataFrame(std_matrix, index=name_list, columns=values_to_fetch)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d7701299-7b3b-43bc-b64f-52388d7bbd08",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/t7/t2tk8_dd3vsc811yh2mdwbbh0000gp/T/ipykernel_26411/520738583.py:53: UserWarning: This figure includes Axes that are not compatible with tight_layout, so results might be incorrect.\n",
      "  fig.tight_layout()\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x700 with 11 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "index = 0\n",
    "mean_arr = results_matrix.mean(1)[index]\n",
    "key_list = list(hps.keys())[1:]\n",
    "\n",
    "fig, axs = plt.subplots(4, 4, figsize=(7,7))\n",
    "for i in range(0,len(key_list)-1):\n",
    "    for j in range(i+1,len(key_list)-1):\n",
    "        fig.delaxes(axs[i, j])\n",
    "\n",
    "fig.subplots_adjust(right=0.8)\n",
    "\n",
    "cmap = 'Blues'\n",
    "vmax = mean_arr.max()\n",
    "vmin = mean_arr.min()\n",
    "\n",
    "for i in range(0,len(key_list)-1):\n",
    "    for j in range(i+1,len(key_list)):\n",
    "        index_list = list(range(len(key_list)))\n",
    "        index_list.remove(i)\n",
    "        index_list.remove(j)\n",
    "        index_list = index_list[::-1]\n",
    "        new_arr = mean_arr.mean(index_list[0])\n",
    "        for k in range(1,len(index_list)):\n",
    "            new_arr = new_arr.mean(index_list[k])\n",
    "        df = pd.DataFrame(new_arr.T, index=hps[key_list[j]], columns=hps[key_list[i]])\n",
    "        if i != 0:\n",
    "            yticklabels = False\n",
    "        else:\n",
    "            yticklabels = True\n",
    "            axs[j-1,i].set(xlabel=key_list[j])\n",
    "        if j+1 != len(key_list):\n",
    "            xticklabels = False\n",
    "        else:\n",
    "            xticklabels = True\n",
    "            axs[j-1,i].set(ylabel=key_list[i])\n",
    "            \n",
    "        if j == 1 and i == 0:\n",
    "            cbar_ax = fig.add_axes([1, 0.12, 0.05, 0.85])\n",
    "            cbar = True\n",
    "        else:\n",
    "            cbar_ax = None\n",
    "            cbar = False\n",
    "        \n",
    "        sn.heatmap(df, annot=True, ax=axs[j-1,i], cbar=cbar, vmin=vmin, vmax=vmax,\n",
    "                cmap=cmap, xticklabels=xticklabels, yticklabels=yticklabels, cbar_ax=cbar_ax)\n",
    "        if j+1 == len(key_list):\n",
    "            axs[j-1, i].set(xlabel=key_list[i])\n",
    "        if i == 0:\n",
    "            axs[j-1, i].set(ylabel=key_list[j])\n",
    "\n",
    "fig.suptitle(f\"Comparison of the {values_to_fetch[index].replace(\"_\", \" \")} for the dataset {dataset}\")\n",
    "# # Layout so plots do not overlap\n",
    "fig.tight_layout()\n",
    "fig.align_labels()\n",
    "\n",
    "plt.savefig(f\"./results/mnist/{dataset}_{values_to_fetch[index]}_{\"ignore_lr_\" if ignore_lr else \"\"}heatmaps.jpg\",bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "2f59cc53-d214-4de7-8bf5-e9acbc1f9b8f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Validation error: 0.009287539683282376 KL bound 0.11196740996092558\n",
      "The best bound for run 42 happens at 0 datapoints with a bound value of 0.015240831966857101\n",
      "The best bound for run 4 happens at 0 datapoints with a bound value of 0.01383238976994503\n",
      "The best bound for run 3 happens at 0 datapoints with a bound value of 0.015403745304398616\n",
      "The best bound for run 2 happens at 0 datapoints with a bound value of 0.015766104069127068\n",
      "The best bound for run 1 happens at 0 datapoints with a bound value of 0.015396679472358257\n",
      "Mean value of the best bounds taken separately : 0.015127950116537215\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Comparison of the behavior of the bound on mnist')"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from bounds.real_valued_bounds import compute_epsilon, kl_bound\n",
    "from copy import deepcopy\n",
    "\n",
    "new_config = deepcopy(hps)\n",
    "new_config.pop(\"seed\", None)\n",
    "\n",
    "best_params = correct_type_of_entry(mean_df.index[mean_df['validation_error'].argmin()].split())\n",
    "\n",
    "print(\"Validation error:\", mean_df['validation_error'].min(), \"KL bound\", mean_df.loc[mean_df.index[mean_df['validation_error'].argmin()]]['kl_bound'])\n",
    "# model_type = \"cnn\"\n",
    "# training_lr = 0.01\n",
    "# dropout_probability = 0.1\n",
    "\n",
    "# '-', '--', '-.', ':', 'None', ' ', '', 'solid', 'dashed', 'dashdot', 'dotted'\n",
    "# linestyles = [\"solid\", \"dotted\", \"dashed\", \"dashdot\", \"dashdotdotted\"]\n",
    "linestyles = ['-', '--', '-.', ':', (0, (3, 1, 1, 1, 1, 1))]\n",
    "colors = [\"#000000\", \"#56B4E9\", \"#009E73\", \"#0072B2\", \"#D55E00\"]\n",
    "\n",
    "plt.figure(figsize=(8,5))\n",
    "run_list = []\n",
    "\n",
    "for run in runs:\n",
    "    counter = 0\n",
    "    for idx, key_item in enumerate(new_config.items()):\n",
    "        key = key_item[0]\n",
    "        if run.config[key] == best_params[idx]:\n",
    "            counter += 1\n",
    "    if counter == len(new_config.keys()):\n",
    "        run_list.append(run)\n",
    "\n",
    "all_seeds_bounds = []\n",
    "max_length_run = 0\n",
    "max_comp_sizes = []\n",
    "counter = 0\n",
    "for run in run_list:\n",
    "    history = run.scan_history(keys=[\"kl_bound\"])\n",
    "    losses = [row[\"kl_bound\"] for row in history]\n",
    "    val_sizes = run.config['data_groupsize'] * run.config['log_iterations']\n",
    "    compression_sizes = np.array(list(range(0,len(losses)-1))) * val_sizes\n",
    "    compression_sizes = np.append(compression_sizes, run.summary['compression_set_size'])\n",
    "\n",
    "    if compression_sizes.shape[0] > max_length_run:\n",
    "        max_length_run = compression_sizes.shape[0]\n",
    "        max_comp_sizes = compression_sizes\n",
    "        \n",
    "    plt.plot(compression_sizes, losses,'d',\n",
    "             markevery=[np.array(losses).argmin()],\n",
    "             linestyle=linestyles[counter], #ms=8, #mew=1.2,\n",
    "             color=colors[counter],\n",
    "             label=f\"Seed {run.config['seed']} (min bound : {min(losses):.4f})\")\n",
    "    best_idx = np.array(losses).argmin()\n",
    "    print(f\"The best bound for run {run.config['seed']} happens at {compression_sizes[best_idx]} datapoints with a bound value of {losses[best_idx]}\")\n",
    "    all_seeds_bounds.append(losses)\n",
    "    counter += 1\n",
    "\n",
    "print(f\"Mean value of the best bounds taken separately : {sum([min(l) for l in all_seeds_bounds])/5}\")\n",
    "plt.legend(markerscale=0)\n",
    "plt.title(f\"Comparison of the behavior of the bound on {dataset}\")\n",
    "# plt.savefig(f\"./results/{dataset}_min_bound_comparison.jpg\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "1e5761cb-dc18-461b-a07d-afe6e1599e26",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.5, 100.0, 0.001, 0.1, 0.0001]\n",
      "Validation error: 0.0107±0.0010 Test error : 0.0114±0.0007  KL bound : 0.0711±0.0022  binomial : 0.0708±0.0022  P2L bound : 0.0152±0.0006  compression set size : 326.4000±12.8000  Complement loss: 0.0014±0.0006 Validation loss: 0.0535±0.0095 Test loss : 0.0467±0.0056  CE KL bound: 0.8279±0.0256 CE Catoni bound : 0.8269±0.0256  Runtime : 270.8941±57.6181\n"
     ]
    }
   ],
   "source": [
    "val_of_interest = \"kl_bound\"\n",
    "best_params = correct_type_of_entry(mean_df.index[mean_df[val_of_interest].argmin()].split())\n",
    "print(best_params)\n",
    "best_val_arr = mean_df.loc[mean_df.index[mean_df[val_of_interest].argmin()]]\n",
    "std_val_arr = std_df.loc[mean_df.index[mean_df[val_of_interest].argmin()]]\n",
    "print(f\"Validation error: {best_val_arr['validation_error']:.4f}±{std_val_arr['validation_error']:.4f} \"\n",
    "      f'Test error : {best_val_arr['test_error']:.4f}±{std_val_arr['test_error']:.4f} ', \n",
    "      f\"KL bound : {best_val_arr['kl_bound']:.4f}±{std_val_arr['kl_bound']:.4f} \",\n",
    "    f'binomial : {best_val_arr['binomial_approximation_shah']:.4f}±{std_val_arr['binomial_approximation_shah']:.4f} ',\n",
    "     f'P2L bound : {best_val_arr['p2l_bound']:.4f}±{std_val_arr['p2l_bound']:.4f} ', \n",
    "      f'compression set size : {best_val_arr['compression_set_size']:.4f}±{std_val_arr['compression_set_size']:.4f} ',\n",
    "f\"Complement loss: {best_val_arr['complement_loss']:.4f}±{std_val_arr['complement_loss']:.4f} \"\n",
    "    f\"Validation loss: {best_val_arr['validation_loss']:.4f}±{std_val_arr['validation_loss']:.4f} \"\n",
    "      f'Test loss : {best_val_arr['test_loss']:.4f}±{std_val_arr['test_loss']:.4f} ', \n",
    "    f\"CE KL bound: {best_val_arr['CE_kl_bound']:.4f}±{std_val_arr['CE_kl_bound']:.4f} \"\n",
