07_metricGrisoni.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd, numpy as np, re, os\n",
    "from evaluation import Evaluator, compilerPatterns, ResultsFromMeanStdDF\n",
    "from matplotlib import pyplot as plt\n",
    "plt.rcParams.update({'font.size': 12})"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "## This cell would need for testing phase - Ignore this\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plotLossOnAx(ax, lossDF, label, **kwargs):\n",
    "   stdKwargs = dict(alpha=0.2, edgecolor='#1B2ACC', facecolor='#089FFF',\n",
    "                            linewidth=4, antialiased=True)\n",
    "   stdKwargs.update(kwargs)\n",
    "   ax.plot(lossDF['mean'], label = label )\n",
    "   ax.fill_between(lossDF.index, lossDF['mean']-lossDF['std'], \n",
    "                  lossDF['mean']+lossDF['std'], **stdKwargs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "paths = [ os.path.join(dataset.name, experiment.name) \n",
    "         for dataset in os.scandir('experiments') if dataset.is_dir() and 'grisoni' in dataset.name \n",
    "         for experiment in os.scandir(dataset.path)  ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['grisoniAug5_SELFIES-RNN/2hl_256hu_512bs_0.001lr',\n",
       " 'grisoniAug5_SELFIES-RNN/2hl_512hu_256bs_0.001lr',\n",
       " 'grisoniAug5_SELFIES-RNN/2hl_512hu_512bs_0.001lr',\n",
       " 'grisoniAug5_SELFIES-RNN/3hl_256hu_256bs_0.001lr',\n",
       " 'grisoniAug5_SELFIES-RNN/2hl_256hu_256bs_0.001lr',\n",
       " 'grisoniAug5_SELFIES-RNN/3hl_256hu_512bs_0.001lr',\n",
       " 'grisoniAug5_SELFIES-RNN/3hl_512hu_256bs_0.001lr',\n",
       " 'grisoniAug5_SELFIES-RNN/3hl_512hu_512bs_0.001lr',\n",
       " 'grisoniAug5_SMILES-RNN/2hl_256hu_512bs_0.001lr',\n",
       " 'grisoniAug5_SMILES-RNN/2hl_512hu_256bs_0.001lr',\n",
       " 'grisoniAug5_SMILES-RNN/2hl_512hu_512bs_0.001lr',\n",
       " 'grisoniAug5_SMILES-RNN/3hl_256hu_256bs_0.001lr',\n",
       " 'grisoniAug5_SMILES-RNN/2hl_256hu_256bs_0.001lr',\n",
       " 'grisoniAug5_SMILES-RNN/3hl_256hu_512bs_0.001lr',\n",
       " 'grisoniAug5_SMILES-RNN/3hl_512hu_256bs_0.001lr',\n",
       " 'grisoniAug5_SMILES-RNN/3hl_512hu_512bs_0.001lr',\n",
       " 'grisoniAug5_fragSMILES-RNN/2hl_256hu_512bs_300es_0.001lr',\n",
       " 'grisoniAug5_fragSMILES-RNN/2hl_512hu_256bs_300es_0.001lr',\n",
       " 'grisoniAug5_fragSMILES-RNN/2hl_512hu_512bs_300es_0.001lr',\n",
       " 'grisoniAug5_fragSMILES-RNN/3hl_256hu_256bs_300es_0.001lr',\n",
       " 'grisoniAug5_fragSMILES-RNN/2hl_256hu_256bs_300es_0.001lr',\n",
       " 'grisoniAug5_fragSMILES-RNN/3hl_256hu_512bs_300es_0.001lr',\n",
       " 'grisoniAug5_fragSMILES-RNN/3hl_512hu_256bs_300es_0.001lr',\n",
       " 'grisoniAug5_fragSMILES-RNN/3hl_512hu_512bs_300es_0.001lr',\n",
       " 'grisoni_SELFIES-RNN/2hl_256hu_512bs_0.001lr',\n",
       " 'grisoni_SELFIES-RNN/2hl_512hu_256bs_0.001lr',\n",
       " 'grisoni_SELFIES-RNN/2hl_512hu_512bs_0.001lr',\n",
       " 'grisoni_SELFIES-RNN/3hl_256hu_256bs_0.001lr',\n",
       " 'grisoni_SELFIES-RNN/2hl_256hu_256bs_0.001lr',\n",
       " 'grisoni_SELFIES-RNN/3hl_256hu_512bs_0.001lr',\n",
       " 'grisoni_SELFIES-RNN/3hl_512hu_256bs_0.001lr',\n",
       " 'grisoni_SELFIES-RNN/3hl_512hu_512bs_0.001lr',\n",
       " 'grisoni_SMILES-RNN/2hl_256hu_512bs_0.001lr',\n",
       " 'grisoni_SMILES-RNN/2hl_512hu_256bs_0.001lr',\n",
       " 'grisoni_SMILES-RNN/2hl_512hu_512bs_0.001lr',\n",
       " 'grisoni_SMILES-RNN/3hl_256hu_256bs_0.001lr',\n",
       " 'grisoni_SMILES-RNN/2hl_256hu_256bs_0.001lr',\n",
       " 'grisoni_SMILES-RNN/3hl_256hu_512bs_0.001lr',\n",
       " 'grisoni_SMILES-RNN/3hl_512hu_256bs_0.001lr',\n",
       " 'grisoni_SMILES-RNN/3hl_512hu_512bs_0.001lr',\n",
       " 'grisoni_fragSMILES-RNN/2hl_512hu_256bs_300es_0.001lr',\n",
       " 'grisoni_fragSMILES-RNN/2hl_256hu_512bs_300es_0.001lr',\n",
       " 'grisoni_fragSMILES-RNN/2hl_512hu_512bs_300es_0.001lr',\n",
       " 'grisoni_fragSMILES-RNN/3hl_256hu_256bs_300es_0.001lr',\n",
       " 'grisoni_fragSMILES-RNN/2hl_512hu_512bs_80es_0.001lr',\n",
       " 'grisoni_fragSMILES-RNN/2hl_256hu_256bs_300es_0.001lr',\n",
       " 'grisoni_fragSMILES-RNN/3hl_256hu_512bs_300es_0.001lr',\n",
       " 'grisoni_fragSMILES-RNN/3hl_512hu_256bs_300es_0.001lr',\n",
       " 'grisoni_fragSMILES-RNN/3hl_512hu_512bs_300es_0.001lr']"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "paths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "evaluators = [Evaluator(path, loadCSV=True) for path in paths]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1500 with 24 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.rcParams.update({'font.size': 7})\n",
    "\n",
    "fig,axes = plt.subplots(nrows=6, ncols=4, figsize=(10,15),gridspec_kw={'wspace':0.7, 'hspace':0.3})\n",
    "onlyToEval = [evaluator for evaluator in evaluators if evaluator.params['hl']==3]\n",
    "for path, evaluator, ax in zip(paths, onlyToEval, axes.flatten()):\n",
    "    train,test = evaluator.getTrainTestLosses()\n",
    "    plotLossOnAx(ax, train, label='Training')\n",
    "    plotLossOnAx(ax, test, label='Validation')\n",
    "    ax.set_title(list((evaluator.dataInfo|evaluator.params).values()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Syntax metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy import stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "genDFs=pd.concat( [evaluator.getResultsGens() for evaluator in evaluators if not evaluator.gens.empty], ignore_index=True )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "genDFs['valid%']=genDFs['valid']/genDFs['amount']\n",
    "genDFs['unique%']=genDFs['unique']/genDFs['valid']\n",
    "genDFs['novel%']=genDFs['novel']/genDFs['unique']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "params=[2,512,-1,0.001,512,6000,10,1.0]\n",
    "paramsFrag = [2,512,300,0.001,512,6000,16,1.0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "params5=[2,512,-1,0.001,256,6000,3,1.0]\n",
    "paramsFrag5 = [2,512,300,0.001,256,6000,4,1.0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_1315920/1376500511.py:1: FutureWarning: Downcasting object dtype arrays on .fillna, .ffill, .bfill is deprecated and will change in a future version. Call result.infer_objects(copy=False) instead. To opt-in to the future behavior, set `pd.set_option('future.no_silent_downcasting', True)`\n",
      "  genGrp=genDFs.fillna(-1).groupby(compilerPatterns.index.drop('fold').tolist(),as_index=True, dropna=False)\n"
     ]
    }
   ],
   "source": [
    "genGrp=genDFs.fillna(-1).groupby(compilerPatterns.index.drop('fold').tolist(),as_index=True, dropna=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "genStats=genGrp.describe().loc(axis=1)[ ( slice(None),['mean','std'] ) ].drop(columns='fold')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "genStats=ResultsFromMeanStdDF(genStats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "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></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>valid</th>\n",
       "      <th>unique</th>\n",
       "      <th>novel</th>\n",
       "      <th>valid%</th>\n",
       "      <th>unique%</th>\n",
       "      <th>novel%</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th>aug</th>\n",
       "      <th>notation</th>\n",
       "      <th>hl</th>\n",
       "      <th>hu</th>\n",
       "      <th>es</th>\n",
       "      <th>lr</th>\n",
       "      <th>bs</th>\n",
       "      <th>amount</th>\n",
       "      <th>epoch</th>\n",
       "      <th>T</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"48\" valign=\"top\">grisoni</th>\n",
       "      <th rowspan=\"24\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"8\" valign=\"top\">SELFIES</th>\n",
       "      <th rowspan=\"4\" valign=\"top\">2</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">256</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>6</th>\n",
       "      <th>1.0</th>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>5999.0 ± 2.0</td>\n",
       "      <td>5999.0 ± 2.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9998 ± 0.0003</td>\n",
       "      <td>0.99997 ± 7e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5999.8 ± 0.4</td>\n",
       "      <td>5999.0 ± 2.0</td>\n",
       "      <td>5998.0 ± 2.0</td>\n",
       "      <td>0.99997 ± 7e-05</td>\n",
       "      <td>0.9999 ± 0.0003</td>\n",
       "      <td>0.9998 ± 0.0001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">512</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>6</th>\n",
       "      <th>1.0</th>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>5998.0 ± 1.0</td>\n",
       "      <td>5978.0 ± 4.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9997 ± 0.0002</td>\n",
       "      <td>0.9967 ± 0.0005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>5999.0 ± 2.0</td>\n",
       "      <td>5971.0 ± 2.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9998 ± 0.0003</td>\n",
       "      <td>0.9953 ± 0.0006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">3</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">256</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>6</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5999.8 ± 0.4</td>\n",
       "      <td>5999.4 ± 0.9</td>\n",
       "      <td>5998.0 ± 2.0</td>\n",
       "      <td>0.99997 ± 7e-05</td>\n",
       "      <td>0.9999 ± 0.0001</td>\n",
       "      <td>0.9998 ± 0.0002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5999.4 ± 0.9</td>\n",
       "      <td>5999.4 ± 0.9</td>\n",
       "      <td>5998.0 ± 2.0</td>\n",
       "      <td>0.9999 ± 0.0001</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9998 ± 0.0003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">512</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>6</th>\n",
       "      <th>1.0</th>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>5999.0 ± 1.0</td>\n",
       "      <td>5986.0 ± 10.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9999 ± 0.0002</td>\n",
       "      <td>0.998 ± 0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>5999.6 ± 0.9</td>\n",
       "      <td>5978.0 ± 9.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9999 ± 0.0001</td>\n",
       "      <td>0.996 ± 0.001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"8\" valign=\"top\">SMILES</th>\n",
       "      <th rowspan=\"4\" valign=\"top\">2</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">256</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>6</th>\n",
       "      <th>1.0</th>\n",
       "      <td>3800.0 ± 1000.0</td>\n",
       "      <td>3800.0 ± 1000.0</td>\n",
       "      <td>3800.0 ± 1000.0</td>\n",
       "      <td>0.6 ± 0.2</td>\n",
       "      <td>0.9998 ± 0.0003</td>\n",
       "      <td>0.9988 ± 0.0005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>4100.0 ± 100.0</td>\n",
       "      <td>4100.0 ± 100.0</td>\n",
       "      <td>4100.0 ± 100.0</td>\n",
       "      <td>0.69 ± 0.02</td>\n",
       "      <td>0.9999 ± 0.0002</td>\n",
       "      <td>0.9985 ± 0.0008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">512</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>6</th>\n",
       "      <th>1.0</th>\n",
       "      <td>4600.0 ± 300.0</td>\n",
       "      <td>4600.0 ± 300.0</td>\n",
       "      <td>4500.0 ± 300.0</td>\n",
       "      <td>0.77 ± 0.05</td>\n",
       "      <td>0.9989 ± 0.0005</td>\n",
       "      <td>0.984 ± 0.006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>4930.0 ± 70.0</td>\n",
       "      <td>4920.0 ± 70.0</td>\n",
       "      <td>4770.0 ± 60.0</td>\n",
       "      <td>0.82 ± 0.01</td>\n",
       "      <td>0.997 ± 0.001</td>\n",
       "      <td>0.97 ± 0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">3</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">256</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>6</th>\n",
       "      <th>1.0</th>\n",
       "      <td>4000.0 ± 1000.0</td>\n",
       "      <td>4000.0 ± 1000.0</td>\n",
       "      <td>4000.0 ± 1000.0</td>\n",
       "      <td>0.7 ± 0.2</td>\n",
       "      <td>0.9997 ± 0.0004</td>\n",
       "      <td>0.9982 ± 0.001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>4500.0 ± 100.0</td>\n",
       "      <td>4500.0 ± 100.0</td>\n",
       "      <td>4500.0 ± 100.0</td>\n",
       "      <td>0.75 ± 0.02</td>\n",
       "      <td>0.9996 ± 0.0002</td>\n",
       "      <td>0.998 ± 0.001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">512</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>6</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5120.0 ± 90.0</td>\n",
       "      <td>5110.0 ± 80.0</td>\n",
       "      <td>5000.0 ± 70.0</td>\n",
       "      <td>0.85 ± 0.01</td>\n",
       "      <td>0.9979 ± 0.0007</td>\n",
       "      <td>0.978 ± 0.003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5090.0 ± 50.0</td>\n",
       "      <td>5080.0 ± 50.0</td>\n",
       "      <td>4910.0 ± 60.0</td>\n",
       "      <td>0.849 ± 0.008</td>\n",
       "      <td>0.9976 ± 0.0009</td>\n",
       "      <td>0.966 ± 0.003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"8\" valign=\"top\">fragSMILES</th>\n",
