nn_inference_eval_5.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import sys\n",
    "sys.path.append('/nas/longleaf/home/siyangj/NiftyNet/')\n",
    "import niftynet as nn\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import SimpleITK as sitk\n",
    "import os\n",
    "import re\n",
    "import nibabel as nib\n",
    "import sklearn as skl\n",
    "import pandas as pd\n",
    "\n",
    "pd.options.display.max_colwidth = 100\n",
    "pd.options.display.max_columns = 10000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Dice(pred,true,labels=[0,1,2,3]):\n",
    "    l = len(labels)\n",
    "    ret = np.zeros(l)\n",
    "    for i in range(l):\n",
    "        lab = labels[i]\n",
    "        p = pred==lab\n",
    "        t = true==lab\n",
    "        ret[i] = 2*np.logical_and(p,t).sum()/(p.sum()+t.sum())\n",
    "    return ret"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def inference_eval(model,inf_dir,data_path,eval_df,test_im,\n",
    "                   seg_suffix='seg/',skip_if_exist=True):\n",
    "    inf_re = re.compile(r'inference.*')\n",
    "    im_re = re.compile(r'.*\\.nii\\.gz')\n",
    "    model_path = '/proj/NIRAL/users/siyangj/model_{}/'.format(model)\n",
    "    \n",
    "    inf_ims = {f[0]:os.path.join(inf_dir,f) for f in os.listdir(inf_dir) \n",
    "            if (os.path.isfile(os.path.join(inf_dir, f)) and im_re.match(f) is not None)}\n",
    "    l = len(eval_df.index)\n",
    "    if len(eval_df.loc[eval_df['inf_dir']==inf_dir])>0:\n",
    "        if skip_if_exist:\n",
    "            return\n",
    "        l = np.where(eval_df['inf_dir']==inf_dir)[0][0]\n",
    "    eval_df.loc[l,'inf_dir']=inf_dir\n",
    "    eval_df.loc[l,'model']=model\n",
    "    eval_df.loc[l,'test_im']=test_im\n",
    "    for im_num,im in inf_ims.items():\n",
    "        true_im = os.path.join(data_path,seg_suffix,im_num+'.nrrd')\n",
    "        true_arr = sitk.GetArrayFromImage(sitk.ReadImage(true_im))\n",
    "        #print(true_arr.shape)\n",
    "        inf_arr = sitk.GetArrayFromImage(sitk.ReadImage(im))\n",
    "        #print(inf_arr.shape)\n",
    "        c_mat = skl.metrics.confusion_matrix(true_arr.flatten(),inf_arr.flatten())\n",
    "        p_mat = c_mat.T / c_mat.astype(np.float).sum(axis=1)\n",
    "        eval_df.loc[l,im_num]=[p_mat.T]\n",
    "        if im_num == test_im:\n",
    "            eval_df.loc[l,'test_11'] = p_mat[1,1]\n",
    "            eval_df.loc[l,'test_12'] = p_mat[2,1]\n",
    "            eval_df.loc[l,'test_21'] = p_mat[1,2]\n",
    "            eval_df.loc[l,'test_22'] = p_mat[2,2]\n",
    "            eval_df.loc[l,'test_11+22'] = eval_df.loc[l,'test_11']+eval_df.loc[l,'test_22']\n",
    "            eval_df.loc[l,'test_12+21'] = eval_df.loc[l,'test_12']+eval_df.loc[l,'test_21']\n",
    "            Dice0123 = Dice(inf_arr,true_arr)\n",
    "            eval_df.loc[l,'Dice0'] = Dice0123[0]\n",
    "            eval_df.loc[l,'Dice1'] = Dice0123[1]\n",
    "            eval_df.loc[l,'Dice2'] = Dice0123[2]\n",
    "            eval_df.loc[l,'Dice3'] = Dice0123[3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def model_eval(model,data_path,eval_df,test_im,seg_suffix='seg/'):\n",
    "    model_path = '/proj/NIRAL/users/siyangj/model_{}/'.format(model)\n",
    "    inf_re = re.compile(r'inference.*')\n",
    "    im_re = re.compile(r'.*\\.nii\\.gz')\n",
    "    inf_dirs = [f for f in os.listdir(model_path) \n",
    "            if (os.path.isdir(os.path.join(model_path, f)) \n",
    "                and inf_re.match(f) is not None)]\n",
    "    for d in inf_dirs:\n",
    "        inf_dir = os.path.join(model_path,d)\n",
    "        inference_eval(model,inf_dir,data_path,eval_df,test_im,seg_suffix=seg_suffix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "eval_df=pd.read_csv('/proj/NIRAL/users/siyangj/inference_eval_4.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "eval_df['model']=eval_df['model'].astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "eval_df['test_im']=eval_df['test_im'].astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>model</th>\n",
       "      <th>inf_dir</th>\n",
       "      <th>test_im</th>\n",
       "      <th>test_11</th>\n",
       "      <th>test_12</th>\n",
       "      <th>test_21</th>\n",
       "      <th>test_22</th>\n",
       "      <th>test_11+22</th>\n",
       "      <th>test_12+21</th>\n",
       "      <th>Dice0</th>\n",
       "      <th>Dice1</th>\n",
       "      <th>Dice2</th>\n",
       "      <th>Dice3</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10112050</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112050/inference_17000_10112050</td>\n",
       "      <td>1</td>\n",
       "      <td>0.918915</td>\n",
       "      <td>0.067399</td>\n",
       "      <td>0.295044</td>\n",
       "      <td>0.649303</td>\n",
       "      <td>1.568218</td>\n",
       "      <td>0.362443</td>\n",
       "      <td>0.993768</td>\n",
       "      <td>0.780229</td>\n",
       "      <td>0.744064</td>\n",
       "      <td>0.805368</td>\n",
       "      <td>[array([[9.99995958e-01, 0.00000000e+00, 0.00000000e+00, 4.04240483e-06],\\n       [6.64359862e-0...</td>\n",
       "      <td>[array([[9.92899343e-01, 2.40386838e-05, 2.32620494e-03, 4.75041374e-03],\\n       [1.10490523e-0...</td>\n",
       "      <td>[array([[9.99996610e-01, 0.00000000e+00, 0.00000000e+00, 3.38991095e-06],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99996543e-01, 0.00000000e+00, 0.00000000e+00, 3.45705559e-06],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99925734e-01, 0.00000000e+00, 1.90424685e-06, 7.23613804e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99957277e-01, 0.00000000e+00, 0.00000000e+00, 4.27234118e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99967062e-01, 0.00000000e+00, 0.00000000e+00, 3.29379214e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99829622e-01, 0.00000000e+00, 7.67470310e-06, 1.62703706e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10112050</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112050/inference</td>\n",
       "      <td>1</td>\n",
       "      <td>0.895175</td>\n",
       "      <td>0.097146</td>\n",
       "      <td>0.224378</td>\n",
       "      <td>0.753510</td>\n",
       "      <td>1.648685</td>\n",
       "      <td>0.321524</td>\n",
       "      <td>0.991811</td>\n",
       "      <td>0.803432</td>\n",
       "      <td>0.796687</td>\n",
       "      <td>0.804782</td>\n",
       "      <td>[array([[9.99987873e-01, 0.00000000e+00, 0.00000000e+00, 1.21272145e-05],\\n       [6.64359862e-0...</td>\n",
       "      <td>[array([[9.92847567e-01, 0.00000000e+00, 6.71788755e-03, 4.34545438e-04],\\n       [1.21539575e-0...</td>\n",
       "      <td>[array([[9.99969491e-01, 0.00000000e+00, 1.01697328e-05, 2.03394657e-05],\\n       [1.98668918e-0...</td>\n",
       "      <td>[array([[9.99970615e-01, 0.00000000e+00, 1.20996946e-05, 1.72852780e-05],\\n       [3.00979689e-0...</td>\n",
       "      <td>[array([[9.99849564e-01, 0.00000000e+00, 2.66594559e-05, 1.23776045e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99917718e-01, 0.00000000e+00, 9.49409151e-06, 7.27880349e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99855701e-01, 0.00000000e+00, 6.11704255e-05, 8.31290398e-05],\\n       [7.04528001e-0...</td>\n",
       "      <td>[array([[9.99889484e-01, 0.00000000e+00, 4.75831592e-05, 6.29325655e-05],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10112050</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112050/inference_13200_10112050</td>\n",
       "      <td>1</td>\n",
       "      <td>0.926572</td>\n",
       "      <td>0.058502</td>\n",
       "      <td>0.308779</td>\n",
       "      <td>0.649881</td>\n",
       "      <td>1.576453</td>\n",
       "      <td>0.367281</td>\n",
       "      <td>0.994202</td>\n",
       "      <td>0.777533</td>\n",
       "      <td>0.752149</td>\n",
       "      <td>0.844704</td>\n",
       "      <td>[array([[9.99981809e-01, 0.00000000e+00, 0.00000000e+00, 1.81908217e-05],\\n       [1.10726644e-0...</td>\n",
       "      <td>[array([[9.95715567e-01, 0.00000000e+00, 3.19899407e-03, 1.08543903e-03],\\n       [7.18188397e-0...</td>\n",
       "      <td>[array([[9.99976271e-01, 0.00000000e+00, 1.01697328e-05, 1.35596438e-05],\\n       [2.33728139e-0...</td>\n",
       "      <td>[array([[9.99987900e-01, 0.00000000e+00, 3.45705559e-06, 8.64263898e-06],\\n       [2.50816407e-0...</td>\n",
       "      <td>[array([[9.99925734e-01, 0.00000000e+00, 1.71382217e-05, 5.71274055e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99933541e-01, 0.00000000e+00, 3.16469717e-06, 6.32939434e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99896481e-01, 0.00000000e+00, 7.84236224e-06, 9.56768194e-05],\\n       [2.81811201e-0...</td>\n",
       "      <td>[array([[9.99926323e-01, 0.00000000e+00, 1.38144656e-05, 5.98626842e-05],\\n       [7.36713374e-0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10112056</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112056/inference_18100_10112056</td>\n",
       "      <td>1</td>\n",
       "      <td>0.825539</td>\n",
       "      <td>0.124662</td>\n",
       "      <td>0.174211</td>\n",
       "      <td>0.737446</td>\n",
       "      <td>1.562985</td>\n",
       "      <td>0.298872</td>\n",
       "      <td>0.992718</td>\n",
       "      <td>0.794706</td>\n",
       "      <td>0.796743</td>\n",
       "      <td>0.786574</td>\n",
       "      <td>[array([[9.99997979e-01, 0.00000000e+00, 0.00000000e+00, 2.02120241e-06],\\n       [1.32871972e-0...</td>\n",
       "      <td>[array([[9.90292070e-01, 7.39651809e-06, 5.56957812e-03, 4.13095535e-03],\\n       [4.97207352e-0...</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [3.50592209e-0...</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [5.01632815e-0...</td>\n",
       "      <td>[array([[9.99969532e-01, 0.00000000e+00, 9.52123426e-06, 2.09467154e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99984177e-01, 0.00000000e+00, 0.00000000e+00, 1.58234859e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99990589e-01, 0.00000000e+00, 0.00000000e+00, 9.41083469e-06],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99930928e-01, 1.53494062e-06, 6.13976248e-06, 6.13976248e-05],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10112056</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112056/inference_16500_10112056</td>\n",
       "      <td>1</td>\n",
       "      <td>0.793855</td>\n",
       "      <td>0.189398</td>\n",
       "      <td>0.090426</td>\n",
       "      <td>0.849576</td>\n",
       "      <td>1.643431</td>\n",
       "      <td>0.279823</td>\n",
       "      <td>0.990101</td>\n",
       "      <td>0.825034</td>\n",
       "      <td>0.838850</td>\n",
       "      <td>0.803951</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [1.16262976e-0...</td>\n",
       "      <td>[array([[9.93474422e-01, 3.69825904e-06, 5.02963230e-03, 1.49224752e-03],\\n       [9.39169443e-0...</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [4.09024243e-0...</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [1.50489844e-0...</td>\n",
       "      <td>[array([[9.99979053e-01, 0.00000000e+00, 1.90424685e-06, 1.90424685e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99990506e-01, 0.00000000e+00, 0.00000000e+00, 9.49409151e-06],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99995295e-01, 0.00000000e+00, 0.00000000e+00, 4.70541735e-06],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99940137e-01, 0.00000000e+00, 9.20964372e-06, 5.06530405e-05],\\n       [7.36713374e-0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>10112056</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112056/inference</td>\n",
       "      <td>1</td>\n",
       "      <td>0.845575</td>\n",
       "      <td>0.147383</td>\n",
       "      <td>0.184313</td>\n",
       "      <td>0.789043</td>\n",
       "      <td>1.634618</td>\n",
       "      <td>0.331696</td>\n",
       "      <td>0.997961</td>\n",
       "      <td>0.792166</td>\n",
       "      <td>0.803326</td>\n",
       "      <td>0.819554</td>\n",
       "      <td>[array([[9.99252155e-01, 0.00000000e+00, 6.06360724e-06, 7.41781286e-04],\\n       [1.99307958e-0...</td>\n",
       "      <td>[array([[9.91715900e-01, 1.84912952e-06, 5.18126092e-03, 3.10099021e-03],\\n       [2.20981045e-0...</td>\n",
       "      <td>[array([[9.99811860e-01, 1.69495547e-06, 8.47477737e-06, 1.77970325e-04],\\n       [7.01184417e-0...</td>\n",
       "      <td>[array([[9.99958515e-01, 0.00000000e+00, 6.91411118e-06, 3.45705559e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99544885e-01, 0.00000000e+00, 8.56911083e-05, 3.69423889e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99761065e-01, 0.00000000e+00, 1.26587887e-05, 2.26275848e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99767866e-01, 0.00000000e+00, 4.07802837e-05, 1.91353639e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99831157e-01, 0.00000000e+00, 4.29783374e-05, 1.25865131e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>10112103</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112103/inference</td>\n",
       "      <td>1</td>\n",
       "      <td>0.721373</td>\n",
       "      <td>0.249500</td>\n",
       "      <td>0.052741</td>\n",
       "      <td>0.869897</td>\n",
       "      <td>1.591270</td>\n",
       "      <td>0.302241</td>\n",
       "      <td>0.993959</td>\n",
       "      <td>0.802792</td>\n",
       "      <td>0.837660</td>\n",
       "      <td>0.810618</td>\n",
       "      <td>[array([[9.99840325e-01, 0.00000000e+00, 0.00000000e+00, 1.59674991e-04],\\n       [3.87543253e-0...</td>\n",
       "      <td>[array([[9.91636387e-01, 0.00000000e+00, 6.45346203e-03, 1.91015080e-03],\\n       [1.65735784e-0...</td>\n",
       "      <td>[array([[9.99993220e-01, 0.00000000e+00, 0.00000000e+00, 6.77982189e-06],\\n       [1.16864070e-0...</td>\n",
       "      <td>[array([[9.99996543e-01, 0.00000000e+00, 0.00000000e+00, 3.45705559e-06],\\n       [5.01632815e-0...</td>\n",
       "      <td>[array([[9.99885745e-01, 1.90424685e-06, 0.00000000e+00, 1.12350564e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99974682e-01, 0.00000000e+00, 0.00000000e+00, 2.53175774e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99987452e-01, 0.00000000e+00, 0.00000000e+00, 1.25477796e-05],\\n       [7.04528001e-0...</td>\n",
       "      <td>[array([[9.99880275e-01, 0.00000000e+00, 0.00000000e+00, 1.19725368e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>10112103</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112103/inference_12300_10112103</td>\n",
       "      <td>1</td>\n",
       "      <td>0.809434</td>\n",
       "      <td>0.176941</td>\n",
       "      <td>0.130890</td>\n",
       "      <td>0.809682</td>\n",
       "      <td>1.619116</td>\n",
       "      <td>0.307831</td>\n",
       "      <td>0.992859</td>\n",
       "      <td>0.807523</td>\n",
       "      <td>0.819547</td>\n",
       "      <td>0.813614</td>\n",
       "      <td>[array([[9.99975746e-01, 0.00000000e+00, 0.00000000e+00, 2.42544290e-05],\\n       [8.85813149e-0...</td>\n",
       "      <td>[array([[9.90759900e-01, 0.00000000e+00, 7.81072310e-03, 1.42937712e-03],\\n       [6.62943136e-0...</td>\n",
       "      <td>[array([[9.99993220e-01, 0.00000000e+00, 0.00000000e+00, 6.77982189e-06],\\n       [2.33728139e-0...</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [1.00326563e-0...</td>\n",
       "      <td>[array([[9.99954298e-01, 0.00000000e+00, 3.80849370e-06, 4.18934307e-05],\\n       [1.40992020e-0...</td>\n",
       "      <td>[array([[9.99984177e-01, 0.00000000e+00, 3.16469717e-06, 1.26587887e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99993726e-01, 0.00000000e+00, 0.00000000e+00, 6.27388980e-06],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99920183e-01, 0.00000000e+00, 3.22337530e-05, 4.75831592e-05],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>10112127</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112127/inference_13500_10112127</td>\n",
       "      <td>1</td>\n",
       "      <td>0.737849</td>\n",
       "      <td>0.235875</td>\n",
       "      <td>0.069771</td>\n",
       "      <td>0.803042</td>\n",
       "      <td>1.540891</td>\n",
       "      <td>0.305646</td>\n",
       "      <td>0.996067</td>\n",
       "      <td>0.802167</td>\n",
       "      <td>0.808498</td>\n",
       "      <td>0.805656</td>\n",
       "      <td>[array([[9.99961146e-01, 0.00000000e+00, 0.00000000e+00, 3.88540112e-05],\\n       [4.42906574e-0...</td>\n",
       "      <td>[array([[9.94207332e-01, 0.00000000e+00, 2.46873096e-03, 3.32393663e-03],\\n       [1.04965997e-0...</td>\n",
       "      <td>[array([[9.99935592e-01, 0.00000000e+00, 0.00000000e+00, 6.44083080e-05],\\n       [1.92825715e-0...</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [5.01632815e-0...</td>\n",
       "      <td>[array([[9.99832690e-01, 0.00000000e+00, 0.00000000e+00, 1.67309808e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99917078e-01, 0.00000000e+00, 0.00000000e+00, 8.29221775e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99933122e-01, 0.00000000e+00, 0.00000000e+00, 6.68783911e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99745686e-01, 0.00000000e+00, 1.65138857e-06, 2.52662451e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10112127</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112127/inference_8700_10112127</td>\n",
       "      <td>1</td>\n",
       "      <td>0.713201</td>\n",
       "      <td>0.271862</td>\n",
       "      <td>0.050931</td>\n",
       "      <td>0.864683</td>\n",
       "      <td>1.577885</td>\n",
       "      <td>0.322792</td>\n",
       "      <td>0.995328</td>\n",
       "      <td>0.797492</td>\n",
       "      <td>0.828315</td>\n",
       "      <td>0.831142</td>\n",
       "      <td>[array([[9.99957056e-01, 0.00000000e+00, 0.00000000e+00, 4.29439071e-05],\\n       [2.49134948e-0...</td>\n",
       "      <td>[array([[9.90071044e-01, 0.00000000e+00, 6.66464200e-03, 3.26431445e-03],\\n       [8.83924181e-0...</td>\n",
       "      <td>[array([[9.99601685e-01, 1.69495547e-06, 1.69495547e-06, 3.94924625e-04],\\n       [4.55769871e-0...</td>\n",
       "      <td>[array([[9.99883783e-01, 0.00000000e+00, 0.00000000e+00, 1.16217351e-04],\\n       [5.01632815e-0...</td>\n",
       "      <td>[array([[9.99806950e-01, 0.00000000e+00, 0.00000000e+00, 1.93049779e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99882527e-01, 0.00000000e+00, 0.00000000e+00, 1.17473085e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99921118e-01, 0.00000000e+00, 1.71483054e-06, 7.71673743e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99734126e-01, 0.00000000e+00, 3.30277714e-05, 2.32845788e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10112130</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112130/inference_7300</td>\n",
       "      <td>1</td>\n",
       "      <td>0.712039</td>\n",
       "      <td>0.238079</td>\n",
       "      <td>0.062735</td>\n",
       "      <td>0.804453</td>\n",
       "      <td>1.516491</td>\n",
       "      <td>0.300814</td>\n",
       "      <td>0.993799</td>\n",
       "      <td>0.789505</td>\n",
       "      <td>0.808477</td>\n",
       "      <td>0.781285</td>\n",
       "      <td>[array([[9.99104313e-01, 0.00000000e+00, 0.00000000e+00, 8.95687205e-04],\\n       [5.20415225e-0...</td>\n",
       "      <td>[array([[9.91621220e-01, 1.86319318e-06, 4.87225016e-03, 3.50466637e-03],\\n       [2.76226307e-0...</td>\n",
       "      <td>[array([[9.99927117e-01, 0.00000000e+00, 0.00000000e+00, 7.28830854e-05],\\n       [1.16864070e-0...</td>\n",
       "      <td>[array([[9.99998265e-01, 0.00000000e+00, 0.00000000e+00, 1.73458732e-06],\\n       [2.00653126e-0...</td>\n",
       "      <td>[array([[9.99240671e-01, 0.00000000e+00, 0.00000000e+00, 7.59329131e-04],\\n       [3.52480050e-0...</td>\n",
       "      <td>[array([[9.99917078e-01, 0.00000000e+00, 0.00000000e+00, 8.29221775e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99933122e-01, 0.00000000e+00, 0.00000000e+00, 6.68783911e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99745686e-01, 0.00000000e+00, 0.00000000e+00, 2.54313840e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>10112055</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112055/inference_18900_10112055</td>\n",
       "      <td>1</td>\n",
       "      <td>0.892019</td>\n",
       "      <td>0.094439</td>\n",
       "      <td>0.250173</td>\n",
       "      <td>0.696974</td>\n",
       "      <td>1.588993</td>\n",
       "      <td>0.344611</td>\n",
       "      <td>0.993306</td>\n",
       "      <td>0.788053</td>\n",
       "      <td>0.773435</td>\n",
       "      <td>0.822208</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [4.42906574e-0...</td>\n",
       "      <td>[array([[9.92322414e-01, 3.69825904e-06, 5.22933829e-03, 2.44454923e-03],\\n       [3.86716829e-0...</td>\n",
       "      <td>[array([[9.99994915e-01, 0.00000000e+00, 0.00000000e+00, 5.08486642e-06],\\n       [1.75296104e-0...</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [5.01632815e-0...</td>\n",
       "      <td>[array([[9.99973341e-01, 0.00000000e+00, 0.00000000e+00, 2.66594559e-05],\\n       [7.04960099e-0...</td>\n",
       "      <td>[array([[9.99987341e-01, 0.00000000e+00, 0.00000000e+00, 1.26587887e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99995295e-01, 0.00000000e+00, 0.00000000e+00, 4.70541735e-06],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99937067e-01, 0.00000000e+00, 1.22795250e-05, 5.06530405e-05],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>10112055</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112055/inference_21900_10112055</td>\n",
       "      <td>1</td>\n",
       "      <td>0.861016</td>\n",
       "      <td>0.125160</td>\n",
       "      <td>0.120896</td>\n",
       "      <td>0.821308</td>\n",
       "      <td>1.682324</td>\n",
       "      <td>0.246056</td>\n",
       "      <td>0.995358</td>\n",
       "      <td>0.843021</td>\n",
       "      <td>0.846834</td>\n",
       "      <td>0.845988</td>\n",
       "      <td>[array([[9.99993936e-01, 0.00000000e+00, 0.00000000e+00, 6.06360724e-06],\\n       [3.32179931e-0...</td>\n",
       "      <td>[array([[9.95284720e-01, 3.69825904e-06, 3.49115654e-03, 1.22042548e-03],\\n       [2.76226307e-0...</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [2.33728139e-0...</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [1.00326563e-0...</td>\n",
       "      <td>[array([[9.99986670e-01, 0.00000000e+00, 0.00000000e+00, 1.33297280e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99988924e-01, 0.00000000e+00, 1.58234859e-06, 9.49409151e-06],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99996863e-01, 0.00000000e+00, 0.00000000e+00, 3.13694490e-06],\\n       [7.04528001e-0...</td>\n",
       "      <td>[array([[9.99969301e-01, 0.00000000e+00, 4.60482186e-06, 2.60939906e-05],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>10112055</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112055/inference</td>\n",
       "      <td>1</td>\n",
       "      <td>0.824332</td>\n",
       "      <td>0.159098</td>\n",
       "      <td>0.141585</td>\n",
       "      <td>0.805546</td>\n",
       "      <td>1.629878</td>\n",
       "      <td>0.300683</td>\n",
       "      <td>0.994605</td>\n",
       "      <td>0.808893</td>\n",
       "      <td>0.818492</td>\n",
       "      <td>0.810333</td>\n",
       "      <td>[array([[9.99939364e-01, 0.00000000e+00, 0.00000000e+00, 6.06360724e-05],\\n       [6.64359862e-0...</td>\n",
       "      <td>[array([[9.90305014e-01, 2.95860724e-05, 6.95272700e-03, 2.71267301e-03],\\n       [2.20981045e-0...</td>\n",
       "      <td>[array([[9.99971186e-01, 1.69495547e-06, 0.00000000e+00, 2.71192876e-05],\\n       [5.84320348e-0...</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [2.00653126e-0...</td>\n",
       "      <td>[array([[9.99771490e-01, 0.00000000e+00, 7.61698741e-06, 2.20892635e-04],\\n       [1.05744015e-0...</td>\n",
       "      <td>[array([[9.99931959e-01, 0.00000000e+00, 4.74704576e-06, 6.32939434e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99973336e-01, 0.00000000e+00, 0.00000000e+00, 2.66640316e-05],\\n       [1.40905600e-0...</td>\n",
       "      <td>[array([[9.99867995e-01, 0.00000000e+00, 5.83277436e-05, 7.36771498e-05],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>10112055</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112055/inference_21900_normal</td>\n",
       "      <td>1</td>\n",
       "      <td>0.861763</td>\n",
       "      <td>0.124861</td>\n",
       "      <td>0.121004</td>\n",
       "      <td>0.822459</td>\n",
       "      <td>1.684223</td>\n",
       "      <td>0.245865</td>\n",
       "      <td>0.991524</td>\n",
       "      <td>0.843291</td>\n",
       "      <td>0.846411</td>\n",
       "      <td>0.833207</td>\n",
       "      <td>[array([[9.91322978e-01, 0.00000000e+00, 1.08538570e-03, 7.59163626e-03],\\n       [3.32179931e-0...</td>\n",
       "      <td>[array([[9.84748380e-01, 4.43791085e-05, 7.87729176e-03, 7.32994943e-03],\\n       [2.20981045e-0...</td>\n",
       "      <td>[array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [2.33728139e-0...</td>\n",
       "      <td>[array([[9.96544673e-01, 0.00000000e+00, 1.75272718e-03, 1.70259988e-03],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[8.89199493e-01, 6.07454746e-04, 3.92788998e-02, 7.09141527e-02],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.16129196e-01, 3.73434266e-04, 2.70961372e-02, 5.64012330e-02],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.14648435e-01, 4.79952569e-04, 3.05444325e-02, 5.43271802e-02],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.29358963e-01, 5.35694277e-04, 2.86005486e-02, 4.15047944e-02],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>10192000</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192000/inference_7200</td>\n",
       "      <td>1</td>\n",
       "      <td>0.882287</td>\n",
       "      <td>0.095098</td>\n",
       "      <td>0.265452</td>\n",
       "      <td>0.683058</td>\n",
       "      <td>1.565344</td>\n",
       "      <td>0.360550</td>\n",
       "      <td>0.995164</td>\n",
       "      <td>0.777106</td>\n",
       "      <td>0.762458</td>\n",
       "      <td>0.836195</td>\n",
       "      <td>[array([[9.98707370e-01, 0.00000000e+00, 1.81457491e-04, 1.11117228e-03],\\n       [2.21453287e-0...</td>\n",
       "      <td>[array([[9.92659423e-01, 0.00000000e+00, 5.34063294e-03, 1.99994414e-03],\\n       [5.52385477e-0...</td>\n",
       "      <td>[array([[9.99981355e-01, 0.00000000e+00, 1.69495547e-06, 1.69495547e-05],\\n       [5.84320348e-0...</td>\n",
       "      <td>[array([[9.99967044e-01, 0.00000000e+00, 2.08142536e-05, 1.21416479e-05],\\n       [3.00979689e-0...</td>\n",
       "      <td>[array([[9.98295605e-01, 0.00000000e+00, 1.96990330e-04, 1.50740427e-03],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99773752e-01, 0.00000000e+00, 4.83584049e-05, 1.77889846e-04],\\n       [7.60438925e-0...</td>\n",
       "      <td>[array([[9.99881684e-01, 0.00000000e+00, 1.20030522e-05, 1.06312748e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99664808e-01, 0.00000000e+00, 1.94840685e-04, 1.40351341e-04],\\n       [7.34824046e-0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>10192000</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192000/inference_7700</td>\n",
       "      <td>1</td>\n",
       "      <td>0.900617</td>\n",
       "      <td>0.084324</td>\n",
       "      <td>0.250964</td>\n",
       "      <td>0.693302</td>\n",
       "      <td>1.593919</td>\n",
       "      <td>0.335288</td>\n",
       "      <td>0.995337</td>\n",
       "      <td>0.792735</td>\n",
       "      <td>0.776372</td>\n",
       "      <td>0.846890</td>\n",
       "      <td>[array([[9.98519796e-01, 0.00000000e+00, 1.40680527e-04, 1.33952328e-03],\\n       [2.76816609e-0...</td>\n",
       "      <td>[array([[9.93545804e-01, 0.00000000e+00, 3.92540246e-03, 2.52879342e-03],\\n       [1.65715643e-0...</td>\n",
       "      <td>[array([[9.99981355e-01, 0.00000000e+00, 1.69495547e-06, 1.69495547e-05],\\n       [7.59616452e-0...</td>\n",
       "      <td>[array([[9.99968779e-01, 0.00000000e+00, 1.04071268e-05, 2.08142536e-05],\\n       [3.51142970e-0...</td>\n",
       "      <td>[array([[9.98396242e-01, 0.00000000e+00, 1.67013541e-04, 1.43674469e-03],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99711577e-01, 0.00000000e+00, 4.14500613e-05, 2.46973282e-04],\\n       [7.60438925e-0...</td>\n",
       "      <td>[array([[9.99881684e-01, 0.00000000e+00, 1.02883305e-05, 1.08027470e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99641691e-01, 0.00000000e+00, 1.02373919e-04, 2.55934798e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>10192002</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192002/inference_5700</td>\n",
       "      <td>1</td>\n",
       "      <td>0.797880</td>\n",
       "      <td>0.149403</td>\n",
       "      <td>0.140641</td>\n",
       "      <td>0.786742</td>\n",
       "      <td>1.584622</td>\n",
       "      <td>0.290044</td>\n",
       "      <td>0.996218</td>\n",
       "      <td>0.797222</td>\n",
       "      <td>0.815309</td>\n",
       "      <td>0.810516</td>\n",
       "      <td>[array([[9.98150765e-01, 0.00000000e+00, 4.68935089e-05, 1.80234182e-03],\\n       [2.21453287e-0...</td>\n",
       "      <td>[array([[9.87728462e-01, 0.00000000e+00, 6.06873179e-03, 6.20280625e-03],\\n       [5.52385477e-0...</td>\n",
       "      <td>[array([[9.99271169e-01, 0.00000000e+00, 3.72890204e-05, 6.91541833e-04],\\n       [2.33728139e-0...</td>\n",
       "      <td>[array([[9.99963575e-01, 0.00000000e+00, 5.20356340e-06, 3.12213804e-05],\\n       [3.51142970e-0...</td>\n",
       "      <td>[array([[9.98132874e-01, 0.00000000e+00, 7.49419735e-05, 1.79218377e-03],\\n       [7.03378326e-0...</td>\n",
       "      <td>[array([[9.99730575e-01, 0.00000000e+00, 2.41792024e-05, 2.45246196e-04],\\n       [7.60438925e-0...</td>\n",
       "      <td>[array([[9.99874825e-01, 0.00000000e+00, 1.71472174e-06, 1.23459966e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99640040e-01, 0.00000000e+00, 3.63262294e-05, 3.23633680e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>10192002</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192002/inference_4600</td>\n",
       "      <td>1</td>\n",
       "      <td>0.751673</td>\n",
       "      <td>0.210691</td>\n",
       "      <td>0.083541</td>\n",
       "      <td>0.868307</td>\n",
       "      <td>1.619980</td>\n",
       "      <td>0.294232</td>\n",
       "      <td>0.993344</td>\n",
       "      <td>0.802652</td>\n",
       "      <td>0.835804</td>\n",
       "      <td>0.798312</td>\n",
       "      <td>[array([[9.99090674e-01, 0.00000000e+00, 4.48546607e-05, 8.64471643e-04],\\n       [8.85813149e-0...</td>\n",
       "      <td>[array([[9.90056144e-01, 0.00000000e+00, 6.54357886e-03, 3.40027746e-03],\\n       [5.52385477e-0...</td>\n",
       "      <td>[array([[9.99688128e-01, 0.00000000e+00, 5.25436197e-05, 2.59328187e-04],\\n       [1.22707273e-0...</td>\n",
       "      <td>[array([[9.99975717e-01, 0.00000000e+00, 1.73452113e-06, 2.25487747e-05],\\n       [1.45473516e-0...</td>\n",
       "      <td>[array([[9.98696010e-01, 0.00000000e+00, 7.92243720e-05, 1.22476597e-03],\\n       [1.54743232e-0...</td>\n",
       "      <td>[array([[9.99830746e-01, 0.00000000e+00, 3.45417178e-05, 1.34712699e-04],\\n       [1.52087785e-0...</td>\n",
       "      <td>[array([[9.99914264e-01, 0.00000000e+00, 1.71472174e-06, 8.40213654e-05],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99701134e-01, 0.00000000e+00, 1.33746572e-04, 1.65119224e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>10192002</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192002/inference_2500</td>\n",