    "      f'CE Catoni bound : {best_val_arr['CE_min_val_catoni_bound']:.4f}±{std_val_arr['CE_min_val_catoni_bound']:.4f} ',\n",
    "    f\"Runtime : {best_val_arr['_runtime']:.4f}±{std_val_arr['_runtime']:.4f}\"\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "66e87958-ef28-451e-81fd-e73c962fea1b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  3%|████▊                                                                                                                                             | 18/540 [02:08<1:01:55,  7.12s/it]"
     ]
    }
   ],
   "source": [
    "from bounds.classical_bounds import binomial_approximation\n",
    "from bounds.bound_utils import zeta, log_binomial_coefficient\n",
    "from time import sleep\n",
    "from tqdm import tqdm\n",
    "\n",
    "new_vals_to_fetch = ['kl_bound',\"compression_set_size\",  \"complement_error\", 'validation_error', 'test_error',]\n",
    "early_stop_result_matrix = np.ones(((len(new_vals_to_fetch)+1,) + size_hyperparams))\n",
    "    \n",
    "for run_idx in tqdm(range(len(runs))):\n",
    "    sleep(5)\n",
    "    run = runs[run_idx]\n",
    "    kl_history = run.scan_history(keys=[\"kl_bound\"])\n",
    "    kl_list = [row[\"kl_bound\"] for row in kl_history]\n",
    "    kl_idx = np.argmin(kl_list)\n",
    "    \n",
    "    val_sizes = run.config['data_groupsize'] * run.config['log_iterations']\n",
    "    compression_sizes = np.array(list(range(1,len(kl_list)))) * val_sizes\n",
    "    compression_sizes = np.append(compression_sizes, run.summary['compression_set_size'])\n",
    "    \n",
    "    for val_to_fetch_idx in range(len(new_vals_to_fetch)):\n",
    "        matrix_idx = tuple([val_to_fetch_idx] + [hps[key].index(run.config[key]) for key in hps.keys()])\n",
    "        val_to_fetch = new_vals_to_fetch[val_to_fetch_idx]\n",
    "        history = run.scan_history(keys=[val_to_fetch])\n",
    "        losses = [row[val_to_fetch] for row in history]\n",
    "        if val_to_fetch == \"validation_error\" and len(losses) == 1:\n",
    "            history = run.scan_history(keys=[\"val_error\"])\n",
    "            losses = [row[\"val_error\"] for row in history]\n",
    "            losses.append(run.summary[\"validation_error\"])\n",
    "        if val_to_fetch == \"compression_set_size\":\n",
    "            losses = compression_sizes\n",
    "        early_stop_result_matrix[matrix_idx] = losses[kl_idx]\n",
    "\n",
    "    matrix_idx = tuple([hps[key].index(run.config[key]) for key in hps.keys()])\n",
    "    complement_error = early_stop_result_matrix[(new_vals_to_fetch.index('complement_error'),) + matrix_idx]\n",
    "    m = early_stop_result_matrix[(new_vals_to_fetch.index('compression_set_size'),) + matrix_idx]\n",
    "    n = run.summary['train_set_size']\n",
    "    k = complement_error * (n-m)\n",
    "    \n",
    "    log_delta = np.log(zeta(m)*run.config['delta']) - log_binomial_coefficient(n,m)\n",
    "    delta_prime = np.exp(log_delta)\n",
    "    early_stop_result_matrix[(-1,) + matrix_idx] = binomial_approximation(k, n-m, log_delta)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "668f3f64-193b-4b10-94e9-77c8dec7bf14",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>kl_bound</th>\n",
       "      <th>compression_set_size</th>\n",
       "      <th>complement_error</th>\n",
       "      <th>validation_error</th>\n",
       "      <th>test_error</th>\n",
       "      <th>binomial_approximation_shah</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.01 0.1 0.001</th>\n",
       "      <td>0.117759</td>\n",
       "      <td>1088.0</td>\n",
       "      <td>0.020230</td>\n",
       "      <td>0.033071</td>\n",
       "      <td>0.032000</td>\n",
       "      <td>0.184228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.01 0.1 0.005</th>\n",
       "      <td>0.131504</td>\n",
       "      <td>896.0</td>\n",
       "      <td>0.032107</td>\n",
       "      <td>0.039915</td>\n",
       "      <td>0.040041</td>\n",
       "      <td>0.213667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.01 0.1 0.0001</th>\n",
       "      <td>0.099341</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.045649</td>\n",
       "      <td>0.050729</td>\n",
       "      <td>0.048380</td>\n",
       "      <td>0.210562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.01 0.2 0.001</th>\n",
       "      <td>0.120586</td>\n",
       "      <td>1088.0</td>\n",
       "      <td>0.021775</td>\n",
       "      <td>0.033424</td>\n",
       "      <td>0.033477</td>\n",
       "      <td>0.189458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.01 0.2 0.005</th>\n",
       "      <td>0.139785</td>\n",
       "      <td>1024.0</td>\n",
       "      <td>0.031543</td>\n",
       "      <td>0.043130</td>\n",
       "      <td>0.042402</td>\n",
       "      <td>0.219838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.01 0.2 0.0001</th>\n",
       "      <td>0.095175</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.042920</td>\n",
       "      <td>0.046858</td>\n",
       "      <td>0.045471</td>\n",
       "      <td>0.203185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.001 0.1 0.001</th>\n",
       "      <td>0.074848</td>\n",
       "      <td>448.0</td>\n",
       "      <td>0.020879</td>\n",
       "      <td>0.025813</td>\n",
       "      <td>0.025093</td>\n",
       "      <td>0.144814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.001 0.1 0.005</th>\n",
       "      <td>0.115413</td>\n",
       "      <td>768.0</td>\n",
       "      <td>0.027059</td>\n",
       "      <td>0.033292</td>\n",
       "      <td>0.031774</td>\n",
       "      <td>0.188517</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.001 0.1 0.0001</th>\n",
       "      <td>0.063331</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.021613</td>\n",
       "      <td>0.024065</td>\n",
       "      <td>0.024659</td>\n",
       "      <td>0.137265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.001 0.2 0.001</th>\n",
       "      <td>0.074606</td>\n",
       "      <td>384.0</td>\n",
       "      <td>0.024325</td>\n",
       "      <td>0.029256</td>\n",
       "      <td>0.027263</td>\n",
       "      <td>0.151583</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.001 0.2 0.005</th>\n",
       "      <td>0.118040</td>\n",
       "      <td>832.0</td>\n",
       "      <td>0.024853</td>\n",
       "      <td>0.029658</td>\n",
       "      <td>0.030336</td>\n",
       "      <td>0.187450</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.001 0.2 0.0001</th>\n",
       "      <td>0.061137</td>\n",
       "      <td>384.0</td>\n",
       "      <td>0.017363</td>\n",
       "      <td>0.021699</td>\n",
       "      <td>0.021958</td>\n",
       "      <td>0.126203</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.0001 0.1 0.001</th>\n",
       "      <td>0.103938</td>\n",
       "      <td>896.0</td>\n",
       "      <td>0.017954</td>\n",
       "      <td>0.024439</td>\n",
       "      <td>0.024254</td>\n",
       "      <td>0.166301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.0001 0.1 0.005</th>\n",
       "      <td>0.434681</td>\n",
       "      <td>8966.4</td>\n",
       "      <td>0.197191</td>\n",
       "      <td>0.379975</td>\n",
       "      <td>0.379509</td>\n",
       "      <td>0.496831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.0001 0.1 0.0001</th>\n",
       "      <td>0.094306</td>\n",
       "      <td>768.0</td>\n",
       "      <td>0.018195</td>\n",
       "      <td>0.027616</td>\n",
       "      <td>0.025940</td>\n",
       "      <td>0.157760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.0001 0.2 0.001</th>\n",
       "      <td>0.105585</td>\n",
       "      <td>832.0</td>\n",
       "      <td>0.020727</td>\n",
       "      <td>0.028009</td>\n",
       "      <td>0.027019</td>\n",
       "      <td>0.172244</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.0001 0.2 0.005</th>\n",
       "      <td>0.130022</td>\n",
       "      <td>896.0</td>\n",
       "      <td>0.029468</td>\n",
       "      <td>0.036109</td>\n",
       "      <td>0.036797</td>\n",
       "      <td>0.206223</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 50 0.0001 0.2 0.0001</th>\n",
       "      <td>0.092215</td>\n",
       "      <td>640.0</td>\n",
       "      <td>0.021686</td>\n",
       "      <td>0.030890</td>\n",
       "      <td>0.029533</td>\n",
       "      <td>0.162008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.01 0.1 0.001</th>\n",
       "      <td>0.115455</td>\n",
       "      <td>1152.0</td>\n",
       "      <td>0.015602</td>\n",
       "      <td>0.028874</td>\n",
       "      <td>0.027850</td>\n",
       "      <td>0.173035</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.01 0.1 0.005</th>\n",