       "      <th rowspan=\"4\" valign=\"top\">2</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">256</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">300</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>8</th>\n",
       "      <th>1.0</th>\n",
       "      <td>4700.0 ± 200.0</td>\n",
       "      <td>4700.0 ± 200.0</td>\n",
       "      <td>4700.0 ± 200.0</td>\n",
       "      <td>0.79 ± 0.03</td>\n",
       "      <td>0.9997 ± 0.0007</td>\n",
       "      <td>0.993 ± 0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>16</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5010.0 ± 70.0</td>\n",
       "      <td>5000.0 ± 70.0</td>\n",
       "      <td>4940.0 ± 60.0</td>\n",
       "      <td>0.83 ± 0.01</td>\n",
       "      <td>0.9992 ± 0.001</td>\n",
       "      <td>0.988 ± 0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">512</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">300</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>8</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5200.0 ± 90.0</td>\n",
       "      <td>5200.0 ± 90.0</td>\n",
       "      <td>5110.0 ± 80.0</td>\n",
       "      <td>0.87 ± 0.01</td>\n",
       "      <td>0.9993 ± 0.0003</td>\n",
       "      <td>0.984 ± 0.003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>16</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5280.0 ± 20.0</td>\n",
       "      <td>5270.0 ± 30.0</td>\n",
       "      <td>5110.0 ± 40.0</td>\n",
       "      <td>0.88 ± 0.004</td>\n",
       "      <td>0.9976 ± 0.0005</td>\n",
       "      <td>0.971 ± 0.004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">3</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">256</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">300</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>8</th>\n",
       "      <th>1.0</th>\n",
       "      <td>4600.0 ± 100.0</td>\n",
       "      <td>4600.0 ± 100.0</td>\n",
       "      <td>4500.0 ± 100.0</td>\n",
       "      <td>0.76 ± 0.02</td>\n",
       "      <td>0.9997 ± 0.0005</td>\n",
       "      <td>0.993 ± 0.001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>16</th>\n",
       "      <th>1.0</th>\n",
       "      <td>4800.0 ± 90.0</td>\n",
       "      <td>4800.0 ± 80.0</td>\n",
       "      <td>4740.0 ± 80.0</td>\n",
       "      <td>0.8 ± 0.01</td>\n",
       "      <td>0.9994 ± 0.0005</td>\n",
       "      <td>0.987 ± 0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">512</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">300</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>8</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5050.0 ± 60.0</td>\n",
       "      <td>5040.0 ± 60.0</td>\n",
       "      <td>4940.0 ± 50.0</td>\n",
       "      <td>0.841 ± 0.01</td>\n",
       "      <td>0.9986 ± 0.0009</td>\n",
       "      <td>0.979 ± 0.003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>16</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5150.0 ± 50.0</td>\n",
       "      <td>5140.0 ± 40.0</td>\n",
       "      <td>4950.0 ± 40.0</td>\n",
       "      <td>0.859 ± 0.008</td>\n",
       "      <td>0.997 ± 0.002</td>\n",
       "      <td>0.963 ± 0.003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"24\" valign=\"top\">5</th>\n",
       "      <th rowspan=\"8\" valign=\"top\">SELFIES</th>\n",
       "      <th rowspan=\"4\" valign=\"top\">2</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">256</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>5999.0 ± 1.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9999 ± 0.0002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5999.8 ± 0.4</td>\n",
       "      <td>5999.0 ± 1.0</td>\n",
       "      <td>5999.0 ± 1.0</td>\n",
       "      <td>0.99997 ± 7e-05</td>\n",
       "      <td>0.9998 ± 0.0002</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">512</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>5997.0 ± 1.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9996 ± 0.0002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>5999.6 ± 0.9</td>\n",
       "      <td>5998.0 ± 2.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9999 ± 0.0001</td>\n",
       "      <td>0.9998 ± 0.0002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">3</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">256</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5999.8 ± 0.4</td>\n",
       "      <td>5999.8 ± 0.4</td>\n",
       "      <td>5999.4 ± 0.9</td>\n",
       "      <td>0.99997 ± 7e-05</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.99993 ± 9e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5999.8 ± 0.4</td>\n",
       "      <td>5999.8 ± 0.4</td>\n",
       "      <td>5999.2 ± 0.8</td>\n",
       "      <td>0.99997 ± 7e-05</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9999 ± 9e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">512</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>5999.6 ± 0.9</td>\n",
       "      <td>5997.8 ± 0.8</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9999 ± 0.0001</td>\n",
       "      <td>0.9997 ± 0.0001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>5999.4 ± 0.9</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9999 ± 0.0001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"8\" valign=\"top\">SMILES</th>\n",
       "      <th rowspan=\"4\" valign=\"top\">2</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">256</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>4900.0 ± 100.0</td>\n",
       "      <td>4900.0 ± 100.0</td>\n",
       "      <td>4900.0 ± 100.0</td>\n",
       "      <td>0.82 ± 0.02</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.99963 ± 9e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>4730.0 ± 80.0</td>\n",
       "      <td>4730.0 ± 80.0</td>\n",
       "      <td>4730.0 ± 80.0</td>\n",
       "      <td>0.79 ± 0.01</td>\n",
       "      <td>0.9998 ± 0.0003</td>\n",
       "      <td>0.9997 ± 0.0002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">512</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5300.0 ± 40.0</td>\n",
       "      <td>5300.0 ± 40.0</td>\n",
       "      <td>5280.0 ± 40.0</td>\n",
       "      <td>0.883 ± 0.007</td>\n",
       "      <td>0.9999 ± 0.0002</td>\n",
       "      <td>0.997 ± 0.0009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5220.0 ± 50.0</td>\n",
       "      <td>5220.0 ± 50.0</td>\n",
       "      <td>5210.0 ± 50.0</td>\n",
       "      <td>0.87 ± 0.009</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9982 ± 0.0006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">3</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">256</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5170.0 ± 30.0</td>\n",
       "      <td>5170.0 ± 30.0</td>\n",
       "      <td>5170.0 ± 30.0</td>\n",
       "      <td>0.862 ± 0.004</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9993 ± 0.0005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5020.0 ± 50.0</td>\n",
       "      <td>5020.0 ± 50.0</td>\n",
       "      <td>5020.0 ± 40.0</td>\n",
       "      <td>0.837 ± 0.008</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9995 ± 0.0004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">512</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">-1</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5470.0 ± 50.0</td>\n",
       "      <td>5470.0 ± 50.0</td>\n",
       "      <td>5450.0 ± 50.0</td>\n",
       "      <td>0.912 ± 0.008</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9969 ± 0.0007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5390.0 ± 30.0</td>\n",
       "      <td>5390.0 ± 30.0</td>\n",
       "      <td>5370.0 ± 30.0</td>\n",
       "      <td>0.898 ± 0.005</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.998 ± 0.0006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"8\" valign=\"top\">fragSMILES</th>\n",
       "      <th rowspan=\"4\" valign=\"top\">2</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">256</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">300</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>4</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5230.0 ± 70.0</td>\n",
       "      <td>5230.0 ± 70.0</td>\n",
       "      <td>5180.0 ± 70.0</td>\n",
       "      <td>0.87 ± 0.01</td>\n",
       "      <td>0.999 ± 0.001</td>\n",
       "      <td>0.991 ± 0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>4</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5070.0 ± 60.0</td>\n",
       "      <td>5070.0 ± 60.0</td>\n",
       "      <td>5050.0 ± 60.0</td>\n",
       "      <td>0.85 ± 0.01</td>\n",
       "      <td>0.9994 ± 0.0006</td>\n",
       "      <td>0.9954 ± 0.0008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">512</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">300</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>4</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5420.0 ± 60.0</td>\n",
       "      <td>5410.0 ± 60.0</td>\n",
       "      <td>5300.0 ± 60.0</td>\n",
       "      <td>0.903 ± 0.009</td>\n",
       "      <td>0.9991 ± 0.0006</td>\n",
       "      <td>0.98 ± 0.003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>4</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5320.0 ± 70.0</td>\n",
       "      <td>5310.0 ± 70.0</td>\n",
       "      <td>5240.0 ± 70.0</td>\n",
       "      <td>0.89 ± 0.01</td>\n",
       "      <td>0.9995 ± 0.0004</td>\n",
       "      <td>0.987 ± 0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"4\" valign=\"top\">3</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">256</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">300</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>4</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5150.0 ± 30.0</td>\n",
       "      <td>5150.0 ± 30.0</td>\n",
       "      <td>5090.0 ± 30.0</td>\n",
       "      <td>0.859 ± 0.005</td>\n",
       "      <td>0.999 ± 0.0008</td>\n",
       "      <td>0.988 ± 0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>4</th>\n",
       "      <th>1.0</th>\n",
       "      <td>4930.0 ± 60.0</td>\n",
       "      <td>4920.0 ± 60.0</td>\n",
       "      <td>4890.0 ± 50.0</td>\n",
       "      <td>0.821 ± 0.009</td>\n",
       "      <td>0.9993 ± 0.0006</td>\n",
       "      <td>0.994 ± 0.001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">512</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">300</th>\n",
       "      <th rowspan=\"2\" valign=\"top\">0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>4</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5390.0 ± 30.0</td>\n",
       "      <td>5380.0 ± 30.0</td>\n",
       "      <td>5240.0 ± 30.0</td>\n",
       "      <td>0.899 ± 0.005</td>\n",
       "      <td>0.9986 ± 0.0007</td>\n",
       "      <td>0.973 ± 0.003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>4</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5200.0 ± 60.0</td>\n",
       "      <td>5190.0 ± 60.0</td>\n",
       "      <td>5120.0 ± 60.0</td>\n",
       "      <td>0.87 ± 0.01</td>\n",
       "      <td>0.9984 ± 0.0009</td>\n",
       "      <td>0.985 ± 0.002</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                         valid  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                      \n",
       "grisoni -1  SELFIES    2  256 -1   0.001 256 6000   6     1.0     6000.0 ± 0.0   \n",
       "                                         512 6000   10    1.0     5999.8 ± 0.4   \n",
       "                          512 -1   0.001 256 6000   6     1.0     6000.0 ± 0.0   \n",
       "                                         512 6000   10    1.0     6000.0 ± 0.0   \n",
       "                       3  256 -1   0.001 256 6000   6     1.0     5999.8 ± 0.4   \n",
       "                                         512 6000   10    1.0     5999.4 ± 0.9   \n",
       "                          512 -1   0.001 256 6000   6     1.0     6000.0 ± 0.0   \n",
       "                                         512 6000   10    1.0     6000.0 ± 0.0   \n",
       "            SMILES     2  256 -1   0.001 256 6000   6     1.0  3800.0 ± 1000.0   \n",
       "                                         512 6000   10    1.0   4100.0 ± 100.0   \n",
       "                          512 -1   0.001 256 6000   6     1.0   4600.0 ± 300.0   \n",
       "                                         512 6000   10    1.0    4930.0 ± 70.0   \n",
       "                       3  256 -1   0.001 256 6000   6     1.0  4000.0 ± 1000.0   \n",
       "                                         512 6000   10    1.0   4500.0 ± 100.0   \n",
       "                          512 -1   0.001 256 6000   6     1.0    5120.0 ± 90.0   \n",
       "                                         512 6000   10    1.0    5090.0 ± 50.0   \n",
       "            fragSMILES 2  256  300 0.001 256 6000   8     1.0   4700.0 ± 200.0   \n",
       "                                         512 6000   16    1.0    5010.0 ± 70.0   \n",
       "                          512  300 0.001 256 6000   8     1.0    5200.0 ± 90.0   \n",
       "                                         512 6000   16    1.0    5280.0 ± 20.0   \n",
       "                       3  256  300 0.001 256 6000   8     1.0   4600.0 ± 100.0   \n",
       "                                         512 6000   16    1.0    4800.0 ± 90.0   \n",