       "      <td>1</td>\n",
       "      <td>0.776913</td>\n",
       "      <td>0.212756</td>\n",
       "      <td>0.153442</td>\n",
       "      <td>0.801818</td>\n",
       "      <td>1.578732</td>\n",
       "      <td>0.366198</td>\n",
       "      <td>0.993348</td>\n",
       "      <td>0.776299</td>\n",
       "      <td>0.787699</td>\n",
       "      <td>0.784175</td>\n",
       "      <td>[array([[9.99213005e-01, 0.00000000e+00, 1.69224402e-04, 6.17771009e-04],\\n       [2.21453287e-0...</td>\n",
       "      <td>[array([[9.90944387e-01, 0.00000000e+00, 6.68137762e-03, 2.37423536e-03],\\n       [5.52385477e-0...</td>\n",
       "      <td>[array([[9.99193201e-01, 0.00000000e+00, 2.71192876e-04, 5.35605930e-04],\\n       [2.45414546e-0...</td>\n",
       "      <td>[array([[9.99987858e-01, 0.00000000e+00, 3.46904227e-06, 8.67260567e-06],\\n       [3.51142970e-0...</td>\n",
       "      <td>[array([[9.98880153e-01, 0.00000000e+00, 1.32754353e-04, 9.87092851e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99856652e-01, 0.00000000e+00, 6.73563496e-05, 7.59917791e-05],\\n       [1.52087785e-0...</td>\n",
       "      <td>[array([[9.99912549e-01, 0.00000000e+00, 1.37177739e-05, 7.37330350e-05],\\n       [3.51506204e-0...</td>\n",
       "      <td>[array([[9.99717646e-01, 0.00000000e+00, 1.89887108e-04, 9.24667656e-05],\\n       [2.20447214e-0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>10192004</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_11100</td>\n",
       "      <td>1</td>\n",
       "      <td>0.805924</td>\n",
       "      <td>0.147715</td>\n",
       "      <td>0.123747</td>\n",
       "      <td>0.801041</td>\n",
       "      <td>1.606965</td>\n",
       "      <td>0.271462</td>\n",
       "      <td>0.995025</td>\n",
       "      <td>0.811626</td>\n",
       "      <td>0.828520</td>\n",
       "      <td>0.822260</td>\n",
       "      <td>[array([[9.98560573e-01, 2.03884821e-06, 4.28158125e-05, 1.39457218e-03],\\n       [1.10726644e-0...</td>\n",
       "      <td>[array([[9.92419206e-01, 0.00000000e+00, 4.84157798e-03, 2.73921585e-03],\\n       [5.52385477e-0...</td>\n",
       "      <td>[array([[9.98855905e-01, 0.00000000e+00, 1.18646883e-04, 1.02544806e-03],\\n       [3.50592209e-0...</td>\n",
       "      <td>[array([[9.99941026e-01, 0.00000000e+00, 6.93808453e-06, 5.20356340e-05],\\n       [3.00979689e-0...</td>\n",
       "      <td>[array([[9.98372689e-01, 0.00000000e+00, 1.22048357e-04, 1.50526307e-03],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99721939e-01, 0.00000000e+00, 2.41792024e-05, 2.53881626e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99876540e-01, 0.00000000e+00, 5.14416523e-06, 1.18315800e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99638389e-01, 0.00000000e+00, 4.95357673e-05, 3.12075334e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>10192004</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_4000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.783851</td>\n",
       "      <td>0.174810</td>\n",
       "      <td>0.108014</td>\n",
       "      <td>0.803988</td>\n",
       "      <td>1.587839</td>\n",
       "      <td>0.282823</td>\n",
       "      <td>0.995970</td>\n",
       "      <td>0.807650</td>\n",
       "      <td>0.824122</td>\n",
       "      <td>0.823838</td>\n",
       "      <td>[array([[9.98132415e-01, 6.11654464e-06, 2.85438750e-05, 1.83292454e-03],\\n       [5.53633218e-0...</td>\n",
       "      <td>[array([[9.91462063e-01, 1.86214538e-06, 4.63301770e-03, 3.90305671e-03],\\n       [5.52385477e-0...</td>\n",
       "      <td>[array([[9.98323689e-01, 0.00000000e+00, 1.28816616e-04, 1.54749435e-03],\\n       [3.50592209e-0...</td>\n",
       "      <td>[array([[9.99913274e-01, 0.00000000e+00, 3.46904227e-06, 8.32570144e-05],\\n       [1.50489844e-0...</td>\n",
       "      <td>[array([[9.98312735e-01, 0.00000000e+00, 1.00636364e-04, 1.58662864e-03],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99702941e-01, 0.00000000e+00, 2.41792024e-05, 2.72879570e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99864537e-01, 0.00000000e+00, 5.14416523e-06, 1.30318853e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99625179e-01, 0.00000000e+00, 1.65119224e-05, 3.58308717e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>10192004</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_3900</td>\n",
       "      <td>1</td>\n",
       "      <td>0.856205</td>\n",
       "      <td>0.096011</td>\n",
       "      <td>0.211923</td>\n",
       "      <td>0.685681</td>\n",
       "      <td>1.541885</td>\n",
       "      <td>0.307934</td>\n",
       "      <td>0.994825</td>\n",
       "      <td>0.791488</td>\n",
       "      <td>0.773357</td>\n",
       "      <td>0.810508</td>\n",
       "      <td>[array([[9.98189503e-01, 2.03884821e-06, 2.85438750e-05, 1.77991449e-03],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.92240440e-01, 1.86214538e-06, 3.60697560e-03, 4.15072205e-03],\\n       [5.52385477e-0...</td>\n",
       "      <td>[array([[9.98333859e-01, 0.00000000e+00, 1.10172106e-04, 1.55596912e-03],\\n       [4.09024243e-0...</td>\n",
       "      <td>[array([[9.99941026e-01, 0.00000000e+00, 1.73452113e-06, 5.72391974e-05],\\n       [1.50489844e-0...</td>\n",
       "      <td>[array([[9.98207816e-01, 0.00000000e+00, 8.13655712e-05, 1.71081820e-03],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99725393e-01, 0.00000000e+00, 2.24521165e-05, 2.52154540e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99859393e-01, 0.00000000e+00, 5.14416523e-06, 1.35463018e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99631784e-01, 0.00000000e+00, 2.80702681e-05, 3.40145602e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>10192004</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_6600</td>\n",
       "      <td>1</td>\n",
       "      <td>0.769910</td>\n",
       "      <td>0.198151</td>\n",
       "      <td>0.084447</td>\n",
       "      <td>0.859092</td>\n",
       "      <td>1.629002</td>\n",
       "      <td>0.282597</td>\n",
       "      <td>0.995134</td>\n",
       "      <td>0.813365</td>\n",
       "      <td>0.843567</td>\n",
       "      <td>0.837692</td>\n",
       "      <td>[array([[9.98468825e-01, 2.03884821e-06, 4.89323571e-05, 1.48020380e-03],\\n       [5.53633218e-0...</td>\n",
       "      <td>[array([[9.91739523e-01, 0.00000000e+00, 5.56222824e-03, 2.69824865e-03],\\n       [5.52385477e-0...</td>\n",
       "      <td>[array([[9.98918618e-01, 0.00000000e+00, 1.66105636e-04, 9.15275956e-04],\\n       [4.09024243e-0...</td>\n",
       "      <td>[array([[9.99937557e-01, 0.00000000e+00, 8.67260567e-06, 5.37701551e-05],\\n       [2.00653126e-0...</td>\n",
       "      <td>[array([[9.98171416e-01, 0.00000000e+00, 1.41319150e-04, 1.68726500e-03],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99725393e-01, 0.00000000e+00, 2.93604601e-05, 2.45246196e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99866252e-01, 0.00000000e+00, 6.85888697e-06, 1.26889409e-04],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.99631784e-01, 0.00000000e+00, 7.26524587e-05, 2.95563412e-04],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>10192004</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_2900</td>\n",
       "      <td>1</td>\n",
       "      <td>0.717924</td>\n",
       "      <td>0.221304</td>\n",
       "      <td>0.074288</td>\n",
       "      <td>0.791369</td>\n",
       "      <td>1.509293</td>\n",
       "      <td>0.295592</td>\n",
       "      <td>0.990993</td>\n",
       "      <td>0.786643</td>\n",
       "      <td>0.805821</td>\n",
       "      <td>0.768787</td>\n",
       "      <td>[array([[9.85075631e-01, 2.03884821e-06, 2.01845973e-04, 1.47204841e-02],\\n       [2.21453287e-0...</td>\n",
       "      <td>[array([[9.77248308e-01, 1.86214538e-06, 4.49149465e-03, 1.82583354e-02],\\n       [5.52385477e-0...</td>\n",
       "      <td>[array([[9.88904821e-01, 0.00000000e+00, 6.33913347e-04, 1.04612652e-02],\\n       [4.67456278e-0...</td>\n",
       "      <td>[array([[9.97557794e-01, 0.00000000e+00, 1.11009353e-04, 2.33119640e-03],\\n       [5.01632815e-0...</td>\n",
       "      <td>[array([[9.90970563e-01, 0.00000000e+00, 5.99535788e-05, 8.96948363e-03],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.97171033e-01, 0.00000000e+00, 3.69596380e-04, 2.45937030e-03],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.96246474e-01, 0.00000000e+00, 2.10910774e-04, 3.54261512e-03],\\n       [0.00000000e+0...</td>\n",
       "      <td>[array([[9.96169234e-01, 0.00000000e+00, 4.39217137e-04, 3.39154887e-03],\\n       [0.00000000e+0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>10202100</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202100/inference_12000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.727369</td>\n",
       "      <td>0.244711</td>\n",
       "      <td>0.052268</td>\n",
       "      <td>0.846252</td>\n",
       "      <td>1.573621</td>\n",
       "      <td>0.296979</td>\n",
       "      <td>0.996687</td>\n",
       "      <td>0.806246</td>\n",
       "      <td>0.830782</td>\n",
       "      <td>0.827865</td>\n",
       "      <td>[[9.97645130e-01 0.00000000e+00 5.91265982e-05 2.29574309e-03]\\n [2.93425606e-04 7.27368858e-01 ...</td>\n",
       "      <td>[[9.90180907e-01 1.86214538e-06 4.87695874e-03 4.94027169e-03]\\n [1.76763353e-04 8.24114940e-01 ...</td>\n",
       "      <td>[[9.99755926e-01 0.00000000e+00 6.77982189e-06 2.37293766e-04]\\n [7.59616452e-05 7.68001449e-01 ...</td>\n",
       "      <td>[[9.99954902e-01 0.00000000e+00 5.20356340e-06 3.98939861e-05]\\n [1.00326563e-04 7.23981560e-01 ...</td>\n",
       "      <td>[[9.97486232e-01 0.00000000e+00 7.70831727e-05 2.43668474e-03]\\n [2.11013498e-05 8.55828544e-01 ...</td>\n",
       "      <td>[[9.99649402e-01 0.00000000e+00 2.59062883e-05 3.24692147e-04]\\n [7.60438925e-06 9.40533676e-01 ...</td>\n",
       "      <td>[[9.99819954e-01 0.00000000e+00 6.85888697e-06 1.73186896e-04]\\n [1.40602482e-05 9.20918134e-01 ...</td>\n",
       "      <td>[[9.99527759e-01 0.00000000e+00 2.31166914e-05 4.49124290e-04]\\n [0.00000000e+00 9.29236444e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>10202100</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202100/inference_5800</td>\n",
       "      <td>1</td>\n",
       "      <td>0.800781</td>\n",
       "      <td>0.179947</td>\n",
       "      <td>0.113441</td>\n",
       "      <td>0.809683</td>\n",
       "      <td>1.610464</td>\n",
       "      <td>0.293389</td>\n",
       "      <td>0.991585</td>\n",
       "      <td>0.811223</td>\n",
       "      <td>0.825079</td>\n",
       "      <td>0.821793</td>\n",
       "      <td>[[9.99500482e-01 2.03884821e-06 3.66992678e-05 4.60779696e-04]\\n [1.43944637e-04 8.00780623e-01 ...</td>\n",
       "      <td>[[9.91167844e-01 3.72429076e-06 5.50450174e-03 3.32392950e-03]\\n [3.31431286e-05 8.34289881e-01 ...</td>\n",
       "      <td>[[9.99938982e-01 0.00000000e+00 0.00000000e+00 6.10183970e-05]\\n [6.36909179e-04 7.45312290e-01 ...</td>\n",
       "      <td>[[9.99993062e-01 0.00000000e+00 1.73452113e-06 5.20356340e-06]\\n [7.02285941e-05 7.37154438e-01 ...</td>\n",
       "      <td>[[9.98745257e-01 0.00000000e+00 9.63539659e-05 1.15838879e-03]\\n [0.00000000e+00 8.55596430e-01 ...</td>\n",
       "      <td>[[9.99856652e-01 0.00000000e+00 2.59062883e-05 1.17441840e-04]\\n [7.22416979e-04 9.08846186e-01 ...</td>\n",
       "      <td>[[9.99915979e-01 0.00000000e+00 0.00000000e+00 8.40213654e-05]\\n [2.46054343e-04 8.87911702e-01 ...</td>\n",
       "      <td>[[9.99725902e-01 0.00000000e+00 4.29309983e-05 2.31166914e-04]\\n [6.61341642e-05 9.07684055e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>10202100</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202100/inference_2000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.801738</td>\n",
       "      <td>0.181940</td>\n",
       "      <td>0.145419</td>\n",
       "      <td>0.815258</td>\n",
       "      <td>1.616996</td>\n",
       "      <td>0.327359</td>\n",
       "      <td>0.993349</td>\n",
       "      <td>0.794069</td>\n",
       "      <td>0.807127</td>\n",
       "      <td>0.788786</td>\n",
       "      <td>[[9.98799118e-01 0.00000000e+00 5.52527866e-04 6.48353732e-04]\\n [4.42906574e-05 8.01738408e-01 ...</td>\n",
       "      <td>[[9.89396944e-01 7.63479605e-05 7.05753098e-03 3.46917684e-03]\\n [1.10477095e-05 5.30058056e-01 ...</td>\n",
       "      <td>[[9.99883048e-01 5.08486642e-06 1.69495547e-05 9.49175065e-05]\\n [7.59616452e-05 8.27210630e-01 ...</td>\n",
       "      <td>[[9.99972248e-01 0.00000000e+00 1.90797325e-05 8.67260567e-06]\\n [1.50489844e-05 7.73914090e-01 ...</td>\n",
       "      <td>[[9.98471184e-01 0.00000000e+00 3.23321086e-04 1.20549517e-03]\\n [2.67283764e-04 8.11867399e-01 ...</td>\n",
       "      <td>[[9.99768570e-01 0.00000000e+00 1.08806411e-04 1.22623098e-04]\\n [4.41054577e-03 8.93477715e-01 ...</td>\n",
       "      <td>[[9.99888543e-01 0.00000000e+00 3.60091566e-05 7.54477567e-05]\\n [2.90344125e-03 8.78568667e-01 ...</td>\n",
       "      <td>[[9.99679669e-01 0.00000000e+00 2.41074068e-04 7.92572277e-05]\\n [1.39616569e-04 9.14936768e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>10202102</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202102/inference_20000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.751585</td>\n",
       "      <td>0.226447</td>\n",
       "      <td>0.080895</td>\n",
       "      <td>0.846833</td>\n",
       "      <td>1.598417</td>\n",
       "      <td>0.307342</td>\n",
       "      <td>0.994971</td>\n",
       "      <td>0.803225</td>\n",
       "      <td>0.828606</td>\n",
       "      <td>0.824020</td>\n",
       "      <td>[[9.98501447e-01 0.00000000e+00 8.56316250e-05 1.41292181e-03]\\n [3.43252595e-04 7.51584775e-01 ...</td>\n",
       "      <td>[[9.92545832e-01 5.58643613e-06 3.76339581e-03 3.68518570e-03]\\n [3.31431286e-05 7.15217667e-01 ...</td>\n",
       "      <td>[[9.99827115e-01 0.00000000e+00 5.08486642e-06 1.67800592e-04]\\n [4.14867447e-04 8.61697217e-01 ...</td>\n",
       "      <td>[[9.99967044e-01 0.00000000e+00 5.20356340e-06 2.77523381e-05]\\n [1.50489844e-05 9.02698283e-01 ...</td>\n",
       "      <td>[[9.98762387e-01 0.00000000e+00 4.71063833e-05 1.19050678e-03]\\n [0.00000000e+00 8.37772823e-01 ...</td>\n",
       "      <td>[[9.99801385e-01 0.00000000e+00 2.41792024e-05 1.74435675e-04]\\n [3.04175570e-05 9.30039619e-01 ...</td>\n",
       "      <td>[[9.99909120e-01 0.00000000e+00 0.00000000e+00 9.08802524e-05]\\n [9.13916131e-05 9.31864037e-01 ...</td>\n",
       "      <td>[[9.99678018e-01 0.00000000e+00 3.30238449e-06 3.18680103e-04]\\n [0.00000000e+00 9.34791714e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>10202102</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202102/inference_11600</td>\n",
       "      <td>1</td>\n",
       "      <td>0.810325</td>\n",
       "      <td>0.179549</td>\n",
       "      <td>0.136749</td>\n",
       "      <td>0.828797</td>\n",
       "      <td>1.639123</td>\n",
       "      <td>0.316298</td>\n",
       "      <td>0.995893</td>\n",
       "      <td>0.802237</td>\n",
       "      <td>0.815528</td>\n",
       "      <td>0.807952</td>\n",
       "      <td>[[9.97922414e-01 0.00000000e+00 6.99324937e-04 1.37826139e-03]\\n [4.98269896e-05 8.10325260e-01 ...</td>\n",
       "      <td>[[9.90517956e-01 0.00000000e+00 7.35919853e-03 2.12284573e-03]\\n [5.52385477e-06 6.42982219e-01 ...</td>\n",
       "      <td>[[9.99832199e-01 0.00000000e+00 2.20344212e-05 1.45766171e-04]\\n [7.01184417e-05 9.01343352e-01 ...</td>\n",
       "      <td>[[9.99963575e-01 0.00000000e+00 1.73452113e-05 1.90797325e-05]\\n [0.00000000e+00 9.22502746e-01 ...</td>\n",
       "      <td>[[9.98250640e-01 2.14119924e-06 2.39814315e-04 1.50740427e-03]\\n [9.00324257e-04 8.49371531e-01 ...</td>\n",
       "      <td>[[9.99746118e-01 0.00000000e+00 8.63542944e-05 1.67527331e-04]\\n [1.51327346e-03 9.64046448e-01 ...</td>\n",
       "      <td>[[9.99895402e-01 0.00000000e+00 2.40061044e-05 8.05919219e-05]\\n [3.51506204e-03 9.40511090e-01 ...</td>\n",
       "      <td>[[9.99658203e-01 0.00000000e+00 1.93189492e-04 1.48607302e-04]\\n [0.00000000e+00 9.55638672e-01 ...</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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>10210130</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_20000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.687646</td>\n",
       "      <td>0.279961</td>\n",
       "      <td>0.035215</td>\n",
       "      <td>0.873425</td>\n",
       "      <td>1.561071</td>\n",
       "      <td>0.315177</td>\n",
       "      <td>0.996453</td>\n",
       "      <td>0.791053</td>\n",
       "      <td>0.833452</td>\n",
       "      <td>0.826392</td>\n",
       "      <td>[[9.97569693e-01 0.00000000e+00 4.89323571e-05 2.38137471e-03]\\n [7.19723183e-05 6.87645675e-01 ...</td>\n",
       "      <td>[[9.87851364e-01 3.72429076e-06 7.65714179e-03 4.48777036e-03]\\n [2.20954191e-05 5.37338496e-01 ...</td>\n",
       "      <td>[[9.99193201e-01 8.47477737e-06 1.59325815e-04 6.38998214e-04]\\n [1.57766494e-04 6.50225840e-01 ...</td>\n",
       "      <td>[[9.99946230e-01 0.00000000e+00 1.56106902e-05 3.81594649e-05]\\n [3.51142970e-05 5.91023782e-01 ...</td>\n",
       "      <td>[[9.97385596e-01 0.00000000e+00 1.13483560e-04 2.50092072e-03]\\n [0.00000000e+00 7.74813429e-01 ...</td>\n",
       "      <td>[[9.99687397e-01 3.45417178e-06 3.28146319e-05 2.76333742e-04]\\n [0.00000000e+00 9.34488187e-01 ...</td>\n",
       "      <td>[[9.99777086e-01 0.00000000e+00 8.57360872e-06 2.14340218e-04]\\n [0.00000000e+00 9.13880980e-01 ...</td>\n",
       "      <td>[[9.99572341e-01 0.00000000e+00 9.90715346e-06 4.17751638e-04]\\n [7.34824046e-06 9.06478944e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>10210130</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_12000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.862400</td>\n",
       "      <td>0.107172</td>\n",
       "      <td>0.171404</td>\n",
       "      <td>0.738618</td>\n",
       "      <td>1.601018</td>\n",
       "      <td>0.278576</td>\n",
       "      <td>0.990123</td>\n",
       "      <td>0.816217</td>\n",
       "      <td>0.805659</td>\n",
       "      <td>0.824301</td>\n",
       "      <td>[[9.98236396e-01 0.00000000e+00 2.03884821e-06 1.76156486e-03]\\n [6.69896194e-04 8.62400000e-01 ...</td>\n",
       "      <td>[[9.90665065e-01 1.86214538e-06 5.99052168e-03 3.34255095e-03]\\n [1.27048660e-04 8.57335403e-01 ...</td>\n",
       "      <td>[[9.99237270e-01 0.00000000e+00 5.59335306e-05 7.06796432e-04]\\n [1.91072754e-03 7.83877433e-01 ...</td>\n",
       "      <td>[[9.99941026e-01 0.00000000e+00 0.00000000e+00 5.89737185e-05]\\n [1.99649860e-03 7.60209482e-01 ...</td>\n",
       "      <td>[[9.95130913e-01 2.14119924e-06 5.13887818e-05 4.81555710e-03]\\n [1.89208770e-03 8.82535820e-01 ...</td>\n",
       "      <td>[[9.99640766e-01 0.00000000e+00 2.07250307e-05 3.38508834e-04]\\n [2.20527288e-04 9.37286602e-01 ...</td>\n",
       "      <td>[[9.99609043e-01 0.00000000e+00 0.00000000e+00 3.90956558e-04]\\n [3.65566452e-04 9.33537207e-01 ...</td>\n",
       "      <td>[[9.99408873e-01 0.00000000e+00 0.00000000e+00 5.91126823e-04]\\n [1.98402493e-04 9.46615033e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>10210130</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_6000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.812136</td>\n",
       "      <td>0.158749</td>\n",
       "      <td>0.172905</td>\n",
       "      <td>0.789296</td>\n",
       "      <td>1.601431</td>\n",
       "      <td>0.331653</td>\n",
       "      <td>0.991836</td>\n",
       "      <td>0.784101</td>\n",
       "      <td>0.790127</td>\n",
       "      <td>0.740799</td>\n",
       "      <td>[[9.98317950e-01 0.00000000e+00 5.52527866e-04 1.12952191e-03]\\n [1.10726644e-04 8.12135640e-01 ...</td>\n",
       "      <td>[[9.71376963e-01 4.76709217e-03 1.08134782e-02 1.30424662e-02]\\n [2.20954191e-05 5.10155607e-01 ...</td>\n",
       "      <td>[[9.98759293e-01 1.01697328e-05 3.96619581e-04 8.33918093e-04]\\n [1.11605186e-03 8.38949626e-01 ...</td>\n",
       "      <td>[[9.99628812e-01 0.00000000e+00 1.28354564e-04 2.42832959e-04]\\n [1.90620470e-04 8.27167430e-01 ...</td>\n",
       "      <td>[[9.97252841e-01 1.07059962e-05 2.56943909e-04 2.47950872e-03]\\n [1.40675665e-05 9.28360917e-01 ...</td>\n",
       "      <td>[[9.99411064e-01 0.00000000e+00 1.34712699e-04 4.54223589e-04]\\n [8.36482818e-05 9.66586314e-01 ...</td>\n",
       "      <td>[[9.99075765e-01 0.00000000e+00 1.16601079e-04 8.07633941e-04]\\n [1.26542233e-04 9.63717530e-01 ...</td>\n",
       "      <td>[[9.99058820e-01 1.65119224e-06 2.74097912e-04 6.65430474e-04]\\n [3.67412023e-05 9.69423971e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>10210132</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_6200</td>\n",
       "      <td>1</td>\n",
       "      <td>0.658901</td>\n",
       "      <td>0.298491</td>\n",
       "      <td>0.048867</td>\n",
       "      <td>0.873030</td>\n",
       "      <td>1.531931</td>\n",
       "      <td>0.347359</td>\n",
       "      <td>0.992602</td>\n",
       "      <td>0.761938</td>\n",
       "      <td>0.822676</td>\n",
       "      <td>0.801886</td>\n",
       "      <td>[[9.98570767e-01 0.00000000e+00 1.14175500e-04 1.31505710e-03]\\n [2.15916955e-04 6.58901038e-01 ...</td>\n",
       "      <td>[[9.77626323e-01 7.44858151e-06 7.19719188e-03 1.51690362e-02]\\n [3.86669834e-05 4.54745820e-01 ...</td>\n",
       "      <td>[[9.98435556e-01 0.00000000e+00 1.33901482e-04 1.43054242e-03]\\n [1.19785671e-03 6.44540403e-01 ...</td>\n",
       "      <td>[[9.99731149e-01 0.00000000e+00 2.08142536e-05 2.48036522e-04]\\n [1.60522501e-04 6.27828582e-01 ...</td>\n",
       "      <td>[[9.96244337e-01 1.07059962e-05 9.42127667e-05 3.65074471e-03]\\n [2.03276336e-03 8.17902385e-01 ...</td>\n",
       "      <td>[[9.99592408e-01 0.00000000e+00 4.83584049e-05 3.59233865e-04]\\n [3.80219463e-05 9.09279636e-01 ...</td>\n",
       "      <td>[[9.99473580e-01 0.00000000e+00 3.08649914e-05 4.95554584e-04]\\n [0.00000000e+00 8.82189181e-01 ...</td>\n",
       "      <td>[[9.99128170e-01 0.00000000e+00 8.42108044e-05 7.87618700e-04]\\n [6.61341642e-05 9.00681182e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>10210132</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_25000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.752758</td>\n",
       "      <td>0.220396</td>\n",
       "      <td>0.075735</td>\n",
       "      <td>0.844798</td>\n",
       "      <td>1.597556</td>\n",
       "      <td>0.296130</td>\n",
       "      <td>0.995467</td>\n",
       "      <td>0.808363</td>\n",
       "      <td>0.829791</td>\n",
       "      <td>0.823269</td>\n",
       "      <td>[[9.98114065e-01 0.00000000e+00 1.24369741e-04 1.76156486e-03]\\n [7.75086505e-05 7.52758478e-01 ...</td>\n",
       "      <td>[[9.88035716e-01 1.48971630e-05 7.72976546e-03 4.21962143e-03]\\n [6.62862572e-05 5.62361558e-01 ...</td>\n",
       "      <td>[[9.99260999e-01 0.00000000e+00 5.08486642e-05 6.88151922e-04]\\n [3.38905802e-04 7.49122059e-01 ...</td>\n",
       "      <td>[[9.99947964e-01 0.00000000e+00 1.04071268e-05 4.16285072e-05]\\n [1.40457188e-04 7.59331624e-01 ...</td>\n",
       "      <td>[[9.98053650e-01 8.56479697e-06 8.99303682e-05 1.84785495e-03]\\n [0.00000000e+00 8.35423539e-01 ...</td>\n",
       "      <td>[[9.99727120e-01 0.00000000e+00 3.45417178e-05 2.38337853e-04]\\n [2.28131678e-05 9.44115343e-01 ...</td>\n",
       "      <td>[[9.99854249e-01 0.00000000e+00 1.02883305e-05 1.35463018e-04]\\n [0.00000000e+00 9.22570213e-01 ...</td>\n",
       "      <td>[[9.99644994e-01 0.00000000e+00 1.48607302e-05 3.40145602e-04]\\n [7.34824046e-06 9.38083726e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>10210132</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_12000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.732130</td>\n",
       "      <td>0.202447</td>\n",
       "      <td>0.080307</td>\n",
       "      <td>0.869290</td>\n",
       "      <td>1.601420</td>\n",
       "      <td>0.282754</td>\n",
       "      <td>0.993884</td>\n",
       "      <td>0.791720</td>\n",
       "      <td>0.840746</td>\n",
       "      <td>0.794139</td>\n",
       "      <td>[[9.96927456e-01 0.00000000e+00 1.08058955e-04 2.96448530e-03]\\n [2.15916955e-04 7.32130104e-01 ...</td>\n",
       "      <td>[[9.88592497e-01 1.86214538e-06 6.65158329e-03 4.75405715e-03]\\n [3.31431286e-05 7.57337060e-01 ...</td>\n",
       "      <td>[[9.98133854e-01 0.00000000e+00 1.15256972e-04 1.75088900e-03]\\n [1.46664407e-03 5.88615102e-01 ...</td>\n",
       "      <td>[[9.99814406e-01 0.00000000e+00 1.38761691e-05 1.71717592e-04]\\n [2.40783751e-04 5.62355467e-01 ...</td>\n",
       "      <td>[[9.96336408e-01 0.00000000e+00 1.43460349e-04 3.52013156e-03]\\n [0.00000000e+00 8.83337671e-01 ...</td>\n",
       "      <td>[[9.99341980e-01 0.00000000e+00 7.77188650e-05 5.80300858e-04]\\n [7.60438925e-06 9.11682623e-01 ...</td>\n",
       "      <td>[[9.99163216e-01 0.00000000e+00 1.37177739e-05 8.23066437e-04]\\n [0.00000000e+00 9.05402650e-01 ...</td>\n",
       "      <td>[[9.99223940e-01 0.00000000e+00 1.79979955e-04 5.96080400e-04]\\n [1.46964809e-05 9.08169039e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>10210134</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210134/inference_11700</td>\n",
       "      <td>1</td>\n",
       "      <td>0.667045</td>\n",
       "      <td>0.269603</td>\n",
       "      <td>0.032461</td>\n",
       "      <td>0.912339</td>\n",
       "      <td>1.579384</td>\n",
       "      <td>0.302064</td>\n",
       "      <td>0.995689</td>\n",
       "      <td>0.777983</td>\n",
       "      <td>0.844907</td>\n",
       "      <td>0.795348</td>\n",
       "      <td>[[9.96544152e-01 0.00000000e+00 5.64760955e-04 2.89108677e-03]\\n [1.71626298e-04 6.67044983e-01 ...</td>\n",
       "      <td>[[9.86985466e-01 0.00000000e+00 8.61428452e-03 4.40024953e-03]\\n [6.07624024e-05 6.34160623e-01 ...</td>\n",
       "      <td>[[9.95881258e-01 0.00000000e+00 8.40697915e-04 3.27804389e-03]\\n [2.57100953e-04 6.71728829e-01 ...</td>\n",
       "      <td>[[9.98933270e-01 0.00000000e+00 5.34232509e-04 5.32497988e-04]\\n [6.52122659e-05 6.92905407e-01 ...</td>\n",
       "      <td>[[9.94745497e-01 0.00000000e+00 6.87324957e-04 4.56717799e-03]\\n [0.00000000e+00 8.55596430e-01 ...</td>\n",
       "      <td>[[9.98466348e-01 0.00000000e+00 3.64415122e-04 1.16923715e-03]\\n [1.52087785e-05 9.07728341e-01 ...</td>\n",
       "      <td>[[9.98245840e-01 0.00000000e+00 3.63521010e-04 1.39063933e-03]\\n [0.00000000e+00 8.88966220e-01 ...</td>\n",
       "      <td>[[9.98198549e-01 0.00000000e+00 4.72240982e-04 1.32920976e-03]\\n [7.34824046e-06 8.92392367e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>10210134</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210134/inference_22000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.732927</td>\n",
       "      <td>0.227139</td>\n",
       "      <td>0.079762</td>\n",
       "      <td>0.874268</td>\n",
       "      <td>1.607196</td>\n",
       "      <td>0.306901</td>\n",
       "      <td>0.995883</td>\n",
       "      <td>0.792475</td>\n",
       "      <td>0.832953</td>\n",
       "      <td>0.808690</td>\n",
       "      <td>[[9.97304643e-01 0.00000000e+00 3.91458857e-04 2.30389848e-03]\\n [4.70588235e-04 7.32927336e-01 ...</td>\n",
       "      <td>[[9.87227545e-01 1.11728723e-05 8.91222778e-03 3.84905450e-03]\\n [5.52385477e-05 6.38220656e-01 ...</td>\n",
       "      <td>[[9.97005014e-01 0.00000000e+00 6.18658748e-04 2.37632757e-03]\\n [1.04593342e-03 6.95569099e-01 ...</td>\n",
       "      <td>[[9.99597591e-01 0.00000000e+00 1.78655677e-04 2.23753226e-04]\\n [9.02939067e-05 7.30387411e-01 ...</td>\n",
       "      <td>[[9.96342832e-01 0.00000000e+00 4.11110255e-04 3.24605805e-03]\\n [7.03378326e-06 8.70564320e-01 ...</td>\n",
       "      <td>[[9.99131276e-01 0.00000000e+00 4.00683926e-04 4.68040276e-04]\\n [0.00000000e+00 9.16846004e-01 ...</td>\n",
       "      <td>[[9.99156357e-01 0.00000000e+00 2.60637705e-04 5.83005393e-04]\\n [0.00000000e+00 9.03673240e-01 ...</td>\n",
       "      <td>[[9.98687302e-01 0.00000000e+00 6.30755437e-04 6.81942397e-04]\\n [1.46964809e-05 9.04392043e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>10210134</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210134/inference_4400</td>\n",
       "      <td>1</td>\n",
       "      <td>0.522508</td>\n",
       "      <td>0.421381</td>\n",
       "      <td>0.012391</td>\n",
       "      <td>0.933210</td>\n",
       "      <td>1.455717</td>\n",
       "      <td>0.433772</td>\n",
       "      <td>0.995112</td>\n",
       "      <td>0.678361</td>\n",
       "      <td>0.816866</td>\n",
       "      <td>0.802167</td>\n",
       "      <td>[[9.96668522e-01 2.03884821e-06 7.56412687e-04 2.57302645e-03]\\n [5.53633218e-06 5.22507958e-01 ...</td>\n",
       "      <td>[[9.88335521e-01 0.00000000e+00 6.67392903e-03 4.99054961e-03]\\n [1.10477095e-05 4.68616219e-01 ...</td>\n",
       "      <td>[[9.97138915e-01 0.00000000e+00 1.02036320e-03 1.84072164e-03]\\n [5.96006755e-04 5.53777923e-01 ...</td>\n",
       "      <td>[[9.99321802e-01 0.00000000e+00 4.83931396e-04 1.94266367e-04]\\n [1.35440860e-04 5.63900496e-01 ...</td>\n",
       "      <td>[[9.96019511e-01 8.56479697e-06 4.56075439e-04 3.51584916e-03]\\n [7.03378326e-06 7.97251198e-01 ...</td>\n",
       "      <td>[[9.99152001e-01 0.00000000e+00 3.47144263e-04 5.00854908e-04]\\n [7.60438925e-06 8.50284784e-01 ...</td>\n",
       "      <td>[[9.99259240e-01 0.00000000e+00 2.72640757e-04 4.68119036e-04]\\n [0.00000000e+00 8.32275300e-01 ...</td>\n",
       "      <td>[[9.99055518e-01 0.00000000e+00 3.99588523e-04 5.44893440e-04]\\n [1.46964809e-05 8.45268101e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>10202106</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202106/inference_2100</td>\n",
       "      <td>1</td>\n",
       "      <td>0.587643</td>\n",
       "      <td>0.369888</td>\n",
       "      <td>0.033761</td>\n",
       "      <td>0.866562</td>\n",
       "      <td>1.454205</td>\n",
       "      <td>0.403649</td>\n",
       "      <td>0.987312</td>\n",
       "      <td>0.717716</td>\n",
       "      <td>0.798132</td>\n",
       "      <td>0.743359</td>\n",
       "      <td>[[9.99429123e-01 0.00000000e+00 1.85535187e-04 3.85342312e-04]\\n [2.04844291e-04 5.87642907e-01 ...</td>\n",
       "      <td>[[9.87702392e-01 3.72429076e-06 5.88251725e-03 6.41136654e-03]\\n [2.87240448e-04 2.52136351e-01 ...</td>\n",
       "      <td>[[9.99766096e-01 6.77982189e-06 6.94931744e-05 1.57630859e-04]\\n [8.12205283e-04 7.43670350e-01 ...</td>\n",
       "      <td>[[9.99963575e-01 0.00000000e+00 6.93808453e-06 2.94868593e-05]\\n [9.02939067e-05 6.89544467e-01 ...</td>\n",
       "      <td>[[9.97950872e-01 1.28471955e-05 1.07059962e-04 1.92922052e-03]\\n [4.92364828e-05 7.49674688e-01 ...</td>\n",
       "      <td>[[9.99768570e-01 0.00000000e+00 5.00854908e-05 1.81344018e-04]\\n [3.49801906e-04 8.71782393e-01 ...</td>\n",
       "      <td>[[9.99855963e-01 0.00000000e+00 8.57360872e-06 1.35463018e-04]\\n [1.82783226e-04 8.46947169e-01 ...</td>\n",
       "      <td>[[9.99653250e-01 0.00000000e+00 1.20537034e-04 2.26213337e-04]\\n [2.42491935e-04 8.59134230e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>10202106</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202106/inference_9000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.563482</td>\n",
       "      <td>0.411848</td>\n",
       "      <td>0.021130</td>\n",
       "      <td>0.900106</td>\n",
       "      <td>1.463589</td>\n",
       "      <td>0.432978</td>\n",
       "      <td>0.993331</td>\n",
       "      <td>0.707024</td>\n",
       "      <td>0.806870</td>\n",