       "      <td>0.153907</td>\n",
       "      <td>1580.8</td>\n",
       "      <td>0.026014</td>\n",
       "      <td>0.035468</td>\n",
       "      <td>0.034512</td>\n",
       "      <td>0.224563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.01 0.1 0.0001</th>\n",
       "      <td>0.077922</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.039945</td>\n",
       "      <td>0.044340</td>\n",
       "      <td>0.040503</td>\n",
       "      <td>0.192964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.01 0.2 0.001</th>\n",
       "      <td>0.088128</td>\n",
       "      <td>832.0</td>\n",
       "      <td>0.021305</td>\n",
       "      <td>0.031189</td>\n",
       "      <td>0.029099</td>\n",
       "      <td>0.171586</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.01 0.2 0.005</th>\n",
       "      <td>0.110550</td>\n",
       "      <td>832.0</td>\n",
       "      <td>0.031997</td>\n",
       "      <td>0.037209</td>\n",
       "      <td>0.035259</td>\n",
       "      <td>0.205906</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.01 0.2 0.0001</th>\n",
       "      <td>0.070988</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.035324</td>\n",
       "      <td>0.038762</td>\n",
       "      <td>0.036697</td>\n",
       "      <td>0.180868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.001 0.1 0.001</th>\n",
       "      <td>0.053247</td>\n",
       "      <td>384.0</td>\n",
       "      <td>0.019118</td>\n",
       "      <td>0.021544</td>\n",
       "      <td>0.021985</td>\n",
       "      <td>0.133308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.001 0.1 0.005</th>\n",
       "      <td>0.077241</td>\n",
       "      <td>384.0</td>\n",
       "      <td>0.032880</td>\n",
       "      <td>0.035428</td>\n",
       "      <td>0.035334</td>\n",
       "      <td>0.175209</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.001 0.1 0.0001</th>\n",
       "      <td>0.048347</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.019735</td>\n",
       "      <td>0.022877</td>\n",
       "      <td>0.021842</td>\n",
       "      <td>0.130816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.001 0.2 0.001</th>\n",
       "      <td>0.054019</td>\n",
       "      <td>448.0</td>\n",
       "      <td>0.016047</td>\n",
       "      <td>0.019832</td>\n",
       "      <td>0.019095</td>\n",
       "      <td>0.127615</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.001 0.2 0.005</th>\n",
       "      <td>0.083390</td>\n",
       "      <td>576.0</td>\n",
       "      <td>0.025677</td>\n",
       "      <td>0.030169</td>\n",
       "      <td>0.028686</td>\n",
       "      <td>0.168592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.001 0.2 0.0001</th>\n",
       "      <td>0.048234</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.019672</td>\n",
       "      <td>0.022537</td>\n",
       "      <td>0.021441</td>\n",
       "      <td>0.130596</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.0001 0.1 0.001</th>\n",
       "      <td>0.079521</td>\n",
       "      <td>704.0</td>\n",
       "      <td>0.021131</td>\n",
       "      <td>0.025157</td>\n",
       "      <td>0.025017</td>\n",
       "      <td>0.162885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.0001 0.1 0.005</th>\n",
       "      <td>0.250623</td>\n",
       "      <td>9120.0</td>\n",
       "      <td>0.022673</td>\n",
       "      <td>0.204468</td>\n",
       "      <td>0.203970</td>\n",
       "      <td>0.334760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.0001 0.1 0.0001</th>\n",
       "      <td>0.068763</td>\n",
       "      <td>576.0</td>\n",
       "      <td>0.020260</td>\n",
       "      <td>0.026051</td>\n",
       "      <td>0.024451</td>\n",
       "      <td>0.151349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.0001 0.2 0.001</th>\n",
       "      <td>0.078986</td>\n",
       "      <td>576.0</td>\n",
       "      <td>0.024602</td>\n",
       "      <td>0.028255</td>\n",
       "      <td>0.027769</td>\n",
       "      <td>0.167891</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.0001 0.2 0.005</th>\n",
       "      <td>0.093192</td>\n",
       "      <td>768.0</td>\n",
       "      <td>0.024512</td>\n",
       "      <td>0.028134</td>\n",
       "      <td>0.026632</td>\n",
       "      <td>0.180173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.1 100 0.0001 0.2 0.0001</th>\n",
       "      <td>0.069466</td>\n",
       "      <td>512.0</td>\n",
       "      <td>0.022938</td>\n",
       "      <td>0.026961</td>\n",
       "      <td>0.025552</td>\n",
       "      <td>0.156919</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.01 0.1 0.001</th>\n",
       "      <td>0.074226</td>\n",
       "      <td>640.0</td>\n",
       "      <td>0.017006</td>\n",
       "      <td>0.026047</td>\n",
       "      <td>0.024949</td>\n",
       "      <td>0.151431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.01 0.1 0.005</th>\n",
       "      <td>0.081996</td>\n",
       "      <td>512.0</td>\n",
       "      <td>0.025804</td>\n",
       "      <td>0.031804</td>\n",
       "      <td>0.031117</td>\n",
       "      <td>0.171902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.01 0.1 0.0001</th>\n",
       "      <td>0.063362</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.027577</td>\n",
       "      <td>0.032627</td>\n",
       "      <td>0.030468</td>\n",
       "      <td>0.161078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.01 0.2 0.001</th>\n",
       "      <td>0.072816</td>\n",
       "      <td>576.0</td>\n",
       "      <td>0.019100</td>\n",
       "      <td>0.027769</td>\n",
       "      <td>0.026206</td>\n",
       "      <td>0.153792</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.01 0.2 0.005</th>\n",
       "      <td>0.079226</td>\n",
       "      <td>448.0</td>\n",
       "      <td>0.027497</td>\n",
       "      <td>0.033653</td>\n",
       "      <td>0.032742</td>\n",
       "      <td>0.172380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.01 0.2 0.0001</th>\n",
       "      <td>0.059263</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.025152</td>\n",
       "      <td>0.030028</td>\n",
       "      <td>0.028848</td>\n",
       "      <td>0.153550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.001 0.1 0.001</th>\n",
       "      <td>0.046822</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.016758</td>\n",
       "      <td>0.018892</td>\n",
       "      <td>0.017517</td>\n",
       "      <td>0.124174</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.001 0.1 0.005</th>\n",
       "      <td>0.075922</td>\n",
       "      <td>448.0</td>\n",
       "      <td>0.024086</td>\n",
       "      <td>0.027144</td>\n",
       "      <td>0.027854</td>\n",
       "      <td>0.160090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.001 0.1 0.0001</th>\n",
       "      <td>0.041409</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.014130</td>\n",
       "      <td>0.017170</td>\n",
       "      <td>0.016325</td>\n",
       "      <td>0.113712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.001 0.2 0.001</th>\n",
       "      <td>0.046743</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.016715</td>\n",
       "      <td>0.019835</td>\n",
       "      <td>0.018267</td>\n",
       "      <td>0.124023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.001 0.2 0.005</th>\n",
       "      <td>0.077265</td>\n",
       "      <td>512.0</td>\n",
       "      <td>0.021416</td>\n",
       "      <td>0.025655</td>\n",
       "      <td>0.024096</td>\n",
       "      <td>0.157148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.001 0.2 0.0001</th>\n",
       "      <td>0.041763</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.014291</td>\n",
       "      <td>0.017665</td>\n",
       "      <td>0.016088</td>\n",
       "      <td>0.114398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.0001 0.1 0.001</th>\n",
       "      <td>0.073441</td>\n",
       "      <td>576.0</td>\n",
       "      <td>0.016952</td>\n",
       "      <td>0.020909</td>\n",
       "      <td>0.021963</td>\n",
       "      <td>0.147841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.0001 0.1 0.005</th>\n",
       "      <td>0.091421</td>\n",
       "      <td>832.0</td>\n",
       "      <td>0.015993</td>\n",
       "      <td>0.021185</td>\n",
       "      <td>0.021990</td>\n",
       "      <td>0.162769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.0001 0.1 0.0001</th>\n",
       "      <td>0.053684</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.020802</td>\n",
       "      <td>0.023269</td>\n",
       "      <td>0.024131</td>\n",