       "                          512  300 0.001 256 6000   8     1.0    5050.0 ± 60.0   \n",
       "                                         512 6000   16    1.0    5150.0 ± 50.0   \n",
       "         5  SELFIES    2  256 -1   0.001 256 6000   3     1.0     6000.0 ± 0.0   \n",
       "                                         512 6000   3     1.0     5999.8 ± 0.4   \n",
       "                          512 -1   0.001 256 6000   3     1.0     6000.0 ± 0.0   \n",
       "                                         512 6000   3     1.0     6000.0 ± 0.0   \n",
       "                       3  256 -1   0.001 256 6000   3     1.0     5999.8 ± 0.4   \n",
       "                                         512 6000   3     1.0     5999.8 ± 0.4   \n",
       "                          512 -1   0.001 256 6000   3     1.0     6000.0 ± 0.0   \n",
       "                                         512 6000   3     1.0     6000.0 ± 0.0   \n",
       "            SMILES     2  256 -1   0.001 256 6000   3     1.0   4900.0 ± 100.0   \n",
       "                                         512 6000   3     1.0    4730.0 ± 80.0   \n",
       "                          512 -1   0.001 256 6000   3     1.0    5300.0 ± 40.0   \n",
       "                                         512 6000   3     1.0    5220.0 ± 50.0   \n",
       "                       3  256 -1   0.001 256 6000   3     1.0    5170.0 ± 30.0   \n",
       "                                         512 6000   3     1.0    5020.0 ± 50.0   \n",
       "                          512 -1   0.001 256 6000   3     1.0    5470.0 ± 50.0   \n",
       "                                         512 6000   3     1.0    5390.0 ± 30.0   \n",
       "            fragSMILES 2  256  300 0.001 256 6000   4     1.0    5230.0 ± 70.0   \n",
       "                                         512 6000   4     1.0    5070.0 ± 60.0   \n",
       "                          512  300 0.001 256 6000   4     1.0    5420.0 ± 60.0   \n",
       "                                         512 6000   4     1.0    5320.0 ± 70.0   \n",
       "                       3  256  300 0.001 256 6000   4     1.0    5150.0 ± 30.0   \n",
       "                                         512 6000   4     1.0    4930.0 ± 60.0   \n",
       "                          512  300 0.001 256 6000   4     1.0    5390.0 ± 30.0   \n",
       "                                         512 6000   4     1.0    5200.0 ± 60.0   \n",
       "\n",
       "                                                                        unique  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                      \n",
       "grisoni -1  SELFIES    2  256 -1   0.001 256 6000   6     1.0     5999.0 ± 2.0   \n",
       "                                         512 6000   10    1.0     5999.0 ± 2.0   \n",
       "                          512 -1   0.001 256 6000   6     1.0     5998.0 ± 1.0   \n",
       "                                         512 6000   10    1.0     5999.0 ± 2.0   \n",
       "                       3  256 -1   0.001 256 6000   6     1.0     5999.4 ± 0.9   \n",
       "                                         512 6000   10    1.0     5999.4 ± 0.9   \n",
       "                          512 -1   0.001 256 6000   6     1.0     5999.0 ± 1.0   \n",
       "                                         512 6000   10    1.0     5999.6 ± 0.9   \n",
       "            SMILES     2  256 -1   0.001 256 6000   6     1.0  3800.0 ± 1000.0   \n",
       "                                         512 6000   10    1.0   4100.0 ± 100.0   \n",
       "                          512 -1   0.001 256 6000   6     1.0   4600.0 ± 300.0   \n",
       "                                         512 6000   10    1.0    4920.0 ± 70.0   \n",
       "                       3  256 -1   0.001 256 6000   6     1.0  4000.0 ± 1000.0   \n",
       "                                         512 6000   10    1.0   4500.0 ± 100.0   \n",
       "                          512 -1   0.001 256 6000   6     1.0    5110.0 ± 80.0   \n",
       "                                         512 6000   10    1.0    5080.0 ± 50.0   \n",
       "            fragSMILES 2  256  300 0.001 256 6000   8     1.0   4700.0 ± 200.0   \n",
       "                                         512 6000   16    1.0    5000.0 ± 70.0   \n",
       "                          512  300 0.001 256 6000   8     1.0    5200.0 ± 90.0   \n",
       "                                         512 6000   16    1.0    5270.0 ± 30.0   \n",
       "                       3  256  300 0.001 256 6000   8     1.0   4600.0 ± 100.0   \n",
       "                                         512 6000   16    1.0    4800.0 ± 80.0   \n",
       "                          512  300 0.001 256 6000   8     1.0    5040.0 ± 60.0   \n",
       "                                         512 6000   16    1.0    5140.0 ± 40.0   \n",
       "         5  SELFIES    2  256 -1   0.001 256 6000   3     1.0     6000.0 ± 0.0   \n",
       "                                         512 6000   3     1.0     5999.0 ± 1.0   \n",
       "                          512 -1   0.001 256 6000   3     1.0     6000.0 ± 0.0   \n",
       "                                         512 6000   3     1.0     5999.6 ± 0.9   \n",
       "                       3  256 -1   0.001 256 6000   3     1.0     5999.8 ± 0.4   \n",
       "                                         512 6000   3     1.0     5999.8 ± 0.4   \n",
       "                          512 -1   0.001 256 6000   3     1.0     5999.6 ± 0.9   \n",
       "                                         512 6000   3     1.0     6000.0 ± 0.0   \n",
       "            SMILES     2  256 -1   0.001 256 6000   3     1.0   4900.0 ± 100.0   \n",
       "                                         512 6000   3     1.0    4730.0 ± 80.0   \n",
       "                          512 -1   0.001 256 6000   3     1.0    5300.0 ± 40.0   \n",
       "                                         512 6000   3     1.0    5220.0 ± 50.0   \n",
       "                       3  256 -1   0.001 256 6000   3     1.0    5170.0 ± 30.0   \n",
       "                                         512 6000   3     1.0    5020.0 ± 50.0   \n",
       "                          512 -1   0.001 256 6000   3     1.0    5470.0 ± 50.0   \n",
       "                                         512 6000   3     1.0    5390.0 ± 30.0   \n",
       "            fragSMILES 2  256  300 0.001 256 6000   4     1.0    5230.0 ± 70.0   \n",
       "                                         512 6000   4     1.0    5070.0 ± 60.0   \n",
       "                          512  300 0.001 256 6000   4     1.0    5410.0 ± 60.0   \n",
       "                                         512 6000   4     1.0    5310.0 ± 70.0   \n",
       "                       3  256  300 0.001 256 6000   4     1.0    5150.0 ± 30.0   \n",
       "                                         512 6000   4     1.0    4920.0 ± 60.0   \n",
       "                          512  300 0.001 256 6000   4     1.0    5380.0 ± 30.0   \n",
       "                                         512 6000   4     1.0    5190.0 ± 60.0   \n",
       "\n",
       "                                                                         novel  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                      \n",
       "grisoni -1  SELFIES    2  256 -1   0.001 256 6000   6     1.0     5999.0 ± 2.0   \n",
       "                                         512 6000   10    1.0     5998.0 ± 2.0   \n",
       "                          512 -1   0.001 256 6000   6     1.0     5978.0 ± 4.0   \n",
       "                                         512 6000   10    1.0     5971.0 ± 2.0   \n",
       "                       3  256 -1   0.001 256 6000   6     1.0     5998.0 ± 2.0   \n",
       "                                         512 6000   10    1.0     5998.0 ± 2.0   \n",
       "                          512 -1   0.001 256 6000   6     1.0    5986.0 ± 10.0   \n",
       "                                         512 6000   10    1.0     5978.0 ± 9.0   \n",
       "            SMILES     2  256 -1   0.001 256 6000   6     1.0  3800.0 ± 1000.0   \n",
       "                                         512 6000   10    1.0   4100.0 ± 100.0   \n",
       "                          512 -1   0.001 256 6000   6     1.0   4500.0 ± 300.0   \n",
       "                                         512 6000   10    1.0    4770.0 ± 60.0   \n",
       "                       3  256 -1   0.001 256 6000   6     1.0  4000.0 ± 1000.0   \n",
       "                                         512 6000   10    1.0   4500.0 ± 100.0   \n",
       "                          512 -1   0.001 256 6000   6     1.0    5000.0 ± 70.0   \n",
       "                                         512 6000   10    1.0    4910.0 ± 60.0   \n",
       "            fragSMILES 2  256  300 0.001 256 6000   8     1.0   4700.0 ± 200.0   \n",
       "                                         512 6000   16    1.0    4940.0 ± 60.0   \n",
       "                          512  300 0.001 256 6000   8     1.0    5110.0 ± 80.0   \n",
       "                                         512 6000   16    1.0    5110.0 ± 40.0   \n",
       "                       3  256  300 0.001 256 6000   8     1.0   4500.0 ± 100.0   \n",
       "                                         512 6000   16    1.0    4740.0 ± 80.0   \n",
       "                          512  300 0.001 256 6000   8     1.0    4940.0 ± 50.0   \n",
       "                                         512 6000   16    1.0    4950.0 ± 40.0   \n",
       "         5  SELFIES    2  256 -1   0.001 256 6000   3     1.0     5999.0 ± 1.0   \n",
       "                                         512 6000   3     1.0     5999.0 ± 1.0   \n",
       "                          512 -1   0.001 256 6000   3     1.0     5997.0 ± 1.0   \n",
       "                                         512 6000   3     1.0     5998.0 ± 2.0   \n",
       "                       3  256 -1   0.001 256 6000   3     1.0     5999.4 ± 0.9   \n",
       "                                         512 6000   3     1.0     5999.2 ± 0.8   \n",
       "                          512 -1   0.001 256 6000   3     1.0     5997.8 ± 0.8   \n",
       "                                         512 6000   3     1.0     5999.4 ± 0.9   \n",
       "            SMILES     2  256 -1   0.001 256 6000   3     1.0   4900.0 ± 100.0   \n",
       "                                         512 6000   3     1.0    4730.0 ± 80.0   \n",
       "                          512 -1   0.001 256 6000   3     1.0    5280.0 ± 40.0   \n",
       "                                         512 6000   3     1.0    5210.0 ± 50.0   \n",
       "                       3  256 -1   0.001 256 6000   3     1.0    5170.0 ± 30.0   \n",
       "                                         512 6000   3     1.0    5020.0 ± 40.0   \n",
       "                          512 -1   0.001 256 6000   3     1.0    5450.0 ± 50.0   \n",
       "                                         512 6000   3     1.0    5370.0 ± 30.0   \n",
       "            fragSMILES 2  256  300 0.001 256 6000   4     1.0    5180.0 ± 70.0   \n",
       "                                         512 6000   4     1.0    5050.0 ± 60.0   \n",
       "                          512  300 0.001 256 6000   4     1.0    5300.0 ± 60.0   \n",
       "                                         512 6000   4     1.0    5240.0 ± 70.0   \n",
       "                       3  256  300 0.001 256 6000   4     1.0    5090.0 ± 30.0   \n",
       "                                         512 6000   4     1.0    4890.0 ± 50.0   \n",
       "                          512  300 0.001 256 6000   4     1.0    5240.0 ± 30.0   \n",
       "                                         512 6000   4     1.0    5120.0 ± 60.0   \n",
       "\n",
       "                                                                        valid%  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                      \n",
       "grisoni -1  SELFIES    2  256 -1   0.001 256 6000   6     1.0        1.0 ± 0.0   \n",
       "                                         512 6000   10    1.0  0.99997 ± 7e-05   \n",
       "                          512 -1   0.001 256 6000   6     1.0        1.0 ± 0.0   \n",
       "                                         512 6000   10    1.0        1.0 ± 0.0   \n",
       "                       3  256 -1   0.001 256 6000   6     1.0  0.99997 ± 7e-05   \n",
       "                                         512 6000   10    1.0  0.9999 ± 0.0001   \n",
       "                          512 -1   0.001 256 6000   6     1.0        1.0 ± 0.0   \n",
       "                                         512 6000   10    1.0        1.0 ± 0.0   \n",
       "            SMILES     2  256 -1   0.001 256 6000   6     1.0        0.6 ± 0.2   \n",
       "                                         512 6000   10    1.0      0.69 ± 0.02   \n",
       "                          512 -1   0.001 256 6000   6     1.0      0.77 ± 0.05   \n",
       "                                         512 6000   10    1.0      0.82 ± 0.01   \n",
       "                       3  256 -1   0.001 256 6000   6     1.0        0.7 ± 0.2   \n",
       "                                         512 6000   10    1.0      0.75 ± 0.02   \n",