       "      <td>0.806477</td>\n",
       "      <td>[[9.98342416e-01 0.00000000e+00 1.16214348e-04 1.54136925e-03]\\n [1.60553633e-04 5.63482353e-01 ...</td>\n",
       "      <td>[[9.88806644e-01 0.00000000e+00 6.01472957e-03 5.17862630e-03]\\n [6.62862572e-05 4.38914452e-01 ...</td>\n",
       "      <td>[[9.99547447e-01 0.00000000e+00 3.72890204e-05 4.15264091e-04]\\n [9.93344591e-05 7.40433215e-01 ...</td>\n",
       "      <td>[[9.99764105e-01 0.00000000e+00 7.97879721e-05 1.56106902e-04]\\n [0.00000000e+00 7.22797707e-01 ...</td>\n",
       "      <td>[[9.96933803e-01 4.28239849e-06 9.20715674e-05 2.96984335e-03]\\n [7.52614809e-04 7.75256557e-01 ...</td>\n",
       "      <td>[[9.99561320e-01 0.00000000e+00 3.10875460e-05 4.07592270e-04]\\n [2.28131678e-05 8.36406774e-01 ...</td>\n",
       "      <td>[[9.99701638e-01 0.00000000e+00 2.22913827e-05 2.76070201e-04]\\n [1.40602482e-05 8.21561391e-01 ...</td>\n",
       "      <td>[[9.99483177e-01 0.00000000e+00 2.31166914e-05 4.93706481e-04]\\n [0.00000000e+00 8.29417946e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>10210050</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210050/inference_47700</td>\n",
       "      <td>1</td>\n",
       "      <td>0.787571</td>\n",
       "      <td>0.184371</td>\n",
       "      <td>0.088806</td>\n",
       "      <td>0.862547</td>\n",
       "      <td>1.650118</td>\n",
       "      <td>0.273177</td>\n",
       "      <td>0.992172</td>\n",
       "      <td>0.821132</td>\n",
       "      <td>0.846978</td>\n",
       "      <td>0.825889</td>\n",
       "      <td>[[9.99400579e-01 0.00000000e+00 1.34563982e-04 4.64857393e-04]\\n [8.02768166e-04 7.87570934e-01 ...</td>\n",
       "      <td>[[9.92761841e-01 0.00000000e+00 6.12459615e-03 1.11356294e-03]\\n [3.25907431e-04 7.10953252e-01 ...</td>\n",
       "      <td>[[9.99969491e-01 6.77982189e-06 3.38991095e-06 2.03394657e-05]\\n [7.30400435e-04 7.63969639e-01 ...</td>\n",
       "      <td>[[9.99986124e-01 0.00000000e+00 8.67260567e-06 5.20356340e-06]\\n [1.05342891e-04 7.22010143e-01 ...</td>\n",
       "      <td>[[9.98663892e-01 0.00000000e+00 1.60589943e-04 1.17551838e-03]\\n [0.00000000e+00 8.79680104e-01 ...</td>\n",
       "      <td>[[9.99877377e-01 0.00000000e+00 3.10875460e-05 9.15355521e-05]\\n [3.04175570e-05 9.51331909e-01 ...</td>\n",
       "      <td>[[9.99921123e-01 0.00000000e+00 6.85888697e-06 7.20183132e-05]\\n [2.10903722e-05 9.31955429e-01 ...</td>\n",
       "      <td>[[9.99727553e-01 0.00000000e+00 1.00722727e-04 1.71723993e-04]\\n [7.34824046e-06 9.43477334e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>10210050</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210050/inference_60000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.854372</td>\n",
       "      <td>0.132274</td>\n",
       "      <td>0.148030</td>\n",
       "      <td>0.797729</td>\n",
       "      <td>1.652101</td>\n",
       "      <td>0.280304</td>\n",
       "      <td>0.995481</td>\n",
       "      <td>0.820110</td>\n",
       "      <td>0.825959</td>\n",
       "      <td>0.833727</td>\n",
       "      <td>[[9.98719603e-01 0.00000000e+00 2.50778330e-04 1.02961835e-03]\\n [1.66089965e-04 8.54372318e-01 ...</td>\n",
       "      <td>[[9.90722792e-01 1.86214538e-06 7.04263382e-03 2.23271231e-03]\\n [7.73339667e-05 7.74096435e-01 ...</td>\n",
       "      <td>[[9.99888133e-01 2.20344212e-05 2.03394657e-05 6.94931744e-05]\\n [1.16864070e-04 8.91222924e-01 ...</td>\n",
       "      <td>[[9.99977451e-01 0.00000000e+00 1.73452113e-05 5.20356340e-06]\\n [5.01632815e-06 8.09956408e-01 ...</td>\n",
       "      <td>[[9.98060073e-01 8.56479697e-06 1.43460349e-04 1.78790137e-03]\\n [0.00000000e+00 8.22861202e-01 ...</td>\n",
       "      <td>[[9.99787568e-01 0.00000000e+00 3.62688036e-05 1.76162761e-04]\\n [0.00000000e+00 9.53559995e-01 ...</td>\n",
       "      <td>[[9.99903976e-01 0.00000000e+00 3.42944349e-06 9.25949742e-05]\\n [0.00000000e+00 9.47217828e-01 ...</td>\n",
       "      <td>[[9.99687925e-01 0.00000000e+00 8.25596122e-05 2.29515722e-04]\\n [0.00000000e+00 9.61333559e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>10210052</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210052/inference_55400</td>\n",
       "      <td>1</td>\n",
       "      <td>0.862876</td>\n",
       "      <td>0.127103</td>\n",
       "      <td>0.177458</td>\n",
       "      <td>0.788410</td>\n",
       "      <td>1.651286</td>\n",
       "      <td>0.304561</td>\n",
       "      <td>0.995355</td>\n",
       "      <td>0.807689</td>\n",
       "      <td>0.810778</td>\n",
       "      <td>0.819124</td>\n",
       "      <td>[[9.97849015e-01 0.00000000e+00 1.95729429e-04 1.95525544e-03]\\n [5.75778547e-04 8.62876125e-01 ...</td>\n",
       "      <td>[[9.89830824e-01 1.35936613e-04 7.74652477e-03 2.28671452e-03]\\n [5.52385477e-05 8.18220987e-01 ...</td>\n",
       "      <td>[[9.99150827e-01 0.00000000e+00 2.00004746e-04 6.49167946e-04]\\n [1.98668918e-04 8.70058841e-01 ...</td>\n",
       "      <td>[[9.99772778e-01 0.00000000e+00 4.50975495e-05 1.82124719e-04]\\n [0.00000000e+00 9.18906039e-01 ...</td>\n",
       "      <td>[[9.96659729e-01 1.71295939e-05 1.75578338e-04 3.14756289e-03]\\n [0.00000000e+00 9.38953795e-01 ...</td>\n",
       "      <td>[[9.99575137e-01 0.00000000e+00 7.77188650e-05 3.47144263e-04]\\n [3.80219463e-05 9.78509996e-01 ...</td>\n",
       "      <td>[[9.99494157e-01 0.00000000e+00 2.22913827e-05 4.83551532e-04]\\n [0.00000000e+00 9.70213364e-01 ...</td>\n",
       "      <td>[[9.99124868e-01 0.00000000e+00 1.78328762e-04 6.96803127e-04]\\n [1.46964809e-05 9.78183074e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>10210052</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210052/inference_70000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.795792</td>\n",
       "      <td>0.186408</td>\n",
       "      <td>0.098338</td>\n",
       "      <td>0.843807</td>\n",
       "      <td>1.639600</td>\n",
       "      <td>0.284747</td>\n",
       "      <td>0.996155</td>\n",
       "      <td>0.818509</td>\n",
       "      <td>0.840425</td>\n",
       "      <td>0.852010</td>\n",
       "      <td>[[9.96902990e-01 0.00000000e+00 1.83496339e-05 3.07866080e-03]\\n [4.92733564e-04 7.95792388e-01 ...</td>\n",
       "      <td>[[9.93245999e-01 1.48971630e-05 5.16559128e-03 1.57351284e-03]\\n [1.32572514e-04 8.73199914e-01 ...</td>\n",
       "      <td>[[9.98396572e-01 0.00000000e+00 4.23738868e-05 1.56105399e-03]\\n [1.81139308e-04 7.55356757e-01 ...</td>\n",
       "      <td>[[9.99373838e-01 0.00000000e+00 0.00000000e+00 6.26162129e-04]\\n [0.00000000e+00 8.29168945e-01 ...</td>\n",
       "      <td>[[9.95216561e-01 0.00000000e+00 7.92243720e-05 4.70421474e-03]\\n [0.00000000e+00 8.78090469e-01 ...</td>\n",
       "      <td>[[9.98872213e-01 0.00000000e+00 3.10875460e-05 1.09669954e-03]\\n [0.00000000e+00 9.19864946e-01 ...</td>\n",
       "      <td>[[9.99123777e-01 0.00000000e+00 1.71472174e-06 8.74508089e-04]\\n [0.00000000e+00 9.06745404e-01 ...</td>\n",
       "      <td>[[9.98928376e-01 0.00000000e+00 6.60476897e-06 1.06501900e-03]\\n [1.46964809e-05 9.24445392e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>10210054</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210054/inference_50000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.829719</td>\n",
       "      <td>0.155244</td>\n",
       "      <td>0.121035</td>\n",
       "      <td>0.832751</td>\n",
       "      <td>1.662470</td>\n",
       "      <td>0.276280</td>\n",
       "      <td>0.995374</td>\n",
       "      <td>0.823699</td>\n",
       "      <td>0.834863</td>\n",
       "      <td>0.831108</td>\n",
       "      <td>[[9.98199697e-01 0.00000000e+00 1.30486286e-04 1.66981669e-03]\\n [5.42560554e-04 8.29719031e-01 ...</td>\n",
       "      <td>[[9.91856838e-01 1.67593084e-05 4.95516885e-03 3.17123358e-03]\\n [1.93334917e-04 8.09581678e-01 ...</td>\n",
       "      <td>[[9.99384731e-01 0.00000000e+00 9.49175065e-05 5.20351330e-04]\\n [1.46080087e-04 8.37880320e-01 ...</td>\n",
       "      <td>[[9.99954902e-01 0.00000000e+00 5.20356340e-06 3.98939861e-05]\\n [3.51142970e-05 8.91747639e-01 ...</td>\n",
       "      <td>[[9.98220663e-01 2.99767894e-05 9.63539659e-05 1.65300582e-03]\\n [0.00000000e+00 9.01090940e-01 ...</td>\n",
       "      <td>[[9.99794477e-01 0.00000000e+00 2.76333742e-05 1.77889846e-04]\\n [6.84395033e-05 9.76487228e-01 ...</td>\n",
       "      <td>[[9.99883399e-01 0.00000000e+00 3.42944349e-06 1.13171635e-04]\\n [1.40602482e-05 9.68153538e-01 ...</td>\n",
       "      <td>[[9.99654901e-01 0.00000000e+00 1.63468032e-04 1.81631147e-04]\\n [3.67412023e-05 9.66036433e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>10210056</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210056/inference_20000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.728963</td>\n",
       "      <td>0.250275</td>\n",
       "      <td>0.077321</td>\n",
       "      <td>0.866446</td>\n",
       "      <td>1.595409</td>\n",
       "      <td>0.327596</td>\n",
       "      <td>0.993784</td>\n",
       "      <td>0.791083</td>\n",
       "      <td>0.827442</td>\n",
       "      <td>0.818046</td>\n",
       "      <td>[[9.97624742e-01 6.72819911e-05 4.09808491e-04 1.89816769e-03]\\n [1.55017301e-04 7.28963322e-01 ...</td>\n",
       "      <td>[[9.89199557e-01 1.48971630e-05 7.28098843e-03 3.50455760e-03]\\n [6.62862572e-05 4.53121807e-01 ...</td>\n",
       "      <td>[[9.99227100e-01 0.00000000e+00 4.74587533e-05 7.25440943e-04]\\n [3.91494633e-04 8.28396800e-01 ...</td>\n",
       "      <td>[[9.99941026e-01 0.00000000e+00 2.08142536e-05 3.81594649e-05]\\n [6.01959378e-05 7.94857260e-01 ...</td>\n",
       "      <td>[[9.96349255e-01 1.07059962e-05 1.43460349e-04 3.49657836e-03]\\n [7.03378326e-06 8.16601135e-01 ...</td>\n",
       "      <td>[[9.99701214e-01 0.00000000e+00 7.42646932e-05 2.24521165e-04]\\n [2.28131678e-05 9.44685673e-01 ...</td>\n",
       "      <td>[[9.99773657e-01 0.00000000e+00 1.02883305e-05 2.16054940e-04]\\n [0.00000000e+00 9.30120567e-01 ...</td>\n",
       "      <td>[[9.99529410e-01 0.00000000e+00 1.93189492e-04 2.77400297e-04]\\n [2.20447214e-05 9.34512481e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>10210056</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210056/inference_10000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.736947</td>\n",
       "      <td>0.245769</td>\n",
       "      <td>0.106176</td>\n",
       "      <td>0.848984</td>\n",
       "      <td>1.585930</td>\n",
       "      <td>0.351945</td>\n",
       "      <td>0.991282</td>\n",
       "      <td>0.778757</td>\n",
       "      <td>0.799489</td>\n",
       "      <td>0.742067</td>\n",
       "      <td>[[9.98829701e-01 2.03884821e-06 5.09712053e-04 6.58547973e-04]\\n [3.32179931e-05 7.36946713e-01 ...</td>\n",
       "      <td>[[9.86106533e-01 1.86214538e-06 7.54727522e-03 6.34432930e-03]\\n [5.52385477e-06 4.24602144e-01 ...</td>\n",
       "      <td>[[9.99364392e-01 0.00000000e+00 3.15261718e-04 3.20346584e-04]\\n [1.05761983e-03 7.91917681e-01 ...</td>\n",
       "      <td>[[9.99939292e-01 0.00000000e+00 4.16285072e-05 1.90797325e-05]\\n [1.15375547e-04 7.85968327e-01 ...</td>\n",
       "      <td>[[9.98355559e-01 1.49883947e-05 4.34663446e-04 1.19478918e-03]\\n [0.00000000e+00 8.21236398e-01 ...</td>\n",
       "      <td>[[9.99792750e-01 0.00000000e+00 1.12260583e-04 9.49897238e-05]\\n [3.19384349e-04 9.09637043e-01 ...</td>\n",
       "      <td>[[9.99754795e-01 0.00000000e+00 8.05919219e-05 1.64613287e-04]\\n [5.62409927e-05 9.11329045e-01 ...</td>\n",
       "      <td>[[9.99644994e-01 0.00000000e+00 2.47678837e-04 1.07327496e-04]\\n [2.20447214e-05 9.26017915e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>10210130</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_55000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.788889</td>\n",
       "      <td>0.194381</td>\n",
       "      <td>0.081139</td>\n",
       "      <td>0.883951</td>\n",
       "      <td>1.672840</td>\n",
       "      <td>0.275520</td>\n",
       "      <td>0.994809</td>\n",
       "      <td>0.827094</td>\n",
       "      <td>0.851354</td>\n",
       "      <td>0.843518</td>\n",
       "      <td>[[9.98894944e-01 0.00000000e+00 2.07962518e-04 8.97093214e-04]\\n [2.60207612e-04 7.88888581e-01 ...</td>\n",
       "      <td>[[9.89007756e-01 0.00000000e+00 7.77073266e-03 3.22151150e-03]\\n [8.28578215e-05 8.35692940e-01 ...</td>\n",
       "      <td>[[9.99644059e-01 2.03394657e-05 1.03392284e-04 2.32208900e-04]\\n [4.84985889e-04 8.08296180e-01 ...</td>\n",
       "      <td>[[9.99949699e-01 1.38761691e-05 3.29559015e-05 3.46904227e-06]\\n [7.52449222e-05 7.43906415e-01 ...</td>\n",
       "      <td>[[9.98325582e-01 0.00000000e+00 1.64872342e-04 1.50954547e-03]\\n [0.00000000e+00 9.00014771e-01 ...</td>\n",
       "      <td>[[9.99811748e-01 6.90834355e-06 5.00854908e-05 1.31258527e-04]\\n [1.52087785e-05 9.49400394e-01 ...</td>\n",
       "      <td>[[9.99900546e-01 0.00000000e+00 1.71472174e-05 8.23066437e-05]\\n [0.00000000e+00 9.35934479e-01 ...</td>\n",
       "      <td>[[9.99692878e-01 0.00000000e+00 1.33746572e-04 1.73375186e-04]\\n [7.34824046e-06 9.42757207e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>10210130</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_45000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.840122</td>\n",
       "      <td>0.147062</td>\n",
       "      <td>0.156935</td>\n",
       "      <td>0.799764</td>\n",
       "      <td>1.639886</td>\n",
       "      <td>0.303997</td>\n",
       "      <td>0.991060</td>\n",
       "      <td>0.809171</td>\n",
       "      <td>0.816202</td>\n",
       "      <td>0.808539</td>\n",
       "      <td>[[9.99435239e-01 0.00000000e+00 1.01942411e-04 4.62818545e-04]\\n [4.42906574e-05 8.40121799e-01 ...</td>\n",
       "      <td>[[9.91810285e-01 4.28293437e-05 6.74469056e-03 1.40219547e-03]\\n [2.76192738e-05 7.90071423e-01 ...</td>\n",
       "      <td>[[9.99883048e-01 1.35596438e-05 5.25436197e-05 5.08486642e-05]\\n [9.40755760e-04 8.87734532e-01 ...</td>\n",
       "      <td>[[9.99987858e-01 0.00000000e+00 1.21416479e-05 0.00000000e+00]\\n [1.60522501e-04 8.11150294e-01 ...</td>\n",
       "      <td>[[9.99025754e-01 2.14119924e-06 1.24189556e-04 8.47914900e-04]\\n [0.00000000e+00 9.01062805e-01 ...</td>\n",
       "      <td>[[9.99905010e-01 0.00000000e+00 3.28146319e-05 6.21750920e-05]\\n [1.52087785e-05 9.80388280e-01 ...</td>\n",
       "      <td>[[9.99933126e-01 0.00000000e+00 6.85888697e-06 6.00152610e-05]\\n [0.00000000e+00 9.70923407e-01 ...</td>\n",
       "      <td>[[9.99752321e-01 0.00000000e+00 1.07327496e-04 1.40351341e-04]\\n [2.20447214e-05 9.78161029e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>10210132</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_45000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.798356</td>\n",
       "      <td>0.177805</td>\n",
       "      <td>0.075448</td>\n",
       "      <td>0.843776</td>\n",
       "      <td>1.642132</td>\n",
       "      <td>0.253253</td>\n",
       "      <td>0.994820</td>\n",
       "      <td>0.835899</td>\n",
       "      <td>0.845961</td>\n",
       "      <td>0.837082</td>\n",
       "      <td>[[9.97728723e-01 0.00000000e+00 3.46604196e-05 2.23661649e-03]\\n [3.87543253e-05 7.98355709e-01 ...</td>\n",
       "      <td>[[9.91923875e-01 0.00000000e+00 6.01845386e-03 2.05767064e-03]\\n [1.32572514e-04 7.58867168e-01 ...</td>\n",
       "      <td>[[9.99162692e-01 0.00000000e+00 8.47477737e-06 8.28833227e-04]\\n [7.12870824e-04 7.77064258e-01 ...</td>\n",
       "      <td>[[9.99956637e-01 0.00000000e+00 5.20356340e-06 3.81594649e-05]\\n [1.50489844e-04 8.25015425e-01 ...</td>\n",
       "      <td>[[9.97689646e-01 2.14119924e-06 9.84951652e-05 2.20971762e-03]\\n [0.00000000e+00 8.66639469e-01 ...</td>\n",
       "      <td>[[9.99711577e-01 0.00000000e+00 2.41792024e-05 2.64244141e-04]\\n [4.56263355e-05 9.64601568e-01 ...</td>\n",
       "      <td>[[9.99816525e-01 0.00000000e+00 3.42944349e-06 1.80045783e-04]\\n [2.10903722e-05 9.45509508e-01 ...</td>\n",
       "      <td>[[9.99615272e-01 0.00000000e+00 6.60476897e-06 3.78123024e-04]\\n [5.14376832e-05 9.55116947e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>10210136</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210136/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.839430</td>\n",
       "      <td>0.149675</td>\n",
       "      <td>0.145136</td>\n",
       "      <td>0.800236</td>\n",
       "      <td>1.639666</td>\n",
       "      <td>0.294811</td>\n",
       "      <td>0.995972</td>\n",
       "      <td>0.815201</td>\n",
       "      <td>0.816830</td>\n",
       "      <td>0.823648</td>\n",
       "      <td>[[9.97426974e-01 2.03884821e-06 6.32042946e-05 2.50778330e-03]\\n [1.66089965e-05 8.39429758e-01 ...</td>\n",
       "      <td>[[9.90603614e-01 0.00000000e+00 7.54541307e-03 1.85097251e-03]\\n [1.10477095e-05 6.74341142e-01 ...</td>\n",
       "      <td>[[9.99435580e-01 1.52545993e-05 7.28830854e-05 4.76282488e-04]\\n [2.92160174e-05 8.80483116e-01 ...</td>\n",
       "      <td>[[9.99963575e-01 0.00000000e+00 1.73452113e-05 1.90797325e-05]\\n [5.51796096e-05 8.57109893e-01 ...</td>\n",
       "      <td>[[9.97488373e-01 2.14119924e-05 8.99303682e-05 2.40028435e-03]\\n [0.00000000e+00 8.67399118e-01 ...</td>\n",
       "      <td>[[9.99713304e-01 0.00000000e+00 4.66313190e-05 2.40064938e-04]\\n [1.52087785e-05 9.75118438e-01 ...</td>\n",
       "      <td>[[9.99850819e-01 0.00000000e+00 5.14416523e-06 1.44036626e-04]\\n [0.00000000e+00 9.68153538e-01 ...</td>\n",
       "      <td>[[9.99623528e-01 0.00000000e+00 3.63262294e-05 3.40145602e-04]\\n [0.00000000e+00 9.71709274e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>10210136</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210136/inference_17800</td>\n",
       "      <td>1</td>\n",
       "      <td>0.688659</td>\n",
       "      <td>0.278057</td>\n",
       "      <td>0.038709</td>\n",
       "      <td>0.861727</td>\n",
       "      <td>1.550385</td>\n",
       "      <td>0.316765</td>\n",
       "      <td>0.995060</td>\n",
       "      <td>0.789468</td>\n",
       "      <td>0.826526</td>\n",
       "      <td>0.808840</td>\n",
       "      <td>[[9.97388235e-01 0.00000000e+00 8.15539286e-05 2.53021063e-03]\\n [9.96539792e-05 6.88658824e-01 ...</td>\n",
       "      <td>[[9.89059896e-01 0.00000000e+00 7.11153320e-03 3.82857090e-03]\\n [7.18101120e-05 4.16697508e-01 ...</td>\n",
       "      <td>[[9.99415240e-01 1.69495547e-06 4.74587533e-05 5.35605930e-04]\\n [5.49261127e-04 7.42992538e-01 ...</td>\n",
       "      <td>[[9.99967044e-01 0.00000000e+00 1.04071268e-05 2.25487747e-05]\\n [1.35440860e-04 7.00956614e-01 ...</td>\n",
       "      <td>[[9.97235712e-01 6.42359773e-06 6.42359773e-05 2.69362865e-03]\\n [1.40675665e-05 8.06992987e-01 ...</td>\n",
       "      <td>[[9.99754754e-01 0.00000000e+00 3.10875460e-05 2.14158650e-04]\\n [7.60438925e-05 9.40373984e-01 ...</td>\n",
       "      <td>[[9.99866252e-01 0.00000000e+00 6.85888697e-06 1.26889409e-04]\\n [4.21807445e-05 9.23758304e-01 ...</td>\n",
       "      <td>[[9.99631784e-01 0.00000000e+00 3.13726526e-05 3.36843218e-04]\\n [1.46964809e-05 9.30529735e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>10210136</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210136/inference_8800</td>\n",
       "      <td>1</td>\n",
       "      <td>0.650680</td>\n",
       "      <td>0.270954</td>\n",
       "      <td>0.061796</td>\n",
       "      <td>0.815973</td>\n",
       "      <td>1.466653</td>\n",
       "      <td>0.332750</td>\n",
       "      <td>0.993133</td>\n",
       "      <td>0.748033</td>\n",
       "      <td>0.802822</td>\n",
       "      <td>0.755055</td>\n",
       "      <td>[[9.96984543e-01 0.00000000e+00 8.76704732e-05 2.92778604e-03]\\n [3.87543253e-05 6.50679585e-01 ...</td>\n",
       "      <td>[[9.89640885e-01 0.00000000e+00 4.98868747e-03 5.37042727e-03]\\n [1.04953241e-04 3.26365911e-01 ...</td>\n",
       "      <td>[[9.99064385e-01 3.38991095e-06 5.08486642e-05 8.81376846e-04]\\n [5.55104330e-04 6.70606934e-01 ...</td>\n",
       "      <td>[[9.99868176e-01 1.73452113e-06 1.04071268e-05 1.19681958e-04]\\n [1.50489844e-04 6.37083707e-01 ...</td>\n",
       "      <td>[[9.95933863e-01 1.07059962e-05 7.28007743e-05 3.98263059e-03]\\n [1.40675665e-05 8.33341539e-01 ...</td>\n",
       "      <td>[[9.99664945e-01 0.00000000e+00 2.41792024e-05 3.10875460e-04]\\n [1.52087785e-05 9.29111883e-01 ...</td>\n",
       "      <td>[[9.99746221e-01 0.00000000e+00 0.00000000e+00 2.53778818e-04]\\n [0.00000000e+00 9.06921157e-01 ...</td>\n",
       "      <td>[[9.99549225e-01 0.00000000e+00 6.60476897e-06 4.44170713e-04]\\n [1.46964809e-05 9.23012485e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>10202100</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202100/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.733049</td>\n",
       "      <td>0.239546</td>\n",
       "      <td>0.055204</td>\n",
       "      <td>0.862342</td>\n",
       "      <td>1.595391</td>\n",
       "      <td>0.294750</td>\n",
       "      <td>0.993889</td>\n",
       "      <td>0.808244</td>\n",
       "      <td>0.838872</td>\n",
       "      <td>0.826014</td>\n",
       "      <td>[[9.98678826e-01 2.03884821e-06 9.17481696e-05 1.22738662e-03]\\n [1.21799308e-04 7.33049135e-01 ...</td>\n",
       "      <td>[[9.90229323e-01 0.00000000e+00 5.30152789e-03 4.46914891e-03]\\n [2.76192738e-05 7.46322494e-01 ...</td>\n",
       "      <td>[[9.99940677e-01 0.00000000e+00 1.69495547e-06 5.76284861e-05]\\n [1.57766494e-04 7.91742385e-01 ...</td>\n",
       "      <td>[[9.99963575e-01 0.00000000e+00 1.04071268e-05 2.60178170e-05]\\n [0.00000000e+00 7.09659943e-01 ...</td>\n",
       "      <td>[[9.98507584e-01 0.00000000e+00 1.11342361e-04 1.38107351e-03]\\n [0.00000000e+00 8.63347659e-01 ...</td>\n",
       "      <td>[[9.99758208e-01 0.00000000e+00 2.24521165e-05 2.19339908e-04]\\n [0.00000000e+00 9.46016441e-01 ...</td>\n",
       "      <td>[[9.99893687e-01 0.00000000e+00 5.14416523e-06 1.01168583e-04]\\n [0.00000000e+00 9.29551127e-01 ...</td>\n",
       "      <td>[[9.99752321e-01 0.00000000e+00 1.65119224e-06 2.46027644e-04]\\n [0.00000000e+00 9.50950495e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>10210054</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210054/inference_57300</td>\n",
       "      <td>1</td>\n",
       "      <td>0.794176</td>\n",
       "      <td>0.185833</td>\n",
       "      <td>0.093085</td>\n",
       "      <td>0.852272</td>\n",
       "      <td>1.646448</td>\n",
       "      <td>0.278917</td>\n",
       "      <td>0.995408</td>\n",
       "      <td>0.821614</td>\n",
       "      <td>0.839846</td>\n",
       "      <td>0.833980</td>\n",
       "      <td>[[9.98220086e-01 0.00000000e+00 5.50489018e-05 1.72486559e-03]\\n [7.80622837e-04 7.94175779e-01 ...</td>\n",
       "      <td>[[9.91530963e-01 2.60700353e-05 4.89930449e-03 3.54366265e-03]\\n [1.54667933e-04 8.09139770e-01 ...</td>\n",
       "      <td>[[9.99394901e-01 0.00000000e+00 9.32225510e-05 5.11876553e-04]\\n [1.40236883e-04 7.94459474e-01 ...</td>\n",
       "      <td>[[9.99960106e-01 0.00000000e+00 3.46904227e-06 3.64249438e-05]\\n [1.50489844e-05 8.57922538e-01 ...</td>\n",
       "      <td>[[9.98152145e-01 8.56479697e-06 1.15624759e-04 1.72366539e-03]\\n [0.00000000e+00 8.68011057e-01 ...</td>\n",
       "      <td>[[9.99787568e-01 0.00000000e+00 2.59062883e-05 1.86525276e-04]\\n [4.56263355e-05 9.67628115e-01 ...</td>\n",
       "      <td>[[9.99876540e-01 0.00000000e+00 3.42944349e-06 1.20030522e-04]\\n [3.51506204e-05 9.53123133e-01 ...</td>\n",
       "      <td>[[9.99663157e-01 0.00000000e+00 1.13932265e-04 2.22910953e-04]\\n [2.93929619e-05 9.54044104e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>10210056</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210056/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.858918</td>\n",
       "      <td>0.123538</td>\n",
       "      <td>0.207265</td>\n",
       "      <td>0.712497</td>\n",
       "      <td>1.571415</td>\n",
       "      <td>0.330803</td>\n",
       "      <td>0.993320</td>\n",
       "      <td>0.792416</td>\n",
       "      <td>0.782156</td>\n",
       "      <td>0.833675</td>\n",
       "      <td>[[9.93547045e-01 1.65146705e-04 2.03884821e-05 6.26741941e-03]\\n [1.55017301e-04 8.58917647e-01 ...</td>\n",
       "      <td>[[9.94096999e-01 0.00000000e+00 3.44310680e-03 2.45989404e-03]\\n [8.83816763e-05 8.53059939e-01 ...</td>\n",
       "      <td>[[9.97411803e-01 1.93224924e-04 3.38991095e-06 2.39158217e-03]\\n [6.89498010e-04 9.17120002e-01 ...</td>\n",
       "      <td>[[9.97899495e-01 2.60178170e-04 1.56106902e-05 1.82471623e-03]\\n [5.01632815e-05 9.00074743e-01 ...</td>\n",
       "      <td>[[9.86970803e-01 1.54166345e-04 3.64003871e-05 1.28386307e-02]\\n [1.05506749e-03 9.12717784e-01 ...</td>\n",
       "      <td>[[9.96972418e-01 2.17612822e-04 2.76333742e-05 2.78233537e-03]\\n [5.32307248e-05 9.69825783e-01 ...</td>\n",
       "      <td>[[9.96752317e-01 1.44036626e-04 3.42944349e-06 3.10021691e-03]\\n [3.51506204e-05 9.65411790e-01 ...</td>\n",
       "      <td>[[9.96552311e-01 1.42002533e-04 6.60476897e-06 3.29908210e-03]\\n [8.81788856e-05 9.74986589e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>10210132</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_55000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.788230</td>\n",
       "      <td>0.190931</td>\n",
       "      <td>0.072818</td>\n",
       "      <td>0.861572</td>\n",
       "      <td>1.649802</td>\n",
       "      <td>0.263749</td>\n",
       "      <td>0.995917</td>\n",
       "      <td>0.831604</td>\n",
       "      <td>0.848871</td>\n",
       "      <td>0.843631</td>\n",
       "      <td>[[9.97620664e-01 0.00000000e+00 1.40680527e-04 2.23865534e-03]\\n [2.15916955e-04 7.88229758e-01 ...</td>\n",
       "      <td>[[9.90113870e-01 3.72429076e-06 7.76700837e-03 2.11539715e-03]\\n [1.38096369e-04 7.68705153e-01 ...</td>\n",
       "      <td>[[9.98938958e-01 0.00000000e+00 5.59335306e-05 1.00510860e-03]\\n [3.27219395e-04 7.47550237e-01 ...</td>\n",
       "      <td>[[9.99934088e-01 0.00000000e+00 2.08142536e-05 4.50975495e-05]\\n [1.05342891e-04 7.90292402e-01 ...</td>\n",
       "      <td>[[9.97850236e-01 0.00000000e+00 1.49883947e-04 1.99988009e-03]\\n [0.00000000e+00 8.59113321e-01 ...</td>\n",
       "      <td>[[9.99659764e-01 1.72708589e-06 3.28146319e-05 3.05694202e-04]\\n [0.00000000e+00 9.63810712e-01 ...</td>\n",
       "      <td>[[9.99833672e-01 0.00000000e+00 5.14416523e-06 1.61183844e-04]\\n [7.03012408e-06 9.48581672e-01 ...</td>\n",
       "      <td>[[9.99628482e-01 0.00000000e+00 6.60476897e-06 3.64913486e-04]\\n [7.34824046e-06 9.49524936e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>10210134</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210134/inference_39300</td>\n",
       "      <td>1</td>\n",
       "      <td>0.713091</td>\n",
       "      <td>0.253929</td>\n",
       "      <td>0.037873</td>\n",
       "      <td>0.886563</td>\n",
       "      <td>1.599653</td>\n",
       "      <td>0.291802</td>\n",
       "      <td>0.995389</td>\n",
       "      <td>0.806570</td>\n",
       "      <td>0.845419</td>\n",
       "      <td>0.829573</td>\n",
       "      <td>[[9.98287368e-01 0.00000000e+00 8.35927768e-05 1.62903972e-03]\\n [6.58823529e-04 7.13090657e-01 ...</td>\n",
       "      <td>[[9.91227433e-01 1.86214538e-06 5.62554119e-03 3.14516354e-03]\\n [5.52385477e-05 6.84521607e-01 ...</td>\n",
       "      <td>[[9.98469455e-01 1.69495547e-06 2.15259345e-04 1.31359049e-03]\\n [4.96672296e-04 6.88814356e-01 ...</td>\n",
       "      <td>[[9.99923681e-01 0.00000000e+00 3.12213804e-05 4.50975495e-05]\\n [7.02285941e-05 7.17199484e-01 ...</td>\n",
       "      <td>[[9.98419795e-01 0.00000000e+00 1.02777564e-04 1.47742748e-03]\\n [0.00000000e+00 8.19555324e-01 ...</td>\n",
       "      <td>[[9.99677035e-01 0.00000000e+00 4.14500613e-05 2.81515000e-04]\\n [4.56263355e-05 9.27651841e-01 ...</td>\n",
       "      <td>[[9.99804522e-01 0.00000000e+00 1.54324957e-05 1.80045783e-04]\\n [4.92108686e-05 9.01198636e-01 ...</td>\n",
       "      <td>[[9.99618575e-01 0.00000000e+00 2.14654992e-05 3.59959909e-04]\\n [4.40894428e-05 9.05744120e-01 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>10210132</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_70000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.861237</td>\n",
       "      <td>0.128304</td>\n",
       "      <td>0.149105</td>\n",
       "      <td>0.802642</td>\n",
       "      <td>1.663880</td>\n",
       "      <td>0.277410</td>\n",
       "      <td>0.995638</td>\n",
       "      <td>0.825284</td>\n",
       "      <td>0.824976</td>\n",
       "      <td>0.832259</td>\n",
       "      <td>[[9.98324067e-01 0.00000000e+00 6.52431428e-05 1.61069009e-03]\\n [6.64359862e-05 8.61237370e-01 ...</td>\n",
       "      <td>[[9.90981630e-01 1.30350176e-05 7.21208905e-03 1.79324600e-03]\\n [8.83816763e-05 8.24656278e-01 ...</td>\n",
       "      <td>[[9.99064385e-01 1.69495547e-06 9.32225510e-05 8.40697915e-04]\\n [2.68787360e-04 8.63269039e-01 ...</td>\n",
       "      <td>[[9.99953168e-01 0.00000000e+00 2.94868593e-05 1.73452113e-05]\\n [1.50489844e-04 9.09781338e-01 ...</td>\n",
       "      <td>[[9.98173557e-01 6.42359773e-06 1.52025146e-04 1.66799421e-03]\\n [1.40675665e-05 9.33143890e-01 ...</td>\n",
       "      <td>[[9.99746118e-01 0.00000000e+00 2.59062883e-05 2.27975337e-04]\\n [0.00000000e+00 9.82190520e-01 ...</td>\n",
       "      <td>[[9.99879969e-01 0.00000000e+00 0.00000000e+00 1.20030522e-04]\\n [7.03012408e-06 9.72322401e-01 ...</td>\n",
       "      <td>[[9.99673064e-01 0.00000000e+00 3.79774216e-05 2.88958643e-04]\\n [2.20447214e-05 9.77661349e-01 ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>73 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       model  \\\n",
       "0   10112050   \n",
       "1   10112050   \n",
       "2   10112050   \n",
       "3   10112056   \n",
       "4   10112056   \n",
       "5   10112056   \n",
       "6   10112103   \n",
       "7   10112103   \n",
       "8   10112127   \n",
       "9   10112127   \n",
       "10  10112130   \n",
       "11  10112055   \n",
       "12  10112055   \n",
       "13  10112055   \n",
       "14  10112055   \n",
       "15  10192000   \n",
       "16  10192000   \n",
       "17  10192002   \n",
       "18  10192002   \n",
       "19  10192002   \n",
       "20  10192004   \n",
       "21  10192004   \n",
       "22  10192004   \n",
       "23  10192004   \n",
       "24  10192004   \n",
       "25  10202100   \n",
       "26  10202100   \n",
       "27  10202100   \n",
       "28  10202102   \n",
       "29  10202102   \n",
       "..       ...   \n",
       "43  10210130   \n",
       "44  10210130   \n",
       "45  10210130   \n",
       "46  10210132   \n",
       "47  10210132   \n",
       "48  10210132   \n",
       "49  10210134   \n",
       "50  10210134   \n",
       "51  10210134   \n",
       "52  10202106   \n",
       "53  10202106   \n",
       "54  10210050   \n",
       "55  10210050   \n",
       "56  10210052   \n",
       "57  10210052   \n",
       "58  10210054   \n",
       "59  10210056   \n",
       "60  10210056   \n",
       "61  10210130   \n",
       "62  10210130   \n",
       "63  10210132   \n",
       "64  10210136   \n",
       "65  10210136   \n",
       "66  10210136   \n",
       "67  10202100   \n",
       "68  10210054   \n",
       "69  10210056   \n",
       "70  10210132   \n",
       "71  10210134   \n",
       "72  10210132   \n",
       "\n",
       "                                                              inf_dir  \\\n",
       "0   /proj/NIRAL/users/siyangj/model_10112050/inference_17000_10112050   \n",
       "1                  /proj/NIRAL/users/siyangj/model_10112050/inference   \n",