       "      <td>0.138587</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.0001 0.2 0.001</th>\n",
       "      <td>0.077219</td>\n",
       "      <td>704.0</td>\n",
       "      <td>0.014637</td>\n",
       "      <td>0.019359</td>\n",
       "      <td>0.019744</td>\n",
       "      <td>0.147829</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.0001 0.2 0.005</th>\n",
       "      <td>0.091659</td>\n",
       "      <td>704.0</td>\n",
       "      <td>0.020296</td>\n",
       "      <td>0.025005</td>\n",
       "      <td>0.025220</td>\n",
       "      <td>0.168678</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 50 0.0001 0.2 0.0001</th>\n",
       "      <td>0.054778</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.021421</td>\n",
       "      <td>0.023945</td>\n",
       "      <td>0.023825</td>\n",
       "      <td>0.140636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.01 0.1 0.001</th>\n",
       "      <td>0.078088</td>\n",
       "      <td>704.0</td>\n",
       "      <td>0.017407</td>\n",
       "      <td>0.026386</td>\n",
       "      <td>0.026481</td>\n",
       "      <td>0.156762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.01 0.1 0.005</th>\n",
       "      <td>0.082902</td>\n",
       "      <td>512.0</td>\n",
       "      <td>0.026878</td>\n",
       "      <td>0.032178</td>\n",
       "      <td>0.032803</td>\n",
       "      <td>0.175797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.01 0.1 0.0001</th>\n",
       "      <td>0.065651</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.029503</td>\n",
       "      <td>0.033482</td>\n",
       "      <td>0.032604</td>\n",
       "      <td>0.167333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.01 0.2 0.001</th>\n",
       "      <td>0.077905</td>\n",
       "      <td>704.0</td>\n",
       "      <td>0.017416</td>\n",
       "      <td>0.026693</td>\n",
       "      <td>0.025377</td>\n",
       "      <td>0.156542</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.01 0.2 0.005</th>\n",
       "      <td>0.082868</td>\n",
       "      <td>512.0</td>\n",
       "      <td>0.026646</td>\n",
       "      <td>0.032819</td>\n",
       "      <td>0.033032</td>\n",
       "      <td>0.175553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.01 0.2 0.0001</th>\n",
       "      <td>0.063622</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.028280</td>\n",
       "      <td>0.031574</td>\n",
       "      <td>0.031647</td>\n",
       "      <td>0.163643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.001 0.1 0.001</th>\n",
       "      <td>0.047163</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.017683</td>\n",
       "      <td>0.020425</td>\n",
       "      <td>0.020581</td>\n",
       "      <td>0.127410</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.001 0.1 0.005</th>\n",
       "      <td>0.075702</td>\n",
       "      <td>512.0</td>\n",
       "      <td>0.021584</td>\n",
       "      <td>0.026281</td>\n",
       "      <td>0.025964</td>\n",
       "      <td>0.157938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.001 0.1 0.0001</th>\n",
       "      <td>0.043795</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.016057</td>\n",
       "      <td>0.019011</td>\n",
       "      <td>0.019608</td>\n",
       "      <td>0.120897</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.001 0.2 0.001</th>\n",
       "      <td>0.046073</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.017133</td>\n",
       "      <td>0.019996</td>\n",
       "      <td>0.020292</td>\n",
       "      <td>0.125309</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.001 0.2 0.005</th>\n",
       "      <td>0.073085</td>\n",
       "      <td>384.0</td>\n",
       "      <td>0.027903</td>\n",
       "      <td>0.028779</td>\n",
       "      <td>0.029734</td>\n",
       "      <td>0.166416</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.001 0.2 0.0001</th>\n",
       "      <td>0.044147</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.016214</td>\n",
       "      <td>0.019623</td>\n",
       "      <td>0.019772</td>\n",
       "      <td>0.121581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.0001 0.1 0.001</th>\n",
       "      <td>0.069967</td>\n",
       "      <td>576.0</td>\n",
       "      <td>0.015340</td>\n",
       "      <td>0.018807</td>\n",
       "      <td>0.020834</td>\n",
       "      <td>0.141800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.0001 0.1 0.005</th>\n",
       "      <td>0.090582</td>\n",
       "      <td>576.0</td>\n",
       "      <td>0.025278</td>\n",
       "      <td>0.027576</td>\n",
       "      <td>0.028495</td>\n",
       "      <td>0.173918</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.0001 0.1 0.0001</th>\n",
       "      <td>0.048299</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.017775</td>\n",
       "      <td>0.020633</td>\n",
       "      <td>0.020684</td>\n",
       "      <td>0.127912</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.0001 0.2 0.001</th>\n",
       "      <td>0.072773</td>\n",
       "      <td>576.0</td>\n",
       "      <td>0.016511</td>\n",
       "      <td>0.019593</td>\n",
       "      <td>0.021217</td>\n",
       "      <td>0.146239</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.0001 0.2 0.005</th>\n",
       "      <td>0.090854</td>\n",
       "      <td>768.0</td>\n",
       "      <td>0.017582</td>\n",
       "      <td>0.022205</td>\n",
       "      <td>0.023084</td>\n",
       "      <td>0.164044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.2 100 0.0001 0.2 0.0001</th>\n",
       "      <td>0.049633</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.018505</td>\n",
       "      <td>0.020876</td>\n",
       "      <td>0.021567</td>\n",
       "      <td>0.130437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.01 0.1 0.001</th>\n",
       "      <td>0.063411</td>\n",
       "      <td>384.0</td>\n",
       "      <td>0.013362</td>\n",
       "      <td>0.019731</td>\n",
       "      <td>0.021220</td>\n",
       "      <td>0.142522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.01 0.1 0.005</th>\n",
       "      <td>0.066313</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.019430</td>\n",
       "      <td>0.022816</td>\n",
       "      <td>0.024583</td>\n",
       "      <td>0.155215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.01 0.1 0.0001</th>\n",
       "      <td>0.055711</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.014035</td>\n",
       "      <td>0.019371</td>\n",
       "      <td>0.021469</td>\n",
       "      <td>0.137028</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.01 0.2 0.001</th>\n",
       "      <td>0.065006</td>\n",
       "      <td>384.0</td>\n",
       "      <td>0.013543</td>\n",
       "      <td>0.020468</td>\n",
       "      <td>0.021380</td>\n",
       "      <td>0.144428</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.01 0.2 0.005</th>\n",
       "      <td>0.070269</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.021997</td>\n",
       "      <td>0.026869</td>\n",
       "      <td>0.028433</td>\n",
       "      <td>0.161395</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.01 0.2 0.0001</th>\n",
       "      <td>0.053471</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.013089</td>\n",
       "      <td>0.018939</td>\n",
       "      <td>0.020692</td>\n",
       "      <td>0.133117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.001 0.1 0.001</th>\n",
       "      <td>0.039792</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.006314</td>\n",
       "      <td>0.009608</td>\n",
       "      <td>0.011236</td>\n",
       "      <td>0.104269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.001 0.1 0.005</th>\n",
       "      <td>0.063547</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.016983</td>\n",
       "      <td>0.017795</td>\n",
       "      <td>0.021487</td>\n",
       "      <td>0.145166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.001 0.1 0.0001</th>\n",
       "      <td>0.035113</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.005142</td>\n",
       "      <td>0.009195</td>\n",
       "      <td>0.010779</td>\n",
       "      <td>0.096942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.001 0.2 0.001</th>\n",
       "      <td>0.039009</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.006082</td>\n",
       "      <td>0.009134</td>\n",
       "      <td>0.011464</td>\n",
       "      <td>0.102977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.001 0.2 0.005</th>\n",
       "      <td>0.058253</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.014178</td>\n",
       "      <td>0.017497</td>\n",
       "      <td>0.019446</td>\n",