       "                          512 -1   0.001 256 6000   6     1.0      0.85 ± 0.01   \n",
       "                                         512 6000   10    1.0    0.849 ± 0.008   \n",
       "            fragSMILES 2  256  300 0.001 256 6000   8     1.0      0.79 ± 0.03   \n",
       "                                         512 6000   16    1.0      0.83 ± 0.01   \n",
       "                          512  300 0.001 256 6000   8     1.0      0.87 ± 0.01   \n",
       "                                         512 6000   16    1.0     0.88 ± 0.004   \n",
       "                       3  256  300 0.001 256 6000   8     1.0      0.76 ± 0.02   \n",
       "                                         512 6000   16    1.0       0.8 ± 0.01   \n",
       "                          512  300 0.001 256 6000   8     1.0     0.841 ± 0.01   \n",
       "                                         512 6000   16    1.0    0.859 ± 0.008   \n",
       "         5  SELFIES    2  256 -1   0.001 256 6000   3     1.0        1.0 ± 0.0   \n",
       "                                         512 6000   3     1.0  0.99997 ± 7e-05   \n",
       "                          512 -1   0.001 256 6000   3     1.0        1.0 ± 0.0   \n",
       "                                         512 6000   3     1.0        1.0 ± 0.0   \n",
       "                       3  256 -1   0.001 256 6000   3     1.0  0.99997 ± 7e-05   \n",
       "                                         512 6000   3     1.0  0.99997 ± 7e-05   \n",
       "                          512 -1   0.001 256 6000   3     1.0        1.0 ± 0.0   \n",
       "                                         512 6000   3     1.0        1.0 ± 0.0   \n",
       "            SMILES     2  256 -1   0.001 256 6000   3     1.0      0.82 ± 0.02   \n",
       "                                         512 6000   3     1.0      0.79 ± 0.01   \n",
       "                          512 -1   0.001 256 6000   3     1.0    0.883 ± 0.007   \n",
       "                                         512 6000   3     1.0     0.87 ± 0.009   \n",
       "                       3  256 -1   0.001 256 6000   3     1.0    0.862 ± 0.004   \n",
       "                                         512 6000   3     1.0    0.837 ± 0.008   \n",
       "                          512 -1   0.001 256 6000   3     1.0    0.912 ± 0.008   \n",
       "                                         512 6000   3     1.0    0.898 ± 0.005   \n",
       "            fragSMILES 2  256  300 0.001 256 6000   4     1.0      0.87 ± 0.01   \n",
       "                                         512 6000   4     1.0      0.85 ± 0.01   \n",
       "                          512  300 0.001 256 6000   4     1.0    0.903 ± 0.009   \n",
       "                                         512 6000   4     1.0      0.89 ± 0.01   \n",
       "                       3  256  300 0.001 256 6000   4     1.0    0.859 ± 0.005   \n",
       "                                         512 6000   4     1.0    0.821 ± 0.009   \n",
       "                          512  300 0.001 256 6000   4     1.0    0.899 ± 0.005   \n",
       "                                         512 6000   4     1.0      0.87 ± 0.01   \n",
       "\n",
       "                                                                       unique%  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                      \n",
       "grisoni -1  SELFIES    2  256 -1   0.001 256 6000   6     1.0  0.9998 ± 0.0003   \n",
       "                                         512 6000   10    1.0  0.9999 ± 0.0003   \n",
       "                          512 -1   0.001 256 6000   6     1.0  0.9997 ± 0.0002   \n",
       "                                         512 6000   10    1.0  0.9998 ± 0.0003   \n",
       "                       3  256 -1   0.001 256 6000   6     1.0  0.9999 ± 0.0001   \n",
       "                                         512 6000   10    1.0        1.0 ± 0.0   \n",
       "                          512 -1   0.001 256 6000   6     1.0  0.9999 ± 0.0002   \n",
       "                                         512 6000   10    1.0  0.9999 ± 0.0001   \n",
       "            SMILES     2  256 -1   0.001 256 6000   6     1.0  0.9998 ± 0.0003   \n",
       "                                         512 6000   10    1.0  0.9999 ± 0.0002   \n",
       "                          512 -1   0.001 256 6000   6     1.0  0.9989 ± 0.0005   \n",
       "                                         512 6000   10    1.0    0.997 ± 0.001   \n",
       "                       3  256 -1   0.001 256 6000   6     1.0  0.9997 ± 0.0004   \n",
       "                                         512 6000   10    1.0  0.9996 ± 0.0002   \n",
       "                          512 -1   0.001 256 6000   6     1.0  0.9979 ± 0.0007   \n",
       "                                         512 6000   10    1.0  0.9976 ± 0.0009   \n",
       "            fragSMILES 2  256  300 0.001 256 6000   8     1.0  0.9997 ± 0.0007   \n",
       "                                         512 6000   16    1.0   0.9992 ± 0.001   \n",
       "                          512  300 0.001 256 6000   8     1.0  0.9993 ± 0.0003   \n",
       "                                         512 6000   16    1.0  0.9976 ± 0.0005   \n",
       "                       3  256  300 0.001 256 6000   8     1.0  0.9997 ± 0.0005   \n",
       "                                         512 6000   16    1.0  0.9994 ± 0.0005   \n",
       "                          512  300 0.001 256 6000   8     1.0  0.9986 ± 0.0009   \n",
       "                                         512 6000   16    1.0    0.997 ± 0.002   \n",
       "         5  SELFIES    2  256 -1   0.001 256 6000   3     1.0        1.0 ± 0.0   \n",
       "                                         512 6000   3     1.0  0.9998 ± 0.0002   \n",
       "                          512 -1   0.001 256 6000   3     1.0        1.0 ± 0.0   \n",
       "                                         512 6000   3     1.0  0.9999 ± 0.0001   \n",
       "                       3  256 -1   0.001 256 6000   3     1.0        1.0 ± 0.0   \n",
       "                                         512 6000   3     1.0        1.0 ± 0.0   \n",
       "                          512 -1   0.001 256 6000   3     1.0  0.9999 ± 0.0001   \n",
       "                                         512 6000   3     1.0        1.0 ± 0.0   \n",
       "            SMILES     2  256 -1   0.001 256 6000   3     1.0        1.0 ± 0.0   \n",
       "                                         512 6000   3     1.0  0.9998 ± 0.0003   \n",
       "                          512 -1   0.001 256 6000   3     1.0  0.9999 ± 0.0002   \n",
       "                                         512 6000   3     1.0        1.0 ± 0.0   \n",
       "                       3  256 -1   0.001 256 6000   3     1.0        1.0 ± 0.0   \n",
       "                                         512 6000   3     1.0        1.0 ± 0.0   \n",
       "                          512 -1   0.001 256 6000   3     1.0        1.0 ± 0.0   \n",
       "                                         512 6000   3     1.0        1.0 ± 0.0   \n",
       "            fragSMILES 2  256  300 0.001 256 6000   4     1.0    0.999 ± 0.001   \n",
       "                                         512 6000   4     1.0  0.9994 ± 0.0006   \n",
       "                          512  300 0.001 256 6000   4     1.0  0.9991 ± 0.0006   \n",
       "                                         512 6000   4     1.0  0.9995 ± 0.0004   \n",
       "                       3  256  300 0.001 256 6000   4     1.0   0.999 ± 0.0008   \n",
       "                                         512 6000   4     1.0  0.9993 ± 0.0006   \n",
       "                          512  300 0.001 256 6000   4     1.0  0.9986 ± 0.0007   \n",
       "                                         512 6000   4     1.0  0.9984 ± 0.0009   \n",
       "\n",
       "                                                                        novel%  \n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                     \n",
       "grisoni -1  SELFIES    2  256 -1   0.001 256 6000   6     1.0  0.99997 ± 7e-05  \n",
       "                                         512 6000   10    1.0  0.9998 ± 0.0001  \n",
       "                          512 -1   0.001 256 6000   6     1.0  0.9967 ± 0.0005  \n",
       "                                         512 6000   10    1.0  0.9953 ± 0.0006  \n",
       "                       3  256 -1   0.001 256 6000   6     1.0  0.9998 ± 0.0002  \n",
       "                                         512 6000   10    1.0  0.9998 ± 0.0003  \n",
       "                          512 -1   0.001 256 6000   6     1.0    0.998 ± 0.002  \n",
       "                                         512 6000   10    1.0    0.996 ± 0.001  \n",
       "            SMILES     2  256 -1   0.001 256 6000   6     1.0  0.9988 ± 0.0005  \n",
       "                                         512 6000   10    1.0  0.9985 ± 0.0008  \n",
       "                          512 -1   0.001 256 6000   6     1.0    0.984 ± 0.006  \n",
       "                                         512 6000   10    1.0     0.97 ± 0.002  \n",
       "                       3  256 -1   0.001 256 6000   6     1.0   0.9982 ± 0.001  \n",
       "                                         512 6000   10    1.0    0.998 ± 0.001  \n",
       "                          512 -1   0.001 256 6000   6     1.0    0.978 ± 0.003  \n",
       "                                         512 6000   10    1.0    0.966 ± 0.003  \n",
       "            fragSMILES 2  256  300 0.001 256 6000   8     1.0    0.993 ± 0.002  \n",
       "                                         512 6000   16    1.0    0.988 ± 0.002  \n",
       "                          512  300 0.001 256 6000   8     1.0    0.984 ± 0.003  \n",
       "                                         512 6000   16    1.0    0.971 ± 0.004  \n",
       "                       3  256  300 0.001 256 6000   8     1.0    0.993 ± 0.001  \n",
       "                                         512 6000   16    1.0    0.987 ± 0.002  \n",
       "                          512  300 0.001 256 6000   8     1.0    0.979 ± 0.003  \n",
       "                                         512 6000   16    1.0    0.963 ± 0.003  \n",
       "         5  SELFIES    2  256 -1   0.001 256 6000   3     1.0  0.9999 ± 0.0002  \n",
       "                                         512 6000   3     1.0        1.0 ± 0.0  \n",
       "                          512 -1   0.001 256 6000   3     1.0  0.9996 ± 0.0002  \n",
       "                                         512 6000   3     1.0  0.9998 ± 0.0002  \n",
       "                       3  256 -1   0.001 256 6000   3     1.0  0.99993 ± 9e-05  \n",
       "                                         512 6000   3     1.0   0.9999 ± 9e-05  \n",
       "                          512 -1   0.001 256 6000   3     1.0  0.9997 ± 0.0001  \n",
       "                                         512 6000   3     1.0  0.9999 ± 0.0001  \n",
       "            SMILES     2  256 -1   0.001 256 6000   3     1.0  0.99963 ± 9e-05  \n",
       "                                         512 6000   3     1.0  0.9997 ± 0.0002  \n",
       "                          512 -1   0.001 256 6000   3     1.0   0.997 ± 0.0009  \n",
       "                                         512 6000   3     1.0  0.9982 ± 0.0006  \n",
       "                       3  256 -1   0.001 256 6000   3     1.0  0.9993 ± 0.0005  \n",
       "                                         512 6000   3     1.0  0.9995 ± 0.0004  \n",
       "                          512 -1   0.001 256 6000   3     1.0  0.9969 ± 0.0007  \n",
       "                                         512 6000   3     1.0   0.998 ± 0.0006  \n",
       "            fragSMILES 2  256  300 0.001 256 6000   4     1.0    0.991 ± 0.002  \n",
       "                                         512 6000   4     1.0  0.9954 ± 0.0008  \n",
       "                          512  300 0.001 256 6000   4     1.0     0.98 ± 0.003  \n",
       "                                         512 6000   4     1.0    0.987 ± 0.002  \n",
       "                       3  256  300 0.001 256 6000   4     1.0    0.988 ± 0.002  \n",
       "                                         512 6000   4     1.0    0.994 ± 0.001  \n",
       "                          512  300 0.001 256 6000   4     1.0    0.973 ± 0.003  \n",
       "                                         512 6000   4     1.0    0.985 ± 0.002  "
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "genStats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>valid</th>\n",
       "      <th>unique</th>\n",
       "      <th>novel</th>\n",
       "      <th>valid%</th>\n",
       "      <th>unique%</th>\n",
       "      <th>novel%</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th>aug</th>\n",
       "      <th>notation</th>\n",
       "      <th>hl</th>\n",
       "      <th>hu</th>\n",
       "      <th>es</th>\n",
       "      <th>lr</th>\n",
       "      <th>bs</th>\n",
       "      <th>amount</th>\n",
       "      <th>epoch</th>\n",
       "      <th>T</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"6\" valign=\"top\">grisoni</th>\n",
       "      <th rowspan=\"3\" valign=\"top\">-1</th>\n",
       "      <th>SELFIES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>-1</th>\n",