       "2   /proj/NIRAL/users/siyangj/model_10112050/inference_13200_10112050   \n",
       "3   /proj/NIRAL/users/siyangj/model_10112056/inference_18100_10112056   \n",
       "4   /proj/NIRAL/users/siyangj/model_10112056/inference_16500_10112056   \n",
       "5                  /proj/NIRAL/users/siyangj/model_10112056/inference   \n",
       "6                  /proj/NIRAL/users/siyangj/model_10112103/inference   \n",
       "7   /proj/NIRAL/users/siyangj/model_10112103/inference_12300_10112103   \n",
       "8   /proj/NIRAL/users/siyangj/model_10112127/inference_13500_10112127   \n",
       "9    /proj/NIRAL/users/siyangj/model_10112127/inference_8700_10112127   \n",
       "10            /proj/NIRAL/users/siyangj/model_10112130/inference_7300   \n",
       "11  /proj/NIRAL/users/siyangj/model_10112055/inference_18900_10112055   \n",
       "12  /proj/NIRAL/users/siyangj/model_10112055/inference_21900_10112055   \n",
       "13                 /proj/NIRAL/users/siyangj/model_10112055/inference   \n",
       "14    /proj/NIRAL/users/siyangj/model_10112055/inference_21900_normal   \n",
       "15            /proj/NIRAL/users/siyangj/model_10192000/inference_7200   \n",
       "16            /proj/NIRAL/users/siyangj/model_10192000/inference_7700   \n",
       "17            /proj/NIRAL/users/siyangj/model_10192002/inference_5700   \n",
       "18            /proj/NIRAL/users/siyangj/model_10192002/inference_4600   \n",
       "19            /proj/NIRAL/users/siyangj/model_10192002/inference_2500   \n",
       "20           /proj/NIRAL/users/siyangj/model_10192004/inference_11100   \n",
       "21            /proj/NIRAL/users/siyangj/model_10192004/inference_4000   \n",
       "22            /proj/NIRAL/users/siyangj/model_10192004/inference_3900   \n",
       "23            /proj/NIRAL/users/siyangj/model_10192004/inference_6600   \n",
       "24            /proj/NIRAL/users/siyangj/model_10192004/inference_2900   \n",
       "25           /proj/NIRAL/users/siyangj/model_10202100/inference_12000   \n",
       "26            /proj/NIRAL/users/siyangj/model_10202100/inference_5800   \n",
       "27            /proj/NIRAL/users/siyangj/model_10202100/inference_2000   \n",
       "28           /proj/NIRAL/users/siyangj/model_10202102/inference_20000   \n",
       "29           /proj/NIRAL/users/siyangj/model_10202102/inference_11600   \n",
       "..                                                                ...   \n",
       "43           /proj/NIRAL/users/siyangj/model_10210130/inference_20000   \n",
       "44           /proj/NIRAL/users/siyangj/model_10210130/inference_12000   \n",
       "45            /proj/NIRAL/users/siyangj/model_10210130/inference_6000   \n",
       "46            /proj/NIRAL/users/siyangj/model_10210132/inference_6200   \n",
       "47           /proj/NIRAL/users/siyangj/model_10210132/inference_25000   \n",
       "48           /proj/NIRAL/users/siyangj/model_10210132/inference_12000   \n",
       "49           /proj/NIRAL/users/siyangj/model_10210134/inference_11700   \n",
       "50           /proj/NIRAL/users/siyangj/model_10210134/inference_22000   \n",
       "51            /proj/NIRAL/users/siyangj/model_10210134/inference_4400   \n",
       "52            /proj/NIRAL/users/siyangj/model_10202106/inference_2100   \n",
       "53            /proj/NIRAL/users/siyangj/model_10202106/inference_9000   \n",
       "54           /proj/NIRAL/users/siyangj/model_10210050/inference_47700   \n",
       "55           /proj/NIRAL/users/siyangj/model_10210050/inference_60000   \n",
       "56           /proj/NIRAL/users/siyangj/model_10210052/inference_55400   \n",
       "57           /proj/NIRAL/users/siyangj/model_10210052/inference_70000   \n",
       "58           /proj/NIRAL/users/siyangj/model_10210054/inference_50000   \n",
       "59           /proj/NIRAL/users/siyangj/model_10210056/inference_20000   \n",
       "60           /proj/NIRAL/users/siyangj/model_10210056/inference_10000   \n",
       "61           /proj/NIRAL/users/siyangj/model_10210130/inference_55000   \n",
       "62           /proj/NIRAL/users/siyangj/model_10210130/inference_45000   \n",
       "63           /proj/NIRAL/users/siyangj/model_10210132/inference_45000   \n",
       "64           /proj/NIRAL/users/siyangj/model_10210136/inference_30000   \n",
       "65           /proj/NIRAL/users/siyangj/model_10210136/inference_17800   \n",
       "66            /proj/NIRAL/users/siyangj/model_10210136/inference_8800   \n",
       "67           /proj/NIRAL/users/siyangj/model_10202100/inference_30000   \n",
       "68           /proj/NIRAL/users/siyangj/model_10210054/inference_57300   \n",
       "69           /proj/NIRAL/users/siyangj/model_10210056/inference_30000   \n",
       "70           /proj/NIRAL/users/siyangj/model_10210132/inference_55000   \n",
       "71           /proj/NIRAL/users/siyangj/model_10210134/inference_39300   \n",
       "72           /proj/NIRAL/users/siyangj/model_10210132/inference_70000   \n",
       "\n",
       "    test_im   test_11   test_12   test_21   test_22  test_11+22  test_12+21  \\\n",
       "0         1  0.918915  0.067399  0.295044  0.649303    1.568218    0.362443   \n",
       "1         1  0.895175  0.097146  0.224378  0.753510    1.648685    0.321524   \n",
       "2         1  0.926572  0.058502  0.308779  0.649881    1.576453    0.367281   \n",
       "3         1  0.825539  0.124662  0.174211  0.737446    1.562985    0.298872   \n",
       "4         1  0.793855  0.189398  0.090426  0.849576    1.643431    0.279823   \n",
       "5         1  0.845575  0.147383  0.184313  0.789043    1.634618    0.331696   \n",
       "6         1  0.721373  0.249500  0.052741  0.869897    1.591270    0.302241   \n",
       "7         1  0.809434  0.176941  0.130890  0.809682    1.619116    0.307831   \n",
       "8         1  0.737849  0.235875  0.069771  0.803042    1.540891    0.305646   \n",
       "9         1  0.713201  0.271862  0.050931  0.864683    1.577885    0.322792   \n",
       "10        1  0.712039  0.238079  0.062735  0.804453    1.516491    0.300814   \n",
       "11        1  0.892019  0.094439  0.250173  0.696974    1.588993    0.344611   \n",
       "12        1  0.861016  0.125160  0.120896  0.821308    1.682324    0.246056   \n",
       "13        1  0.824332  0.159098  0.141585  0.805546    1.629878    0.300683   \n",
       "14        1  0.861763  0.124861  0.121004  0.822459    1.684223    0.245865   \n",
       "15        1  0.882287  0.095098  0.265452  0.683058    1.565344    0.360550   \n",
       "16        1  0.900617  0.084324  0.250964  0.693302    1.593919    0.335288   \n",
       "17        1  0.797880  0.149403  0.140641  0.786742    1.584622    0.290044   \n",
       "18        1  0.751673  0.210691  0.083541  0.868307    1.619980    0.294232   \n",
       "19        1  0.776913  0.212756  0.153442  0.801818    1.578732    0.366198   \n",
       "20        1  0.805924  0.147715  0.123747  0.801041    1.606965    0.271462   \n",
       "21        1  0.783851  0.174810  0.108014  0.803988    1.587839    0.282823   \n",
       "22        1  0.856205  0.096011  0.211923  0.685681    1.541885    0.307934   \n",
       "23        1  0.769910  0.198151  0.084447  0.859092    1.629002    0.282597   \n",
       "24        1  0.717924  0.221304  0.074288  0.791369    1.509293    0.295592   \n",
       "25        1  0.727369  0.244711  0.052268  0.846252    1.573621    0.296979   \n",
       "26        1  0.800781  0.179947  0.113441  0.809683    1.610464    0.293389   \n",
       "27        1  0.801738  0.181940  0.145419  0.815258    1.616996    0.327359   \n",
       "28        1  0.751585  0.226447  0.080895  0.846833    1.598417    0.307342   \n",
       "29        1  0.810325  0.179549  0.136749  0.828797    1.639123    0.316298   \n",
       "..      ...       ...       ...       ...       ...         ...         ...   \n",
       "43        1  0.687646  0.279961  0.035215  0.873425    1.561071    0.315177   \n",
       "44        1  0.862400  0.107172  0.171404  0.738618    1.601018    0.278576   \n",
       "45        1  0.812136  0.158749  0.172905  0.789296    1.601431    0.331653   \n",
       "46        1  0.658901  0.298491  0.048867  0.873030    1.531931    0.347359   \n",
       "47        1  0.752758  0.220396  0.075735  0.844798    1.597556    0.296130   \n",
       "48        1  0.732130  0.202447  0.080307  0.869290    1.601420    0.282754   \n",
       "49        1  0.667045  0.269603  0.032461  0.912339    1.579384    0.302064   \n",
       "50        1  0.732927  0.227139  0.079762  0.874268    1.607196    0.306901   \n",
       "51        1  0.522508  0.421381  0.012391  0.933210    1.455717    0.433772   \n",
       "52        1  0.587643  0.369888  0.033761  0.866562    1.454205    0.403649   \n",
       "53        1  0.563482  0.411848  0.021130  0.900106    1.463589    0.432978   \n",
       "54        1  0.787571  0.184371  0.088806  0.862547    1.650118    0.273177   \n",
       "55        1  0.854372  0.132274  0.148030  0.797729    1.652101    0.280304   \n",
       "56        1  0.862876  0.127103  0.177458  0.788410    1.651286    0.304561   \n",
       "57        1  0.795792  0.186408  0.098338  0.843807    1.639600    0.284747   \n",
       "58        1  0.829719  0.155244  0.121035  0.832751    1.662470    0.276280   \n",
       "59        1  0.728963  0.250275  0.077321  0.866446    1.595409    0.327596   \n",
       "60        1  0.736947  0.245769  0.106176  0.848984    1.585930    0.351945   \n",
       "61        1  0.788889  0.194381  0.081139  0.883951    1.672840    0.275520   \n",
       "62        1  0.840122  0.147062  0.156935  0.799764    1.639886    0.303997   \n",
       "63        1  0.798356  0.177805  0.075448  0.843776    1.642132    0.253253   \n",
       "64        1  0.839430  0.149675  0.145136  0.800236    1.639666    0.294811   \n",
       "65        1  0.688659  0.278057  0.038709  0.861727    1.550385    0.316765   \n",
       "66        1  0.650680  0.270954  0.061796  0.815973    1.466653    0.332750   \n",
       "67        1  0.733049  0.239546  0.055204  0.862342    1.595391    0.294750   \n",
       "68        1  0.794176  0.185833  0.093085  0.852272    1.646448    0.278917   \n",
       "69        1  0.858918  0.123538  0.207265  0.712497    1.571415    0.330803   \n",
       "70        1  0.788230  0.190931  0.072818  0.861572    1.649802    0.263749   \n",
       "71        1  0.713091  0.253929  0.037873  0.886563    1.599653    0.291802   \n",
       "72        1  0.861237  0.128304  0.149105  0.802642    1.663880    0.277410   \n",
       "\n",
       "       Dice0     Dice1     Dice2     Dice3  \\\n",
       "0   0.993768  0.780229  0.744064  0.805368   \n",
       "1   0.991811  0.803432  0.796687  0.804782   \n",
       "2   0.994202  0.777533  0.752149  0.844704   \n",
       "3   0.992718  0.794706  0.796743  0.786574   \n",
       "4   0.990101  0.825034  0.838850  0.803951   \n",
       "5   0.997961  0.792166  0.803326  0.819554   \n",
       "6   0.993959  0.802792  0.837660  0.810618   \n",
       "7   0.992859  0.807523  0.819547  0.813614   \n",
       "8   0.996067  0.802167  0.808498  0.805656   \n",
       "9   0.995328  0.797492  0.828315  0.831142   \n",
       "10  0.993799  0.789505  0.808477  0.781285   \n",
       "11  0.993306  0.788053  0.773435  0.822208   \n",
       "12  0.995358  0.843021  0.846834  0.845988   \n",
       "13  0.994605  0.808893  0.818492  0.810333   \n",
       "14  0.991524  0.843291  0.846411  0.833207   \n",
       "15  0.995164  0.777106  0.762458  0.836195   \n",
       "16  0.995337  0.792735  0.776372  0.846890   \n",
       "17  0.996218  0.797222  0.815309  0.810516   \n",
       "18  0.993344  0.802652  0.835804  0.798312   \n",
       "19  0.993348  0.776299  0.787699  0.784175   \n",
       "20  0.995025  0.811626  0.828520  0.822260   \n",
       "21  0.995970  0.807650  0.824122  0.823838   \n",
       "22  0.994825  0.791488  0.773357  0.810508   \n",
       "23  0.995134  0.813365  0.843567  0.837692   \n",
       "24  0.990993  0.786643  0.805821  0.768787   \n",
       "25  0.996687  0.806246  0.830782  0.827865   \n",
       "26  0.991585  0.811223  0.825079  0.821793   \n",
       "27  0.993349  0.794069  0.807127  0.788786   \n",
       "28  0.994971  0.803225  0.828606  0.824020   \n",
       "29  0.995893  0.802237  0.815528  0.807952   \n",
       "..       ...       ...       ...       ...   \n",
       "43  0.996453  0.791053  0.833452  0.826392   \n",
       "44  0.990123  0.816217  0.805659  0.824301   \n",
       "45  0.991836  0.784101  0.790127  0.740799   \n",
       "46  0.992602  0.761938  0.822676  0.801886   \n",
       "47  0.995467  0.808363  0.829791  0.823269   \n",
       "48  0.993884  0.791720  0.840746  0.794139   \n",
       "49  0.995689  0.777983  0.844907  0.795348   \n",
       "50  0.995883  0.792475  0.832953  0.808690   \n",
       "51  0.995112  0.678361  0.816866  0.802167   \n",
       "52  0.987312  0.717716  0.798132  0.743359   \n",
       "53  0.993331  0.707024  0.806870  0.806477   \n",
       "54  0.992172  0.821132  0.846978  0.825889   \n",
       "55  0.995481  0.820110  0.825959  0.833727   \n",
       "56  0.995355  0.807689  0.810778  0.819124   \n",
       "57  0.996155  0.818509  0.840425  0.852010   \n",
       "58  0.995374  0.823699  0.834863  0.831108   \n",
       "59  0.993784  0.791083  0.827442  0.818046   \n",
       "60  0.991282  0.778757  0.799489  0.742067   \n",
       "61  0.994809  0.827094  0.851354  0.843518   \n",
       "62  0.991060  0.809171  0.816202  0.808539   \n",
       "63  0.994820  0.835899  0.845961  0.837082   \n",
       "64  0.995972  0.815201  0.816830  0.823648   \n",
       "65  0.995060  0.789468  0.826526  0.808840   \n",
       "66  0.993133  0.748033  0.802822  0.755055   \n",
       "67  0.993889  0.808244  0.838872  0.826014   \n",
       "68  0.995408  0.821614  0.839846  0.833980   \n",
       "69  0.993320  0.792416  0.782156  0.833675   \n",
       "70  0.995917  0.831604  0.848871  0.843631   \n",
       "71  0.995389  0.806570  0.845419  0.829573   \n",
       "72  0.995638  0.825284  0.824976  0.832259   \n",
       "\n",
       "                                                                                                      1  \\\n",
       "0   [array([[9.99995958e-01, 0.00000000e+00, 0.00000000e+00, 4.04240483e-06],\\n       [6.64359862e-0...   \n",
       "1   [array([[9.99987873e-01, 0.00000000e+00, 0.00000000e+00, 1.21272145e-05],\\n       [6.64359862e-0...   \n",
       "2   [array([[9.99981809e-01, 0.00000000e+00, 0.00000000e+00, 1.81908217e-05],\\n       [1.10726644e-0...   \n",
       "3   [array([[9.99997979e-01, 0.00000000e+00, 0.00000000e+00, 2.02120241e-06],\\n       [1.32871972e-0...   \n",
       "4   [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [1.16262976e-0...   \n",
       "5   [array([[9.99252155e-01, 0.00000000e+00, 6.06360724e-06, 7.41781286e-04],\\n       [1.99307958e-0...   \n",
       "6   [array([[9.99840325e-01, 0.00000000e+00, 0.00000000e+00, 1.59674991e-04],\\n       [3.87543253e-0...   \n",
       "7   [array([[9.99975746e-01, 0.00000000e+00, 0.00000000e+00, 2.42544290e-05],\\n       [8.85813149e-0...   \n",
       "8   [array([[9.99961146e-01, 0.00000000e+00, 0.00000000e+00, 3.88540112e-05],\\n       [4.42906574e-0...   \n",
       "9   [array([[9.99957056e-01, 0.00000000e+00, 0.00000000e+00, 4.29439071e-05],\\n       [2.49134948e-0...   \n",
       "10  [array([[9.99104313e-01, 0.00000000e+00, 0.00000000e+00, 8.95687205e-04],\\n       [5.20415225e-0...   \n",
       "11  [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [4.42906574e-0...   \n",
       "12  [array([[9.99993936e-01, 0.00000000e+00, 0.00000000e+00, 6.06360724e-06],\\n       [3.32179931e-0...   \n",
       "13  [array([[9.99939364e-01, 0.00000000e+00, 0.00000000e+00, 6.06360724e-05],\\n       [6.64359862e-0...   \n",
       "14  [array([[9.91322978e-01, 0.00000000e+00, 1.08538570e-03, 7.59163626e-03],\\n       [3.32179931e-0...   \n",
       "15  [array([[9.98707370e-01, 0.00000000e+00, 1.81457491e-04, 1.11117228e-03],\\n       [2.21453287e-0...   \n",
       "16  [array([[9.98519796e-01, 0.00000000e+00, 1.40680527e-04, 1.33952328e-03],\\n       [2.76816609e-0...   \n",
       "17  [array([[9.98150765e-01, 0.00000000e+00, 4.68935089e-05, 1.80234182e-03],\\n       [2.21453287e-0...   \n",
       "18  [array([[9.99090674e-01, 0.00000000e+00, 4.48546607e-05, 8.64471643e-04],\\n       [8.85813149e-0...   \n",
       "19  [array([[9.99213005e-01, 0.00000000e+00, 1.69224402e-04, 6.17771009e-04],\\n       [2.21453287e-0...   \n",
       "20  [array([[9.98560573e-01, 2.03884821e-06, 4.28158125e-05, 1.39457218e-03],\\n       [1.10726644e-0...   \n",
       "21  [array([[9.98132415e-01, 6.11654464e-06, 2.85438750e-05, 1.83292454e-03],\\n       [5.53633218e-0...   \n",
       "22  [array([[9.98189503e-01, 2.03884821e-06, 2.85438750e-05, 1.77991449e-03],\\n       [0.00000000e+0...   \n",
       "23  [array([[9.98468825e-01, 2.03884821e-06, 4.89323571e-05, 1.48020380e-03],\\n       [5.53633218e-0...   \n",
       "24  [array([[9.85075631e-01, 2.03884821e-06, 2.01845973e-04, 1.47204841e-02],\\n       [2.21453287e-0...   \n",
       "25  [[9.97645130e-01 0.00000000e+00 5.91265982e-05 2.29574309e-03]\\n [2.93425606e-04 7.27368858e-01 ...   \n",
       "26  [[9.99500482e-01 2.03884821e-06 3.66992678e-05 4.60779696e-04]\\n [1.43944637e-04 8.00780623e-01 ...   \n",
       "27  [[9.98799118e-01 0.00000000e+00 5.52527866e-04 6.48353732e-04]\\n [4.42906574e-05 8.01738408e-01 ...   \n",
       "28  [[9.98501447e-01 0.00000000e+00 8.56316250e-05 1.41292181e-03]\\n [3.43252595e-04 7.51584775e-01 ...   \n",
       "29  [[9.97922414e-01 0.00000000e+00 6.99324937e-04 1.37826139e-03]\\n [4.98269896e-05 8.10325260e-01 ...   \n",
       "..                                                                                                  ...   \n",
       "43  [[9.97569693e-01 0.00000000e+00 4.89323571e-05 2.38137471e-03]\\n [7.19723183e-05 6.87645675e-01 ...   \n",
       "44  [[9.98236396e-01 0.00000000e+00 2.03884821e-06 1.76156486e-03]\\n [6.69896194e-04 8.62400000e-01 ...   \n",
       "45  [[9.98317950e-01 0.00000000e+00 5.52527866e-04 1.12952191e-03]\\n [1.10726644e-04 8.12135640e-01 ...   \n",
       "46  [[9.98570767e-01 0.00000000e+00 1.14175500e-04 1.31505710e-03]\\n [2.15916955e-04 6.58901038e-01 ...   \n",
       "47  [[9.98114065e-01 0.00000000e+00 1.24369741e-04 1.76156486e-03]\\n [7.75086505e-05 7.52758478e-01 ...   \n",
       "48  [[9.96927456e-01 0.00000000e+00 1.08058955e-04 2.96448530e-03]\\n [2.15916955e-04 7.32130104e-01 ...   \n",
       "49  [[9.96544152e-01 0.00000000e+00 5.64760955e-04 2.89108677e-03]\\n [1.71626298e-04 6.67044983e-01 ...   \n",
       "50  [[9.97304643e-01 0.00000000e+00 3.91458857e-04 2.30389848e-03]\\n [4.70588235e-04 7.32927336e-01 ...   \n",
       "51  [[9.96668522e-01 2.03884821e-06 7.56412687e-04 2.57302645e-03]\\n [5.53633218e-06 5.22507958e-01 ...   \n",
       "52  [[9.99429123e-01 0.00000000e+00 1.85535187e-04 3.85342312e-04]\\n [2.04844291e-04 5.87642907e-01 ...   \n",
       "53  [[9.98342416e-01 0.00000000e+00 1.16214348e-04 1.54136925e-03]\\n [1.60553633e-04 5.63482353e-01 ...   \n",
       "54  [[9.99400579e-01 0.00000000e+00 1.34563982e-04 4.64857393e-04]\\n [8.02768166e-04 7.87570934e-01 ...   \n",
       "55  [[9.98719603e-01 0.00000000e+00 2.50778330e-04 1.02961835e-03]\\n [1.66089965e-04 8.54372318e-01 ...   \n",
       "56  [[9.97849015e-01 0.00000000e+00 1.95729429e-04 1.95525544e-03]\\n [5.75778547e-04 8.62876125e-01 ...   \n",
       "57  [[9.96902990e-01 0.00000000e+00 1.83496339e-05 3.07866080e-03]\\n [4.92733564e-04 7.95792388e-01 ...   \n",
       "58  [[9.98199697e-01 0.00000000e+00 1.30486286e-04 1.66981669e-03]\\n [5.42560554e-04 8.29719031e-01 ...   \n",
       "59  [[9.97624742e-01 6.72819911e-05 4.09808491e-04 1.89816769e-03]\\n [1.55017301e-04 7.28963322e-01 ...   \n",
       "60  [[9.98829701e-01 2.03884821e-06 5.09712053e-04 6.58547973e-04]\\n [3.32179931e-05 7.36946713e-01 ...   \n",
       "61  [[9.98894944e-01 0.00000000e+00 2.07962518e-04 8.97093214e-04]\\n [2.60207612e-04 7.88888581e-01 ...   \n",
       "62  [[9.99435239e-01 0.00000000e+00 1.01942411e-04 4.62818545e-04]\\n [4.42906574e-05 8.40121799e-01 ...   \n",
       "63  [[9.97728723e-01 0.00000000e+00 3.46604196e-05 2.23661649e-03]\\n [3.87543253e-05 7.98355709e-01 ...   \n",
       "64  [[9.97426974e-01 2.03884821e-06 6.32042946e-05 2.50778330e-03]\\n [1.66089965e-05 8.39429758e-01 ...   \n",
       "65  [[9.97388235e-01 0.00000000e+00 8.15539286e-05 2.53021063e-03]\\n [9.96539792e-05 6.88658824e-01 ...   \n",
       "66  [[9.96984543e-01 0.00000000e+00 8.76704732e-05 2.92778604e-03]\\n [3.87543253e-05 6.50679585e-01 ...   \n",
       "67  [[9.98678826e-01 2.03884821e-06 9.17481696e-05 1.22738662e-03]\\n [1.21799308e-04 7.33049135e-01 ...   \n",
       "68  [[9.98220086e-01 0.00000000e+00 5.50489018e-05 1.72486559e-03]\\n [7.80622837e-04 7.94175779e-01 ...   \n",
       "69  [[9.93547045e-01 1.65146705e-04 2.03884821e-05 6.26741941e-03]\\n [1.55017301e-04 8.58917647e-01 ...   \n",
       "70  [[9.97620664e-01 0.00000000e+00 1.40680527e-04 2.23865534e-03]\\n [2.15916955e-04 7.88229758e-01 ...   \n",
       "71  [[9.98287368e-01 0.00000000e+00 8.35927768e-05 1.62903972e-03]\\n [6.58823529e-04 7.13090657e-01 ...   \n",
       "72  [[9.98324067e-01 0.00000000e+00 6.52431428e-05 1.61069009e-03]\\n [6.64359862e-05 8.61237370e-01 ...   \n",
       "\n",
       "                                                                                                      2  \\\n",
       "0   [array([[9.92899343e-01, 2.40386838e-05, 2.32620494e-03, 4.75041374e-03],\\n       [1.10490523e-0...   \n",
       "1   [array([[9.92847567e-01, 0.00000000e+00, 6.71788755e-03, 4.34545438e-04],\\n       [1.21539575e-0...   \n",
       "2   [array([[9.95715567e-01, 0.00000000e+00, 3.19899407e-03, 1.08543903e-03],\\n       [7.18188397e-0...   \n",
       "3   [array([[9.90292070e-01, 7.39651809e-06, 5.56957812e-03, 4.13095535e-03],\\n       [4.97207352e-0...   \n",
       "4   [array([[9.93474422e-01, 3.69825904e-06, 5.02963230e-03, 1.49224752e-03],\\n       [9.39169443e-0...   \n",
       "5   [array([[9.91715900e-01, 1.84912952e-06, 5.18126092e-03, 3.10099021e-03],\\n       [2.20981045e-0...   \n",
       "6   [array([[9.91636387e-01, 0.00000000e+00, 6.45346203e-03, 1.91015080e-03],\\n       [1.65735784e-0...   \n",
       "7   [array([[9.90759900e-01, 0.00000000e+00, 7.81072310e-03, 1.42937712e-03],\\n       [6.62943136e-0...   \n",
       "8   [array([[9.94207332e-01, 0.00000000e+00, 2.46873096e-03, 3.32393663e-03],\\n       [1.04965997e-0...   \n",
       "9   [array([[9.90071044e-01, 0.00000000e+00, 6.66464200e-03, 3.26431445e-03],\\n       [8.83924181e-0...   \n",
       "10  [array([[9.91621220e-01, 1.86319318e-06, 4.87225016e-03, 3.50466637e-03],\\n       [2.76226307e-0...   \n",
       "11  [array([[9.92322414e-01, 3.69825904e-06, 5.22933829e-03, 2.44454923e-03],\\n       [3.86716829e-0...   \n",
       "12  [array([[9.95284720e-01, 3.69825904e-06, 3.49115654e-03, 1.22042548e-03],\\n       [2.76226307e-0...   \n",
       "13  [array([[9.90305014e-01, 2.95860724e-05, 6.95272700e-03, 2.71267301e-03],\\n       [2.20981045e-0...   \n",
       "14  [array([[9.84748380e-01, 4.43791085e-05, 7.87729176e-03, 7.32994943e-03],\\n       [2.20981045e-0...   \n",
       "15  [array([[9.92659423e-01, 0.00000000e+00, 5.34063294e-03, 1.99994414e-03],\\n       [5.52385477e-0...   \n",
       "16  [array([[9.93545804e-01, 0.00000000e+00, 3.92540246e-03, 2.52879342e-03],\\n       [1.65715643e-0...   \n",
       "17  [array([[9.87728462e-01, 0.00000000e+00, 6.06873179e-03, 6.20280625e-03],\\n       [5.52385477e-0...   \n",
       "18  [array([[9.90056144e-01, 0.00000000e+00, 6.54357886e-03, 3.40027746e-03],\\n       [5.52385477e-0...   \n",
       "19  [array([[9.90944387e-01, 0.00000000e+00, 6.68137762e-03, 2.37423536e-03],\\n       [5.52385477e-0...   \n",
       "20  [array([[9.92419206e-01, 0.00000000e+00, 4.84157798e-03, 2.73921585e-03],\\n       [5.52385477e-0...   \n",
       "21  [array([[9.91462063e-01, 1.86214538e-06, 4.63301770e-03, 3.90305671e-03],\\n       [5.52385477e-0...   \n",
       "22  [array([[9.92240440e-01, 1.86214538e-06, 3.60697560e-03, 4.15072205e-03],\\n       [5.52385477e-0...   \n",
       "23  [array([[9.91739523e-01, 0.00000000e+00, 5.56222824e-03, 2.69824865e-03],\\n       [5.52385477e-0...   \n",
       "24  [array([[9.77248308e-01, 1.86214538e-06, 4.49149465e-03, 1.82583354e-02],\\n       [5.52385477e-0...   \n",
       "25  [[9.90180907e-01 1.86214538e-06 4.87695874e-03 4.94027169e-03]\\n [1.76763353e-04 8.24114940e-01 ...   \n",
       "26  [[9.91167844e-01 3.72429076e-06 5.50450174e-03 3.32392950e-03]\\n [3.31431286e-05 8.34289881e-01 ...   \n",
       "27  [[9.89396944e-01 7.63479605e-05 7.05753098e-03 3.46917684e-03]\\n [1.10477095e-05 5.30058056e-01 ...   \n",
       "28  [[9.92545832e-01 5.58643613e-06 3.76339581e-03 3.68518570e-03]\\n [3.31431286e-05 7.15217667e-01 ...   \n",
       "29  [[9.90517956e-01 0.00000000e+00 7.35919853e-03 2.12284573e-03]\\n [5.52385477e-06 6.42982219e-01 ...   \n",
       "..                                                                                                  ...   \n",
       "43  [[9.87851364e-01 3.72429076e-06 7.65714179e-03 4.48777036e-03]\\n [2.20954191e-05 5.37338496e-01 ...   \n",
       "44  [[9.90665065e-01 1.86214538e-06 5.99052168e-03 3.34255095e-03]\\n [1.27048660e-04 8.57335403e-01 ...   \n",
       "45  [[9.71376963e-01 4.76709217e-03 1.08134782e-02 1.30424662e-02]\\n [2.20954191e-05 5.10155607e-01 ...   \n",
       "46  [[9.77626323e-01 7.44858151e-06 7.19719188e-03 1.51690362e-02]\\n [3.86669834e-05 4.54745820e-01 ...   \n",
       "47  [[9.88035716e-01 1.48971630e-05 7.72976546e-03 4.21962143e-03]\\n [6.62862572e-05 5.62361558e-01 ...   \n",
       "48  [[9.88592497e-01 1.86214538e-06 6.65158329e-03 4.75405715e-03]\\n [3.31431286e-05 7.57337060e-01 ...   \n",
       "49  [[9.86985466e-01 0.00000000e+00 8.61428452e-03 4.40024953e-03]\\n [6.07624024e-05 6.34160623e-01 ...   \n",
       "50  [[9.87227545e-01 1.11728723e-05 8.91222778e-03 3.84905450e-03]\\n [5.52385477e-05 6.38220656e-01 ...   \n",
       "51  [[9.88335521e-01 0.00000000e+00 6.67392903e-03 4.99054961e-03]\\n [1.10477095e-05 4.68616219e-01 ...   \n",
       "52  [[9.87702392e-01 3.72429076e-06 5.88251725e-03 6.41136654e-03]\\n [2.87240448e-04 2.52136351e-01 ...   \n",
       "53  [[9.88806644e-01 0.00000000e+00 6.01472957e-03 5.17862630e-03]\\n [6.62862572e-05 4.38914452e-01 ...   \n",
       "54  [[9.92761841e-01 0.00000000e+00 6.12459615e-03 1.11356294e-03]\\n [3.25907431e-04 7.10953252e-01 ...   \n",
       "55  [[9.90722792e-01 1.86214538e-06 7.04263382e-03 2.23271231e-03]\\n [7.73339667e-05 7.74096435e-01 ...   \n",
       "56  [[9.89830824e-01 1.35936613e-04 7.74652477e-03 2.28671452e-03]\\n [5.52385477e-05 8.18220987e-01 ...   \n",
       "57  [[9.93245999e-01 1.48971630e-05 5.16559128e-03 1.57351284e-03]\\n [1.32572514e-04 8.73199914e-01 ...   \n",
       "58  [[9.91856838e-01 1.67593084e-05 4.95516885e-03 3.17123358e-03]\\n [1.93334917e-04 8.09581678e-01 ...   \n",
       "59  [[9.89199557e-01 1.48971630e-05 7.28098843e-03 3.50455760e-03]\\n [6.62862572e-05 4.53121807e-01 ...   \n",
       "60  [[9.86106533e-01 1.86214538e-06 7.54727522e-03 6.34432930e-03]\\n [5.52385477e-06 4.24602144e-01 ...   \n",
       "61  [[9.89007756e-01 0.00000000e+00 7.77073266e-03 3.22151150e-03]\\n [8.28578215e-05 8.35692940e-01 ...   \n",
       "62  [[9.91810285e-01 4.28293437e-05 6.74469056e-03 1.40219547e-03]\\n [2.76192738e-05 7.90071423e-01 ...   \n",
       "63  [[9.91923875e-01 0.00000000e+00 6.01845386e-03 2.05767064e-03]\\n [1.32572514e-04 7.58867168e-01 ...   \n",
       "64  [[9.90603614e-01 0.00000000e+00 7.54541307e-03 1.85097251e-03]\\n [1.10477095e-05 6.74341142e-01 ...   \n",
       "65  [[9.89059896e-01 0.00000000e+00 7.11153320e-03 3.82857090e-03]\\n [7.18101120e-05 4.16697508e-01 ...   \n",
       "66  [[9.89640885e-01 0.00000000e+00 4.98868747e-03 5.37042727e-03]\\n [1.04953241e-04 3.26365911e-01 ...   \n",
       "67  [[9.90229323e-01 0.00000000e+00 5.30152789e-03 4.46914891e-03]\\n [2.76192738e-05 7.46322494e-01 ...   \n",
       "68  [[9.91530963e-01 2.60700353e-05 4.89930449e-03 3.54366265e-03]\\n [1.54667933e-04 8.09139770e-01 ...   \n",
       "69  [[9.94096999e-01 0.00000000e+00 3.44310680e-03 2.45989404e-03]\\n [8.83816763e-05 8.53059939e-01 ...   \n",
       "70  [[9.90113870e-01 3.72429076e-06 7.76700837e-03 2.11539715e-03]\\n [1.38096369e-04 7.68705153e-01 ...   \n",
       "71  [[9.91227433e-01 1.86214538e-06 5.62554119e-03 3.14516354e-03]\\n [5.52385477e-05 6.84521607e-01 ...   \n",
       "72  [[9.90981630e-01 1.30350176e-05 7.21208905e-03 1.79324600e-03]\\n [8.83816763e-05 8.24656278e-01 ...   \n",
       "\n",
       "                                                                                                      3  \\\n",
       "0   [array([[9.99996610e-01, 0.00000000e+00, 0.00000000e+00, 3.38991095e-06],\\n       [0.00000000e+0...   \n",
       "1   [array([[9.99969491e-01, 0.00000000e+00, 1.01697328e-05, 2.03394657e-05],\\n       [1.98668918e-0...   \n",
       "2   [array([[9.99976271e-01, 0.00000000e+00, 1.01697328e-05, 1.35596438e-05],\\n       [2.33728139e-0...   \n",
       "3   [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [3.50592209e-0...   \n",
       "4   [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [4.09024243e-0...   \n",
       "5   [array([[9.99811860e-01, 1.69495547e-06, 8.47477737e-06, 1.77970325e-04],\\n       [7.01184417e-0...   \n",
       "6   [array([[9.99993220e-01, 0.00000000e+00, 0.00000000e+00, 6.77982189e-06],\\n       [1.16864070e-0...   \n",
       "7   [array([[9.99993220e-01, 0.00000000e+00, 0.00000000e+00, 6.77982189e-06],\\n       [2.33728139e-0...   \n",
       "8   [array([[9.99935592e-01, 0.00000000e+00, 0.00000000e+00, 6.44083080e-05],\\n       [1.92825715e-0...   \n",
       "9   [array([[9.99601685e-01, 1.69495547e-06, 1.69495547e-06, 3.94924625e-04],\\n       [4.55769871e-0...   \n",
       "10  [array([[9.99927117e-01, 0.00000000e+00, 0.00000000e+00, 7.28830854e-05],\\n       [1.16864070e-0...   \n",
       "11  [array([[9.99994915e-01, 0.00000000e+00, 0.00000000e+00, 5.08486642e-06],\\n       [1.75296104e-0...   \n",
       "12  [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [2.33728139e-0...   \n",
       "13  [array([[9.99971186e-01, 1.69495547e-06, 0.00000000e+00, 2.71192876e-05],\\n       [5.84320348e-0...   \n",
       "14  [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [2.33728139e-0...   \n",
       "15  [array([[9.99981355e-01, 0.00000000e+00, 1.69495547e-06, 1.69495547e-05],\\n       [5.84320348e-0...   \n",
       "16  [array([[9.99981355e-01, 0.00000000e+00, 1.69495547e-06, 1.69495547e-05],\\n       [7.59616452e-0...   \n",
       "17  [array([[9.99271169e-01, 0.00000000e+00, 3.72890204e-05, 6.91541833e-04],\\n       [2.33728139e-0...   \n",
       "18  [array([[9.99688128e-01, 0.00000000e+00, 5.25436197e-05, 2.59328187e-04],\\n       [1.22707273e-0...   \n",
       "19  [array([[9.99193201e-01, 0.00000000e+00, 2.71192876e-04, 5.35605930e-04],\\n       [2.45414546e-0...   \n",
       "20  [array([[9.98855905e-01, 0.00000000e+00, 1.18646883e-04, 1.02544806e-03],\\n       [3.50592209e-0...   \n",
       "21  [array([[9.98323689e-01, 0.00000000e+00, 1.28816616e-04, 1.54749435e-03],\\n       [3.50592209e-0...   \n",
       "22  [array([[9.98333859e-01, 0.00000000e+00, 1.10172106e-04, 1.55596912e-03],\\n       [4.09024243e-0...   \n",
       "23  [array([[9.98918618e-01, 0.00000000e+00, 1.66105636e-04, 9.15275956e-04],\\n       [4.09024243e-0...   \n",
       "24  [array([[9.88904821e-01, 0.00000000e+00, 6.33913347e-04, 1.04612652e-02],\\n       [4.67456278e-0...   \n",
       "25  [[9.99755926e-01 0.00000000e+00 6.77982189e-06 2.37293766e-04]\\n [7.59616452e-05 7.68001449e-01 ...   \n",