       "      <td>0.136166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.001 0.2 0.0001</th>\n",
       "      <td>0.035683</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.005242</td>\n",
       "      <td>0.008857</td>\n",
       "      <td>0.010673</td>\n",
       "      <td>0.097737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.0001 0.1 0.001</th>\n",
       "      <td>0.068864</td>\n",
       "      <td>448.0</td>\n",
       "      <td>0.009357</td>\n",
       "      <td>0.014876</td>\n",
       "      <td>0.016578</td>\n",
       "      <td>0.137071</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.0001 0.1 0.005</th>\n",
       "      <td>0.097264</td>\n",
       "      <td>576.0</td>\n",
       "      <td>0.015587</td>\n",
       "      <td>0.020369</td>\n",
       "      <td>0.022964</td>\n",
       "      <td>0.170995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.0001 0.1 0.0001</th>\n",
       "      <td>0.039304</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.006121</td>\n",
       "      <td>0.011401</td>\n",
       "      <td>0.013548</td>\n",
       "      <td>0.103198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.0001 0.2 0.001</th>\n",
       "      <td>0.063186</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.016758</td>\n",
       "      <td>0.020169</td>\n",
       "      <td>0.021577</td>\n",
       "      <td>0.144297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.0001 0.2 0.005</th>\n",
       "      <td>0.098811</td>\n",
       "      <td>576.0</td>\n",
       "      <td>0.014880</td>\n",
       "      <td>0.021575</td>\n",
       "      <td>0.022936</td>\n",
       "      <td>0.171363</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 50 0.0001 0.2 0.0001</th>\n",
       "      <td>0.039431</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.006159</td>\n",
       "      <td>0.011121</td>\n",
       "      <td>0.012887</td>\n",
       "      <td>0.103408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.01 0.1 0.001</th>\n",
       "      <td>0.072984</td>\n",
       "      <td>448.0</td>\n",
       "      <td>0.013637</td>\n",
       "      <td>0.022207</td>\n",
       "      <td>0.024129</td>\n",
       "      <td>0.154063</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.01 0.1 0.005</th>\n",
       "      <td>0.072826</td>\n",
       "      <td>448.0</td>\n",
       "      <td>0.013586</td>\n",
       "      <td>0.021167</td>\n",
       "      <td>0.024339</td>\n",
       "      <td>0.153824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.01 0.1 0.0001</th>\n",
       "      <td>0.061265</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.017713</td>\n",
       "      <td>0.022762</td>\n",
       "      <td>0.026238</td>\n",
       "      <td>0.150752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.01 0.2 0.001</th>\n",
       "      <td>0.069900</td>\n",
       "      <td>384.0</td>\n",
       "      <td>0.016867</td>\n",
       "      <td>0.023235</td>\n",
       "      <td>0.026334</td>\n",
       "      <td>0.156846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.01 0.2 0.005</th>\n",
       "      <td>0.073239</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.024316</td>\n",
       "      <td>0.027808</td>\n",
       "      <td>0.030572</td>\n",
       "      <td>0.170788</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.01 0.2 0.0001</th>\n",
       "      <td>0.056883</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.015655</td>\n",
       "      <td>0.021295</td>\n",
       "      <td>0.024752</td>\n",
       "      <td>0.143140</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.001 0.1 0.001</th>\n",
       "      <td>0.037341</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.005388</td>\n",
       "      <td>0.008861</td>\n",
       "      <td>0.011760</td>\n",
       "      <td>0.099429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.001 0.1 0.005</th>\n",
       "      <td>0.055987</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.012888</td>\n",
       "      <td>0.016313</td>\n",
       "      <td>0.018238</td>\n",
       "      <td>0.131466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.001 0.1 0.0001</th>\n",
       "      <td>0.034969</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.004842</td>\n",
       "      <td>0.008955</td>\n",
       "      <td>0.011496</td>\n",
       "      <td>0.095795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.001 0.2 0.001</th>\n",
       "      <td>0.037566</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.005442</td>\n",
       "      <td>0.009087</td>\n",
       "      <td>0.011469</td>\n",
       "      <td>0.099780</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.001 0.2 0.005</th>\n",
       "      <td>0.050248</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.010170</td>\n",
       "      <td>0.012786</td>\n",
       "      <td>0.015312</td>\n",
       "      <td>0.121501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.001 0.2 0.0001</th>\n",
       "      <td>0.035325</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.004907</td>\n",
       "      <td>0.009032</td>\n",
       "      <td>0.011698</td>\n",
       "      <td>0.096302</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.0001 0.1 0.001</th>\n",
       "      <td>0.056399</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.013009</td>\n",
       "      <td>0.015225</td>\n",
       "      <td>0.017534</td>\n",
       "      <td>0.131493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.0001 0.1 0.005</th>\n",
       "      <td>0.101803</td>\n",
       "      <td>640.0</td>\n",
       "      <td>0.012740</td>\n",
       "      <td>0.019806</td>\n",
       "      <td>0.021231</td>\n",
       "      <td>0.170578</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.0001 0.1 0.0001</th>\n",
       "      <td>0.037800</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.005339</td>\n",
       "      <td>0.010347</td>\n",
       "      <td>0.012220</td>\n",
       "      <td>0.099455</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.0001 0.2 0.001</th>\n",
       "      <td>0.054422</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.012023</td>\n",
       "      <td>0.015056</td>\n",
       "      <td>0.017081</td>\n",
       "      <td>0.128085</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.0001 0.2 0.005</th>\n",
       "      <td>0.098353</td>\n",
       "      <td>576.0</td>\n",
       "      <td>0.014392</td>\n",
       "      <td>0.020508</td>\n",
       "      <td>0.022418</td>\n",
       "      <td>0.169332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5 100 0.0001 0.2 0.0001</th>\n",
       "      <td>0.038264</td>\n",
       "      <td>320.0</td>\n",
       "      <td>0.005492</td>\n",
       "      <td>0.010279</td>\n",
       "      <td>0.012224</td>\n",
       "      <td>0.100241</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           kl_bound  compression_set_size  complement_error  \\\n",
       "0.1 50 0.01 0.1 0.001      0.117759                1088.0          0.020230   \n",
       "0.1 50 0.01 0.1 0.005      0.131504                 896.0          0.032107   \n",
       "0.1 50 0.01 0.1 0.0001     0.099341                 320.0          0.045649   \n",
       "0.1 50 0.01 0.2 0.001      0.120586                1088.0          0.021775   \n",
       "0.1 50 0.01 0.2 0.005      0.139785                1024.0          0.031543   \n",
       "0.1 50 0.01 0.2 0.0001     0.095175                 320.0          0.042920   \n",
       "0.1 50 0.001 0.1 0.001     0.074848                 448.0          0.020879   \n",
       "0.1 50 0.001 0.1 0.005     0.115413                 768.0          0.027059   \n",
       "0.1 50 0.001 0.1 0.0001    0.063331                 320.0          0.021613   \n",
       "0.1 50 0.001 0.2 0.001     0.074606                 384.0          0.024325   \n",
       "0.1 50 0.001 0.2 0.005     0.118040                 832.0          0.024853   \n",
       "0.1 50 0.001 0.2 0.0001    0.061137                 384.0          0.017363   \n",
       "0.1 50 0.0001 0.1 0.001    0.103938                 896.0          0.017954   \n",
       "0.1 50 0.0001 0.1 0.005    0.434681                8966.4          0.197191   \n",
       "0.1 50 0.0001 0.1 0.0001   0.094306                 768.0          0.018195   \n",
       "0.1 50 0.0001 0.2 0.001    0.105585                 832.0          0.020727   \n",
       "0.1 50 0.0001 0.2 0.005    0.130022                 896.0          0.029468   \n",
       "0.1 50 0.0001 0.2 0.0001   0.092215                 640.0          0.021686   \n",
       "0.1 100 0.01 0.1 0.001     0.115455                1152.0          0.015602   \n",
       "0.1 100 0.01 0.1 0.005     0.153907                1580.8          0.026014   \n",