       "      <th>0.001</th>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>5999.0 ± 2.0</td>\n",
       "      <td>5971.0 ± 2.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9998 ± 0.0003</td>\n",
       "      <td>0.9953 ± 0.0006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>-1</th>\n",
       "      <th>0.001</th>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>4930.0 ± 70.0</td>\n",
       "      <td>4920.0 ± 70.0</td>\n",
       "      <td>4770.0 ± 60.0</td>\n",
       "      <td>0.82 ± 0.01</td>\n",
       "      <td>0.997 ± 0.001</td>\n",
       "      <td>0.97 ± 0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fragSMILES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>300</th>\n",
       "      <th>0.001</th>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>16</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5280.0 ± 20.0</td>\n",
       "      <td>5270.0 ± 30.0</td>\n",
       "      <td>5110.0 ± 40.0</td>\n",
       "      <td>0.88 ± 0.004</td>\n",
       "      <td>0.9976 ± 0.0005</td>\n",
       "      <td>0.971 ± 0.004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">5</th>\n",
       "      <th>SELFIES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>-1</th>\n",
       "      <th>0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>6000.0 ± 0.0</td>\n",
       "      <td>5997.0 ± 1.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9996 ± 0.0002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>-1</th>\n",
       "      <th>0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5300.0 ± 40.0</td>\n",
       "      <td>5300.0 ± 40.0</td>\n",
       "      <td>5280.0 ± 40.0</td>\n",
       "      <td>0.883 ± 0.007</td>\n",
       "      <td>0.9999 ± 0.0002</td>\n",
       "      <td>0.997 ± 0.0009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fragSMILES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>300</th>\n",
       "      <th>0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>4</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5420.0 ± 60.0</td>\n",
       "      <td>5410.0 ± 60.0</td>\n",
       "      <td>5300.0 ± 60.0</td>\n",
       "      <td>0.903 ± 0.009</td>\n",
       "      <td>0.9991 ± 0.0006</td>\n",
       "      <td>0.98 ± 0.003</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                       valid  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                    \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0   6000.0 ± 0.0   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0  4930.0 ± 70.0   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0  5280.0 ± 20.0   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0   6000.0 ± 0.0   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  5300.0 ± 40.0   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  5420.0 ± 60.0   \n",
       "\n",
       "                                                                      unique  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                    \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0   5999.0 ± 2.0   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0  4920.0 ± 70.0   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0  5270.0 ± 30.0   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0   6000.0 ± 0.0   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  5300.0 ± 40.0   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  5410.0 ± 60.0   \n",
       "\n",
       "                                                                       novel  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                    \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0   5971.0 ± 2.0   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0  4770.0 ± 60.0   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0  5110.0 ± 40.0   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0   5997.0 ± 1.0   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  5280.0 ± 40.0   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  5300.0 ± 60.0   \n",
       "\n",
       "                                                                      valid%  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                    \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0      1.0 ± 0.0   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0    0.82 ± 0.01   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0   0.88 ± 0.004   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0      1.0 ± 0.0   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  0.883 ± 0.007   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  0.903 ± 0.009   \n",
       "\n",
       "                                                                       unique%  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                      \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0  0.9998 ± 0.0003   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0    0.997 ± 0.001   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0  0.9976 ± 0.0005   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0        1.0 ± 0.0   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  0.9999 ± 0.0002   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  0.9991 ± 0.0006   \n",
       "\n",
       "                                                                        novel%  \n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                     \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0  0.9953 ± 0.0006  \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0     0.97 ± 0.002  \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0    0.971 ± 0.004  \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  0.9996 ± 0.0002  \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0   0.997 ± 0.0009  \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0     0.98 ± 0.003  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "genStats.query(' ( aug==5 and hl==2 and hu==512 and bs==256 ) or ( aug!=5 and hl==2 and hu==512 and bs==512 ) ')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/scipy/stats/_axis_nan_policy.py:523: RuntimeWarning: Precision loss occurred in moment calculation due to catastrophic cancellation. This occurs when the data are nearly identical. Results may be unreliable.\n",
      "  res = hypotest_fun_out(*samples, **kwds)\n"
     ]
    }
   ],
   "source": [
    "genPvals = pd.DataFrame.from_dict(\n",
    "                {notation+'x1': \n",
    "                {metric:stats.ttest_ind(\n",
    "                genGrp.get_group(('grisoni',-1,notation,*params))[metric],\n",
    "                genGrp.get_group(('grisoni',-1,'fragSMILES',*paramsFrag))[metric],\n",
    "                ).pvalue for metric in genStats.columns} for notation in ['SMILES','SELFIES']\n",
    "                } |\n",
    "                {notation+'x5': \n",
    "                {metric:stats.ttest_ind(\n",
    "                genGrp.get_group(('grisoni',5,notation,*params5))[metric],\n",
    "                genGrp.get_group(('grisoni',5,'fragSMILES',*paramsFrag5))[metric],\n",
    "                ).pvalue for metric in genStats.columns} for notation in ['SMILES','SELFIES']\n",
    "                }\n",
    "                , orient='index'\n",
    "            )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>valid</th>\n",
       "      <th>unique</th>\n",
       "      <th>novel</th>\n",
       "      <th>valid%</th>\n",
       "      <th>unique%</th>\n",
       "      <th>novel%</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>SMILESx1</th>\n",
       "      <td>5.836981e-06</td>\n",
       "      <td>4.036319e-06</td>\n",
       "      <td>6.399316e-06</td>\n",
       "      <td>5.836981e-06</td>\n",
       "      <td>0.517643</td>\n",
       "      <td>8.938461e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SELFIESx1</th>\n",
       "      <td>2.642000e-12</td>\n",
       "      <td>3.675778e-12</td>\n",
       "      <td>8.476472e-11</td>\n",
       "      <td>2.642000e-12</td>\n",
       "      <td>0.000030</td>\n",
       "      <td>1.109582e-06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILESx5</th>\n",
       "      <td>5.479541e-03</td>\n",
       "      <td>6.740004e-03</td>\n",
       "      <td>4.799941e-01</td>\n",
       "      <td>5.479541e-03</td>\n",
       "      <td>0.020852</td>\n",
       "      <td>1.086421e-06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SELFIESx5</th>\n",
       "      <td>1.224551e-08</td>\n",
       "      <td>1.288487e-08</td>\n",
       "      <td>3.641024e-09</td>\n",
       "      <td>1.224551e-08</td>\n",
       "      <td>0.011307</td>\n",
       "      <td>2.598607e-07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  valid        unique         novel        valid%   unique%  \\\n",
       "SMILESx1   5.836981e-06  4.036319e-06  6.399316e-06  5.836981e-06  0.517643   \n",
       "SELFIESx1  2.642000e-12  3.675778e-12  8.476472e-11  2.642000e-12  0.000030   \n",
       "SMILESx5   5.479541e-03  6.740004e-03  4.799941e-01  5.479541e-03  0.020852   \n",
       "SELFIESx5  1.224551e-08  1.288487e-08  3.641024e-09  1.224551e-08  0.011307   \n",
       "\n",
       "                 novel%  \n",
       "SMILESx1   8.938461e-01  \n",
       "SELFIESx1  1.109582e-06  \n",
       "SMILESx5   1.086421e-06  \n",
       "SELFIESx5  2.598607e-07  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "genPvals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>valid</th>\n",
       "      <th>unique</th>\n",
       "      <th>novel</th>\n",
       "      <th>valid%</th>\n",
       "      <th>unique%</th>\n",
       "      <th>novel%</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>SMILESx1</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SELFIESx1</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILESx5</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SELFIESx5</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           valid  unique  novel  valid%  unique%  novel%\n",
       "SMILESx1    True    True   True    True    False   False\n",
       "SELFIESx1   True    True   True    True     True    True\n",
       "SMILESx5    True    True  False    True     True    True\n",
       "SELFIESx5   True    True   True    True     True    True"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "genPvals.map(lambda x: x<0.05)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "genStats.to_csv('results/07_grisoniSyntax.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Chiral Metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "genChirDFs=pd.concat( [evaluator.getChiralResultsGens() for evaluator in evaluators if not evaluator.gens.empty], ignore_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>dataset</th>\n",
       "      <th>aug</th>\n",
       "      <th>notation</th>\n",
       "      <th>hl</th>\n",
       "      <th>hu</th>\n",
       "      <th>bs</th>\n",
       "      <th>es</th>\n",
       "      <th>lr</th>\n",
       "      <th>fold</th>\n",
       "      <th>amount</th>\n",
       "      <th>epoch</th>\n",
       "      <th>T</th>\n",
       "      <th>chirals</th>\n",
       "      <th>invalid</th>\n",
       "      <th>valid</th>\n",
       "      <th>unique</th>\n",
       "      <th>novel</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>grisoni</td>\n",
       "      <td>5</td>\n",
       "      <td>SELFIES</td>\n",
       "      <td>2</td>\n",
       "      <td>256</td>\n",
       "      <td>512</td>\n",
       "      <td>None</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0</td>\n",
       "      <td>6000</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1694</td>\n",
       "      <td>647</td>\n",
       "      <td>1047</td>\n",
       "      <td>1047</td>\n",
       "      <td>1047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>grisoni</td>\n",
       "      <td>5</td>\n",
       "      <td>SELFIES</td>\n",
       "      <td>2</td>\n",
       "      <td>256</td>\n",
       "      <td>512</td>\n",
       "      <td>None</td>\n",
       "      <td>0.001</td>\n",
       "      <td>1</td>\n",
       "      <td>6000</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1820</td>\n",
       "      <td>753</td>\n",
       "      <td>1067</td>\n",
       "      <td>1067</td>\n",
       "      <td>1067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>grisoni</td>\n",
       "      <td>5</td>\n",
       "      <td>SELFIES</td>\n",
       "      <td>2</td>\n",
       "      <td>256</td>\n",
       "      <td>512</td>\n",
       "      <td>None</td>\n",
       "      <td>0.001</td>\n",
       "      <td>2</td>\n",
       "      <td>6000</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1818</td>\n",
       "      <td>725</td>\n",
       "      <td>1093</td>\n",
       "      <td>1093</td>\n",
       "      <td>1093</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>grisoni</td>\n",
       "      <td>5</td>\n",
       "      <td>SELFIES</td>\n",
       "      <td>2</td>\n",
       "      <td>256</td>\n",
       "      <td>512</td>\n",
       "      <td>None</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3</td>\n",
       "      <td>6000</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1860</td>\n",