       "26  [[9.99938982e-01 0.00000000e+00 0.00000000e+00 6.10183970e-05]\\n [6.36909179e-04 7.45312290e-01 ...   \n",
       "27  [[9.99883048e-01 5.08486642e-06 1.69495547e-05 9.49175065e-05]\\n [7.59616452e-05 8.27210630e-01 ...   \n",
       "28  [[9.99827115e-01 0.00000000e+00 5.08486642e-06 1.67800592e-04]\\n [4.14867447e-04 8.61697217e-01 ...   \n",
       "29  [[9.99832199e-01 0.00000000e+00 2.20344212e-05 1.45766171e-04]\\n [7.01184417e-05 9.01343352e-01 ...   \n",
       "..                                                                                                  ...   \n",
       "43  [[9.99193201e-01 8.47477737e-06 1.59325815e-04 6.38998214e-04]\\n [1.57766494e-04 6.50225840e-01 ...   \n",
       "44  [[9.99237270e-01 0.00000000e+00 5.59335306e-05 7.06796432e-04]\\n [1.91072754e-03 7.83877433e-01 ...   \n",
       "45  [[9.98759293e-01 1.01697328e-05 3.96619581e-04 8.33918093e-04]\\n [1.11605186e-03 8.38949626e-01 ...   \n",
       "46  [[9.98435556e-01 0.00000000e+00 1.33901482e-04 1.43054242e-03]\\n [1.19785671e-03 6.44540403e-01 ...   \n",
       "47  [[9.99260999e-01 0.00000000e+00 5.08486642e-05 6.88151922e-04]\\n [3.38905802e-04 7.49122059e-01 ...   \n",
       "48  [[9.98133854e-01 0.00000000e+00 1.15256972e-04 1.75088900e-03]\\n [1.46664407e-03 5.88615102e-01 ...   \n",
       "49  [[9.95881258e-01 0.00000000e+00 8.40697915e-04 3.27804389e-03]\\n [2.57100953e-04 6.71728829e-01 ...   \n",
       "50  [[9.97005014e-01 0.00000000e+00 6.18658748e-04 2.37632757e-03]\\n [1.04593342e-03 6.95569099e-01 ...   \n",
       "51  [[9.97138915e-01 0.00000000e+00 1.02036320e-03 1.84072164e-03]\\n [5.96006755e-04 5.53777923e-01 ...   \n",
       "52  [[9.99766096e-01 6.77982189e-06 6.94931744e-05 1.57630859e-04]\\n [8.12205283e-04 7.43670350e-01 ...   \n",
       "53  [[9.99547447e-01 0.00000000e+00 3.72890204e-05 4.15264091e-04]\\n [9.93344591e-05 7.40433215e-01 ...   \n",
       "54  [[9.99969491e-01 6.77982189e-06 3.38991095e-06 2.03394657e-05]\\n [7.30400435e-04 7.63969639e-01 ...   \n",
       "55  [[9.99888133e-01 2.20344212e-05 2.03394657e-05 6.94931744e-05]\\n [1.16864070e-04 8.91222924e-01 ...   \n",
       "56  [[9.99150827e-01 0.00000000e+00 2.00004746e-04 6.49167946e-04]\\n [1.98668918e-04 8.70058841e-01 ...   \n",
       "57  [[9.98396572e-01 0.00000000e+00 4.23738868e-05 1.56105399e-03]\\n [1.81139308e-04 7.55356757e-01 ...   \n",
       "58  [[9.99384731e-01 0.00000000e+00 9.49175065e-05 5.20351330e-04]\\n [1.46080087e-04 8.37880320e-01 ...   \n",
       "59  [[9.99227100e-01 0.00000000e+00 4.74587533e-05 7.25440943e-04]\\n [3.91494633e-04 8.28396800e-01 ...   \n",
       "60  [[9.99364392e-01 0.00000000e+00 3.15261718e-04 3.20346584e-04]\\n [1.05761983e-03 7.91917681e-01 ...   \n",
       "61  [[9.99644059e-01 2.03394657e-05 1.03392284e-04 2.32208900e-04]\\n [4.84985889e-04 8.08296180e-01 ...   \n",
       "62  [[9.99883048e-01 1.35596438e-05 5.25436197e-05 5.08486642e-05]\\n [9.40755760e-04 8.87734532e-01 ...   \n",
       "63  [[9.99162692e-01 0.00000000e+00 8.47477737e-06 8.28833227e-04]\\n [7.12870824e-04 7.77064258e-01 ...   \n",
       "64  [[9.99435580e-01 1.52545993e-05 7.28830854e-05 4.76282488e-04]\\n [2.92160174e-05 8.80483116e-01 ...   \n",
       "65  [[9.99415240e-01 1.69495547e-06 4.74587533e-05 5.35605930e-04]\\n [5.49261127e-04 7.42992538e-01 ...   \n",
       "66  [[9.99064385e-01 3.38991095e-06 5.08486642e-05 8.81376846e-04]\\n [5.55104330e-04 6.70606934e-01 ...   \n",
       "67  [[9.99940677e-01 0.00000000e+00 1.69495547e-06 5.76284861e-05]\\n [1.57766494e-04 7.91742385e-01 ...   \n",
       "68  [[9.99394901e-01 0.00000000e+00 9.32225510e-05 5.11876553e-04]\\n [1.40236883e-04 7.94459474e-01 ...   \n",
       "69  [[9.97411803e-01 1.93224924e-04 3.38991095e-06 2.39158217e-03]\\n [6.89498010e-04 9.17120002e-01 ...   \n",
       "70  [[9.98938958e-01 0.00000000e+00 5.59335306e-05 1.00510860e-03]\\n [3.27219395e-04 7.47550237e-01 ...   \n",
       "71  [[9.98469455e-01 1.69495547e-06 2.15259345e-04 1.31359049e-03]\\n [4.96672296e-04 6.88814356e-01 ...   \n",
       "72  [[9.99064385e-01 1.69495547e-06 9.32225510e-05 8.40697915e-04]\\n [2.68787360e-04 8.63269039e-01 ...   \n",
       "\n",
       "                                                                                                      4  \\\n",
       "0   [array([[9.99996543e-01, 0.00000000e+00, 0.00000000e+00, 3.45705559e-06],\\n       [0.00000000e+0...   \n",
       "1   [array([[9.99970615e-01, 0.00000000e+00, 1.20996946e-05, 1.72852780e-05],\\n       [3.00979689e-0...   \n",
       "2   [array([[9.99987900e-01, 0.00000000e+00, 3.45705559e-06, 8.64263898e-06],\\n       [2.50816407e-0...   \n",
       "3   [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [5.01632815e-0...   \n",
       "4   [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [1.50489844e-0...   \n",
       "5   [array([[9.99958515e-01, 0.00000000e+00, 6.91411118e-06, 3.45705559e-05],\\n       [0.00000000e+0...   \n",
       "6   [array([[9.99996543e-01, 0.00000000e+00, 0.00000000e+00, 3.45705559e-06],\\n       [5.01632815e-0...   \n",
       "7   [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [1.00326563e-0...   \n",
       "8   [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [5.01632815e-0...   \n",
       "9   [array([[9.99883783e-01, 0.00000000e+00, 0.00000000e+00, 1.16217351e-04],\\n       [5.01632815e-0...   \n",
       "10  [array([[9.99998265e-01, 0.00000000e+00, 0.00000000e+00, 1.73458732e-06],\\n       [2.00653126e-0...   \n",
       "11  [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [5.01632815e-0...   \n",
       "12  [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [1.00326563e-0...   \n",
       "13  [array([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],\\n       [2.00653126e-0...   \n",
       "14  [array([[9.96544673e-01, 0.00000000e+00, 1.75272718e-03, 1.70259988e-03],\\n       [0.00000000e+0...   \n",
       "15  [array([[9.99967044e-01, 0.00000000e+00, 2.08142536e-05, 1.21416479e-05],\\n       [3.00979689e-0...   \n",
       "16  [array([[9.99968779e-01, 0.00000000e+00, 1.04071268e-05, 2.08142536e-05],\\n       [3.51142970e-0...   \n",
       "17  [array([[9.99963575e-01, 0.00000000e+00, 5.20356340e-06, 3.12213804e-05],\\n       [3.51142970e-0...   \n",
       "18  [array([[9.99975717e-01, 0.00000000e+00, 1.73452113e-06, 2.25487747e-05],\\n       [1.45473516e-0...   \n",
       "19  [array([[9.99987858e-01, 0.00000000e+00, 3.46904227e-06, 8.67260567e-06],\\n       [3.51142970e-0...   \n",
       "20  [array([[9.99941026e-01, 0.00000000e+00, 6.93808453e-06, 5.20356340e-05],\\n       [3.00979689e-0...   \n",
       "21  [array([[9.99913274e-01, 0.00000000e+00, 3.46904227e-06, 8.32570144e-05],\\n       [1.50489844e-0...   \n",
       "22  [array([[9.99941026e-01, 0.00000000e+00, 1.73452113e-06, 5.72391974e-05],\\n       [1.50489844e-0...   \n",
       "23  [array([[9.99937557e-01, 0.00000000e+00, 8.67260567e-06, 5.37701551e-05],\\n       [2.00653126e-0...   \n",
       "24  [array([[9.97557794e-01, 0.00000000e+00, 1.11009353e-04, 2.33119640e-03],\\n       [5.01632815e-0...   \n",
       "25  [[9.99954902e-01 0.00000000e+00 5.20356340e-06 3.98939861e-05]\\n [1.00326563e-04 7.23981560e-01 ...   \n",
       "26  [[9.99993062e-01 0.00000000e+00 1.73452113e-06 5.20356340e-06]\\n [7.02285941e-05 7.37154438e-01 ...   \n",
       "27  [[9.99972248e-01 0.00000000e+00 1.90797325e-05 8.67260567e-06]\\n [1.50489844e-05 7.73914090e-01 ...   \n",
       "28  [[9.99967044e-01 0.00000000e+00 5.20356340e-06 2.77523381e-05]\\n [1.50489844e-05 9.02698283e-01 ...   \n",
       "29  [[9.99963575e-01 0.00000000e+00 1.73452113e-05 1.90797325e-05]\\n [0.00000000e+00 9.22502746e-01 ...   \n",
       "..                                                                                                  ...   \n",
       "43  [[9.99946230e-01 0.00000000e+00 1.56106902e-05 3.81594649e-05]\\n [3.51142970e-05 5.91023782e-01 ...   \n",
       "44  [[9.99941026e-01 0.00000000e+00 0.00000000e+00 5.89737185e-05]\\n [1.99649860e-03 7.60209482e-01 ...   \n",
       "45  [[9.99628812e-01 0.00000000e+00 1.28354564e-04 2.42832959e-04]\\n [1.90620470e-04 8.27167430e-01 ...   \n",
       "46  [[9.99731149e-01 0.00000000e+00 2.08142536e-05 2.48036522e-04]\\n [1.60522501e-04 6.27828582e-01 ...   \n",
       "47  [[9.99947964e-01 0.00000000e+00 1.04071268e-05 4.16285072e-05]\\n [1.40457188e-04 7.59331624e-01 ...   \n",
       "48  [[9.99814406e-01 0.00000000e+00 1.38761691e-05 1.71717592e-04]\\n [2.40783751e-04 5.62355467e-01 ...   \n",
       "49  [[9.98933270e-01 0.00000000e+00 5.34232509e-04 5.32497988e-04]\\n [6.52122659e-05 6.92905407e-01 ...   \n",
       "50  [[9.99597591e-01 0.00000000e+00 1.78655677e-04 2.23753226e-04]\\n [9.02939067e-05 7.30387411e-01 ...   \n",
       "51  [[9.99321802e-01 0.00000000e+00 4.83931396e-04 1.94266367e-04]\\n [1.35440860e-04 5.63900496e-01 ...   \n",
       "52  [[9.99963575e-01 0.00000000e+00 6.93808453e-06 2.94868593e-05]\\n [9.02939067e-05 6.89544467e-01 ...   \n",
       "53  [[9.99764105e-01 0.00000000e+00 7.97879721e-05 1.56106902e-04]\\n [0.00000000e+00 7.22797707e-01 ...   \n",
       "54  [[9.99986124e-01 0.00000000e+00 8.67260567e-06 5.20356340e-06]\\n [1.05342891e-04 7.22010143e-01 ...   \n",
       "55  [[9.99977451e-01 0.00000000e+00 1.73452113e-05 5.20356340e-06]\\n [5.01632815e-06 8.09956408e-01 ...   \n",
       "56  [[9.99772778e-01 0.00000000e+00 4.50975495e-05 1.82124719e-04]\\n [0.00000000e+00 9.18906039e-01 ...   \n",
       "57  [[9.99373838e-01 0.00000000e+00 0.00000000e+00 6.26162129e-04]\\n [0.00000000e+00 8.29168945e-01 ...   \n",
       "58  [[9.99954902e-01 0.00000000e+00 5.20356340e-06 3.98939861e-05]\\n [3.51142970e-05 8.91747639e-01 ...   \n",
       "59  [[9.99941026e-01 0.00000000e+00 2.08142536e-05 3.81594649e-05]\\n [6.01959378e-05 7.94857260e-01 ...   \n",
       "60  [[9.99939292e-01 0.00000000e+00 4.16285072e-05 1.90797325e-05]\\n [1.15375547e-04 7.85968327e-01 ...   \n",
       "61  [[9.99949699e-01 1.38761691e-05 3.29559015e-05 3.46904227e-06]\\n [7.52449222e-05 7.43906415e-01 ...   \n",
       "62  [[9.99987858e-01 0.00000000e+00 1.21416479e-05 0.00000000e+00]\\n [1.60522501e-04 8.11150294e-01 ...   \n",
       "63  [[9.99956637e-01 0.00000000e+00 5.20356340e-06 3.81594649e-05]\\n [1.50489844e-04 8.25015425e-01 ...   \n",
       "64  [[9.99963575e-01 0.00000000e+00 1.73452113e-05 1.90797325e-05]\\n [5.51796096e-05 8.57109893e-01 ...   \n",
       "65  [[9.99967044e-01 0.00000000e+00 1.04071268e-05 2.25487747e-05]\\n [1.35440860e-04 7.00956614e-01 ...   \n",
       "66  [[9.99868176e-01 1.73452113e-06 1.04071268e-05 1.19681958e-04]\\n [1.50489844e-04 6.37083707e-01 ...   \n",
       "67  [[9.99963575e-01 0.00000000e+00 1.04071268e-05 2.60178170e-05]\\n [0.00000000e+00 7.09659943e-01 ...   \n",
       "68  [[9.99960106e-01 0.00000000e+00 3.46904227e-06 3.64249438e-05]\\n [1.50489844e-05 8.57922538e-01 ...   \n",
       "69  [[9.97899495e-01 2.60178170e-04 1.56106902e-05 1.82471623e-03]\\n [5.01632815e-05 9.00074743e-01 ...   \n",
       "70  [[9.99934088e-01 0.00000000e+00 2.08142536e-05 4.50975495e-05]\\n [1.05342891e-04 7.90292402e-01 ...   \n",
       "71  [[9.99923681e-01 0.00000000e+00 3.12213804e-05 4.50975495e-05]\\n [7.02285941e-05 7.17199484e-01 ...   \n",
       "72  [[9.99953168e-01 0.00000000e+00 2.94868593e-05 1.73452113e-05]\\n [1.50489844e-04 9.09781338e-01 ...   \n",
       "\n",
       "                                                                                                      5  \\\n",
       "0   [array([[9.99925734e-01, 0.00000000e+00, 1.90424685e-06, 7.23613804e-05],\\n       [0.00000000e+0...   \n",
       "1   [array([[9.99849564e-01, 0.00000000e+00, 2.66594559e-05, 1.23776045e-04],\\n       [0.00000000e+0...   \n",
       "2   [array([[9.99925734e-01, 0.00000000e+00, 1.71382217e-05, 5.71274055e-05],\\n       [0.00000000e+0...   \n",
       "3   [array([[9.99969532e-01, 0.00000000e+00, 9.52123426e-06, 2.09467154e-05],\\n       [0.00000000e+0...   \n",
       "4   [array([[9.99979053e-01, 0.00000000e+00, 1.90424685e-06, 1.90424685e-05],\\n       [0.00000000e+0...   \n",
       "5   [array([[9.99544885e-01, 0.00000000e+00, 8.56911083e-05, 3.69423889e-04],\\n       [0.00000000e+0...   \n",
       "6   [array([[9.99885745e-01, 1.90424685e-06, 0.00000000e+00, 1.12350564e-04],\\n       [0.00000000e+0...   \n",
       "7   [array([[9.99954298e-01, 0.00000000e+00, 3.80849370e-06, 4.18934307e-05],\\n       [1.40992020e-0...   \n",
       "8   [array([[9.99832690e-01, 0.00000000e+00, 0.00000000e+00, 1.67309808e-04],\\n       [0.00000000e+0...   \n",
       "9   [array([[9.99806950e-01, 0.00000000e+00, 0.00000000e+00, 1.93049779e-04],\\n       [0.00000000e+0...   \n",
       "10  [array([[9.99240671e-01, 0.00000000e+00, 0.00000000e+00, 7.59329131e-04],\\n       [3.52480050e-0...   \n",
       "11  [array([[9.99973341e-01, 0.00000000e+00, 0.00000000e+00, 2.66594559e-05],\\n       [7.04960099e-0...   \n",
       "12  [array([[9.99986670e-01, 0.00000000e+00, 0.00000000e+00, 1.33297280e-05],\\n       [0.00000000e+0...   \n",
       "13  [array([[9.99771490e-01, 0.00000000e+00, 7.61698741e-06, 2.20892635e-04],\\n       [1.05744015e-0...   \n",
       "14  [array([[8.89199493e-01, 6.07454746e-04, 3.92788998e-02, 7.09141527e-02],\\n       [0.00000000e+0...   \n",
       "15  [array([[9.98295605e-01, 0.00000000e+00, 1.96990330e-04, 1.50740427e-03],\\n       [0.00000000e+0...   \n",
       "16  [array([[9.98396242e-01, 0.00000000e+00, 1.67013541e-04, 1.43674469e-03],\\n       [0.00000000e+0...   \n",
       "17  [array([[9.98132874e-01, 0.00000000e+00, 7.49419735e-05, 1.79218377e-03],\\n       [7.03378326e-0...   \n",
       "18  [array([[9.98696010e-01, 0.00000000e+00, 7.92243720e-05, 1.22476597e-03],\\n       [1.54743232e-0...   \n",
       "19  [array([[9.98880153e-01, 0.00000000e+00, 1.32754353e-04, 9.87092851e-04],\\n       [0.00000000e+0...   \n",
       "20  [array([[9.98372689e-01, 0.00000000e+00, 1.22048357e-04, 1.50526307e-03],\\n       [0.00000000e+0...   \n",
       "21  [array([[9.98312735e-01, 0.00000000e+00, 1.00636364e-04, 1.58662864e-03],\\n       [0.00000000e+0...   \n",
       "22  [array([[9.98207816e-01, 0.00000000e+00, 8.13655712e-05, 1.71081820e-03],\\n       [0.00000000e+0...   \n",
       "23  [array([[9.98171416e-01, 0.00000000e+00, 1.41319150e-04, 1.68726500e-03],\\n       [0.00000000e+0...   \n",
       "24  [array([[9.90970563e-01, 0.00000000e+00, 5.99535788e-05, 8.96948363e-03],\\n       [0.00000000e+0...   \n",
       "25  [[9.97486232e-01 0.00000000e+00 7.70831727e-05 2.43668474e-03]\\n [2.11013498e-05 8.55828544e-01 ...   \n",
       "26  [[9.98745257e-01 0.00000000e+00 9.63539659e-05 1.15838879e-03]\\n [0.00000000e+00 8.55596430e-01 ...   \n",
       "27  [[9.98471184e-01 0.00000000e+00 3.23321086e-04 1.20549517e-03]\\n [2.67283764e-04 8.11867399e-01 ...   \n",
       "28  [[9.98762387e-01 0.00000000e+00 4.71063833e-05 1.19050678e-03]\\n [0.00000000e+00 8.37772823e-01 ...   \n",
       "29  [[9.98250640e-01 2.14119924e-06 2.39814315e-04 1.50740427e-03]\\n [9.00324257e-04 8.49371531e-01 ...   \n",
       "..                                                                                                  ...   \n",
       "43  [[9.97385596e-01 0.00000000e+00 1.13483560e-04 2.50092072e-03]\\n [0.00000000e+00 7.74813429e-01 ...   \n",
       "44  [[9.95130913e-01 2.14119924e-06 5.13887818e-05 4.81555710e-03]\\n [1.89208770e-03 8.82535820e-01 ...   \n",
       "45  [[9.97252841e-01 1.07059962e-05 2.56943909e-04 2.47950872e-03]\\n [1.40675665e-05 9.28360917e-01 ...   \n",
       "46  [[9.96244337e-01 1.07059962e-05 9.42127667e-05 3.65074471e-03]\\n [2.03276336e-03 8.17902385e-01 ...   \n",
       "47  [[9.98053650e-01 8.56479697e-06 8.99303682e-05 1.84785495e-03]\\n [0.00000000e+00 8.35423539e-01 ...   \n",
       "48  [[9.96336408e-01 0.00000000e+00 1.43460349e-04 3.52013156e-03]\\n [0.00000000e+00 8.83337671e-01 ...   \n",
       "49  [[9.94745497e-01 0.00000000e+00 6.87324957e-04 4.56717799e-03]\\n [0.00000000e+00 8.55596430e-01 ...   \n",
       "50  [[9.96342832e-01 0.00000000e+00 4.11110255e-04 3.24605805e-03]\\n [7.03378326e-06 8.70564320e-01 ...   \n",
       "51  [[9.96019511e-01 8.56479697e-06 4.56075439e-04 3.51584916e-03]\\n [7.03378326e-06 7.97251198e-01 ...   \n",
       "52  [[9.97950872e-01 1.28471955e-05 1.07059962e-04 1.92922052e-03]\\n [4.92364828e-05 7.49674688e-01 ...   \n",
       "53  [[9.96933803e-01 4.28239849e-06 9.20715674e-05 2.96984335e-03]\\n [7.52614809e-04 7.75256557e-01 ...   \n",
       "54  [[9.98663892e-01 0.00000000e+00 1.60589943e-04 1.17551838e-03]\\n [0.00000000e+00 8.79680104e-01 ...   \n",
       "55  [[9.98060073e-01 8.56479697e-06 1.43460349e-04 1.78790137e-03]\\n [0.00000000e+00 8.22861202e-01 ...   \n",
       "56  [[9.96659729e-01 1.71295939e-05 1.75578338e-04 3.14756289e-03]\\n [0.00000000e+00 9.38953795e-01 ...   \n",
       "57  [[9.95216561e-01 0.00000000e+00 7.92243720e-05 4.70421474e-03]\\n [0.00000000e+00 8.78090469e-01 ...   \n",
       "58  [[9.98220663e-01 2.99767894e-05 9.63539659e-05 1.65300582e-03]\\n [0.00000000e+00 9.01090940e-01 ...   \n",
       "59  [[9.96349255e-01 1.07059962e-05 1.43460349e-04 3.49657836e-03]\\n [7.03378326e-06 8.16601135e-01 ...   \n",
       "60  [[9.98355559e-01 1.49883947e-05 4.34663446e-04 1.19478918e-03]\\n [0.00000000e+00 8.21236398e-01 ...   \n",
       "61  [[9.98325582e-01 0.00000000e+00 1.64872342e-04 1.50954547e-03]\\n [0.00000000e+00 9.00014771e-01 ...   \n",
       "62  [[9.99025754e-01 2.14119924e-06 1.24189556e-04 8.47914900e-04]\\n [0.00000000e+00 9.01062805e-01 ...   \n",
       "63  [[9.97689646e-01 2.14119924e-06 9.84951652e-05 2.20971762e-03]\\n [0.00000000e+00 8.66639469e-01 ...   \n",
       "64  [[9.97488373e-01 2.14119924e-05 8.99303682e-05 2.40028435e-03]\\n [0.00000000e+00 8.67399118e-01 ...   \n",
       "65  [[9.97235712e-01 6.42359773e-06 6.42359773e-05 2.69362865e-03]\\n [1.40675665e-05 8.06992987e-01 ...   \n",
       "66  [[9.95933863e-01 1.07059962e-05 7.28007743e-05 3.98263059e-03]\\n [1.40675665e-05 8.33341539e-01 ...   \n",
       "67  [[9.98507584e-01 0.00000000e+00 1.11342361e-04 1.38107351e-03]\\n [0.00000000e+00 8.63347659e-01 ...   \n",
       "68  [[9.98152145e-01 8.56479697e-06 1.15624759e-04 1.72366539e-03]\\n [0.00000000e+00 8.68011057e-01 ...   \n",
       "69  [[9.86970803e-01 1.54166345e-04 3.64003871e-05 1.28386307e-02]\\n [1.05506749e-03 9.12717784e-01 ...   \n",
       "70  [[9.97850236e-01 0.00000000e+00 1.49883947e-04 1.99988009e-03]\\n [0.00000000e+00 8.59113321e-01 ...   \n",
       "71  [[9.98419795e-01 0.00000000e+00 1.02777564e-04 1.47742748e-03]\\n [0.00000000e+00 8.19555324e-01 ...   \n",
       "72  [[9.98173557e-01 6.42359773e-06 1.52025146e-04 1.66799421e-03]\\n [1.40675665e-05 9.33143890e-01 ...   \n",
       "\n",
       "                                                                                                      6  \\\n",
       "0   [array([[9.99957277e-01, 0.00000000e+00, 0.00000000e+00, 4.27234118e-05],\\n       [0.00000000e+0...   \n",
       "1   [array([[9.99917718e-01, 0.00000000e+00, 9.49409151e-06, 7.27880349e-05],\\n       [0.00000000e+0...   \n",
       "2   [array([[9.99933541e-01, 0.00000000e+00, 3.16469717e-06, 6.32939434e-05],\\n       [0.00000000e+0...   \n",
       "3   [array([[9.99984177e-01, 0.00000000e+00, 0.00000000e+00, 1.58234859e-05],\\n       [0.00000000e+0...   \n",
       "4   [array([[9.99990506e-01, 0.00000000e+00, 0.00000000e+00, 9.49409151e-06],\\n       [0.00000000e+0...   \n",
       "5   [array([[9.99761065e-01, 0.00000000e+00, 1.26587887e-05, 2.26275848e-04],\\n       [0.00000000e+0...   \n",
       "6   [array([[9.99974682e-01, 0.00000000e+00, 0.00000000e+00, 2.53175774e-05],\\n       [0.00000000e+0...   \n",
       "7   [array([[9.99984177e-01, 0.00000000e+00, 3.16469717e-06, 1.26587887e-05],\\n       [0.00000000e+0...   \n",
       "8   [array([[9.99917078e-01, 0.00000000e+00, 0.00000000e+00, 8.29221775e-05],\\n       [0.00000000e+0...   \n",
       "9   [array([[9.99882527e-01, 0.00000000e+00, 0.00000000e+00, 1.17473085e-04],\\n       [0.00000000e+0...   \n",
       "10  [array([[9.99917078e-01, 0.00000000e+00, 0.00000000e+00, 8.29221775e-05],\\n       [0.00000000e+0...   \n",
       "11  [array([[9.99987341e-01, 0.00000000e+00, 0.00000000e+00, 1.26587887e-05],\\n       [0.00000000e+0...   \n",
       "12  [array([[9.99988924e-01, 0.00000000e+00, 1.58234859e-06, 9.49409151e-06],\\n       [0.00000000e+0...   \n",
       "13  [array([[9.99931959e-01, 0.00000000e+00, 4.74704576e-06, 6.32939434e-05],\\n       [0.00000000e+0...   \n",
       "14  [array([[9.16129196e-01, 3.73434266e-04, 2.70961372e-02, 5.64012330e-02],\\n       [0.00000000e+0...   \n",
       "15  [array([[9.99773752e-01, 0.00000000e+00, 4.83584049e-05, 1.77889846e-04],\\n       [7.60438925e-0...   \n",
       "16  [array([[9.99711577e-01, 0.00000000e+00, 4.14500613e-05, 2.46973282e-04],\\n       [7.60438925e-0...   \n",
       "17  [array([[9.99730575e-01, 0.00000000e+00, 2.41792024e-05, 2.45246196e-04],\\n       [7.60438925e-0...   \n",
       "18  [array([[9.99830746e-01, 0.00000000e+00, 3.45417178e-05, 1.34712699e-04],\\n       [1.52087785e-0...   \n",
       "19  [array([[9.99856652e-01, 0.00000000e+00, 6.73563496e-05, 7.59917791e-05],\\n       [1.52087785e-0...   \n",
       "20  [array([[9.99721939e-01, 0.00000000e+00, 2.41792024e-05, 2.53881626e-04],\\n       [0.00000000e+0...   \n",
       "21  [array([[9.99702941e-01, 0.00000000e+00, 2.41792024e-05, 2.72879570e-04],\\n       [0.00000000e+0...   \n",
       "22  [array([[9.99725393e-01, 0.00000000e+00, 2.24521165e-05, 2.52154540e-04],\\n       [0.00000000e+0...   \n",
       "23  [array([[9.99725393e-01, 0.00000000e+00, 2.93604601e-05, 2.45246196e-04],\\n       [0.00000000e+0...   \n",
       "24  [array([[9.97171033e-01, 0.00000000e+00, 3.69596380e-04, 2.45937030e-03],\\n       [0.00000000e+0...   \n",
       "25  [[9.99649402e-01 0.00000000e+00 2.59062883e-05 3.24692147e-04]\\n [7.60438925e-06 9.40533676e-01 ...   \n",
       "26  [[9.99856652e-01 0.00000000e+00 2.59062883e-05 1.17441840e-04]\\n [7.22416979e-04 9.08846186e-01 ...   \n",
       "27  [[9.99768570e-01 0.00000000e+00 1.08806411e-04 1.22623098e-04]\\n [4.41054577e-03 8.93477715e-01 ...   \n",
       "28  [[9.99801385e-01 0.00000000e+00 2.41792024e-05 1.74435675e-04]\\n [3.04175570e-05 9.30039619e-01 ...   \n",
       "29  [[9.99746118e-01 0.00000000e+00 8.63542944e-05 1.67527331e-04]\\n [1.51327346e-03 9.64046448e-01 ...   \n",
       "..                                                                                                  ...   \n",
       "43  [[9.99687397e-01 3.45417178e-06 3.28146319e-05 2.76333742e-04]\\n [0.00000000e+00 9.34488187e-01 ...   \n",
       "44  [[9.99640766e-01 0.00000000e+00 2.07250307e-05 3.38508834e-04]\\n [2.20527288e-04 9.37286602e-01 ...   \n",
       "45  [[9.99411064e-01 0.00000000e+00 1.34712699e-04 4.54223589e-04]\\n [8.36482818e-05 9.66586314e-01 ...   \n",
       "46  [[9.99592408e-01 0.00000000e+00 4.83584049e-05 3.59233865e-04]\\n [3.80219463e-05 9.09279636e-01 ...   \n",
       "47  [[9.99727120e-01 0.00000000e+00 3.45417178e-05 2.38337853e-04]\\n [2.28131678e-05 9.44115343e-01 ...   \n",
       "48  [[9.99341980e-01 0.00000000e+00 7.77188650e-05 5.80300858e-04]\\n [7.60438925e-06 9.11682623e-01 ...   \n",
       "49  [[9.98466348e-01 0.00000000e+00 3.64415122e-04 1.16923715e-03]\\n [1.52087785e-05 9.07728341e-01 ...   \n",
       "50  [[9.99131276e-01 0.00000000e+00 4.00683926e-04 4.68040276e-04]\\n [0.00000000e+00 9.16846004e-01 ...   \n",
       "51  [[9.99152001e-01 0.00000000e+00 3.47144263e-04 5.00854908e-04]\\n [7.60438925e-06 8.50284784e-01 ...   \n",
       "52  [[9.99768570e-01 0.00000000e+00 5.00854908e-05 1.81344018e-04]\\n [3.49801906e-04 8.71782393e-01 ...   \n",
       "53  [[9.99561320e-01 0.00000000e+00 3.10875460e-05 4.07592270e-04]\\n [2.28131678e-05 8.36406774e-01 ...   \n",
       "54  [[9.99877377e-01 0.00000000e+00 3.10875460e-05 9.15355521e-05]\\n [3.04175570e-05 9.51331909e-01 ...   \n",
       "55  [[9.99787568e-01 0.00000000e+00 3.62688036e-05 1.76162761e-04]\\n [0.00000000e+00 9.53559995e-01 ...   \n",
       "56  [[9.99575137e-01 0.00000000e+00 7.77188650e-05 3.47144263e-04]\\n [3.80219463e-05 9.78509996e-01 ...   \n",
       "57  [[9.98872213e-01 0.00000000e+00 3.10875460e-05 1.09669954e-03]\\n [0.00000000e+00 9.19864946e-01 ...   \n",
       "58  [[9.99794477e-01 0.00000000e+00 2.76333742e-05 1.77889846e-04]\\n [6.84395033e-05 9.76487228e-01 ...   \n",
       "59  [[9.99701214e-01 0.00000000e+00 7.42646932e-05 2.24521165e-04]\\n [2.28131678e-05 9.44685673e-01 ...   \n",
       "60  [[9.99792750e-01 0.00000000e+00 1.12260583e-04 9.49897238e-05]\\n [3.19384349e-04 9.09637043e-01 ...   \n",
       "61  [[9.99811748e-01 6.90834355e-06 5.00854908e-05 1.31258527e-04]\\n [1.52087785e-05 9.49400394e-01 ...   \n",
       "62  [[9.99905010e-01 0.00000000e+00 3.28146319e-05 6.21750920e-05]\\n [1.52087785e-05 9.80388280e-01 ...   \n",
       "63  [[9.99711577e-01 0.00000000e+00 2.41792024e-05 2.64244141e-04]\\n [4.56263355e-05 9.64601568e-01 ...   \n",
       "64  [[9.99713304e-01 0.00000000e+00 4.66313190e-05 2.40064938e-04]\\n [1.52087785e-05 9.75118438e-01 ...   \n",
       "65  [[9.99754754e-01 0.00000000e+00 3.10875460e-05 2.14158650e-04]\\n [7.60438925e-05 9.40373984e-01 ...   \n",
       "66  [[9.99664945e-01 0.00000000e+00 2.41792024e-05 3.10875460e-04]\\n [1.52087785e-05 9.29111883e-01 ...   \n",
       "67  [[9.99758208e-01 0.00000000e+00 2.24521165e-05 2.19339908e-04]\\n [0.00000000e+00 9.46016441e-01 ...   \n",
       "68  [[9.99787568e-01 0.00000000e+00 2.59062883e-05 1.86525276e-04]\\n [4.56263355e-05 9.67628115e-01 ...   \n",
       "69  [[9.96972418e-01 2.17612822e-04 2.76333742e-05 2.78233537e-03]\\n [5.32307248e-05 9.69825783e-01 ...   \n",
       "70  [[9.99659764e-01 1.72708589e-06 3.28146319e-05 3.05694202e-04]\\n [0.00000000e+00 9.63810712e-01 ...   \n",
       "71  [[9.99677035e-01 0.00000000e+00 4.14500613e-05 2.81515000e-04]\\n [4.56263355e-05 9.27651841e-01 ...   \n",
       "72  [[9.99746118e-01 0.00000000e+00 2.59062883e-05 2.27975337e-04]\\n [0.00000000e+00 9.82190520e-01 ...   \n",
       "\n",
       "                                                                                                      7  \\\n",
       "0   [array([[9.99967062e-01, 0.00000000e+00, 0.00000000e+00, 3.29379214e-05],\\n       [0.00000000e+0...   \n",
       "1   [array([[9.99855701e-01, 0.00000000e+00, 6.11704255e-05, 8.31290398e-05],\\n       [7.04528001e-0...   \n",
       "2   [array([[9.99896481e-01, 0.00000000e+00, 7.84236224e-06, 9.56768194e-05],\\n       [2.81811201e-0...   \n",
       "3   [array([[9.99990589e-01, 0.00000000e+00, 0.00000000e+00, 9.41083469e-06],\\n       [0.00000000e+0...   \n",
       "4   [array([[9.99995295e-01, 0.00000000e+00, 0.00000000e+00, 4.70541735e-06],\\n       [0.00000000e+0...   \n",
       "5   [array([[9.99767866e-01, 0.00000000e+00, 4.07802837e-05, 1.91353639e-04],\\n       [0.00000000e+0...   \n",
       "6   [array([[9.99987452e-01, 0.00000000e+00, 0.00000000e+00, 1.25477796e-05],\\n       [7.04528001e-0...   \n",
       "7   [array([[9.99993726e-01, 0.00000000e+00, 0.00000000e+00, 6.27388980e-06],\\n       [0.00000000e+0...   \n",
       "8   [array([[9.99933122e-01, 0.00000000e+00, 0.00000000e+00, 6.68783911e-05],\\n       [0.00000000e+0...   \n",
       "9   [array([[9.99921118e-01, 0.00000000e+00, 1.71483054e-06, 7.71673743e-05],\\n       [0.00000000e+0...   \n",
       "10  [array([[9.99933122e-01, 0.00000000e+00, 0.00000000e+00, 6.68783911e-05],\\n       [0.00000000e+0...   \n",
       "11  [array([[9.99995295e-01, 0.00000000e+00, 0.00000000e+00, 4.70541735e-06],\\n       [0.00000000e+0...   \n",
       "12  [array([[9.99996863e-01, 0.00000000e+00, 0.00000000e+00, 3.13694490e-06],\\n       [7.04528001e-0...   \n",
       "13  [array([[9.99973336e-01, 0.00000000e+00, 0.00000000e+00, 2.66640316e-05],\\n       [1.40905600e-0...   \n",
       "14  [array([[9.14648435e-01, 4.79952569e-04, 3.05444325e-02, 5.43271802e-02],\\n       [0.00000000e+0...   \n",
       "15  [array([[9.99881684e-01, 0.00000000e+00, 1.20030522e-05, 1.06312748e-04],\\n       [0.00000000e+0...   \n",
       "16  [array([[9.99881684e-01, 0.00000000e+00, 1.02883305e-05, 1.08027470e-04],\\n       [0.00000000e+0...   \n",
       "17  [array([[9.99874825e-01, 0.00000000e+00, 1.71472174e-06, 1.23459966e-04],\\n       [0.00000000e+0...   \n",
       "18  [array([[9.99914264e-01, 0.00000000e+00, 1.71472174e-06, 8.40213654e-05],\\n       [0.00000000e+0...   \n",
       "19  [array([[9.99912549e-01, 0.00000000e+00, 1.37177739e-05, 7.37330350e-05],\\n       [3.51506204e-0...   \n",
       "20  [array([[9.99876540e-01, 0.00000000e+00, 5.14416523e-06, 1.18315800e-04],\\n       [0.00000000e+0...   \n",
       "21  [array([[9.99864537e-01, 0.00000000e+00, 5.14416523e-06, 1.30318853e-04],\\n       [0.00000000e+0...   \n",
       "22  [array([[9.99859393e-01, 0.00000000e+00, 5.14416523e-06, 1.35463018e-04],\\n       [0.00000000e+0...   \n",
       "23  [array([[9.99866252e-01, 0.00000000e+00, 6.85888697e-06, 1.26889409e-04],\\n       [0.00000000e+0...   \n",
       "24  [array([[9.96246474e-01, 0.00000000e+00, 2.10910774e-04, 3.54261512e-03],\\n       [0.00000000e+0...   \n",
       "25  [[9.99819954e-01 0.00000000e+00 6.85888697e-06 1.73186896e-04]\\n [1.40602482e-05 9.20918134e-01 ...   \n",
       "26  [[9.99915979e-01 0.00000000e+00 0.00000000e+00 8.40213654e-05]\\n [2.46054343e-04 8.87911702e-01 ...   \n",
       "27  [[9.99888543e-01 0.00000000e+00 3.60091566e-05 7.54477567e-05]\\n [2.90344125e-03 8.78568667e-01 ...   \n",
       "28  [[9.99909120e-01 0.00000000e+00 0.00000000e+00 9.08802524e-05]\\n [9.13916131e-05 9.31864037e-01 ...   \n",
       "29  [[9.99895402e-01 0.00000000e+00 2.40061044e-05 8.05919219e-05]\\n [3.51506204e-03 9.40511090e-01 ...   \n",
       "..                                                                                                  ...   \n",
       "43  [[9.99777086e-01 0.00000000e+00 8.57360872e-06 2.14340218e-04]\\n [0.00000000e+00 9.13880980e-01 ...   \n",
       "44  [[9.99609043e-01 0.00000000e+00 0.00000000e+00 3.90956558e-04]\\n [3.65566452e-04 9.33537207e-01 ...   \n",
       "45  [[9.99075765e-01 0.00000000e+00 1.16601079e-04 8.07633941e-04]\\n [1.26542233e-04 9.63717530e-01 ...   \n",
       "46  [[9.99473580e-01 0.00000000e+00 3.08649914e-05 4.95554584e-04]\\n [0.00000000e+00 8.82189181e-01 ...   \n",
       "47  [[9.99854249e-01 0.00000000e+00 1.02883305e-05 1.35463018e-04]\\n [0.00000000e+00 9.22570213e-01 ...   \n",
       "48  [[9.99163216e-01 0.00000000e+00 1.37177739e-05 8.23066437e-04]\\n [0.00000000e+00 9.05402650e-01 ...   \n",
       "49  [[9.98245840e-01 0.00000000e+00 3.63521010e-04 1.39063933e-03]\\n [0.00000000e+00 8.88966220e-01 ...   \n",
       "50  [[9.99156357e-01 0.00000000e+00 2.60637705e-04 5.83005393e-04]\\n [0.00000000e+00 9.03673240e-01 ...   \n",
       "51  [[9.99259240e-01 0.00000000e+00 2.72640757e-04 4.68119036e-04]\\n [0.00000000e+00 8.32275300e-01 ...   \n",
       "52  [[9.99855963e-01 0.00000000e+00 8.57360872e-06 1.35463018e-04]\\n [1.82783226e-04 8.46947169e-01 ...   \n",
       "53  [[9.99701638e-01 0.00000000e+00 2.22913827e-05 2.76070201e-04]\\n [1.40602482e-05 8.21561391e-01 ...   \n",
       "54  [[9.99921123e-01 0.00000000e+00 6.85888697e-06 7.20183132e-05]\\n [2.10903722e-05 9.31955429e-01 ...   \n",
       "55  [[9.99903976e-01 0.00000000e+00 3.42944349e-06 9.25949742e-05]\\n [0.00000000e+00 9.47217828e-01 ...   \n",
       "56  [[9.99494157e-01 0.00000000e+00 2.22913827e-05 4.83551532e-04]\\n [0.00000000e+00 9.70213364e-01 ...   \n",
       "57  [[9.99123777e-01 0.00000000e+00 1.71472174e-06 8.74508089e-04]\\n [0.00000000e+00 9.06745404e-01 ...   \n",
       "58  [[9.99883399e-01 0.00000000e+00 3.42944349e-06 1.13171635e-04]\\n [1.40602482e-05 9.68153538e-01 ...   \n",
       "59  [[9.99773657e-01 0.00000000e+00 1.02883305e-05 2.16054940e-04]\\n [0.00000000e+00 9.30120567e-01 ...   \n",