       "0.1 100 0.01 0.1 0.0001    0.077922                 320.0          0.039945   \n",
       "0.1 100 0.01 0.2 0.001     0.088128                 832.0          0.021305   \n",
       "0.1 100 0.01 0.2 0.005     0.110550                 832.0          0.031997   \n",
       "0.1 100 0.01 0.2 0.0001    0.070988                 320.0          0.035324   \n",
       "0.1 100 0.001 0.1 0.001    0.053247                 384.0          0.019118   \n",
       "0.1 100 0.001 0.1 0.005    0.077241                 384.0          0.032880   \n",
       "0.1 100 0.001 0.1 0.0001   0.048347                 320.0          0.019735   \n",
       "0.1 100 0.001 0.2 0.001    0.054019                 448.0          0.016047   \n",
       "0.1 100 0.001 0.2 0.005    0.083390                 576.0          0.025677   \n",
       "0.1 100 0.001 0.2 0.0001   0.048234                 320.0          0.019672   \n",
       "0.1 100 0.0001 0.1 0.001   0.079521                 704.0          0.021131   \n",
       "0.1 100 0.0001 0.1 0.005   0.250623                9120.0          0.022673   \n",
       "0.1 100 0.0001 0.1 0.0001  0.068763                 576.0          0.020260   \n",
       "0.1 100 0.0001 0.2 0.001   0.078986                 576.0          0.024602   \n",
       "0.1 100 0.0001 0.2 0.005   0.093192                 768.0          0.024512   \n",
       "0.1 100 0.0001 0.2 0.0001  0.069466                 512.0          0.022938   \n",
       "0.2 50 0.01 0.1 0.001      0.074226                 640.0          0.017006   \n",
       "0.2 50 0.01 0.1 0.005      0.081996                 512.0          0.025804   \n",
       "0.2 50 0.01 0.1 0.0001     0.063362                 320.0          0.027577   \n",
       "0.2 50 0.01 0.2 0.001      0.072816                 576.0          0.019100   \n",
       "0.2 50 0.01 0.2 0.005      0.079226                 448.0          0.027497   \n",
       "0.2 50 0.01 0.2 0.0001     0.059263                 320.0          0.025152   \n",
       "0.2 50 0.001 0.1 0.001     0.046822                 320.0          0.016758   \n",
       "0.2 50 0.001 0.1 0.005     0.075922                 448.0          0.024086   \n",
       "0.2 50 0.001 0.1 0.0001    0.041409                 320.0          0.014130   \n",
       "0.2 50 0.001 0.2 0.001     0.046743                 320.0          0.016715   \n",
       "0.2 50 0.001 0.2 0.005     0.077265                 512.0          0.021416   \n",
       "0.2 50 0.001 0.2 0.0001    0.041763                 320.0          0.014291   \n",
       "0.2 50 0.0001 0.1 0.001    0.073441                 576.0          0.016952   \n",
       "0.2 50 0.0001 0.1 0.005    0.091421                 832.0          0.015993   \n",
       "0.2 50 0.0001 0.1 0.0001   0.053684                 320.0          0.020802   \n",
       "0.2 50 0.0001 0.2 0.001    0.077219                 704.0          0.014637   \n",
       "0.2 50 0.0001 0.2 0.005    0.091659                 704.0          0.020296   \n",
       "0.2 50 0.0001 0.2 0.0001   0.054778                 320.0          0.021421   \n",
       "0.2 100 0.01 0.1 0.001     0.078088                 704.0          0.017407   \n",
       "0.2 100 0.01 0.1 0.005     0.082902                 512.0          0.026878   \n",
       "0.2 100 0.01 0.1 0.0001    0.065651                 320.0          0.029503   \n",
       "0.2 100 0.01 0.2 0.001     0.077905                 704.0          0.017416   \n",
       "0.2 100 0.01 0.2 0.005     0.082868                 512.0          0.026646   \n",
       "0.2 100 0.01 0.2 0.0001    0.063622                 320.0          0.028280   \n",
       "0.2 100 0.001 0.1 0.001    0.047163                 320.0          0.017683   \n",
       "0.2 100 0.001 0.1 0.005    0.075702                 512.0          0.021584   \n",
       "0.2 100 0.001 0.1 0.0001   0.043795                 320.0          0.016057   \n",
       "0.2 100 0.001 0.2 0.001    0.046073                 320.0          0.017133   \n",
       "0.2 100 0.001 0.2 0.005    0.073085                 384.0          0.027903   \n",
       "0.2 100 0.001 0.2 0.0001   0.044147                 320.0          0.016214   \n",
       "0.2 100 0.0001 0.1 0.001   0.069967                 576.0          0.015340   \n",
       "0.2 100 0.0001 0.1 0.005   0.090582                 576.0          0.025278   \n",
       "0.2 100 0.0001 0.1 0.0001  0.048299                 320.0          0.017775   \n",
       "0.2 100 0.0001 0.2 0.001   0.072773                 576.0          0.016511   \n",
       "0.2 100 0.0001 0.2 0.005   0.090854                 768.0          0.017582   \n",
       "0.2 100 0.0001 0.2 0.0001  0.049633                 320.0          0.018505   \n",
       "0.5 50 0.01 0.1 0.001      0.063411                 384.0          0.013362   \n",
       "0.5 50 0.01 0.1 0.005      0.066313                 320.0          0.019430   \n",
       "0.5 50 0.01 0.1 0.0001     0.055711                 320.0          0.014035   \n",
       "0.5 50 0.01 0.2 0.001      0.065006                 384.0          0.013543   \n",
       "0.5 50 0.01 0.2 0.005      0.070269                 320.0          0.021997   \n",
       "0.5 50 0.01 0.2 0.0001     0.053471                 320.0          0.013089   \n",
       "0.5 50 0.001 0.1 0.001     0.039792                 320.0          0.006314   \n",
       "0.5 50 0.001 0.1 0.005     0.063547                 320.0          0.016983   \n",
       "0.5 50 0.001 0.1 0.0001    0.035113                 320.0          0.005142   \n",
       "0.5 50 0.001 0.2 0.001     0.039009                 320.0          0.006082   \n",
       "0.5 50 0.001 0.2 0.005     0.058253                 320.0          0.014178   \n",
       "0.5 50 0.001 0.2 0.0001    0.035683                 320.0          0.005242   \n",
       "0.5 50 0.0001 0.1 0.001    0.068864                 448.0          0.009357   \n",
       "0.5 50 0.0001 0.1 0.005    0.097264                 576.0          0.015587   \n",
       "0.5 50 0.0001 0.1 0.0001   0.039304                 320.0          0.006121   \n",
       "0.5 50 0.0001 0.2 0.001    0.063186                 320.0          0.016758   \n",
       "0.5 50 0.0001 0.2 0.005    0.098811                 576.0          0.014880   \n",
       "0.5 50 0.0001 0.2 0.0001   0.039431                 320.0          0.006159   \n",
       "0.5 100 0.01 0.1 0.001     0.072984                 448.0          0.013637   \n",
       "0.5 100 0.01 0.1 0.005     0.072826                 448.0          0.013586   \n",
       "0.5 100 0.01 0.1 0.0001    0.061265                 320.0          0.017713   \n",
       "0.5 100 0.01 0.2 0.001     0.069900                 384.0          0.016867   \n",
       "0.5 100 0.01 0.2 0.005     0.073239                 320.0          0.024316   \n",
       "0.5 100 0.01 0.2 0.0001    0.056883                 320.0          0.015655   \n",
       "0.5 100 0.001 0.1 0.001    0.037341                 320.0          0.005388   \n",
       "0.5 100 0.001 0.1 0.005    0.055987                 320.0          0.012888   \n",
       "0.5 100 0.001 0.1 0.0001   0.034969                 320.0          0.004842   \n",
       "0.5 100 0.001 0.2 0.001    0.037566                 320.0          0.005442   \n",
       "0.5 100 0.001 0.2 0.005    0.050248                 320.0          0.010170   \n",
       "0.5 100 0.001 0.2 0.0001   0.035325                 320.0          0.004907   \n",
       "0.5 100 0.0001 0.1 0.001   0.056399                 320.0          0.013009   \n",
       "0.5 100 0.0001 0.1 0.005   0.101803                 640.0          0.012740   \n",
       "0.5 100 0.0001 0.1 0.0001  0.037800                 320.0          0.005339   \n",
       "0.5 100 0.0001 0.2 0.001   0.054422                 320.0          0.012023   \n",
       "0.5 100 0.0001 0.2 0.005   0.098353                 576.0          0.014392   \n",
       "0.5 100 0.0001 0.2 0.0001  0.038264                 320.0          0.005492   \n",
       "\n",
       "                           validation_error  test_error  \\\n",
       "0.1 50 0.01 0.1 0.001              0.033071    0.032000   \n",
       "0.1 50 0.01 0.1 0.005              0.039915    0.040041   \n",
       "0.1 50 0.01 0.1 0.0001             0.050729    0.048380   \n",
       "0.1 50 0.01 0.2 0.001              0.033424    0.033477   \n",
       "0.1 50 0.01 0.2 0.005              0.043130    0.042402   \n",
       "0.1 50 0.01 0.2 0.0001             0.046858    0.045471   \n",
       "0.1 50 0.001 0.1 0.001             0.025813    0.025093   \n",
       "0.1 50 0.001 0.1 0.005             0.033292    0.031774   \n",
       "0.1 50 0.001 0.1 0.0001            0.024065    0.024659   \n",
       "0.1 50 0.001 0.2 0.001             0.029256    0.027263   \n",