       "      <td>766</td>\n",
       "      <td>1094</td>\n",
       "      <td>1094</td>\n",
       "      <td>1094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>grisoni</td>\n",
       "      <td>5</td>\n",
       "      <td>SELFIES</td>\n",
       "      <td>2</td>\n",
       "      <td>256</td>\n",
       "      <td>512</td>\n",
       "      <td>None</td>\n",
       "      <td>0.001</td>\n",
       "      <td>4</td>\n",
       "      <td>6000</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1867</td>\n",
       "      <td>733</td>\n",
       "      <td>1134</td>\n",
       "      <td>1134</td>\n",
       "      <td>1134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>235</th>\n",
       "      <td>grisoni</td>\n",
       "      <td>None</td>\n",
       "      <td>fragSMILES</td>\n",
       "      <td>3</td>\n",
       "      <td>512</td>\n",
       "      <td>512</td>\n",
       "      <td>300</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0</td>\n",
       "      <td>6000</td>\n",
       "      <td>16</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1751</td>\n",
       "      <td>347</td>\n",
       "      <td>1404</td>\n",
       "      <td>1400</td>\n",
       "      <td>1356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>236</th>\n",
       "      <td>grisoni</td>\n",
       "      <td>None</td>\n",
       "      <td>fragSMILES</td>\n",
       "      <td>3</td>\n",
       "      <td>512</td>\n",
       "      <td>512</td>\n",
       "      <td>300</td>\n",
       "      <td>0.001</td>\n",
       "      <td>1</td>\n",
       "      <td>6000</td>\n",
       "      <td>16</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1762</td>\n",
       "      <td>400</td>\n",
       "      <td>1362</td>\n",
       "      <td>1360</td>\n",
       "      <td>1311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>237</th>\n",
       "      <td>grisoni</td>\n",
       "      <td>None</td>\n",
       "      <td>fragSMILES</td>\n",
       "      <td>3</td>\n",
       "      <td>512</td>\n",
       "      <td>512</td>\n",
       "      <td>300</td>\n",
       "      <td>0.001</td>\n",
       "      <td>2</td>\n",
       "      <td>6000</td>\n",
       "      <td>16</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1928</td>\n",
       "      <td>373</td>\n",
       "      <td>1555</td>\n",
       "      <td>1545</td>\n",
       "      <td>1476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>238</th>\n",
       "      <td>grisoni</td>\n",
       "      <td>None</td>\n",
       "      <td>fragSMILES</td>\n",
       "      <td>3</td>\n",
       "      <td>512</td>\n",
       "      <td>512</td>\n",
       "      <td>300</td>\n",
       "      <td>0.001</td>\n",
       "      <td>3</td>\n",
       "      <td>6000</td>\n",
       "      <td>16</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1795</td>\n",
       "      <td>417</td>\n",
       "      <td>1378</td>\n",
       "      <td>1374</td>\n",
       "      <td>1325</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>239</th>\n",
       "      <td>grisoni</td>\n",
       "      <td>None</td>\n",
       "      <td>fragSMILES</td>\n",
       "      <td>3</td>\n",
       "      <td>512</td>\n",
       "      <td>512</td>\n",
       "      <td>300</td>\n",
       "      <td>0.001</td>\n",
       "      <td>4</td>\n",
       "      <td>6000</td>\n",
       "      <td>16</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1732</td>\n",
       "      <td>374</td>\n",
       "      <td>1358</td>\n",
       "      <td>1354</td>\n",
       "      <td>1296</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>240 rows × 17 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     dataset   aug    notation  hl   hu   bs    es     lr  fold  amount  \\\n",
       "0    grisoni     5     SELFIES   2  256  512  None  0.001     0    6000   \n",
       "1    grisoni     5     SELFIES   2  256  512  None  0.001     1    6000   \n",
       "2    grisoni     5     SELFIES   2  256  512  None  0.001     2    6000   \n",
       "3    grisoni     5     SELFIES   2  256  512  None  0.001     3    6000   \n",
       "4    grisoni     5     SELFIES   2  256  512  None  0.001     4    6000   \n",
       "..       ...   ...         ...  ..  ...  ...   ...    ...   ...     ...   \n",
       "235  grisoni  None  fragSMILES   3  512  512   300  0.001     0    6000   \n",
       "236  grisoni  None  fragSMILES   3  512  512   300  0.001     1    6000   \n",
       "237  grisoni  None  fragSMILES   3  512  512   300  0.001     2    6000   \n",
       "238  grisoni  None  fragSMILES   3  512  512   300  0.001     3    6000   \n",
       "239  grisoni  None  fragSMILES   3  512  512   300  0.001     4    6000   \n",
       "\n",
       "     epoch    T  chirals  invalid  valid  unique  novel  \n",
       "0        3  1.0     1694      647   1047    1047   1047  \n",
       "1        3  1.0     1820      753   1067    1067   1067  \n",
       "2        3  1.0     1818      725   1093    1093   1093  \n",
       "3        3  1.0     1860      766   1094    1094   1094  \n",
       "4        3  1.0     1867      733   1134    1134   1134  \n",
       "..     ...  ...      ...      ...    ...     ...    ...  \n",
       "235     16  1.0     1751      347   1404    1400   1356  \n",
       "236     16  1.0     1762      400   1362    1360   1311  \n",
       "237     16  1.0     1928      373   1555    1545   1476  \n",
       "238     16  1.0     1795      417   1378    1374   1325  \n",
       "239     16  1.0     1732      374   1358    1354   1296  \n",
       "\n",
       "[240 rows x 17 columns]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "genChirDFs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "genChirDFs['invalid%']=genChirDFs['invalid']/genChirDFs['chirals']\n",
    "genChirDFs['valid%']=genChirDFs['valid']/genChirDFs['chirals']\n",
    "genChirDFs['unique%']=genChirDFs['unique']/genChirDFs['valid']\n",
    "genChirDFs['novel%']=genChirDFs['novel']/genChirDFs['unique']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_1315920/3061462175.py:1: FutureWarning: Downcasting object dtype arrays on .fillna, .ffill, .bfill is deprecated and will change in a future version. Call result.infer_objects(copy=False) instead. To opt-in to the future behavior, set `pd.set_option('future.no_silent_downcasting', True)`\n",
      "  genChirGrp=genChirDFs.fillna(-1).groupby(compilerPatterns.index.drop('fold').tolist(),as_index=True, dropna=False)\n"
     ]
    }
   ],
   "source": [
    "genChirGrp=genChirDFs.fillna(-1).groupby(compilerPatterns.index.drop('fold').tolist(),as_index=True, dropna=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "genChirStats=genChirGrp.describe().loc(axis=1)[ ( slice(None),['mean','std'] ) ].drop(columns='fold')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "genChirStats=ResultsFromMeanStdDF(genChirStats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "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></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>chirals</th>\n",
       "      <th>invalid</th>\n",
       "      <th>valid</th>\n",
       "      <th>unique</th>\n",
       "      <th>novel</th>\n",
       "      <th>invalid%</th>\n",
       "      <th>valid%</th>\n",
       "      <th>unique%</th>\n",
       "      <th>novel%</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th>aug</th>\n",
       "      <th>notation</th>\n",
       "      <th>hl</th>\n",
       "      <th>hu</th>\n",
       "      <th>es</th>\n",
       "      <th>lr</th>\n",
       "      <th>bs</th>\n",
       "      <th>amount</th>\n",
       "      <th>epoch</th>\n",
       "      <th>T</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"6\" valign=\"top\">grisoni</th>\n",
       "      <th rowspan=\"3\" valign=\"top\">-1</th>\n",
       "      <th>SELFIES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>-1</th>\n",
       "      <th>0.001</th>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1820.0 ± 40.0</td>\n",
       "      <td>670.0 ± 40.0</td>\n",
       "      <td>1150.0 ± 20.0</td>\n",
       "      <td>1150.0 ± 20.0</td>\n",
       "      <td>1140.0 ± 20.0</td>\n",
       "      <td>0.37 ± 0.01</td>\n",
       "      <td>0.63 ± 0.01</td>\n",
       "      <td>0.9996 ± 0.0008</td>\n",
       "      <td>0.9938 ± 0.0007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>-1</th>\n",
       "      <th>0.001</th>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1770.0 ± 70.0</td>\n",
       "      <td>400.0 ± 40.0</td>\n",
       "      <td>1370.0 ± 40.0</td>\n",
       "      <td>1370.0 ± 40.0</td>\n",
       "      <td>1320.0 ± 40.0</td>\n",
       "      <td>0.22 ± 0.02</td>\n",
       "      <td>0.78 ± 0.02</td>\n",
       "      <td>0.996 ± 0.001</td>\n",
       "      <td>0.964 ± 0.003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fragSMILES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>300</th>\n",
       "      <th>0.001</th>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>16</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1770.0 ± 90.0</td>\n",
       "      <td>330.0 ± 30.0</td>\n",
       "      <td>1440.0 ± 70.0</td>\n",
       "      <td>1440.0 ± 60.0</td>\n",
       "      <td>1400.0 ± 60.0</td>\n",
       "      <td>0.19 ± 0.01</td>\n",
       "      <td>0.81 ± 0.01</td>\n",
       "      <td>0.996 ± 0.002</td>\n",
       "      <td>0.97 ± 0.008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">5</th>\n",
       "      <th>SELFIES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>-1</th>\n",
       "      <th>0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1900.0 ± 100.0</td>\n",
       "      <td>520.0 ± 40.0</td>\n",
       "      <td>1380.0 ± 80.0</td>\n",
       "      <td>1380.0 ± 80.0</td>\n",
       "      <td>1370.0 ± 80.0</td>\n",
       "      <td>0.274 ± 0.009</td>\n",
       "      <td>0.726 ± 0.009</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.9991 ± 0.0006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>-1</th>\n",
       "      <th>0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>1800.0 ± 200.0</td>\n",
       "      <td>320.0 ± 50.0</td>\n",
       "      <td>1500.0 ± 100.0</td>\n",
       "      <td>1500.0 ± 100.0</td>\n",
       "      <td>1500.0 ± 100.0</td>\n",
       "      <td>0.17 ± 0.01</td>\n",
       "      <td>0.83 ± 0.01</td>\n",
       "      <td>1.0 ± 0.0</td>\n",
       "      <td>0.997 ± 0.001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fragSMILES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>300</th>\n",
       "      <th>0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>4</th>\n",
       "      <th>1.0</th>\n",
       "      <td>2000.0 ± 100.0</td>\n",
       "      <td>290.0 ± 30.0</td>\n",
       "      <td>1700.0 ± 100.0</td>\n",
       "      <td>1700.0 ± 100.0</td>\n",
       "      <td>1600.0 ± 100.0</td>\n",
       "      <td>0.15 ± 0.01</td>\n",
       "      <td>0.85 ± 0.01</td>\n",
       "      <td>0.9988 ± 0.0009</td>\n",
       "      <td>0.977 ± 0.003</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                      chirals  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                     \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0   1820.0 ± 40.0   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0   1770.0 ± 70.0   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0   1770.0 ± 90.0   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  1900.0 ± 100.0   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  1800.0 ± 200.0   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  2000.0 ± 100.0   \n",
       "\n",
       "                                                                    invalid  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                   \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0  670.0 ± 40.0   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0  400.0 ± 40.0   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0  330.0 ± 30.0   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  520.0 ± 40.0   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  320.0 ± 50.0   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  290.0 ± 30.0   \n",
       "\n",
       "                                                                        valid  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                     \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0   1150.0 ± 20.0   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0   1370.0 ± 40.0   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0   1440.0 ± 70.0   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0   1380.0 ± 80.0   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  1500.0 ± 100.0   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  1700.0 ± 100.0   \n",
       "\n",
       "                                                                       unique  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                     \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0   1150.0 ± 20.0   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0   1370.0 ± 40.0   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0   1440.0 ± 60.0   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0   1380.0 ± 80.0   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  1500.0 ± 100.0   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  1700.0 ± 100.0   \n",