       "60  [[9.99754795e-01 0.00000000e+00 8.05919219e-05 1.64613287e-04]\\n [5.62409927e-05 9.11329045e-01 ...   \n",
       "61  [[9.99900546e-01 0.00000000e+00 1.71472174e-05 8.23066437e-05]\\n [0.00000000e+00 9.35934479e-01 ...   \n",
       "62  [[9.99933126e-01 0.00000000e+00 6.85888697e-06 6.00152610e-05]\\n [0.00000000e+00 9.70923407e-01 ...   \n",
       "63  [[9.99816525e-01 0.00000000e+00 3.42944349e-06 1.80045783e-04]\\n [2.10903722e-05 9.45509508e-01 ...   \n",
       "64  [[9.99850819e-01 0.00000000e+00 5.14416523e-06 1.44036626e-04]\\n [0.00000000e+00 9.68153538e-01 ...   \n",
       "65  [[9.99866252e-01 0.00000000e+00 6.85888697e-06 1.26889409e-04]\\n [4.21807445e-05 9.23758304e-01 ...   \n",
       "66  [[9.99746221e-01 0.00000000e+00 0.00000000e+00 2.53778818e-04]\\n [0.00000000e+00 9.06921157e-01 ...   \n",
       "67  [[9.99893687e-01 0.00000000e+00 5.14416523e-06 1.01168583e-04]\\n [0.00000000e+00 9.29551127e-01 ...   \n",
       "68  [[9.99876540e-01 0.00000000e+00 3.42944349e-06 1.20030522e-04]\\n [3.51506204e-05 9.53123133e-01 ...   \n",
       "69  [[9.96752317e-01 1.44036626e-04 3.42944349e-06 3.10021691e-03]\\n [3.51506204e-05 9.65411790e-01 ...   \n",
       "70  [[9.99833672e-01 0.00000000e+00 5.14416523e-06 1.61183844e-04]\\n [7.03012408e-06 9.48581672e-01 ...   \n",
       "71  [[9.99804522e-01 0.00000000e+00 1.54324957e-05 1.80045783e-04]\\n [4.92108686e-05 9.01198636e-01 ...   \n",
       "72  [[9.99879969e-01 0.00000000e+00 0.00000000e+00 1.20030522e-04]\\n [7.03012408e-06 9.72322401e-01 ...   \n",
       "\n",
       "                                                                                                      8  \n",
       "0   [array([[9.99829622e-01, 0.00000000e+00, 7.67470310e-06, 1.62703706e-04],\\n       [0.00000000e+0...  \n",
       "1   [array([[9.99889484e-01, 0.00000000e+00, 4.75831592e-05, 6.29325655e-05],\\n       [0.00000000e+0...  \n",
       "2   [array([[9.99926323e-01, 0.00000000e+00, 1.38144656e-05, 5.98626842e-05],\\n       [7.36713374e-0...  \n",
       "3   [array([[9.99930928e-01, 1.53494062e-06, 6.13976248e-06, 6.13976248e-05],\\n       [0.00000000e+0...  \n",
       "4   [array([[9.99940137e-01, 0.00000000e+00, 9.20964372e-06, 5.06530405e-05],\\n       [7.36713374e-0...  \n",
       "5   [array([[9.99831157e-01, 0.00000000e+00, 4.29783374e-05, 1.25865131e-04],\\n       [0.00000000e+0...  \n",
       "6   [array([[9.99880275e-01, 0.00000000e+00, 0.00000000e+00, 1.19725368e-04],\\n       [0.00000000e+0...  \n",
       "7   [array([[9.99920183e-01, 0.00000000e+00, 3.22337530e-05, 4.75831592e-05],\\n       [0.00000000e+0...  \n",
       "8   [array([[9.99745686e-01, 0.00000000e+00, 1.65138857e-06, 2.52662451e-04],\\n       [0.00000000e+0...  \n",
       "9   [array([[9.99734126e-01, 0.00000000e+00, 3.30277714e-05, 2.32845788e-04],\\n       [0.00000000e+0...  \n",
       "10  [array([[9.99745686e-01, 0.00000000e+00, 0.00000000e+00, 2.54313840e-04],\\n       [0.00000000e+0...  \n",
       "11  [array([[9.99937067e-01, 0.00000000e+00, 1.22795250e-05, 5.06530405e-05],\\n       [0.00000000e+0...  \n",
       "12  [array([[9.99969301e-01, 0.00000000e+00, 4.60482186e-06, 2.60939906e-05],\\n       [0.00000000e+0...  \n",
       "13  [array([[9.99867995e-01, 0.00000000e+00, 5.83277436e-05, 7.36771498e-05],\\n       [0.00000000e+0...  \n",
       "14  [array([[9.29358963e-01, 5.35694277e-04, 2.86005486e-02, 4.15047944e-02],\\n       [0.00000000e+0...  \n",
       "15  [array([[9.99664808e-01, 0.00000000e+00, 1.94840685e-04, 1.40351341e-04],\\n       [7.34824046e-0...  \n",
       "16  [array([[9.99641691e-01, 0.00000000e+00, 1.02373919e-04, 2.55934798e-04],\\n       [0.00000000e+0...  \n",
       "17  [array([[9.99640040e-01, 0.00000000e+00, 3.63262294e-05, 3.23633680e-04],\\n       [0.00000000e+0...  \n",
       "18  [array([[9.99701134e-01, 0.00000000e+00, 1.33746572e-04, 1.65119224e-04],\\n       [0.00000000e+0...  \n",
       "19  [array([[9.99717646e-01, 0.00000000e+00, 1.89887108e-04, 9.24667656e-05],\\n       [2.20447214e-0...  \n",
       "20  [array([[9.99638389e-01, 0.00000000e+00, 4.95357673e-05, 3.12075334e-04],\\n       [0.00000000e+0...  \n",
       "21  [array([[9.99625179e-01, 0.00000000e+00, 1.65119224e-05, 3.58308717e-04],\\n       [0.00000000e+0...  \n",
       "22  [array([[9.99631784e-01, 0.00000000e+00, 2.80702681e-05, 3.40145602e-04],\\n       [0.00000000e+0...  \n",
       "23  [array([[9.99631784e-01, 0.00000000e+00, 7.26524587e-05, 2.95563412e-04],\\n       [0.00000000e+0...  \n",
       "24  [array([[9.96169234e-01, 0.00000000e+00, 4.39217137e-04, 3.39154887e-03],\\n       [0.00000000e+0...  \n",
       "25  [[9.99527759e-01 0.00000000e+00 2.31166914e-05 4.49124290e-04]\\n [0.00000000e+00 9.29236444e-01 ...  \n",
       "26  [[9.99725902e-01 0.00000000e+00 4.29309983e-05 2.31166914e-04]\\n [6.61341642e-05 9.07684055e-01 ...  \n",
       "27  [[9.99679669e-01 0.00000000e+00 2.41074068e-04 7.92572277e-05]\\n [1.39616569e-04 9.14936768e-01 ...  \n",
       "28  [[9.99678018e-01 0.00000000e+00 3.30238449e-06 3.18680103e-04]\\n [0.00000000e+00 9.34791714e-01 ...  \n",
       "29  [[9.99658203e-01 0.00000000e+00 1.93189492e-04 1.48607302e-04]\\n [0.00000000e+00 9.55638672e-01 ...  \n",
       "..                                                                                                  ...  \n",
       "43  [[9.99572341e-01 0.00000000e+00 9.90715346e-06 4.17751638e-04]\\n [7.34824046e-06 9.06478944e-01 ...  \n",
       "44  [[9.99408873e-01 0.00000000e+00 0.00000000e+00 5.91126823e-04]\\n [1.98402493e-04 9.46615033e-01 ...  \n",
       "45  [[9.99058820e-01 1.65119224e-06 2.74097912e-04 6.65430474e-04]\\n [3.67412023e-05 9.69423971e-01 ...  \n",
       "46  [[9.99128170e-01 0.00000000e+00 8.42108044e-05 7.87618700e-04]\\n [6.61341642e-05 9.00681182e-01 ...  \n",
       "47  [[9.99644994e-01 0.00000000e+00 1.48607302e-05 3.40145602e-04]\\n [7.34824046e-06 9.38083726e-01 ...  \n",
       "48  [[9.99223940e-01 0.00000000e+00 1.79979955e-04 5.96080400e-04]\\n [1.46964809e-05 9.08169039e-01 ...  \n",
       "49  [[9.98198549e-01 0.00000000e+00 4.72240982e-04 1.32920976e-03]\\n [7.34824046e-06 8.92392367e-01 ...  \n",
       "50  [[9.98687302e-01 0.00000000e+00 6.30755437e-04 6.81942397e-04]\\n [1.46964809e-05 9.04392043e-01 ...  \n",
       "51  [[9.99055518e-01 0.00000000e+00 3.99588523e-04 5.44893440e-04]\\n [1.46964809e-05 8.45268101e-01 ...  \n",
       "52  [[9.99653250e-01 0.00000000e+00 1.20537034e-04 2.26213337e-04]\\n [2.42491935e-04 8.59134230e-01 ...  \n",
       "53  [[9.99483177e-01 0.00000000e+00 2.31166914e-05 4.93706481e-04]\\n [0.00000000e+00 8.29417946e-01 ...  \n",
       "54  [[9.99727553e-01 0.00000000e+00 1.00722727e-04 1.71723993e-04]\\n [7.34824046e-06 9.43477334e-01 ...  \n",
       "55  [[9.99687925e-01 0.00000000e+00 8.25596122e-05 2.29515722e-04]\\n [0.00000000e+00 9.61333559e-01 ...  \n",
       "56  [[9.99124868e-01 0.00000000e+00 1.78328762e-04 6.96803127e-04]\\n [1.46964809e-05 9.78183074e-01 ...  \n",
       "57  [[9.98928376e-01 0.00000000e+00 6.60476897e-06 1.06501900e-03]\\n [1.46964809e-05 9.24445392e-01 ...  \n",
       "58  [[9.99654901e-01 0.00000000e+00 1.63468032e-04 1.81631147e-04]\\n [3.67412023e-05 9.66036433e-01 ...  \n",
       "59  [[9.99529410e-01 0.00000000e+00 1.93189492e-04 2.77400297e-04]\\n [2.20447214e-05 9.34512481e-01 ...  \n",
       "60  [[9.99644994e-01 0.00000000e+00 2.47678837e-04 1.07327496e-04]\\n [2.20447214e-05 9.26017915e-01 ...  \n",
       "61  [[9.99692878e-01 0.00000000e+00 1.33746572e-04 1.73375186e-04]\\n [7.34824046e-06 9.42757207e-01 ...  \n",
       "62  [[9.99752321e-01 0.00000000e+00 1.07327496e-04 1.40351341e-04]\\n [2.20447214e-05 9.78161029e-01 ...  \n",
       "63  [[9.99615272e-01 0.00000000e+00 6.60476897e-06 3.78123024e-04]\\n [5.14376832e-05 9.55116947e-01 ...  \n",
       "64  [[9.99623528e-01 0.00000000e+00 3.63262294e-05 3.40145602e-04]\\n [0.00000000e+00 9.71709274e-01 ...  \n",
       "65  [[9.99631784e-01 0.00000000e+00 3.13726526e-05 3.36843218e-04]\\n [1.46964809e-05 9.30529735e-01 ...  \n",
       "66  [[9.99549225e-01 0.00000000e+00 6.60476897e-06 4.44170713e-04]\\n [1.46964809e-05 9.23012485e-01 ...  \n",
       "67  [[9.99752321e-01 0.00000000e+00 1.65119224e-06 2.46027644e-04]\\n [0.00000000e+00 9.50950495e-01 ...  \n",
       "68  [[9.99663157e-01 0.00000000e+00 1.13932265e-04 2.22910953e-04]\\n [2.93929619e-05 9.54044104e-01 ...  \n",
       "69  [[9.96552311e-01 1.42002533e-04 6.60476897e-06 3.29908210e-03]\\n [8.81788856e-05 9.74986589e-01 ...  \n",
       "70  [[9.99628482e-01 0.00000000e+00 6.60476897e-06 3.64913486e-04]\\n [7.34824046e-06 9.49524936e-01 ...  \n",
       "71  [[9.99618575e-01 0.00000000e+00 2.14654992e-05 3.59959909e-04]\\n [4.40894428e-05 9.05744120e-01 ...  \n",
       "72  [[9.99673064e-01 0.00000000e+00 3.79774216e-05 2.88958643e-04]\\n [2.20447214e-05 9.77661349e-01 ...  \n",
       "\n",
       "[73 rows x 21 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "eval_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "inference_eval('10210132','/proj/NIRAL/users/siyangj/model_10210132/inference_60000',\n",
    "               '/proj/NIRAL/users/siyangj/myData/min_normal/',eval_df,'1')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "eval_df_sort = eval_df.sort_values('test_11+22',ascending=False).iloc[:,0:13]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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>model</th>\n",
       "      <th>inf_dir</th>\n",
       "      <th>test_im</th>\n",
       "      <th>test_11</th>\n",
       "      <th>test_12</th>\n",
       "      <th>test_21</th>\n",
       "      <th>test_22</th>\n",
       "      <th>test_11+22</th>\n",
       "      <th>test_12+21</th>\n",
       "      <th>Dice0</th>\n",
       "      <th>Dice1</th>\n",
       "      <th>Dice2</th>\n",
       "      <th>Dice3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>10112055.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112055/inference_21900_normal</td>\n",
       "      <td>1</td>\n",
       "      <td>0.861763</td>\n",
       "      <td>0.124861</td>\n",
       "      <td>0.121004</td>\n",
       "      <td>0.822459</td>\n",
       "      <td>1.684223</td>\n",
       "      <td>0.245865</td>\n",
       "      <td>0.991524</td>\n",
       "      <td>0.843291</td>\n",
       "      <td>0.846411</td>\n",
       "      <td>0.833207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>10112055.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112055/inference_21900_10112055</td>\n",
       "      <td>1</td>\n",
       "      <td>0.861016</td>\n",
       "      <td>0.125160</td>\n",
       "      <td>0.120896</td>\n",
       "      <td>0.821308</td>\n",
       "      <td>1.682324</td>\n",
       "      <td>0.246056</td>\n",
       "      <td>0.995358</td>\n",
       "      <td>0.843021</td>\n",
       "      <td>0.846834</td>\n",
       "      <td>0.845988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>10210050.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210050/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.850768</td>\n",
       "      <td>0.138990</td>\n",
       "      <td>0.139546</td>\n",
       "      <td>0.831239</td>\n",
       "      <td>1.682007</td>\n",
       "      <td>0.278536</td>\n",
       "      <td>0.992319</td>\n",
       "      <td>0.825083</td>\n",
       "      <td>0.831247</td>\n",
       "      <td>0.809081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>10210130.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_55000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.788889</td>\n",
       "      <td>0.194381</td>\n",
       "      <td>0.081139</td>\n",
       "      <td>0.883951</td>\n",
       "      <td>1.672840</td>\n",
       "      <td>0.275520</td>\n",
       "      <td>0.994809</td>\n",
       "      <td>0.827094</td>\n",
       "      <td>0.851354</td>\n",
       "      <td>0.843518</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>10210132.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_70000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.861237</td>\n",
       "      <td>0.128304</td>\n",
       "      <td>0.149105</td>\n",
       "      <td>0.802642</td>\n",
       "      <td>1.663880</td>\n",
       "      <td>0.277410</td>\n",
       "      <td>0.995638</td>\n",
       "      <td>0.825284</td>\n",
       "      <td>0.824976</td>\n",
       "      <td>0.832259</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>10210054.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210054/inference_50000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.829719</td>\n",
       "      <td>0.155244</td>\n",
       "      <td>0.121035</td>\n",
       "      <td>0.832751</td>\n",
       "      <td>1.662470</td>\n",
       "      <td>0.276280</td>\n",
       "      <td>0.995374</td>\n",
       "      <td>0.823699</td>\n",
       "      <td>0.834863</td>\n",
       "      <td>0.831108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>10210132</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_60000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.840177</td>\n",
       "      <td>0.136249</td>\n",
       "      <td>0.108549</td>\n",
       "      <td>0.813160</td>\n",
       "      <td>1.653337</td>\n",
       "      <td>0.244798</td>\n",
       "      <td>0.991953</td>\n",
       "      <td>0.839596</td>\n",
       "      <td>0.841971</td>\n",
       "      <td>0.827433</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>10210050.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210050/inference_60000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.854372</td>\n",
       "      <td>0.132274</td>\n",
       "      <td>0.148030</td>\n",
       "      <td>0.797729</td>\n",
       "      <td>1.652101</td>\n",
       "      <td>0.280304</td>\n",
       "      <td>0.995481</td>\n",
       "      <td>0.820110</td>\n",
       "      <td>0.825959</td>\n",
       "      <td>0.833727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>10210052.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210052/inference_55400</td>\n",
       "      <td>1</td>\n",
       "      <td>0.862876</td>\n",
       "      <td>0.127103</td>\n",
       "      <td>0.177458</td>\n",
       "      <td>0.788410</td>\n",
       "      <td>1.651286</td>\n",
       "      <td>0.304561</td>\n",
       "      <td>0.995355</td>\n",
       "      <td>0.807689</td>\n",
       "      <td>0.810778</td>\n",
       "      <td>0.819124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>10210050.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210050/inference_47700</td>\n",
       "      <td>1</td>\n",
       "      <td>0.787571</td>\n",
       "      <td>0.184371</td>\n",
       "      <td>0.088806</td>\n",
       "      <td>0.862547</td>\n",
       "      <td>1.650118</td>\n",
       "      <td>0.273177</td>\n",
       "      <td>0.992172</td>\n",
       "      <td>0.821132</td>\n",
       "      <td>0.846978</td>\n",
       "      <td>0.825889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>10210132.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_55000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.788230</td>\n",
       "      <td>0.190931</td>\n",
       "      <td>0.072818</td>\n",
       "      <td>0.861572</td>\n",
       "      <td>1.649802</td>\n",
       "      <td>0.263749</td>\n",
       "      <td>0.995917</td>\n",
       "      <td>0.831604</td>\n",
       "      <td>0.848871</td>\n",
       "      <td>0.843631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10112050.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112050/inference</td>\n",
       "      <td>1</td>\n",
       "      <td>0.895175</td>\n",
       "      <td>0.097146</td>\n",
       "      <td>0.224378</td>\n",
       "      <td>0.753510</td>\n",
       "      <td>1.648685</td>\n",
       "      <td>0.321524</td>\n",
       "      <td>0.991811</td>\n",
       "      <td>0.803432</td>\n",
       "      <td>0.796687</td>\n",
       "      <td>0.804782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>10210054.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210054/inference_57300</td>\n",
       "      <td>1</td>\n",
       "      <td>0.794176</td>\n",
       "      <td>0.185833</td>\n",
       "      <td>0.093085</td>\n",
       "      <td>0.852272</td>\n",
       "      <td>1.646448</td>\n",
       "      <td>0.278917</td>\n",
       "      <td>0.995408</td>\n",
       "      <td>0.821614</td>\n",
       "      <td>0.839846</td>\n",
       "      <td>0.833980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10112056.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112056/inference_16500_10112056</td>\n",
       "      <td>1</td>\n",
       "      <td>0.793855</td>\n",
       "      <td>0.189398</td>\n",
       "      <td>0.090426</td>\n",
       "      <td>0.849576</td>\n",
       "      <td>1.643431</td>\n",
       "      <td>0.279823</td>\n",
       "      <td>0.990101</td>\n",
       "      <td>0.825034</td>\n",
       "      <td>0.838850</td>\n",
       "      <td>0.803951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>10210132.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_45000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.798356</td>\n",
       "      <td>0.177805</td>\n",
       "      <td>0.075448</td>\n",
       "      <td>0.843776</td>\n",
       "      <td>1.642132</td>\n",
       "      <td>0.253253</td>\n",
       "      <td>0.994820</td>\n",
       "      <td>0.835899</td>\n",
       "      <td>0.845961</td>\n",
       "      <td>0.837082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>10210130.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_45000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.840122</td>\n",
       "      <td>0.147062</td>\n",
       "      <td>0.156935</td>\n",
       "      <td>0.799764</td>\n",
       "      <td>1.639886</td>\n",
       "      <td>0.303997</td>\n",
       "      <td>0.991060</td>\n",
       "      <td>0.809171</td>\n",
       "      <td>0.816202</td>\n",
       "      <td>0.808539</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>10210136.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210136/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.839430</td>\n",
       "      <td>0.149675</td>\n",
       "      <td>0.145136</td>\n",
       "      <td>0.800236</td>\n",
       "      <td>1.639666</td>\n",
       "      <td>0.294811</td>\n",
       "      <td>0.995972</td>\n",
       "      <td>0.815201</td>\n",
       "      <td>0.816830</td>\n",
       "      <td>0.823648</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>10210052.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210052/inference_70000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.795792</td>\n",
       "      <td>0.186408</td>\n",
       "      <td>0.098338</td>\n",
       "      <td>0.843807</td>\n",
       "      <td>1.639600</td>\n",
       "      <td>0.284747</td>\n",
       "      <td>0.996155</td>\n",
       "      <td>0.818509</td>\n",
       "      <td>0.840425</td>\n",
       "      <td>0.852010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>10202102.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202102/inference_11600</td>\n",
       "      <td>1</td>\n",
       "      <td>0.810325</td>\n",
       "      <td>0.179549</td>\n",
       "      <td>0.136749</td>\n",
       "      <td>0.828797</td>\n",
       "      <td>1.639123</td>\n",
       "      <td>0.316298</td>\n",
       "      <td>0.995893</td>\n",
       "      <td>0.802237</td>\n",
       "      <td>0.815528</td>\n",
       "      <td>0.807952</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>10112056.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112056/inference</td>\n",
       "      <td>1</td>\n",
       "      <td>0.845575</td>\n",
       "      <td>0.147383</td>\n",
       "      <td>0.184313</td>\n",
       "      <td>0.789043</td>\n",
       "      <td>1.634618</td>\n",
       "      <td>0.331696</td>\n",
       "      <td>0.997961</td>\n",
       "      <td>0.792166</td>\n",
       "      <td>0.803326</td>\n",
       "      <td>0.819554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>10210052.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210052/inference_20000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.841932</td>\n",
       "      <td>0.141248</td>\n",
       "      <td>0.142003</td>\n",
       "      <td>0.788619</td>\n",
       "      <td>1.630551</td>\n",
       "      <td>0.283251</td>\n",
       "      <td>0.996288</td>\n",
       "      <td>0.818332</td>\n",
       "      <td>0.814715</td>\n",
       "      <td>0.815823</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>10112055.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112055/inference</td>\n",
       "      <td>1</td>\n",
       "      <td>0.824332</td>\n",
       "      <td>0.159098</td>\n",
       "      <td>0.141585</td>\n",
       "      <td>0.805546</td>\n",
       "      <td>1.629878</td>\n",
       "      <td>0.300683</td>\n",
       "      <td>0.994605</td>\n",
       "      <td>0.808893</td>\n",
       "      <td>0.818492</td>\n",
       "      <td>0.810333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>10192004.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_6600</td>\n",
       "      <td>1</td>\n",
       "      <td>0.769910</td>\n",
       "      <td>0.198151</td>\n",
       "      <td>0.084447</td>\n",
       "      <td>0.859092</td>\n",
       "      <td>1.629002</td>\n",
       "      <td>0.282597</td>\n",
       "      <td>0.995134</td>\n",
       "      <td>0.813365</td>\n",
       "      <td>0.843567</td>\n",
       "      <td>0.837692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>10210052.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210052/inference_10000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.847114</td>\n",
       "      <td>0.142494</td>\n",
       "      <td>0.188487</td>\n",
       "      <td>0.776054</td>\n",
       "      <td>1.623168</td>\n",
       "      <td>0.330981</td>\n",
       "      <td>0.994722</td>\n",
       "      <td>0.796587</td>\n",
       "      <td>0.780807</td>\n",
       "      <td>0.751619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>10192002.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192002/inference_4600</td>\n",
       "      <td>1</td>\n",
       "      <td>0.751673</td>\n",
       "      <td>0.210691</td>\n",
       "      <td>0.083541</td>\n",
       "      <td>0.868307</td>\n",
       "      <td>1.619980</td>\n",
       "      <td>0.294232</td>\n",
       "      <td>0.993344</td>\n",
       "      <td>0.802652</td>\n",
       "      <td>0.835804</td>\n",
       "      <td>0.798312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>10112103.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112103/inference_12300_10112103</td>\n",
       "      <td>1</td>\n",
       "      <td>0.809434</td>\n",
       "      <td>0.176941</td>\n",
       "      <td>0.130890</td>\n",
       "      <td>0.809682</td>\n",
       "      <td>1.619116</td>\n",
       "      <td>0.307831</td>\n",
       "      <td>0.992859</td>\n",
       "      <td>0.807523</td>\n",
       "      <td>0.819547</td>\n",
       "      <td>0.813614</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>10202100.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202100/inference_2000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.801738</td>\n",
       "      <td>0.181940</td>\n",
       "      <td>0.145419</td>\n",
       "      <td>0.815258</td>\n",
       "      <td>1.616996</td>\n",
       "      <td>0.327359</td>\n",
       "      <td>0.993349</td>\n",
       "      <td>0.794069</td>\n",
       "      <td>0.807127</td>\n",
       "      <td>0.788786</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>10202100.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202100/inference_5800</td>\n",
       "      <td>1</td>\n",
       "      <td>0.800781</td>\n",
       "      <td>0.179947</td>\n",
       "      <td>0.113441</td>\n",
       "      <td>0.809683</td>\n",
       "      <td>1.610464</td>\n",
       "      <td>0.293389</td>\n",
       "      <td>0.991585</td>\n",
       "      <td>0.811223</td>\n",
       "      <td>0.825079</td>\n",
       "      <td>0.821793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>10210054.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210054/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.766677</td>\n",
       "      <td>0.198644</td>\n",
       "      <td>0.095696</td>\n",
       "      <td>0.842744</td>\n",
       "      <td>1.609420</td>\n",
       "      <td>0.294340</td>\n",
       "      <td>0.995408</td>\n",
       "      <td>0.804334</td>\n",
       "      <td>0.833286</td>\n",
       "      <td>0.827636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>10210134.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210134/inference_22000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.732927</td>\n",
       "      <td>0.227139</td>\n",
       "      <td>0.079762</td>\n",
       "      <td>0.874268</td>\n",
       "      <td>1.607196</td>\n",
       "      <td>0.306901</td>\n",
       "      <td>0.995883</td>\n",
       "      <td>0.792475</td>\n",
       "      <td>0.832953</td>\n",
       "      <td>0.808690</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>10202104.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202104/inference_10000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.857190</td>\n",
       "      <td>0.128941</td>\n",
       "      <td>0.210557</td>\n",
       "      <td>0.731070</td>\n",
       "      <td>1.588261</td>\n",
       "      <td>0.339498</td>\n",
       "      <td>0.994264</td>\n",
       "      <td>0.790629</td>\n",
       "      <td>0.776794</td>\n",
       "      <td>0.798461</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>10192004.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_4000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.783851</td>\n",
       "      <td>0.174810</td>\n",
       "      <td>0.108014</td>\n",
       "      <td>0.803988</td>\n",
       "      <td>1.587839</td>\n",
       "      <td>0.282823</td>\n",
       "      <td>0.995970</td>\n",
       "      <td>0.807650</td>\n",
       "      <td>0.824122</td>\n",
       "      <td>0.823838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>10210056.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210056/inference_10000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.736947</td>\n",
       "      <td>0.245769</td>\n",
       "      <td>0.106176</td>\n",
       "      <td>0.848984</td>\n",
       "      <td>1.585930</td>\n",
       "      <td>0.351945</td>\n",
       "      <td>0.991282</td>\n",
       "      <td>0.778757</td>\n",
       "      <td>0.799489</td>\n",
       "      <td>0.742067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>10192002.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192002/inference_5700</td>\n",
       "      <td>1</td>\n",
       "      <td>0.797880</td>\n",
       "      <td>0.149403</td>\n",
       "      <td>0.140641</td>\n",
       "      <td>0.786742</td>\n",
       "      <td>1.584622</td>\n",
       "      <td>0.290044</td>\n",
       "      <td>0.996218</td>\n",
       "      <td>0.797222</td>\n",
       "      <td>0.815309</td>\n",
       "      <td>0.810516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>10210052.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210052/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.903158</td>\n",
       "      <td>0.080006</td>\n",
       "      <td>0.256125</td>\n",
       "      <td>0.680945</td>\n",
       "      <td>1.584104</td>\n",
       "      <td>0.336130</td>\n",
       "      <td>0.996178</td>\n",
       "      <td>0.789520</td>\n",
       "      <td>0.764866</td>\n",
       "      <td>0.822953</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>10202104.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202104/inference_19600</td>\n",
       "      <td>1</td>\n",
       "      <td>0.773348</td>\n",
       "      <td>0.195172</td>\n",
       "      <td>0.097723</td>\n",
       "      <td>0.808526</td>\n",
       "      <td>1.581874</td>\n",
       "      <td>0.292896</td>\n",
       "      <td>0.995460</td>\n",
       "      <td>0.807557</td>\n",
       "      <td>0.819218</td>\n",
       "      <td>0.813403</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>10210134.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210134/inference_11700</td>\n",
       "      <td>1</td>\n",
       "      <td>0.667045</td>\n",
       "      <td>0.269603</td>\n",
       "      <td>0.032461</td>\n",
       "      <td>0.912339</td>\n",
       "      <td>1.579384</td>\n",
       "      <td>0.302064</td>\n",
       "      <td>0.995689</td>\n",
       "      <td>0.777983</td>\n",
       "      <td>0.844907</td>\n",
       "      <td>0.795348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>10192002.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192002/inference_2500</td>\n",
       "      <td>1</td>\n",
       "      <td>0.776913</td>\n",
       "      <td>0.212756</td>\n",
       "      <td>0.153442</td>\n",
       "      <td>0.801818</td>\n",
       "      <td>1.578732</td>\n",
       "      <td>0.366198</td>\n",
       "      <td>0.993348</td>\n",
       "      <td>0.776299</td>\n",
       "      <td>0.787699</td>\n",
       "      <td>0.784175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10112127.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112127/inference_8700_10112127</td>\n",
       "      <td>1</td>\n",
       "      <td>0.713201</td>\n",
       "      <td>0.271862</td>\n",
       "      <td>0.050931</td>\n",
       "      <td>0.864683</td>\n",
       "      <td>1.577885</td>\n",
       "      <td>0.322792</td>\n",
       "      <td>0.995328</td>\n",
       "      <td>0.797492</td>\n",
       "      <td>0.828315</td>\n",
       "      <td>0.831142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10112050.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112050/inference_13200_10112050</td>\n",
       "      <td>1</td>\n",
       "      <td>0.926572</td>\n",
       "      <td>0.058502</td>\n",
       "      <td>0.308779</td>\n",
       "      <td>0.649881</td>\n",
       "      <td>1.576453</td>\n",
       "      <td>0.367281</td>\n",
       "      <td>0.994202</td>\n",
       "      <td>0.777533</td>\n",
       "      <td>0.752149</td>\n",
       "      <td>0.844704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>10202100.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202100/inference_12000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.727369</td>\n",
       "      <td>0.244711</td>\n",
       "      <td>0.052268</td>\n",
       "      <td>0.846252</td>\n",
       "      <td>1.573621</td>\n",
       "      <td>0.296979</td>\n",
       "      <td>0.996687</td>\n",
       "      <td>0.806246</td>\n",
       "      <td>0.830782</td>\n",
       "      <td>0.827865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>10210054.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210054/inference_18900</td>\n",
       "      <td>1</td>\n",
       "      <td>0.867145</td>\n",
       "      <td>0.109188</td>\n",
       "      <td>0.208108</td>\n",
       "      <td>0.705932</td>\n",
       "      <td>1.573077</td>\n",
       "      <td>0.317296</td>\n",
       "      <td>0.993086</td>\n",
       "      <td>0.797152</td>\n",
       "      <td>0.782825</td>\n",
       "      <td>0.825374</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>10210056.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210056/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.858918</td>\n",
       "      <td>0.123538</td>\n",
       "      <td>0.207265</td>\n",
       "      <td>0.712497</td>\n",
       "      <td>1.571415</td>\n",
       "      <td>0.330803</td>\n",
       "      <td>0.993320</td>\n",
       "      <td>0.792416</td>\n",
       "      <td>0.782156</td>\n",