       "0.1 50 0.001 0.2 0.005             0.029658    0.030336   \n",
       "0.1 50 0.001 0.2 0.0001            0.021699    0.021958   \n",
       "0.1 50 0.0001 0.1 0.001            0.024439    0.024254   \n",
       "0.1 50 0.0001 0.1 0.005            0.379975    0.379509   \n",
       "0.1 50 0.0001 0.1 0.0001           0.027616    0.025940   \n",
       "0.1 50 0.0001 0.2 0.001            0.028009    0.027019   \n",
       "0.1 50 0.0001 0.2 0.005            0.036109    0.036797   \n",
       "0.1 50 0.0001 0.2 0.0001           0.030890    0.029533   \n",
       "0.1 100 0.01 0.1 0.001             0.028874    0.027850   \n",
       "0.1 100 0.01 0.1 0.005             0.035468    0.034512   \n",
       "0.1 100 0.01 0.1 0.0001            0.044340    0.040503   \n",
       "0.1 100 0.01 0.2 0.001             0.031189    0.029099   \n",
       "0.1 100 0.01 0.2 0.005             0.037209    0.035259   \n",
       "0.1 100 0.01 0.2 0.0001            0.038762    0.036697   \n",
       "0.1 100 0.001 0.1 0.001            0.021544    0.021985   \n",
       "0.1 100 0.001 0.1 0.005            0.035428    0.035334   \n",
       "0.1 100 0.001 0.1 0.0001           0.022877    0.021842   \n",
       "0.1 100 0.001 0.2 0.001            0.019832    0.019095   \n",
       "0.1 100 0.001 0.2 0.005            0.030169    0.028686   \n",
       "0.1 100 0.001 0.2 0.0001           0.022537    0.021441   \n",
       "0.1 100 0.0001 0.1 0.001           0.025157    0.025017   \n",
       "0.1 100 0.0001 0.1 0.005           0.204468    0.203970   \n",
       "0.1 100 0.0001 0.1 0.0001          0.026051    0.024451   \n",
       "0.1 100 0.0001 0.2 0.001           0.028255    0.027769   \n",
       "0.1 100 0.0001 0.2 0.005           0.028134    0.026632   \n",
       "0.1 100 0.0001 0.2 0.0001          0.026961    0.025552   \n",
       "0.2 50 0.01 0.1 0.001              0.026047    0.024949   \n",
       "0.2 50 0.01 0.1 0.005              0.031804    0.031117   \n",
       "0.2 50 0.01 0.1 0.0001             0.032627    0.030468   \n",
       "0.2 50 0.01 0.2 0.001              0.027769    0.026206   \n",
       "0.2 50 0.01 0.2 0.005              0.033653    0.032742   \n",
       "0.2 50 0.01 0.2 0.0001             0.030028    0.028848   \n",
       "0.2 50 0.001 0.1 0.001             0.018892    0.017517   \n",
       "0.2 50 0.001 0.1 0.005             0.027144    0.027854   \n",
       "0.2 50 0.001 0.1 0.0001            0.017170    0.016325   \n",
       "0.2 50 0.001 0.2 0.001             0.019835    0.018267   \n",
       "0.2 50 0.001 0.2 0.005             0.025655    0.024096   \n",
       "0.2 50 0.001 0.2 0.0001            0.017665    0.016088   \n",
       "0.2 50 0.0001 0.1 0.001            0.020909    0.021963   \n",
       "0.2 50 0.0001 0.1 0.005            0.021185    0.021990   \n",
       "0.2 50 0.0001 0.1 0.0001           0.023269    0.024131   \n",
       "0.2 50 0.0001 0.2 0.001            0.019359    0.019744   \n",
       "0.2 50 0.0001 0.2 0.005            0.025005    0.025220   \n",
       "0.2 50 0.0001 0.2 0.0001           0.023945    0.023825   \n",
       "0.2 100 0.01 0.1 0.001             0.026386    0.026481   \n",
       "0.2 100 0.01 0.1 0.005             0.032178    0.032803   \n",
       "0.2 100 0.01 0.1 0.0001            0.033482    0.032604   \n",
       "0.2 100 0.01 0.2 0.001             0.026693    0.025377   \n",
       "0.2 100 0.01 0.2 0.005             0.032819    0.033032   \n",
       "0.2 100 0.01 0.2 0.0001            0.031574    0.031647   \n",
       "0.2 100 0.001 0.1 0.001            0.020425    0.020581   \n",
       "0.2 100 0.001 0.1 0.005            0.026281    0.025964   \n",
       "0.2 100 0.001 0.1 0.0001           0.019011    0.019608   \n",
       "0.2 100 0.001 0.2 0.001            0.019996    0.020292   \n",
       "0.2 100 0.001 0.2 0.005            0.028779    0.029734   \n",
       "0.2 100 0.001 0.2 0.0001           0.019623    0.019772   \n",
       "0.2 100 0.0001 0.1 0.001           0.018807    0.020834   \n",
       "0.2 100 0.0001 0.1 0.005           0.027576    0.028495   \n",
       "0.2 100 0.0001 0.1 0.0001          0.020633    0.020684   \n",
       "0.2 100 0.0001 0.2 0.001           0.019593    0.021217   \n",
       "0.2 100 0.0001 0.2 0.005           0.022205    0.023084   \n",
       "0.2 100 0.0001 0.2 0.0001          0.020876    0.021567   \n",
       "0.5 50 0.01 0.1 0.001              0.019731    0.021220   \n",
       "0.5 50 0.01 0.1 0.005              0.022816    0.024583   \n",
       "0.5 50 0.01 0.1 0.0001             0.019371    0.021469   \n",
       "0.5 50 0.01 0.2 0.001              0.020468    0.021380   \n",
       "0.5 50 0.01 0.2 0.005              0.026869    0.028433   \n",
       "0.5 50 0.01 0.2 0.0001             0.018939    0.020692   \n",
       "0.5 50 0.001 0.1 0.001             0.009608    0.011236   \n",
       "0.5 50 0.001 0.1 0.005             0.017795    0.021487   \n",
       "0.5 50 0.001 0.1 0.0001            0.009195    0.010779   \n",
       "0.5 50 0.001 0.2 0.001             0.009134    0.011464   \n",
       "0.5 50 0.001 0.2 0.005             0.017497    0.019446   \n",
       "0.5 50 0.001 0.2 0.0001            0.008857    0.010673   \n",
       "0.5 50 0.0001 0.1 0.001            0.014876    0.016578   \n",
       "0.5 50 0.0001 0.1 0.005            0.020369    0.022964   \n",
       "0.5 50 0.0001 0.1 0.0001           0.011401    0.013548   \n",
       "0.5 50 0.0001 0.2 0.001            0.020169    0.021577   \n",
       "0.5 50 0.0001 0.2 0.005            0.021575    0.022936   \n",
       "0.5 50 0.0001 0.2 0.0001           0.011121    0.012887   \n",
       "0.5 100 0.01 0.1 0.001             0.022207    0.024129   \n",
       "0.5 100 0.01 0.1 0.005             0.021167    0.024339   \n",
       "0.5 100 0.01 0.1 0.0001            0.022762    0.026238   \n",
       "0.5 100 0.01 0.2 0.001             0.023235    0.026334   \n",
       "0.5 100 0.01 0.2 0.005             0.027808    0.030572   \n",
       "0.5 100 0.01 0.2 0.0001            0.021295    0.024752   \n",
       "0.5 100 0.001 0.1 0.001            0.008861    0.011760   \n",
       "0.5 100 0.001 0.1 0.005            0.016313    0.018238   \n",
       "0.5 100 0.001 0.1 0.0001           0.008955    0.011496   \n",
       "0.5 100 0.001 0.2 0.001            0.009087    0.011469   \n",
       "0.5 100 0.001 0.2 0.005            0.012786    0.015312   \n",
       "0.5 100 0.001 0.2 0.0001           0.009032    0.011698   \n",
       "0.5 100 0.0001 0.1 0.001           0.015225    0.017534   \n",
       "0.5 100 0.0001 0.1 0.005           0.019806    0.021231   \n",
       "0.5 100 0.0001 0.1 0.0001          0.010347    0.012220   \n",
       "0.5 100 0.0001 0.2 0.001           0.015056    0.017081   \n",
       "0.5 100 0.0001 0.2 0.005           0.020508    0.022418   \n",
       "0.5 100 0.0001 0.2 0.0001          0.010279    0.012224   \n",
       "\n",
       "                           binomial_approximation_shah  \n",
       "0.1 50 0.01 0.1 0.001                         0.184228  \n",
       "0.1 50 0.01 0.1 0.005                         0.213667  \n",
       "0.1 50 0.01 0.1 0.0001                        0.210562  \n",
       "0.1 50 0.01 0.2 0.001                         0.189458  \n",
       "0.1 50 0.01 0.2 0.005                         0.219838  \n",
       "0.1 50 0.01 0.2 0.0001                        0.203185  \n",
       "0.1 50 0.001 0.1 0.001                        0.144814  \n",
       "0.1 50 0.001 0.1 0.005                        0.188517  \n",
       "0.1 50 0.001 0.1 0.0001                       0.137265  \n",
       "0.1 50 0.001 0.2 0.001                        0.151583  \n",
       "0.1 50 0.001 0.2 0.005                        0.187450  \n",
       "0.1 50 0.001 0.2 0.0001                       0.126203  \n",
       "0.1 50 0.0001 0.1 0.001                       0.166301  \n",
       "0.1 50 0.0001 0.1 0.005                       0.496831  \n",
       "0.1 50 0.0001 0.1 0.0001                      0.157760  \n",
       "0.1 50 0.0001 0.2 0.001                       0.172244  \n",
       "0.1 50 0.0001 0.2 0.005                       0.206223  \n",
       "0.1 50 0.0001 0.2 0.0001                      0.162008  \n",
       "0.1 100 0.01 0.1 0.001                        0.173035  \n",
       "0.1 100 0.01 0.1 0.005                        0.224563  \n",
       "0.1 100 0.01 0.1 0.0001                       0.192964  \n",
       "0.1 100 0.01 0.2 0.001                        0.171586  \n",
       "0.1 100 0.01 0.2 0.005                        0.205906  \n",
       "0.1 100 0.01 0.2 0.0001                       0.180868  \n",
       "0.1 100 0.001 0.1 0.001                       0.133308  \n",
       "0.1 100 0.001 0.1 0.005                       0.175209  \n",
       "0.1 100 0.001 0.1 0.0001                      0.130816  \n",