       "\n",
       "                                                                        novel  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                     \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0   1140.0 ± 20.0   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0   1320.0 ± 40.0   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0   1400.0 ± 60.0   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0   1370.0 ± 80.0   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  1500.0 ± 100.0   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  1600.0 ± 100.0   \n",
       "\n",
       "                                                                    invalid%  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                    \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0    0.37 ± 0.01   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0    0.22 ± 0.02   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0    0.19 ± 0.01   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  0.274 ± 0.009   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0    0.17 ± 0.01   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0    0.15 ± 0.01   \n",
       "\n",
       "                                                                      valid%  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                    \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0    0.63 ± 0.01   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0    0.78 ± 0.02   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0    0.81 ± 0.01   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  0.726 ± 0.009   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0    0.83 ± 0.01   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0    0.85 ± 0.01   \n",
       "\n",
       "                                                                       unique%  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                      \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0  0.9996 ± 0.0008   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0    0.996 ± 0.001   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0    0.996 ± 0.002   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0        1.0 ± 0.0   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0        1.0 ± 0.0   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  0.9988 ± 0.0009   \n",
       "\n",
       "                                                                        novel%  \n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                     \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0  0.9938 ± 0.0007  \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0    0.964 ± 0.003  \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0     0.97 ± 0.008  \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  0.9991 ± 0.0006  \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0    0.997 ± 0.001  \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0    0.977 ± 0.003  "
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "genChirStats.query(' ( aug==5 and hl==2 and hu==512 and bs==256 ) or ( aug!=5 and hl==2 and hu==512 and bs==512 ) ')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3/dist-packages/scipy/stats/_axis_nan_policy.py:523: RuntimeWarning: Precision loss occurred in moment calculation due to catastrophic cancellation. This occurs when the data are nearly identical. Results may be unreliable.\n",
      "  res = hypotest_fun_out(*samples, **kwds)\n"
     ]
    }
   ],
   "source": [
    "genChirPvals = pd.DataFrame.from_dict(\n",
    "                {notation+'x1': \n",
    "                {metric:stats.ttest_ind(\n",
    "                genChirGrp.get_group(('grisoni',-1,notation,*params))[metric],\n",
    "                genChirGrp.get_group(('grisoni',-1,'fragSMILES',*paramsFrag))[metric],\n",
    "                ).pvalue for metric in genChirStats.columns} for notation in ['SELFIES','SMILES']\n",
    "                } |\n",
    "                {notation+'x5': \n",
    "                {metric:stats.ttest_ind(\n",
    "                genChirGrp.get_group(('grisoni',5,notation,*params5))[metric],\n",
    "                genChirGrp.get_group(('grisoni',5,'fragSMILES',*paramsFrag5))[metric],\n",
    "                ).pvalue for metric in genChirStats.columns} for notation in ['SELFIES','SMILES']\n",
    "                }\n",
    "                , orient='index'\n",
    "            )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "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>chirals</th>\n",
       "      <th>invalid</th>\n",
       "      <th>valid</th>\n",
       "      <th>unique</th>\n",
       "      <th>novel</th>\n",
       "      <th>invalid%</th>\n",
       "      <th>valid%</th>\n",
       "      <th>unique%</th>\n",
       "      <th>novel%</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>SELFIESx1</th>\n",
       "      <td>0.335176</td>\n",
       "      <td>4.502110e-07</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>0.000013</td>\n",
       "      <td>2.852547e-08</td>\n",
       "      <td>2.852547e-08</td>\n",
       "      <td>0.003060</td>\n",
       "      <td>1.320359e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILESx1</th>\n",
       "      <td>0.918940</td>\n",
       "      <td>1.951212e-02</td>\n",
       "      <td>0.075640</td>\n",
       "      <td>0.070225</td>\n",
       "      <td>0.039030</td>\n",
       "      <td>2.784226e-03</td>\n",
       "      <td>2.784226e-03</td>\n",
       "      <td>0.947904</td>\n",
       "      <td>1.124316e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SELFIESx5</th>\n",
       "      <td>0.418123</td>\n",
       "      <td>1.126787e-05</td>\n",
       "      <td>0.001292</td>\n",
       "      <td>0.001373</td>\n",
       "      <td>0.002686</td>\n",
       "      <td>8.873469e-08</td>\n",
       "      <td>8.873469e-08</td>\n",
       "      <td>0.013913</td>\n",
       "      <td>5.176287e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILESx5</th>\n",
       "      <td>0.238558</td>\n",
       "      <td>4.282225e-01</td>\n",
       "      <td>0.073160</td>\n",
       "      <td>0.076939</td>\n",
       "      <td>0.152536</td>\n",
       "      <td>3.088840e-02</td>\n",
       "      <td>3.088840e-02</td>\n",
       "      <td>0.013913</td>\n",
       "      <td>1.895506e-06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            chirals       invalid     valid    unique     novel      invalid%  \\\n",
       "SELFIESx1  0.335176  4.502110e-07  0.000010  0.000010  0.000013  2.852547e-08   \n",
       "SMILESx1   0.918940  1.951212e-02  0.075640  0.070225  0.039030  2.784226e-03   \n",
       "SELFIESx5  0.418123  1.126787e-05  0.001292  0.001373  0.002686  8.873469e-08   \n",
       "SMILESx5   0.238558  4.282225e-01  0.073160  0.076939  0.152536  3.088840e-02   \n",
       "\n",
       "                 valid%   unique%        novel%  \n",
       "SELFIESx1  2.852547e-08  0.003060  1.320359e-04  \n",
       "SMILESx1   2.784226e-03  0.947904  1.124316e-01  \n",
       "SELFIESx5  8.873469e-08  0.013913  5.176287e-07  \n",
       "SMILESx5   3.088840e-02  0.013913  1.895506e-06  "
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "genChirPvals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "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>chirals</th>\n",
       "      <th>invalid</th>\n",
       "      <th>valid</th>\n",
       "      <th>unique</th>\n",
       "      <th>novel</th>\n",
       "      <th>invalid%</th>\n",
       "      <th>valid%</th>\n",
       "      <th>unique%</th>\n",
       "      <th>novel%</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>SELFIESx1</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILESx1</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SELFIESx5</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILESx5</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           chirals  invalid  valid  unique  novel  invalid%  valid%  unique%  \\\n",
       "SELFIESx1    False     True   True    True   True      True    True     True   \n",
       "SMILESx1     False     True  False   False   True      True    True    False   \n",
       "SELFIESx5    False     True   True    True   True      True    True     True   \n",
       "SMILESx5     False    False  False   False  False      True    True     True   \n",
       "\n",
       "           novel%  \n",
       "SELFIESx1    True  \n",
       "SMILESx1    False  \n",
       "SELFIESx5    True  \n",
       "SMILESx5     True  "
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "genChirPvals.map(lambda x: x<0.05)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "genChirStats.to_csv('results/07_grisoniChirals.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Metrics of Novels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "novels = pd.concat( [evaluator.getResultsNovels() for evaluator in evaluators], ignore_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_1315920/2569388527.py:1: FutureWarning: Downcasting object dtype arrays on .fillna, .ffill, .bfill is deprecated and will change in a future version. Call result.infer_objects(copy=False) instead. To opt-in to the future behavior, set `pd.set_option('future.no_silent_downcasting', True)`\n",
      "  novGrp=novels.fillna(-1).groupby(compilerPatterns.index.drop('fold').tolist(), as_index=True, dropna=False)\n"
     ]
    }
   ],
   "source": [
    "novGrp=novels.fillna(-1).groupby(compilerPatterns.index.drop('fold').tolist(), as_index=True, dropna=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "novStats=novGrp.describe().loc(axis=1)[ ( slice(None),['mean','std'] ) ].drop(columns='fold')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "novStats = ResultsFromMeanStdDF(novStats)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>FCD/Test</th>\n",
       "      <th>SNN/Test</th>\n",
       "      <th>Scaf/Test</th>\n",
       "      <th>IntDiv</th>\n",
       "      <th>IntDiv2</th>\n",
       "      <th>Filters</th>\n",
       "      <th>logP</th>\n",
       "      <th>SA</th>\n",
       "      <th>QED</th>\n",
       "      <th>weight</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th>aug</th>\n",
       "      <th>notation</th>\n",
       "      <th>hl</th>\n",
       "      <th>hu</th>\n",
       "      <th>es</th>\n",
       "      <th>lr</th>\n",
       "      <th>bs</th>\n",
       "      <th>amount</th>\n",
       "      <th>epoch</th>\n",
       "      <th>T</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"6\" valign=\"top\">grisoni</th>\n",
       "      <th rowspan=\"3\" valign=\"top\">-1</th>\n",
       "      <th>SELFIES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>-1</th>\n",
       "      <th>0.001</th>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>5.5 ± 0.2</td>\n",
       "      <td>0.377 ± 0.005</td>\n",
       "      <td>0.25 ± 0.04</td>\n",
       "      <td>0.88 ± 0.001</td>\n",
       "      <td>0.875 ± 0.001</td>\n",
       "      <td>0.578 ± 0.008</td>\n",
       "      <td>0.2 ± 0.09</td>\n",
       "      <td>0.74 ± 0.04</td>\n",
       "      <td>0.019 ± 0.003</td>\n",
       "      <td>5.0 ± 3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>-1</th>\n",
       "      <th>0.001</th>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>10</th>\n",
       "      <th>1.0</th>\n",
       "      <td>0.8 ± 0.1</td>\n",
       "      <td>0.486 ± 0.003</td>\n",
       "      <td>0.38 ± 0.04</td>\n",
       "      <td>0.869 ± 0.001</td>\n",
       "      <td>0.863 ± 0.001</td>\n",
       "      <td>0.849 ± 0.005</td>\n",
       "      <td>0.08 ± 0.03</td>\n",
       "      <td>0.05 ± 0.03</td>\n",
       "      <td>0.02 ± 0.01</td>\n",
       "      <td>14.0 ± 4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fragSMILES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>300</th>\n",
       "      <th>0.001</th>\n",
       "      <th>512</th>\n",
       "      <th>6000</th>\n",
       "      <th>16</th>\n",
       "      <th>1.0</th>\n",
       "      <td>0.69 ± 0.05</td>\n",
       "      <td>0.51 ± 0.003</td>\n",
       "      <td>0.38 ± 0.02</td>\n",
       "      <td>0.8689 ± 0.0008</td>\n",
       "      <td>0.8634 ± 0.0008</td>\n",
       "      <td>0.845 ± 0.003</td>\n",
       "      <td>0.11 ± 0.06</td>\n",
       "      <td>0.05 ± 0.03</td>\n",
       "      <td>0.01 ± 0.01</td>\n",
       "      <td>9.0 ± 5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"3\" valign=\"top\">5</th>\n",
       "      <th>SELFIES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>-1</th>\n",
       "      <th>0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>3.4 ± 0.1</td>\n",
       "      <td>0.375 ± 0.001</td>\n",
       "      <td>0.25 ± 0.02</td>\n",
       "      <td>0.8764 ± 0.0006</td>\n",
       "      <td>0.8715 ± 0.0006</td>\n",
       "      <td>0.68 ± 0.01</td>\n",
       "      <td>0.12 ± 0.05</td>\n",
       "      <td>0.53 ± 0.02</td>\n",
       "      <td>0.017 ± 0.005</td>\n",