       "      <td>0.833675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10112050.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112050/inference_17000_10112050</td>\n",
       "      <td>1</td>\n",
       "      <td>0.918915</td>\n",
       "      <td>0.067399</td>\n",
       "      <td>0.295044</td>\n",
       "      <td>0.649303</td>\n",
       "      <td>1.568218</td>\n",
       "      <td>0.362443</td>\n",
       "      <td>0.993768</td>\n",
       "      <td>0.780229</td>\n",
       "      <td>0.744064</td>\n",
       "      <td>0.805368</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>10192000.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192000/inference_7200</td>\n",
       "      <td>1</td>\n",
       "      <td>0.882287</td>\n",
       "      <td>0.095098</td>\n",
       "      <td>0.265452</td>\n",
       "      <td>0.683058</td>\n",
       "      <td>1.565344</td>\n",
       "      <td>0.360550</td>\n",
       "      <td>0.995164</td>\n",
       "      <td>0.777106</td>\n",
       "      <td>0.762458</td>\n",
       "      <td>0.836195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10112056.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112056/inference_18100_10112056</td>\n",
       "      <td>1</td>\n",
       "      <td>0.825539</td>\n",
       "      <td>0.124662</td>\n",
       "      <td>0.174211</td>\n",
       "      <td>0.737446</td>\n",
       "      <td>1.562985</td>\n",
       "      <td>0.298872</td>\n",
       "      <td>0.992718</td>\n",
       "      <td>0.794706</td>\n",
       "      <td>0.796743</td>\n",
       "      <td>0.786574</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>10210130.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_20000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.687646</td>\n",
       "      <td>0.279961</td>\n",
       "      <td>0.035215</td>\n",
       "      <td>0.873425</td>\n",
       "      <td>1.561071</td>\n",
       "      <td>0.315177</td>\n",
       "      <td>0.996453</td>\n",
       "      <td>0.791053</td>\n",
       "      <td>0.833452</td>\n",
       "      <td>0.826392</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>10210136.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210136/inference_17800</td>\n",
       "      <td>1</td>\n",
       "      <td>0.688659</td>\n",
       "      <td>0.278057</td>\n",
       "      <td>0.038709</td>\n",
       "      <td>0.861727</td>\n",
       "      <td>1.550385</td>\n",
       "      <td>0.316765</td>\n",
       "      <td>0.995060</td>\n",
       "      <td>0.789468</td>\n",
       "      <td>0.826526</td>\n",
       "      <td>0.808840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>10192004.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_3900</td>\n",
       "      <td>1</td>\n",
       "      <td>0.856205</td>\n",
       "      <td>0.096011</td>\n",
       "      <td>0.211923</td>\n",
       "      <td>0.685681</td>\n",
       "      <td>1.541885</td>\n",
       "      <td>0.307934</td>\n",
       "      <td>0.994825</td>\n",
       "      <td>0.791488</td>\n",
       "      <td>0.773357</td>\n",
       "      <td>0.810508</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>10112127.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112127/inference_13500_10112127</td>\n",
       "      <td>1</td>\n",
       "      <td>0.737849</td>\n",
       "      <td>0.235875</td>\n",
       "      <td>0.069771</td>\n",
       "      <td>0.803042</td>\n",
       "      <td>1.540891</td>\n",
       "      <td>0.305646</td>\n",
       "      <td>0.996067</td>\n",
       "      <td>0.802167</td>\n",
       "      <td>0.808498</td>\n",
       "      <td>0.805656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>10210132.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_6200</td>\n",
       "      <td>1</td>\n",
       "      <td>0.658901</td>\n",
       "      <td>0.298491</td>\n",
       "      <td>0.048867</td>\n",
       "      <td>0.873030</td>\n",
       "      <td>1.531931</td>\n",
       "      <td>0.347359</td>\n",
       "      <td>0.992602</td>\n",
       "      <td>0.761938</td>\n",
       "      <td>0.822676</td>\n",
       "      <td>0.801886</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>10210054.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210054/inference_10000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.612606</td>\n",
       "      <td>0.369583</td>\n",
       "      <td>0.025029</td>\n",
       "      <td>0.916149</td>\n",
       "      <td>1.528755</td>\n",
       "      <td>0.394613</td>\n",
       "      <td>0.995426</td>\n",
       "      <td>0.742947</td>\n",
       "      <td>0.825055</td>\n",
       "      <td>0.837594</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10112130.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112130/inference_7300</td>\n",
       "      <td>1</td>\n",
       "      <td>0.712039</td>\n",
       "      <td>0.238079</td>\n",
       "      <td>0.062735</td>\n",
       "      <td>0.804453</td>\n",
       "      <td>1.516491</td>\n",
       "      <td>0.300814</td>\n",
       "      <td>0.993799</td>\n",
       "      <td>0.789505</td>\n",
       "      <td>0.808477</td>\n",
       "      <td>0.781285</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>10192004.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_2900</td>\n",
       "      <td>1</td>\n",
       "      <td>0.717924</td>\n",
       "      <td>0.221304</td>\n",
       "      <td>0.074288</td>\n",
       "      <td>0.791369</td>\n",
       "      <td>1.509293</td>\n",
       "      <td>0.295592</td>\n",
       "      <td>0.990993</td>\n",
       "      <td>0.786643</td>\n",
       "      <td>0.805821</td>\n",
       "      <td>0.768787</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>10210050.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210050/inference_10000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.578580</td>\n",
       "      <td>0.393927</td>\n",
       "      <td>0.014724</td>\n",
       "      <td>0.908180</td>\n",
       "      <td>1.486760</td>\n",
       "      <td>0.408650</td>\n",
       "      <td>0.995960</td>\n",
       "      <td>0.723195</td>\n",
       "      <td>0.817109</td>\n",
       "      <td>0.825188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>10210136.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210136/inference_8800</td>\n",
       "      <td>1</td>\n",
       "      <td>0.650680</td>\n",
       "      <td>0.270954</td>\n",
       "      <td>0.061796</td>\n",
       "      <td>0.815973</td>\n",
       "      <td>1.466653</td>\n",
       "      <td>0.332750</td>\n",
       "      <td>0.993133</td>\n",
       "      <td>0.748033</td>\n",
       "      <td>0.802822</td>\n",
       "      <td>0.755055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>10202106.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202106/inference_9000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.563482</td>\n",
       "      <td>0.411848</td>\n",
       "      <td>0.021130</td>\n",
       "      <td>0.900106</td>\n",
       "      <td>1.463589</td>\n",
       "      <td>0.432978</td>\n",
       "      <td>0.993331</td>\n",
       "      <td>0.707024</td>\n",
       "      <td>0.806870</td>\n",
       "      <td>0.806477</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>10210134.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210134/inference_4400</td>\n",
       "      <td>1</td>\n",
       "      <td>0.522508</td>\n",
       "      <td>0.421381</td>\n",
       "      <td>0.012391</td>\n",
       "      <td>0.933210</td>\n",
       "      <td>1.455717</td>\n",
       "      <td>0.433772</td>\n",
       "      <td>0.995112</td>\n",
       "      <td>0.678361</td>\n",
       "      <td>0.816866</td>\n",
       "      <td>0.802167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>10202106.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202106/inference_2100</td>\n",
       "      <td>1</td>\n",
       "      <td>0.587643</td>\n",
       "      <td>0.369888</td>\n",
       "      <td>0.033761</td>\n",
       "      <td>0.866562</td>\n",
       "      <td>1.454205</td>\n",
       "      <td>0.403649</td>\n",
       "      <td>0.987312</td>\n",
       "      <td>0.717716</td>\n",
       "      <td>0.798132</td>\n",
       "      <td>0.743359</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>10202104.0</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202104/inference_2300</td>\n",
       "      <td>1</td>\n",
       "      <td>0.532673</td>\n",
       "      <td>0.418215</td>\n",
       "      <td>0.014700</td>\n",
       "      <td>0.865243</td>\n",
       "      <td>1.397916</td>\n",
       "      <td>0.432915</td>\n",
       "      <td>0.995076</td>\n",
       "      <td>0.685615</td>\n",
       "      <td>0.792096</td>\n",
       "      <td>0.784187</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>74 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         model  \\\n",
       "14  10112055.0   \n",
       "12  10112055.0   \n",
       "35  10210050.0   \n",
       "61  10210130.0   \n",
       "72  10210132.0   \n",
       "58  10210054.0   \n",
       "73    10210132   \n",
       "55  10210050.0   \n",
       "56  10210052.0   \n",
       "54  10210050.0   \n",
       "70  10210132.0   \n",
       "1   10112050.0   \n",
       "68  10210054.0   \n",
       "4   10112056.0   \n",
       "63  10210132.0   \n",
       "62  10210130.0   \n",
       "64  10210136.0   \n",
       "57  10210052.0   \n",
       "29  10202102.0   \n",
       "5   10112056.0   \n",
       "37  10210052.0   \n",
       "13  10112055.0   \n",
       "23  10192004.0   \n",
       "39  10210052.0   \n",
       "18  10192002.0   \n",
       "7   10112103.0   \n",
       "27  10202100.0   \n",
       "26  10202100.0   \n",
       "40  10210054.0   \n",
       "50  10210134.0   \n",
       "..         ...   \n",
       "31  10202104.0   \n",
       "21  10192004.0   \n",
       "60  10210056.0   \n",
       "17  10192002.0   \n",
       "38  10210052.0   \n",
       "32  10202104.0   \n",
       "49  10210134.0   \n",
       "19  10192002.0   \n",
       "9   10112127.0   \n",
       "2   10112050.0   \n",
       "25  10202100.0   \n",
       "42  10210054.0   \n",
       "69  10210056.0   \n",
       "0   10112050.0   \n",
       "15  10192000.0   \n",
       "3   10112056.0   \n",
       "43  10210130.0   \n",
       "65  10210136.0   \n",
       "22  10192004.0   \n",
       "8   10112127.0   \n",
       "46  10210132.0   \n",
       "41  10210054.0   \n",
       "10  10112130.0   \n",
       "24  10192004.0   \n",
       "36  10210050.0   \n",
       "66  10210136.0   \n",
       "53  10202106.0   \n",
       "51  10210134.0   \n",
       "52  10202106.0   \n",
       "33  10202104.0   \n",
       "\n",
       "                                                              inf_dir test_im  \\\n",
       "14    /proj/NIRAL/users/siyangj/model_10112055/inference_21900_normal       1   \n",
       "12  /proj/NIRAL/users/siyangj/model_10112055/inference_21900_10112055       1   \n",
       "35           /proj/NIRAL/users/siyangj/model_10210050/inference_30000       1   \n",
       "61           /proj/NIRAL/users/siyangj/model_10210130/inference_55000       1   \n",
       "72           /proj/NIRAL/users/siyangj/model_10210132/inference_70000       1   \n",
       "58           /proj/NIRAL/users/siyangj/model_10210054/inference_50000       1   \n",
       "73           /proj/NIRAL/users/siyangj/model_10210132/inference_60000       1   \n",
       "55           /proj/NIRAL/users/siyangj/model_10210050/inference_60000       1   \n",
       "56           /proj/NIRAL/users/siyangj/model_10210052/inference_55400       1   \n",
       "54           /proj/NIRAL/users/siyangj/model_10210050/inference_47700       1   \n",
       "70           /proj/NIRAL/users/siyangj/model_10210132/inference_55000       1   \n",
       "1                  /proj/NIRAL/users/siyangj/model_10112050/inference       1   \n",
       "68           /proj/NIRAL/users/siyangj/model_10210054/inference_57300       1   \n",
       "4   /proj/NIRAL/users/siyangj/model_10112056/inference_16500_10112056       1   \n",
       "63           /proj/NIRAL/users/siyangj/model_10210132/inference_45000       1   \n",
       "62           /proj/NIRAL/users/siyangj/model_10210130/inference_45000       1   \n",
       "64           /proj/NIRAL/users/siyangj/model_10210136/inference_30000       1   \n",
       "57           /proj/NIRAL/users/siyangj/model_10210052/inference_70000       1   \n",
       "29           /proj/NIRAL/users/siyangj/model_10202102/inference_11600       1   \n",
       "5                  /proj/NIRAL/users/siyangj/model_10112056/inference       1   \n",
       "37           /proj/NIRAL/users/siyangj/model_10210052/inference_20000       1   \n",
       "13                 /proj/NIRAL/users/siyangj/model_10112055/inference       1   \n",
       "23            /proj/NIRAL/users/siyangj/model_10192004/inference_6600       1   \n",
       "39           /proj/NIRAL/users/siyangj/model_10210052/inference_10000       1   \n",
       "18            /proj/NIRAL/users/siyangj/model_10192002/inference_4600       1   \n",
       "7   /proj/NIRAL/users/siyangj/model_10112103/inference_12300_10112103       1   \n",
       "27            /proj/NIRAL/users/siyangj/model_10202100/inference_2000       1   \n",
       "26            /proj/NIRAL/users/siyangj/model_10202100/inference_5800       1   \n",
       "40           /proj/NIRAL/users/siyangj/model_10210054/inference_30000       1   \n",
       "50           /proj/NIRAL/users/siyangj/model_10210134/inference_22000       1   \n",
       "..                                                                ...     ...   \n",
       "31           /proj/NIRAL/users/siyangj/model_10202104/inference_10000       1   \n",
       "21            /proj/NIRAL/users/siyangj/model_10192004/inference_4000       1   \n",
       "60           /proj/NIRAL/users/siyangj/model_10210056/inference_10000       1   \n",
       "17            /proj/NIRAL/users/siyangj/model_10192002/inference_5700       1   \n",
       "38           /proj/NIRAL/users/siyangj/model_10210052/inference_30000       1   \n",
       "32           /proj/NIRAL/users/siyangj/model_10202104/inference_19600       1   \n",
       "49           /proj/NIRAL/users/siyangj/model_10210134/inference_11700       1   \n",
       "19            /proj/NIRAL/users/siyangj/model_10192002/inference_2500       1   \n",
       "9    /proj/NIRAL/users/siyangj/model_10112127/inference_8700_10112127       1   \n",
       "2   /proj/NIRAL/users/siyangj/model_10112050/inference_13200_10112050       1   \n",
       "25           /proj/NIRAL/users/siyangj/model_10202100/inference_12000       1   \n",
       "42           /proj/NIRAL/users/siyangj/model_10210054/inference_18900       1   \n",
       "69           /proj/NIRAL/users/siyangj/model_10210056/inference_30000       1   \n",
       "0   /proj/NIRAL/users/siyangj/model_10112050/inference_17000_10112050       1   \n",
       "15            /proj/NIRAL/users/siyangj/model_10192000/inference_7200       1   \n",
       "3   /proj/NIRAL/users/siyangj/model_10112056/inference_18100_10112056       1   \n",
       "43           /proj/NIRAL/users/siyangj/model_10210130/inference_20000       1   \n",
       "65           /proj/NIRAL/users/siyangj/model_10210136/inference_17800       1   \n",
       "22            /proj/NIRAL/users/siyangj/model_10192004/inference_3900       1   \n",
       "8   /proj/NIRAL/users/siyangj/model_10112127/inference_13500_10112127       1   \n",
       "46            /proj/NIRAL/users/siyangj/model_10210132/inference_6200       1   \n",
       "41           /proj/NIRAL/users/siyangj/model_10210054/inference_10000       1   \n",
       "10            /proj/NIRAL/users/siyangj/model_10112130/inference_7300       1   \n",
       "24            /proj/NIRAL/users/siyangj/model_10192004/inference_2900       1   \n",
       "36           /proj/NIRAL/users/siyangj/model_10210050/inference_10000       1   \n",
       "66            /proj/NIRAL/users/siyangj/model_10210136/inference_8800       1   \n",
       "53            /proj/NIRAL/users/siyangj/model_10202106/inference_9000       1   \n",
       "51            /proj/NIRAL/users/siyangj/model_10210134/inference_4400       1   \n",
       "52            /proj/NIRAL/users/siyangj/model_10202106/inference_2100       1   \n",
       "33            /proj/NIRAL/users/siyangj/model_10202104/inference_2300       1   \n",
       "\n",
       "     test_11   test_12   test_21   test_22  test_11+22  test_12+21     Dice0  \\\n",
       "14  0.861763  0.124861  0.121004  0.822459    1.684223    0.245865  0.991524   \n",
       "12  0.861016  0.125160  0.120896  0.821308    1.682324    0.246056  0.995358   \n",
       "35  0.850768  0.138990  0.139546  0.831239    1.682007    0.278536  0.992319   \n",
       "61  0.788889  0.194381  0.081139  0.883951    1.672840    0.275520  0.994809   \n",
       "72  0.861237  0.128304  0.149105  0.802642    1.663880    0.277410  0.995638   \n",
       "58  0.829719  0.155244  0.121035  0.832751    1.662470    0.276280  0.995374   \n",
       "73  0.840177  0.136249  0.108549  0.813160    1.653337    0.244798  0.991953   \n",
       "55  0.854372  0.132274  0.148030  0.797729    1.652101    0.280304  0.995481   \n",
       "56  0.862876  0.127103  0.177458  0.788410    1.651286    0.304561  0.995355   \n",
       "54  0.787571  0.184371  0.088806  0.862547    1.650118    0.273177  0.992172   \n",
       "70  0.788230  0.190931  0.072818  0.861572    1.649802    0.263749  0.995917   \n",
       "1   0.895175  0.097146  0.224378  0.753510    1.648685    0.321524  0.991811   \n",
       "68  0.794176  0.185833  0.093085  0.852272    1.646448    0.278917  0.995408   \n",
       "4   0.793855  0.189398  0.090426  0.849576    1.643431    0.279823  0.990101   \n",
       "63  0.798356  0.177805  0.075448  0.843776    1.642132    0.253253  0.994820   \n",
       "62  0.840122  0.147062  0.156935  0.799764    1.639886    0.303997  0.991060   \n",
       "64  0.839430  0.149675  0.145136  0.800236    1.639666    0.294811  0.995972   \n",
       "57  0.795792  0.186408  0.098338  0.843807    1.639600    0.284747  0.996155   \n",
       "29  0.810325  0.179549  0.136749  0.828797    1.639123    0.316298  0.995893   \n",
       "5   0.845575  0.147383  0.184313  0.789043    1.634618    0.331696  0.997961   \n",
       "37  0.841932  0.141248  0.142003  0.788619    1.630551    0.283251  0.996288   \n",
       "13  0.824332  0.159098  0.141585  0.805546    1.629878    0.300683  0.994605   \n",
       "23  0.769910  0.198151  0.084447  0.859092    1.629002    0.282597  0.995134   \n",
       "39  0.847114  0.142494  0.188487  0.776054    1.623168    0.330981  0.994722   \n",
       "18  0.751673  0.210691  0.083541  0.868307    1.619980    0.294232  0.993344   \n",
       "7   0.809434  0.176941  0.130890  0.809682    1.619116    0.307831  0.992859   \n",
       "27  0.801738  0.181940  0.145419  0.815258    1.616996    0.327359  0.993349   \n",
       "26  0.800781  0.179947  0.113441  0.809683    1.610464    0.293389  0.991585   \n",
       "40  0.766677  0.198644  0.095696  0.842744    1.609420    0.294340  0.995408   \n",
       "50  0.732927  0.227139  0.079762  0.874268    1.607196    0.306901  0.995883   \n",
       "..       ...       ...       ...       ...         ...         ...       ...   \n",
       "31  0.857190  0.128941  0.210557  0.731070    1.588261    0.339498  0.994264   \n",
       "21  0.783851  0.174810  0.108014  0.803988    1.587839    0.282823  0.995970   \n",
       "60  0.736947  0.245769  0.106176  0.848984    1.585930    0.351945  0.991282   \n",
       "17  0.797880  0.149403  0.140641  0.786742    1.584622    0.290044  0.996218   \n",
       "38  0.903158  0.080006  0.256125  0.680945    1.584104    0.336130  0.996178   \n",
       "32  0.773348  0.195172  0.097723  0.808526    1.581874    0.292896  0.995460   \n",
       "49  0.667045  0.269603  0.032461  0.912339    1.579384    0.302064  0.995689   \n",
       "19  0.776913  0.212756  0.153442  0.801818    1.578732    0.366198  0.993348   \n",
       "9   0.713201  0.271862  0.050931  0.864683    1.577885    0.322792  0.995328   \n",
       "2   0.926572  0.058502  0.308779  0.649881    1.576453    0.367281  0.994202   \n",
       "25  0.727369  0.244711  0.052268  0.846252    1.573621    0.296979  0.996687   \n",
       "42  0.867145  0.109188  0.208108  0.705932    1.573077    0.317296  0.993086   \n",
       "69  0.858918  0.123538  0.207265  0.712497    1.571415    0.330803  0.993320   \n",
       "0   0.918915  0.067399  0.295044  0.649303    1.568218    0.362443  0.993768   \n",
       "15  0.882287  0.095098  0.265452  0.683058    1.565344    0.360550  0.995164   \n",
       "3   0.825539  0.124662  0.174211  0.737446    1.562985    0.298872  0.992718   \n",
       "43  0.687646  0.279961  0.035215  0.873425    1.561071    0.315177  0.996453   \n",
       "65  0.688659  0.278057  0.038709  0.861727    1.550385    0.316765  0.995060   \n",
       "22  0.856205  0.096011  0.211923  0.685681    1.541885    0.307934  0.994825   \n",
       "8   0.737849  0.235875  0.069771  0.803042    1.540891    0.305646  0.996067   \n",
       "46  0.658901  0.298491  0.048867  0.873030    1.531931    0.347359  0.992602   \n",
       "41  0.612606  0.369583  0.025029  0.916149    1.528755    0.394613  0.995426   \n",
       "10  0.712039  0.238079  0.062735  0.804453    1.516491    0.300814  0.993799   \n",
       "24  0.717924  0.221304  0.074288  0.791369    1.509293    0.295592  0.990993   \n",
       "36  0.578580  0.393927  0.014724  0.908180    1.486760    0.408650  0.995960   \n",
       "66  0.650680  0.270954  0.061796  0.815973    1.466653    0.332750  0.993133   \n",
       "53  0.563482  0.411848  0.021130  0.900106    1.463589    0.432978  0.993331   \n",
       "51  0.522508  0.421381  0.012391  0.933210    1.455717    0.433772  0.995112   \n",
       "52  0.587643  0.369888  0.033761  0.866562    1.454205    0.403649  0.987312   \n",
       "33  0.532673  0.418215  0.014700  0.865243    1.397916    0.432915  0.995076   \n",
       "\n",
       "       Dice1     Dice2     Dice3  \n",
       "14  0.843291  0.846411  0.833207  \n",
       "12  0.843021  0.846834  0.845988  \n",
       "35  0.825083  0.831247  0.809081  \n",
       "61  0.827094  0.851354  0.843518  \n",
       "72  0.825284  0.824976  0.832259  \n",
       "58  0.823699  0.834863  0.831108  \n",
       "73  0.839596  0.841971  0.827433  \n",
       "55  0.820110  0.825959  0.833727  \n",
       "56  0.807689  0.810778  0.819124  \n",
       "54  0.821132  0.846978  0.825889  \n",
       "70  0.831604  0.848871  0.843631  \n",
       "1   0.803432  0.796687  0.804782  \n",
       "68  0.821614  0.839846  0.833980  \n",
       "4   0.825034  0.838850  0.803951  \n",
       "63  0.835899  0.845961  0.837082  \n",
       "62  0.809171  0.816202  0.808539  \n",
       "64  0.815201  0.816830  0.823648  \n",
       "57  0.818509  0.840425  0.852010  \n",
       "29  0.802237  0.815528  0.807952  \n",
       "5   0.792166  0.803326  0.819554  \n",
       "37  0.818332  0.814715  0.815823  \n",
       "13  0.808893  0.818492  0.810333  \n",
       "23  0.813365  0.843567  0.837692  \n",
       "39  0.796587  0.780807  0.751619  \n",
       "18  0.802652  0.835804  0.798312  \n",
       "7   0.807523  0.819547  0.813614  \n",
       "27  0.794069  0.807127  0.788786  \n",
       "26  0.811223  0.825079  0.821793  \n",
       "40  0.804334  0.833286  0.827636  \n",
       "50  0.792475  0.832953  0.808690  \n",
       "..       ...       ...       ...  \n",
       "31  0.790629  0.776794  0.798461  \n",
       "21  0.807650  0.824122  0.823838  \n",
       "60  0.778757  0.799489  0.742067  \n",
       "17  0.797222  0.815309  0.810516  \n",
       "38  0.789520  0.764866  0.822953  \n",
       "32  0.807557  0.819218  0.813403  \n",
       "49  0.777983  0.844907  0.795348  \n",
       "19  0.776299  0.787699  0.784175  \n",
       "9   0.797492  0.828315  0.831142  \n",
       "2   0.777533  0.752149  0.844704  \n",
       "25  0.806246  0.830782  0.827865  \n",
       "42  0.797152  0.782825  0.825374  \n",
       "69  0.792416  0.782156  0.833675  \n",
       "0   0.780229  0.744064  0.805368  \n",
       "15  0.777106  0.762458  0.836195  \n",
       "3   0.794706  0.796743  0.786574  \n",
       "43  0.791053  0.833452  0.826392  \n",
       "65  0.789468  0.826526  0.808840  \n",
       "22  0.791488  0.773357  0.810508  \n",
       "8   0.802167  0.808498  0.805656  \n",
       "46  0.761938  0.822676  0.801886  \n",
       "41  0.742947  0.825055  0.837594  \n",
       "10  0.789505  0.808477  0.781285  \n",
       "24  0.786643  0.805821  0.768787  \n",
       "36  0.723195  0.817109  0.825188  \n",
       "66  0.748033  0.802822  0.755055  \n",
       "53  0.707024  0.806870  0.806477  \n",
       "51  0.678361  0.816866  0.802167  \n",
       "52  0.717716  0.798132  0.743359  \n",
       "33  0.685615  0.792096  0.784187  \n",
       "\n",
       "[74 rows x 13 columns]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "eval_df_sort"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "eval_df_sort.drop([14,12,1,4,5,13],inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>model</th>\n",
       "      <th>inf_dir</th>\n",
       "      <th>test_im</th>\n",
       "      <th>test_11</th>\n",
       "      <th>test_12</th>\n",
       "      <th>test_21</th>\n",
       "      <th>test_22</th>\n",
       "      <th>test_11+22</th>\n",
       "      <th>test_12+21</th>\n",
       "      <th>Dice0</th>\n",
       "      <th>Dice1</th>\n",
       "      <th>Dice2</th>\n",
       "      <th>Dice3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>10210050</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210050/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.850768</td>\n",
       "      <td>0.138990</td>\n",
       "      <td>0.139546</td>\n",
       "      <td>0.831239</td>\n",
       "      <td>1.682007</td>\n",
       "      <td>0.278536</td>\n",
       "      <td>0.992319</td>\n",
       "      <td>0.825083</td>\n",
       "      <td>0.831247</td>\n",
       "      <td>0.809081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>10210130</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_55000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.788889</td>\n",
       "      <td>0.194381</td>\n",
       "      <td>0.081139</td>\n",
       "      <td>0.883951</td>\n",
       "      <td>1.672840</td>\n",
       "      <td>0.275520</td>\n",
       "      <td>0.994809</td>\n",
       "      <td>0.827094</td>\n",
       "      <td>0.851354</td>\n",
       "      <td>0.843518</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>10210054</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210054/inference_50000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.829719</td>\n",
       "      <td>0.155244</td>\n",
       "      <td>0.121035</td>\n",
       "      <td>0.832751</td>\n",
       "      <td>1.662470</td>\n",
       "      <td>0.276280</td>\n",
       "      <td>0.995374</td>\n",
       "      <td>0.823699</td>\n",
       "      <td>0.834863</td>\n",
       "      <td>0.831108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>10210050</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210050/inference_60000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.854372</td>\n",
       "      <td>0.132274</td>\n",
       "      <td>0.148030</td>\n",
       "      <td>0.797729</td>\n",
       "      <td>1.652101</td>\n",
       "      <td>0.280304</td>\n",
       "      <td>0.995481</td>\n",
       "      <td>0.820110</td>\n",
       "      <td>0.825959</td>\n",
       "      <td>0.833727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>10210052</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210052/inference_55400</td>\n",
       "      <td>1</td>\n",
       "      <td>0.862876</td>\n",
       "      <td>0.127103</td>\n",
       "      <td>0.177458</td>\n",
       "      <td>0.788410</td>\n",
       "      <td>1.651286</td>\n",
       "      <td>0.304561</td>\n",
       "      <td>0.995355</td>\n",
       "      <td>0.807689</td>\n",
       "      <td>0.810778</td>\n",
       "      <td>0.819124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>10210050</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210050/inference_47700</td>\n",
       "      <td>1</td>\n",
       "      <td>0.787571</td>\n",
       "      <td>0.184371</td>\n",
       "      <td>0.088806</td>\n",
       "      <td>0.862547</td>\n",
       "      <td>1.650118</td>\n",
       "      <td>0.273177</td>\n",
       "      <td>0.992172</td>\n",
       "      <td>0.821132</td>\n",
       "      <td>0.846978</td>\n",
       "      <td>0.825889</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>10210132</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_55000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.788230</td>\n",
       "      <td>0.190931</td>\n",
       "      <td>0.072818</td>\n",
       "      <td>0.861572</td>\n",
       "      <td>1.649802</td>\n",
       "      <td>0.263749</td>\n",
       "      <td>0.995917</td>\n",
       "      <td>0.831604</td>\n",
       "      <td>0.848871</td>\n",
       "      <td>0.843631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>10210054</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210054/inference_57300</td>\n",
       "      <td>1</td>\n",
       "      <td>0.794176</td>\n",
       "      <td>0.185833</td>\n",
       "      <td>0.093085</td>\n",
       "      <td>0.852272</td>\n",
       "      <td>1.646448</td>\n",
       "      <td>0.278917</td>\n",
       "      <td>0.995408</td>\n",
       "      <td>0.821614</td>\n",
       "      <td>0.839846</td>\n",
       "      <td>0.833980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>10210132</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_45000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.798356</td>\n",
       "      <td>0.177805</td>\n",
       "      <td>0.075448</td>\n",
       "      <td>0.843776</td>\n",
       "      <td>1.642132</td>\n",
       "      <td>0.253253</td>\n",
       "      <td>0.994820</td>\n",
       "      <td>0.835899</td>\n",
       "      <td>0.845961</td>\n",
       "      <td>0.837082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>10210130</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_45000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.840122</td>\n",
       "      <td>0.147062</td>\n",
       "      <td>0.156935</td>\n",
       "      <td>0.799764</td>\n",
       "      <td>1.639886</td>\n",
       "      <td>0.303997</td>\n",
       "      <td>0.991060</td>\n",
       "      <td>0.809171</td>\n",
       "      <td>0.816202</td>\n",
       "      <td>0.808539</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>10210136</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210136/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.839430</td>\n",
       "      <td>0.149675</td>\n",
       "      <td>0.145136</td>\n",
       "      <td>0.800236</td>\n",
       "      <td>1.639666</td>\n",
       "      <td>0.294811</td>\n",
       "      <td>0.995972</td>\n",
       "      <td>0.815201</td>\n",
       "      <td>0.816830</td>\n",
       "      <td>0.823648</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>10210052</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210052/inference_70000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.795792</td>\n",
       "      <td>0.186408</td>\n",
       "      <td>0.098338</td>\n",
       "      <td>0.843807</td>\n",
       "      <td>1.639600</td>\n",
       "      <td>0.284747</td>\n",
       "      <td>0.996155</td>\n",
       "      <td>0.818509</td>\n",
       "      <td>0.840425</td>\n",
       "      <td>0.852010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>10202102</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202102/inference_11600</td>\n",