       "0.1 100 0.001 0.2 0.001                       0.127615  \n",
       "0.1 100 0.001 0.2 0.005                       0.168592  \n",
       "0.1 100 0.001 0.2 0.0001                      0.130596  \n",
       "0.1 100 0.0001 0.1 0.001                      0.162885  \n",
       "0.1 100 0.0001 0.1 0.005                      0.334760  \n",
       "0.1 100 0.0001 0.1 0.0001                     0.151349  \n",
       "0.1 100 0.0001 0.2 0.001                      0.167891  \n",
       "0.1 100 0.0001 0.2 0.005                      0.180173  \n",
       "0.1 100 0.0001 0.2 0.0001                     0.156919  \n",
       "0.2 50 0.01 0.1 0.001                         0.151431  \n",
       "0.2 50 0.01 0.1 0.005                         0.171902  \n",
       "0.2 50 0.01 0.1 0.0001                        0.161078  \n",
       "0.2 50 0.01 0.2 0.001                         0.153792  \n",
       "0.2 50 0.01 0.2 0.005                         0.172380  \n",
       "0.2 50 0.01 0.2 0.0001                        0.153550  \n",
       "0.2 50 0.001 0.1 0.001                        0.124174  \n",
       "0.2 50 0.001 0.1 0.005                        0.160090  \n",
       "0.2 50 0.001 0.1 0.0001                       0.113712  \n",
       "0.2 50 0.001 0.2 0.001                        0.124023  \n",
       "0.2 50 0.001 0.2 0.005                        0.157148  \n",
       "0.2 50 0.001 0.2 0.0001                       0.114398  \n",
       "0.2 50 0.0001 0.1 0.001                       0.147841  \n",
       "0.2 50 0.0001 0.1 0.005                       0.162769  \n",
       "0.2 50 0.0001 0.1 0.0001                      0.138587  \n",
       "0.2 50 0.0001 0.2 0.001                       0.147829  \n",
       "0.2 50 0.0001 0.2 0.005                       0.168678  \n",
       "0.2 50 0.0001 0.2 0.0001                      0.140636  \n",
       "0.2 100 0.01 0.1 0.001                        0.156762  \n",
       "0.2 100 0.01 0.1 0.005                        0.175797  \n",
       "0.2 100 0.01 0.1 0.0001                       0.167333  \n",
       "0.2 100 0.01 0.2 0.001                        0.156542  \n",
       "0.2 100 0.01 0.2 0.005                        0.175553  \n",
       "0.2 100 0.01 0.2 0.0001                       0.163643  \n",
       "0.2 100 0.001 0.1 0.001                       0.127410  \n",
       "0.2 100 0.001 0.1 0.005                       0.157938  \n",
       "0.2 100 0.001 0.1 0.0001                      0.120897  \n",
       "0.2 100 0.001 0.2 0.001                       0.125309  \n",
       "0.2 100 0.001 0.2 0.005                       0.166416  \n",
       "0.2 100 0.001 0.2 0.0001                      0.121581  \n",
       "0.2 100 0.0001 0.1 0.001                      0.141800  \n",
       "0.2 100 0.0001 0.1 0.005                      0.173918  \n",
       "0.2 100 0.0001 0.1 0.0001                     0.127912  \n",
       "0.2 100 0.0001 0.2 0.001                      0.146239  \n",
       "0.2 100 0.0001 0.2 0.005                      0.164044  \n",
       "0.2 100 0.0001 0.2 0.0001                     0.130437  \n",
       "0.5 50 0.01 0.1 0.001                         0.142522  \n",
       "0.5 50 0.01 0.1 0.005                         0.155215  \n",
       "0.5 50 0.01 0.1 0.0001                        0.137028  \n",
       "0.5 50 0.01 0.2 0.001                         0.144428  \n",
       "0.5 50 0.01 0.2 0.005                         0.161395  \n",
       "0.5 50 0.01 0.2 0.0001                        0.133117  \n",
       "0.5 50 0.001 0.1 0.001                        0.104269  \n",
       "0.5 50 0.001 0.1 0.005                        0.145166  \n",
       "0.5 50 0.001 0.1 0.0001                       0.096942  \n",
       "0.5 50 0.001 0.2 0.001                        0.102977  \n",
       "0.5 50 0.001 0.2 0.005                        0.136166  \n",
       "0.5 50 0.001 0.2 0.0001                       0.097737  \n",
       "0.5 50 0.0001 0.1 0.001                       0.137071  \n",
       "0.5 50 0.0001 0.1 0.005                       0.170995  \n",
       "0.5 50 0.0001 0.1 0.0001                      0.103198  \n",
       "0.5 50 0.0001 0.2 0.001                       0.144297  \n",
       "0.5 50 0.0001 0.2 0.005                       0.171363  \n",
       "0.5 50 0.0001 0.2 0.0001                      0.103408  \n",
       "0.5 100 0.01 0.1 0.001                        0.154063  \n",
       "0.5 100 0.01 0.1 0.005                        0.153824  \n",
       "0.5 100 0.01 0.1 0.0001                       0.150752  \n",
       "0.5 100 0.01 0.2 0.001                        0.156846  \n",
       "0.5 100 0.01 0.2 0.005                        0.170788  \n",
       "0.5 100 0.01 0.2 0.0001                       0.143140  \n",
       "0.5 100 0.001 0.1 0.001                       0.099429  \n",
       "0.5 100 0.001 0.1 0.005                       0.131466  \n",
       "0.5 100 0.001 0.1 0.0001                      0.095795  \n",
       "0.5 100 0.001 0.2 0.001                       0.099780  \n",
       "0.5 100 0.001 0.2 0.005                       0.121501  \n",
       "0.5 100 0.001 0.2 0.0001                      0.096302  \n",
       "0.5 100 0.0001 0.1 0.001                      0.131493  \n",
       "0.5 100 0.0001 0.1 0.005                      0.170578  \n",
       "0.5 100 0.0001 0.1 0.0001                     0.099455  \n",
       "0.5 100 0.0001 0.2 0.001                      0.128085  \n",
       "0.5 100 0.0001 0.2 0.005                      0.169332  \n",
       "0.5 100 0.0001 0.2 0.0001                     0.100241  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "early_stop_matrix = early_stop_result_matrix.mean(1).reshape(early_stop_result_matrix.shape[0],np.prod(early_stop_result_matrix.shape[2:])).T\n",
    "l = deepcopy(new_vals_to_fetch)\n",
    "l.append(\"binomial_approximation_shah\")\n",
    "mean_early_stop_df = pd.DataFrame(early_stop_matrix, index=name_list, columns=l)\n",
    "mean_early_stop_df.to_csv(f\"./results/mnist/mean_{project}.csv\")\n",
    "mean_early_stop_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "8e1d7d37-fea7-4d5c-a1ed-fc66f935994b",
   "metadata": {},
   "outputs": [],
   "source": [
    "early_stop_std = early_stop_result_matrix.std(1).reshape(early_stop_result_matrix.shape[0],np.prod(early_stop_result_matrix.shape[2:])).T\n",
    "std_early_stop_df = pd.DataFrame(early_stop_std, index=name_list, columns=l)\n",
    "mean_early_stop_df.to_csv(f\"./results/mnist/std_{project}.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "44ecebab-61d1-414e-878c-e9cdb611dfd6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.5, 100.0, 0.001, 0.1, 0.0001]\n",
      "Complement error: 0.0048±0.0044  Validation error: 0.0090±0.0027 Test error : 0.0115±0.0012  KL bound : 0.0350±0.0296  binomial : 0.0958±0.0221  compression set size : 320.0000±0.0000 \n"
     ]
    }
   ],
   "source": [
    "val_of_interest = \"kl_bound\"\n",
    "best_params = correct_type_of_entry(mean_early_stop_df.index[mean_early_stop_df[val_of_interest].argmin()].split())\n",
    "print(best_params)\n",
    "best_val_arr = mean_early_stop_df.loc[mean_early_stop_df.index[mean_early_stop_df[val_of_interest].argmin()]]\n",
    "std_val_arr = std_early_stop_df.loc[mean_early_stop_df.index[mean_early_stop_df[val_of_interest].argmin()]]\n",
    "print(f\"Complement error: {best_val_arr['complement_error']:.4f}±{std_val_arr['complement_error']:.4f} \",\n",
    "    f\"Validation error: {best_val_arr['validation_error']:.4f}±{std_val_arr['validation_error']:.4f} \"\n",
    "      f'Test error : {best_val_arr['test_error']:.4f}±{std_val_arr['test_error']:.4f} ', \n",
    "      f\"KL bound : {best_val_arr['kl_bound']:.4f}±{std_val_arr['kl_bound']:.4f} \",\n",
    "    f'binomial : {best_val_arr['binomial_approximation_shah']:.4f}±{std_val_arr['binomial_approximation_shah']:.4f} ',\n",
    "     # f'P2L bound : {best_val_arr['p2l_bound']:.4f}±{std_val_arr['p2l_bound']:.4f} ', \n",
    "      f'compression set size : {best_val_arr['compression_set_size']:.4f}±{std_val_arr['compression_set_size']:.4f} ',\n",
    "    # f\"Validation loss: {best_val_arr['validation_loss']:.4f}±{std_val_arr['validation_loss']:.4f} \"\n",
    "      # f'Test loss : {best_val_arr['test_loss']:.4f}±{std_val_arr['test_loss']:.4f} ', \n",
    "    # f\"CE KL bound: {best_val_arr['CE_kl_bound']:.4f}±{std_val_arr['CE_kl_bound']:.4f} \"\n",
    "      # f'CE Catoni bound : {best_val_arr['CE_min_val_catoni_bound']:.4f}±{std_val_arr['CE_min_val_catoni_bound']:.4f} ',\n",
    "    # f\"Runtime : {best_val_arr['_runtime']:.4f}±{std_val_arr['_runtime']:.4f}\"\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b4729dc4-2e71-4d27-8088-108976d3fb2e",
   "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": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}