       "      <td>5.0 ± 2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>-1</th>\n",
       "      <th>0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>3</th>\n",
       "      <th>1.0</th>\n",
       "      <td>0.99 ± 0.07</td>\n",
       "      <td>0.428 ± 0.001</td>\n",
       "      <td>0.3 ± 0.04</td>\n",
       "      <td>0.872 ± 0.001</td>\n",
       "      <td>0.867 ± 0.001</td>\n",
       "      <td>0.841 ± 0.006</td>\n",
       "      <td>0.11 ± 0.04</td>\n",
       "      <td>0.06 ± 0.02</td>\n",
       "      <td>0.02 ± 0.02</td>\n",
       "      <td>15.0 ± 9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fragSMILES</th>\n",
       "      <th>2</th>\n",
       "      <th>512</th>\n",
       "      <th>300</th>\n",
       "      <th>0.001</th>\n",
       "      <th>256</th>\n",
       "      <th>6000</th>\n",
       "      <th>4</th>\n",
       "      <th>1.0</th>\n",
       "      <td>0.72 ± 0.06</td>\n",
       "      <td>0.5 ± 0.006</td>\n",
       "      <td>0.38 ± 0.02</td>\n",
       "      <td>0.869 ± 0.002</td>\n",
       "      <td>0.863 ± 0.002</td>\n",
       "      <td>0.85 ± 0.01</td>\n",
       "      <td>0.15 ± 0.07</td>\n",
       "      <td>0.05 ± 0.02</td>\n",
       "      <td>0.015 ± 0.007</td>\n",
       "      <td>7.0 ± 4.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                  FCD/Test  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                  \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0    5.5 ± 0.2   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0    0.8 ± 0.1   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0  0.69 ± 0.05   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0    3.4 ± 0.1   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  0.99 ± 0.07   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  0.72 ± 0.06   \n",
       "\n",
       "                                                                    SNN/Test  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                    \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0  0.377 ± 0.005   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0  0.486 ± 0.003   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0   0.51 ± 0.003   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  0.375 ± 0.001   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  0.428 ± 0.001   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0    0.5 ± 0.006   \n",
       "\n",
       "                                                                 Scaf/Test  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                  \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0  0.25 ± 0.04   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0  0.38 ± 0.04   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0  0.38 ± 0.02   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  0.25 ± 0.02   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0   0.3 ± 0.04   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  0.38 ± 0.02   \n",
       "\n",
       "                                                                        IntDiv  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                      \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0     0.88 ± 0.001   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0    0.869 ± 0.001   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0  0.8689 ± 0.0008   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  0.8764 ± 0.0006   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0    0.872 ± 0.001   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0    0.869 ± 0.002   \n",
       "\n",
       "                                                                       IntDiv2  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                      \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0    0.875 ± 0.001   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0    0.863 ± 0.001   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0  0.8634 ± 0.0008   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  0.8715 ± 0.0006   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0    0.867 ± 0.001   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0    0.863 ± 0.002   \n",
       "\n",
       "                                                                     Filters  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                    \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0  0.578 ± 0.008   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0  0.849 ± 0.005   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0  0.845 ± 0.003   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0    0.68 ± 0.01   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  0.841 ± 0.006   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0    0.85 ± 0.01   \n",
       "\n",
       "                                                                      logP  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                  \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0   0.2 ± 0.09   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0  0.08 ± 0.03   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0  0.11 ± 0.06   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  0.12 ± 0.05   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  0.11 ± 0.04   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  0.15 ± 0.07   \n",
       "\n",
       "                                                                        SA  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                  \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0  0.74 ± 0.04   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0  0.05 ± 0.03   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0  0.05 ± 0.03   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  0.53 ± 0.02   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  0.06 ± 0.02   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  0.05 ± 0.02   \n",
       "\n",
       "                                                                         QED  \\\n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                    \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0  0.019 ± 0.003   \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0    0.02 ± 0.01   \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0    0.01 ± 0.01   \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0  0.017 ± 0.005   \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0    0.02 ± 0.02   \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0  0.015 ± 0.007   \n",
       "\n",
       "                                                                   weight  \n",
       "dataset aug notation   hl hu  es   lr    bs  amount epoch T                \n",
       "grisoni -1  SELFIES    2  512 -1   0.001 512 6000   10    1.0   5.0 ± 3.0  \n",
       "            SMILES     2  512 -1   0.001 512 6000   10    1.0  14.0 ± 4.0  \n",
       "            fragSMILES 2  512  300 0.001 512 6000   16    1.0   9.0 ± 5.0  \n",
       "         5  SELFIES    2  512 -1   0.001 256 6000   3     1.0   5.0 ± 2.0  \n",
       "            SMILES     2  512 -1   0.001 256 6000   3     1.0  15.0 ± 9.0  \n",
       "            fragSMILES 2  512  300 0.001 256 6000   4     1.0   7.0 ± 4.0  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "novStats.query(' ( aug==5 and hl==2 and hu==512 and bs==256 ) or ( aug!=5 and hl==2 and hu==512 and bs==512 ) ')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "novStats.T.to_csv('results/07_grisoniNovelMetrics.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "novPvals = pd.DataFrame.from_dict(\n",
    "                {notation+'x1': \n",
    "                {metric:stats.ttest_ind(\n",
    "                novGrp.get_group(('grisoni',-1,notation,*params))[metric],\n",
    "                novGrp.get_group(('grisoni',-1,'fragSMILES',*paramsFrag))[metric],\n",
    "                ).pvalue for metric in novStats.columns} for notation in ['SELFIES','SMILES']\n",
    "                } |\n",
    "                {notation+'x5': \n",
    "                {metric:stats.ttest_ind(\n",
    "                novGrp.get_group(('grisoni',5,notation,*params5))[metric],\n",
    "                novGrp.get_group(('grisoni',5,'fragSMILES',*paramsFrag5))[metric],\n",
    "                ).pvalue for metric in novStats.columns} for notation in ['SELFIES','SMILES']\n",
    "                }\n",
    "                , orient='index'\n",
    "            )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>FCD/Test</th>\n",
       "      <th>SNN/Test</th>\n",
       "      <th>Scaf/Test</th>\n",
       "      <th>IntDiv</th>\n",
       "      <th>IntDiv2</th>\n",
       "      <th>Filters</th>\n",
       "      <th>logP</th>\n",
       "      <th>SA</th>\n",
       "      <th>QED</th>\n",
       "      <th>weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>SELFIESx1</th>\n",
       "      <td>2.511745e-11</td>\n",
       "      <td>1.289718e-11</td>\n",
       "      <td>0.000269</td>\n",
       "      <td>2.613503e-07</td>\n",
       "      <td>1.119802e-07</td>\n",
       "      <td>1.970335e-12</td>\n",
       "      <td>0.103788</td>\n",
       "      <td>1.747794e-09</td>\n",
       "      <td>0.306073</td>\n",
       "      <td>0.110658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILESx1</th>\n",
       "      <td>3.568206e-02</td>\n",
       "      <td>5.545198e-07</td>\n",
       "      <td>0.810828</td>\n",
       "      <td>7.489774e-01</td>\n",
       "      <td>9.099706e-01</td>\n",
       "      <td>2.416050e-01</td>\n",
       "      <td>0.339026</td>\n",
       "      <td>8.190494e-01</td>\n",
       "      <td>0.202524</td>\n",
       "      <td>0.138269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SELFIESx5</th>\n",
       "      <td>4.726027e-11</td>\n",
       "      <td>7.497243e-11</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>4.327195e-05</td>\n",
       "      <td>1.461162e-05</td>\n",
       "      <td>4.268837e-08</td>\n",
       "      <td>0.498604</td>\n",
       "      <td>1.874516e-10</td>\n",
       "      <td>0.616179</td>\n",
       "      <td>0.315130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILESx5</th>\n",
       "      <td>1.253824e-04</td>\n",
       "      <td>5.887502e-09</td>\n",
       "      <td>0.002278</td>\n",
       "      <td>1.289053e-02</td>\n",
       "      <td>1.076266e-02</td>\n",
       "      <td>5.598158e-01</td>\n",
       "      <td>0.329676</td>\n",
       "      <td>2.509228e-01</td>\n",
       "      <td>0.555078</td>\n",
       "      <td>0.130280</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               FCD/Test      SNN/Test  Scaf/Test        IntDiv       IntDiv2  \\\n",
       "SELFIESx1  2.511745e-11  1.289718e-11   0.000269  2.613503e-07  1.119802e-07   \n",
       "SMILESx1   3.568206e-02  5.545198e-07   0.810828  7.489774e-01  9.099706e-01   \n",
       "SELFIESx5  4.726027e-11  7.497243e-11   0.000010  4.327195e-05  1.461162e-05   \n",
       "SMILESx5   1.253824e-04  5.887502e-09   0.002278  1.289053e-02  1.076266e-02   \n",
       "\n",
       "                Filters      logP            SA       QED    weight  \n",
       "SELFIESx1  1.970335e-12  0.103788  1.747794e-09  0.306073  0.110658  \n",
       "SMILESx1   2.416050e-01  0.339026  8.190494e-01  0.202524  0.138269  \n",
       "SELFIESx5  4.268837e-08  0.498604  1.874516e-10  0.616179  0.315130  \n",
       "SMILESx5   5.598158e-01  0.329676  2.509228e-01  0.555078  0.130280  "
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "novPvals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>FCD/Test</th>\n",
       "      <th>SNN/Test</th>\n",
       "      <th>Scaf/Test</th>\n",
       "      <th>IntDiv</th>\n",
       "      <th>IntDiv2</th>\n",
       "      <th>Filters</th>\n",
       "      <th>logP</th>\n",
       "      <th>SA</th>\n",
       "      <th>QED</th>\n",
       "      <th>weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>SELFIESx1</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILESx1</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SELFIESx5</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SMILESx5</th>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           FCD/Test  SNN/Test  Scaf/Test  IntDiv  IntDiv2  Filters   logP  \\\n",
       "SELFIESx1      True      True       True    True     True     True  False   \n",
       "SMILESx1       True      True      False   False    False    False  False   \n",
       "SELFIESx5      True      True       True    True     True     True  False   \n",
       "SMILESx5       True      True       True    True     True    False  False   \n",
       "\n",
       "              SA    QED  weight  \n",
       "SELFIESx1   True  False   False  \n",
       "SMILESx1   False  False   False  \n",
       "SELFIESx5   True  False   False  \n",
       "SMILESx5   False  False   False  "
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "novPvals.map(lambda x: x<0.05)"
   ]
  }
 ],
 "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": "3.11.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}