       "      <td>1</td>\n",
       "      <td>0.810325</td>\n",
       "      <td>0.179549</td>\n",
       "      <td>0.136749</td>\n",
       "      <td>0.828797</td>\n",
       "      <td>1.639123</td>\n",
       "      <td>0.316298</td>\n",
       "      <td>0.995893</td>\n",
       "      <td>0.802237</td>\n",
       "      <td>0.815528</td>\n",
       "      <td>0.807952</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>10210052</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210052/inference_20000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.841932</td>\n",
       "      <td>0.141248</td>\n",
       "      <td>0.142003</td>\n",
       "      <td>0.788619</td>\n",
       "      <td>1.630551</td>\n",
       "      <td>0.283251</td>\n",
       "      <td>0.996288</td>\n",
       "      <td>0.818332</td>\n",
       "      <td>0.814715</td>\n",
       "      <td>0.815823</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>10192004</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_6600</td>\n",
       "      <td>1</td>\n",
       "      <td>0.769910</td>\n",
       "      <td>0.198151</td>\n",
       "      <td>0.084447</td>\n",
       "      <td>0.859092</td>\n",
       "      <td>1.629002</td>\n",
       "      <td>0.282597</td>\n",
       "      <td>0.995134</td>\n",
       "      <td>0.813365</td>\n",
       "      <td>0.843567</td>\n",
       "      <td>0.837692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>10210052</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210052/inference_10000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.847114</td>\n",
       "      <td>0.142494</td>\n",
       "      <td>0.188487</td>\n",
       "      <td>0.776054</td>\n",
       "      <td>1.623168</td>\n",
       "      <td>0.330981</td>\n",
       "      <td>0.994722</td>\n",
       "      <td>0.796587</td>\n",
       "      <td>0.780807</td>\n",
       "      <td>0.751619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>10192002</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192002/inference_4600</td>\n",
       "      <td>1</td>\n",
       "      <td>0.751673</td>\n",
       "      <td>0.210691</td>\n",
       "      <td>0.083541</td>\n",
       "      <td>0.868307</td>\n",
       "      <td>1.619980</td>\n",
       "      <td>0.294232</td>\n",
       "      <td>0.993344</td>\n",
       "      <td>0.802652</td>\n",
       "      <td>0.835804</td>\n",
       "      <td>0.798312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>10112103</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112103/inference_12300_10112103</td>\n",
       "      <td>1</td>\n",
       "      <td>0.809434</td>\n",
       "      <td>0.176941</td>\n",
       "      <td>0.130890</td>\n",
       "      <td>0.809682</td>\n",
       "      <td>1.619116</td>\n",
       "      <td>0.307831</td>\n",
       "      <td>0.992859</td>\n",
       "      <td>0.807523</td>\n",
       "      <td>0.819547</td>\n",
       "      <td>0.813614</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>10202100</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202100/inference_2000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.801738</td>\n",
       "      <td>0.181940</td>\n",
       "      <td>0.145419</td>\n",
       "      <td>0.815258</td>\n",
       "      <td>1.616996</td>\n",
       "      <td>0.327359</td>\n",
       "      <td>0.993349</td>\n",
       "      <td>0.794069</td>\n",
       "      <td>0.807127</td>\n",
       "      <td>0.788786</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>10202100</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202100/inference_5800</td>\n",
       "      <td>1</td>\n",
       "      <td>0.800781</td>\n",
       "      <td>0.179947</td>\n",
       "      <td>0.113441</td>\n",
       "      <td>0.809683</td>\n",
       "      <td>1.610464</td>\n",
       "      <td>0.293389</td>\n",
       "      <td>0.991585</td>\n",
       "      <td>0.811223</td>\n",
       "      <td>0.825079</td>\n",
       "      <td>0.821793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>10210054</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210054/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.766677</td>\n",
       "      <td>0.198644</td>\n",
       "      <td>0.095696</td>\n",
       "      <td>0.842744</td>\n",
       "      <td>1.609420</td>\n",
       "      <td>0.294340</td>\n",
       "      <td>0.995408</td>\n",
       "      <td>0.804334</td>\n",
       "      <td>0.833286</td>\n",
       "      <td>0.827636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>10210134</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210134/inference_22000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.732927</td>\n",
       "      <td>0.227139</td>\n",
       "      <td>0.079762</td>\n",
       "      <td>0.874268</td>\n",
       "      <td>1.607196</td>\n",
       "      <td>0.306901</td>\n",
       "      <td>0.995883</td>\n",
       "      <td>0.792475</td>\n",
       "      <td>0.832953</td>\n",
       "      <td>0.808690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>10192004</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_11100</td>\n",
       "      <td>1</td>\n",
       "      <td>0.805924</td>\n",
       "      <td>0.147715</td>\n",
       "      <td>0.123747</td>\n",
       "      <td>0.801041</td>\n",
       "      <td>1.606965</td>\n",
       "      <td>0.271462</td>\n",
       "      <td>0.995025</td>\n",
       "      <td>0.811626</td>\n",
       "      <td>0.828520</td>\n",
       "      <td>0.822260</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>10210130</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_6000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.812136</td>\n",
       "      <td>0.158749</td>\n",
       "      <td>0.172905</td>\n",
       "      <td>0.789296</td>\n",
       "      <td>1.601431</td>\n",
       "      <td>0.331653</td>\n",
       "      <td>0.991836</td>\n",
       "      <td>0.784101</td>\n",
       "      <td>0.790127</td>\n",
       "      <td>0.740799</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>10210132</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_12000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.732130</td>\n",
       "      <td>0.202447</td>\n",
       "      <td>0.080307</td>\n",
       "      <td>0.869290</td>\n",
       "      <td>1.601420</td>\n",
       "      <td>0.282754</td>\n",
       "      <td>0.993884</td>\n",
       "      <td>0.791720</td>\n",
       "      <td>0.840746</td>\n",
       "      <td>0.794139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>10210130</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_12000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.862400</td>\n",
       "      <td>0.107172</td>\n",
       "      <td>0.171404</td>\n",
       "      <td>0.738618</td>\n",
       "      <td>1.601018</td>\n",
       "      <td>0.278576</td>\n",
       "      <td>0.990123</td>\n",
       "      <td>0.816217</td>\n",
       "      <td>0.805659</td>\n",
       "      <td>0.824301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>10210134</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210134/inference_39300</td>\n",
       "      <td>1</td>\n",
       "      <td>0.713091</td>\n",
       "      <td>0.253929</td>\n",
       "      <td>0.037873</td>\n",
       "      <td>0.886563</td>\n",
       "      <td>1.599653</td>\n",
       "      <td>0.291802</td>\n",
       "      <td>0.995389</td>\n",
       "      <td>0.806570</td>\n",
       "      <td>0.845419</td>\n",
       "      <td>0.829573</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>10202102</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202102/inference_20000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.751585</td>\n",
       "      <td>0.226447</td>\n",
       "      <td>0.080895</td>\n",
       "      <td>0.846833</td>\n",
       "      <td>1.598417</td>\n",
       "      <td>0.307342</td>\n",
       "      <td>0.994971</td>\n",
       "      <td>0.803225</td>\n",
       "      <td>0.828606</td>\n",
       "      <td>0.824020</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>10210132</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_25000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.752758</td>\n",
       "      <td>0.220396</td>\n",
       "      <td>0.075735</td>\n",
       "      <td>0.844798</td>\n",
       "      <td>1.597556</td>\n",
       "      <td>0.296130</td>\n",
       "      <td>0.995467</td>\n",
       "      <td>0.808363</td>\n",
       "      <td>0.829791</td>\n",
       "      <td>0.823269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>10210050</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210050/inference_20000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.695402</td>\n",
       "      <td>0.264897</td>\n",
       "      <td>0.055630</td>\n",
       "      <td>0.900725</td>\n",
       "      <td>1.596127</td>\n",
       "      <td>0.320527</td>\n",
       "      <td>0.995051</td>\n",
       "      <td>0.783159</td>\n",
       "      <td>0.840469</td>\n",
       "      <td>0.826030</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>10202104</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202104/inference_10000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.857190</td>\n",
       "      <td>0.128941</td>\n",
       "      <td>0.210557</td>\n",
       "      <td>0.731070</td>\n",
       "      <td>1.588261</td>\n",
       "      <td>0.339498</td>\n",
       "      <td>0.994264</td>\n",
       "      <td>0.790629</td>\n",
       "      <td>0.776794</td>\n",
       "      <td>0.798461</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>10192004</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_4000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.783851</td>\n",
       "      <td>0.174810</td>\n",
       "      <td>0.108014</td>\n",
       "      <td>0.803988</td>\n",
       "      <td>1.587839</td>\n",
       "      <td>0.282823</td>\n",
       "      <td>0.995970</td>\n",
       "      <td>0.807650</td>\n",
       "      <td>0.824122</td>\n",
       "      <td>0.823838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>10210056</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210056/inference_10000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.736947</td>\n",
       "      <td>0.245769</td>\n",
       "      <td>0.106176</td>\n",
       "      <td>0.848984</td>\n",
       "      <td>1.585930</td>\n",
       "      <td>0.351945</td>\n",
       "      <td>0.991282</td>\n",
       "      <td>0.778757</td>\n",
       "      <td>0.799489</td>\n",
       "      <td>0.742067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>10192002</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192002/inference_5700</td>\n",
       "      <td>1</td>\n",
       "      <td>0.797880</td>\n",
       "      <td>0.149403</td>\n",
       "      <td>0.140641</td>\n",
       "      <td>0.786742</td>\n",
       "      <td>1.584622</td>\n",
       "      <td>0.290044</td>\n",
       "      <td>0.996218</td>\n",
       "      <td>0.797222</td>\n",
       "      <td>0.815309</td>\n",
       "      <td>0.810516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>10210052</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210052/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.903158</td>\n",
       "      <td>0.080006</td>\n",
       "      <td>0.256125</td>\n",
       "      <td>0.680945</td>\n",
       "      <td>1.584104</td>\n",
       "      <td>0.336130</td>\n",
       "      <td>0.996178</td>\n",
       "      <td>0.789520</td>\n",
       "      <td>0.764866</td>\n",
       "      <td>0.822953</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>10202104</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202104/inference_19600</td>\n",
       "      <td>1</td>\n",
       "      <td>0.773348</td>\n",
       "      <td>0.195172</td>\n",
       "      <td>0.097723</td>\n",
       "      <td>0.808526</td>\n",
       "      <td>1.581874</td>\n",
       "      <td>0.292896</td>\n",
       "      <td>0.995460</td>\n",
       "      <td>0.807557</td>\n",
       "      <td>0.819218</td>\n",
       "      <td>0.813403</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>10210134</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210134/inference_11700</td>\n",
       "      <td>1</td>\n",
       "      <td>0.667045</td>\n",
       "      <td>0.269603</td>\n",
       "      <td>0.032461</td>\n",
       "      <td>0.912339</td>\n",
       "      <td>1.579384</td>\n",
       "      <td>0.302064</td>\n",
       "      <td>0.995689</td>\n",
       "      <td>0.777983</td>\n",
       "      <td>0.844907</td>\n",
       "      <td>0.795348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>10192002</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192002/inference_2500</td>\n",
       "      <td>1</td>\n",
       "      <td>0.776913</td>\n",
       "      <td>0.212756</td>\n",
       "      <td>0.153442</td>\n",
       "      <td>0.801818</td>\n",
       "      <td>1.578732</td>\n",
       "      <td>0.366198</td>\n",
       "      <td>0.993348</td>\n",
       "      <td>0.776299</td>\n",
       "      <td>0.787699</td>\n",
       "      <td>0.784175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10112127</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112127/inference_8700_10112127</td>\n",
       "      <td>1</td>\n",
       "      <td>0.713201</td>\n",
       "      <td>0.271862</td>\n",
       "      <td>0.050931</td>\n",
       "      <td>0.864683</td>\n",
       "      <td>1.577885</td>\n",
       "      <td>0.322792</td>\n",
       "      <td>0.995328</td>\n",
       "      <td>0.797492</td>\n",
       "      <td>0.828315</td>\n",
       "      <td>0.831142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10112050</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112050/inference_13200_10112050</td>\n",
       "      <td>1</td>\n",
       "      <td>0.926572</td>\n",
       "      <td>0.058502</td>\n",
       "      <td>0.308779</td>\n",
       "      <td>0.649881</td>\n",
       "      <td>1.576453</td>\n",
       "      <td>0.367281</td>\n",
       "      <td>0.994202</td>\n",
       "      <td>0.777533</td>\n",
       "      <td>0.752149</td>\n",
       "      <td>0.844704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>10202100</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202100/inference_12000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.727369</td>\n",
       "      <td>0.244711</td>\n",
       "      <td>0.052268</td>\n",
       "      <td>0.846252</td>\n",
       "      <td>1.573621</td>\n",
       "      <td>0.296979</td>\n",
       "      <td>0.996687</td>\n",
       "      <td>0.806246</td>\n",
       "      <td>0.830782</td>\n",
       "      <td>0.827865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>10210054</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210054/inference_18900</td>\n",
       "      <td>1</td>\n",
       "      <td>0.867145</td>\n",
       "      <td>0.109188</td>\n",
       "      <td>0.208108</td>\n",
       "      <td>0.705932</td>\n",
       "      <td>1.573077</td>\n",
       "      <td>0.317296</td>\n",
       "      <td>0.993086</td>\n",
       "      <td>0.797152</td>\n",
       "      <td>0.782825</td>\n",
       "      <td>0.825374</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>10210056</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210056/inference_30000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.858918</td>\n",
       "      <td>0.123538</td>\n",
       "      <td>0.207265</td>\n",
       "      <td>0.712497</td>\n",
       "      <td>1.571415</td>\n",
       "      <td>0.330803</td>\n",
       "      <td>0.993320</td>\n",
       "      <td>0.792416</td>\n",
       "      <td>0.782156</td>\n",
       "      <td>0.833675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10112050</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112050/inference_17000_10112050</td>\n",
       "      <td>1</td>\n",
       "      <td>0.918915</td>\n",
       "      <td>0.067399</td>\n",
       "      <td>0.295044</td>\n",
       "      <td>0.649303</td>\n",
       "      <td>1.568218</td>\n",
       "      <td>0.362443</td>\n",
       "      <td>0.993768</td>\n",
       "      <td>0.780229</td>\n",
       "      <td>0.744064</td>\n",
       "      <td>0.805368</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>10192000</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192000/inference_7200</td>\n",
       "      <td>1</td>\n",
       "      <td>0.882287</td>\n",
       "      <td>0.095098</td>\n",
       "      <td>0.265452</td>\n",
       "      <td>0.683058</td>\n",
       "      <td>1.565344</td>\n",
       "      <td>0.360550</td>\n",
       "      <td>0.995164</td>\n",
       "      <td>0.777106</td>\n",
       "      <td>0.762458</td>\n",
       "      <td>0.836195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10112056</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112056/inference_18100_10112056</td>\n",
       "      <td>1</td>\n",
       "      <td>0.825539</td>\n",
       "      <td>0.124662</td>\n",
       "      <td>0.174211</td>\n",
       "      <td>0.737446</td>\n",
       "      <td>1.562985</td>\n",
       "      <td>0.298872</td>\n",
       "      <td>0.992718</td>\n",
       "      <td>0.794706</td>\n",
       "      <td>0.796743</td>\n",
       "      <td>0.786574</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>10210130</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210130/inference_20000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.687646</td>\n",
       "      <td>0.279961</td>\n",
       "      <td>0.035215</td>\n",
       "      <td>0.873425</td>\n",
       "      <td>1.561071</td>\n",
       "      <td>0.315177</td>\n",
       "      <td>0.996453</td>\n",
       "      <td>0.791053</td>\n",
       "      <td>0.833452</td>\n",
       "      <td>0.826392</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>10210136</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210136/inference_17800</td>\n",
       "      <td>1</td>\n",
       "      <td>0.688659</td>\n",
       "      <td>0.278057</td>\n",
       "      <td>0.038709</td>\n",
       "      <td>0.861727</td>\n",
       "      <td>1.550385</td>\n",
       "      <td>0.316765</td>\n",
       "      <td>0.995060</td>\n",
       "      <td>0.789468</td>\n",
       "      <td>0.826526</td>\n",
       "      <td>0.808840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>10192004</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_3900</td>\n",
       "      <td>1</td>\n",
       "      <td>0.856205</td>\n",
       "      <td>0.096011</td>\n",
       "      <td>0.211923</td>\n",
       "      <td>0.685681</td>\n",
       "      <td>1.541885</td>\n",
       "      <td>0.307934</td>\n",
       "      <td>0.994825</td>\n",
       "      <td>0.791488</td>\n",
       "      <td>0.773357</td>\n",
       "      <td>0.810508</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>10112127</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112127/inference_13500_10112127</td>\n",
       "      <td>1</td>\n",
       "      <td>0.737849</td>\n",
       "      <td>0.235875</td>\n",
       "      <td>0.069771</td>\n",
       "      <td>0.803042</td>\n",
       "      <td>1.540891</td>\n",
       "      <td>0.305646</td>\n",
       "      <td>0.996067</td>\n",
       "      <td>0.802167</td>\n",
       "      <td>0.808498</td>\n",
       "      <td>0.805656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>10210132</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210132/inference_6200</td>\n",
       "      <td>1</td>\n",
       "      <td>0.658901</td>\n",
       "      <td>0.298491</td>\n",
       "      <td>0.048867</td>\n",
       "      <td>0.873030</td>\n",
       "      <td>1.531931</td>\n",
       "      <td>0.347359</td>\n",
       "      <td>0.992602</td>\n",
       "      <td>0.761938</td>\n",
       "      <td>0.822676</td>\n",
       "      <td>0.801886</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>10210054</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210054/inference_10000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.612606</td>\n",
       "      <td>0.369583</td>\n",
       "      <td>0.025029</td>\n",
       "      <td>0.916149</td>\n",
       "      <td>1.528755</td>\n",
       "      <td>0.394613</td>\n",
       "      <td>0.995426</td>\n",
       "      <td>0.742947</td>\n",
       "      <td>0.825055</td>\n",
       "      <td>0.837594</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10112130</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10112130/inference_7300</td>\n",
       "      <td>1</td>\n",
       "      <td>0.712039</td>\n",
       "      <td>0.238079</td>\n",
       "      <td>0.062735</td>\n",
       "      <td>0.804453</td>\n",
       "      <td>1.516491</td>\n",
       "      <td>0.300814</td>\n",
       "      <td>0.993799</td>\n",
       "      <td>0.789505</td>\n",
       "      <td>0.808477</td>\n",
       "      <td>0.781285</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>10192004</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10192004/inference_2900</td>\n",
       "      <td>1</td>\n",
       "      <td>0.717924</td>\n",
       "      <td>0.221304</td>\n",
       "      <td>0.074288</td>\n",
       "      <td>0.791369</td>\n",
       "      <td>1.509293</td>\n",
       "      <td>0.295592</td>\n",
       "      <td>0.990993</td>\n",
       "      <td>0.786643</td>\n",
       "      <td>0.805821</td>\n",
       "      <td>0.768787</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>10210050</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210050/inference_10000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.578580</td>\n",
       "      <td>0.393927</td>\n",
       "      <td>0.014724</td>\n",
       "      <td>0.908180</td>\n",
       "      <td>1.486760</td>\n",
       "      <td>0.408650</td>\n",
       "      <td>0.995960</td>\n",
       "      <td>0.723195</td>\n",
       "      <td>0.817109</td>\n",
       "      <td>0.825188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>10210136</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210136/inference_8800</td>\n",
       "      <td>1</td>\n",
       "      <td>0.650680</td>\n",
       "      <td>0.270954</td>\n",
       "      <td>0.061796</td>\n",
       "      <td>0.815973</td>\n",
       "      <td>1.466653</td>\n",
       "      <td>0.332750</td>\n",
       "      <td>0.993133</td>\n",
       "      <td>0.748033</td>\n",
       "      <td>0.802822</td>\n",
       "      <td>0.755055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>10202106</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202106/inference_9000</td>\n",
       "      <td>1</td>\n",
       "      <td>0.563482</td>\n",
       "      <td>0.411848</td>\n",
       "      <td>0.021130</td>\n",
       "      <td>0.900106</td>\n",
       "      <td>1.463589</td>\n",
       "      <td>0.432978</td>\n",
       "      <td>0.993331</td>\n",
       "      <td>0.707024</td>\n",
       "      <td>0.806870</td>\n",
       "      <td>0.806477</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>10210134</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10210134/inference_4400</td>\n",
       "      <td>1</td>\n",
       "      <td>0.522508</td>\n",
       "      <td>0.421381</td>\n",
       "      <td>0.012391</td>\n",
       "      <td>0.933210</td>\n",
       "      <td>1.455717</td>\n",
       "      <td>0.433772</td>\n",
       "      <td>0.995112</td>\n",
       "      <td>0.678361</td>\n",
       "      <td>0.816866</td>\n",
       "      <td>0.802167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>10202106</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202106/inference_2100</td>\n",
       "      <td>1</td>\n",
       "      <td>0.587643</td>\n",
       "      <td>0.369888</td>\n",
       "      <td>0.033761</td>\n",
       "      <td>0.866562</td>\n",
       "      <td>1.454205</td>\n",
       "      <td>0.403649</td>\n",
       "      <td>0.987312</td>\n",
       "      <td>0.717716</td>\n",
       "      <td>0.798132</td>\n",
       "      <td>0.743359</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>10202104</td>\n",
       "      <td>/proj/NIRAL/users/siyangj/model_10202104/inference_2300</td>\n",
       "      <td>1</td>\n",
       "      <td>0.532673</td>\n",
       "      <td>0.418215</td>\n",
       "      <td>0.014700</td>\n",
       "      <td>0.865243</td>\n",
       "      <td>1.397916</td>\n",
       "      <td>0.432915</td>\n",
       "      <td>0.995076</td>\n",
       "      <td>0.685615</td>\n",
       "      <td>0.792096</td>\n",
       "      <td>0.784187</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>66 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       model  \\\n",
       "35  10210050   \n",
       "61  10210130   \n",
       "58  10210054   \n",
       "55  10210050   \n",
       "56  10210052   \n",
       "54  10210050   \n",
       "70  10210132   \n",
       "68  10210054   \n",
       "63  10210132   \n",
       "62  10210130   \n",
       "64  10210136   \n",
       "57  10210052   \n",
       "29  10202102   \n",
       "37  10210052   \n",
       "23  10192004   \n",
       "39  10210052   \n",
       "18  10192002   \n",
       "7   10112103   \n",
       "27  10202100   \n",
       "26  10202100   \n",
       "40  10210054   \n",
       "50  10210134   \n",
       "20  10192004   \n",
       "45  10210130   \n",
       "48  10210132   \n",
       "44  10210130   \n",
       "71  10210134   \n",
       "28  10202102   \n",
       "47  10210132   \n",
       "34  10210050   \n",
       "..       ...   \n",
       "31  10202104   \n",
       "21  10192004   \n",
       "60  10210056   \n",
       "17  10192002   \n",
       "38  10210052   \n",
       "32  10202104   \n",
       "49  10210134   \n",
       "19  10192002   \n",
       "9   10112127   \n",
       "2   10112050   \n",
       "25  10202100   \n",
       "42  10210054   \n",
       "69  10210056   \n",
       "0   10112050   \n",
       "15  10192000   \n",
       "3   10112056   \n",
       "43  10210130   \n",
       "65  10210136   \n",
       "22  10192004   \n",
       "8   10112127   \n",
       "46  10210132   \n",
       "41  10210054   \n",
       "10  10112130   \n",
       "24  10192004   \n",
       "36  10210050   \n",
       "66  10210136   \n",
       "53  10202106   \n",
       "51  10210134   \n",
       "52  10202106   \n",
       "33  10202104   \n",
       "\n",
       "                                                              inf_dir  \\\n",
       "35           /proj/NIRAL/users/siyangj/model_10210050/inference_30000   \n",
       "61           /proj/NIRAL/users/siyangj/model_10210130/inference_55000   \n",
       "58           /proj/NIRAL/users/siyangj/model_10210054/inference_50000   \n",
       "55           /proj/NIRAL/users/siyangj/model_10210050/inference_60000   \n",
       "56           /proj/NIRAL/users/siyangj/model_10210052/inference_55400   \n",
       "54           /proj/NIRAL/users/siyangj/model_10210050/inference_47700   \n",
       "70           /proj/NIRAL/users/siyangj/model_10210132/inference_55000   \n",
       "68           /proj/NIRAL/users/siyangj/model_10210054/inference_57300   \n",
       "63           /proj/NIRAL/users/siyangj/model_10210132/inference_45000   \n",
       "62           /proj/NIRAL/users/siyangj/model_10210130/inference_45000   \n",
       "64           /proj/NIRAL/users/siyangj/model_10210136/inference_30000   \n",
       "57           /proj/NIRAL/users/siyangj/model_10210052/inference_70000   \n",
       "29           /proj/NIRAL/users/siyangj/model_10202102/inference_11600   \n",
       "37           /proj/NIRAL/users/siyangj/model_10210052/inference_20000   \n",
       "23            /proj/NIRAL/users/siyangj/model_10192004/inference_6600   \n",
       "39           /proj/NIRAL/users/siyangj/model_10210052/inference_10000   \n",
       "18            /proj/NIRAL/users/siyangj/model_10192002/inference_4600   \n",
       "7   /proj/NIRAL/users/siyangj/model_10112103/inference_12300_10112103   \n",
       "27            /proj/NIRAL/users/siyangj/model_10202100/inference_2000   \n",
       "26            /proj/NIRAL/users/siyangj/model_10202100/inference_5800   \n",
       "40           /proj/NIRAL/users/siyangj/model_10210054/inference_30000   \n",
       "50           /proj/NIRAL/users/siyangj/model_10210134/inference_22000   \n",
       "20           /proj/NIRAL/users/siyangj/model_10192004/inference_11100   \n",
       "45            /proj/NIRAL/users/siyangj/model_10210130/inference_6000   \n",
       "48           /proj/NIRAL/users/siyangj/model_10210132/inference_12000   \n",
       "44           /proj/NIRAL/users/siyangj/model_10210130/inference_12000   \n",
       "71           /proj/NIRAL/users/siyangj/model_10210134/inference_39300   \n",
       "28           /proj/NIRAL/users/siyangj/model_10202102/inference_20000   \n",
       "47           /proj/NIRAL/users/siyangj/model_10210132/inference_25000   \n",
       "34           /proj/NIRAL/users/siyangj/model_10210050/inference_20000   \n",
       "..                                                                ...   \n",
       "31           /proj/NIRAL/users/siyangj/model_10202104/inference_10000   \n",
       "21            /proj/NIRAL/users/siyangj/model_10192004/inference_4000   \n",
       "60           /proj/NIRAL/users/siyangj/model_10210056/inference_10000   \n",
       "17            /proj/NIRAL/users/siyangj/model_10192002/inference_5700   \n",
       "38           /proj/NIRAL/users/siyangj/model_10210052/inference_30000   \n",
       "32           /proj/NIRAL/users/siyangj/model_10202104/inference_19600   \n",
       "49           /proj/NIRAL/users/siyangj/model_10210134/inference_11700   \n",
       "19            /proj/NIRAL/users/siyangj/model_10192002/inference_2500   \n",
       "9    /proj/NIRAL/users/siyangj/model_10112127/inference_8700_10112127   \n",
       "2   /proj/NIRAL/users/siyangj/model_10112050/inference_13200_10112050   \n",
       "25           /proj/NIRAL/users/siyangj/model_10202100/inference_12000   \n",
       "42           /proj/NIRAL/users/siyangj/model_10210054/inference_18900   \n",
       "69           /proj/NIRAL/users/siyangj/model_10210056/inference_30000   \n",
       "0   /proj/NIRAL/users/siyangj/model_10112050/inference_17000_10112050   \n",
       "15            /proj/NIRAL/users/siyangj/model_10192000/inference_7200   \n",
       "3   /proj/NIRAL/users/siyangj/model_10112056/inference_18100_10112056   \n",
       "43           /proj/NIRAL/users/siyangj/model_10210130/inference_20000   \n",
       "65           /proj/NIRAL/users/siyangj/model_10210136/inference_17800   \n",
       "22            /proj/NIRAL/users/siyangj/model_10192004/inference_3900   \n",
       "8   /proj/NIRAL/users/siyangj/model_10112127/inference_13500_10112127   \n",
       "46            /proj/NIRAL/users/siyangj/model_10210132/inference_6200   \n",
       "41           /proj/NIRAL/users/siyangj/model_10210054/inference_10000   \n",
       "10            /proj/NIRAL/users/siyangj/model_10112130/inference_7300   \n",
       "24            /proj/NIRAL/users/siyangj/model_10192004/inference_2900   \n",
       "36           /proj/NIRAL/users/siyangj/model_10210050/inference_10000   \n",
       "66            /proj/NIRAL/users/siyangj/model_10210136/inference_8800   \n",
       "53            /proj/NIRAL/users/siyangj/model_10202106/inference_9000   \n",
       "51            /proj/NIRAL/users/siyangj/model_10210134/inference_4400   \n",
       "52            /proj/NIRAL/users/siyangj/model_10202106/inference_2100   \n",
       "33            /proj/NIRAL/users/siyangj/model_10202104/inference_2300   \n",
       "\n",
       "    test_im   test_11   test_12   test_21   test_22  test_11+22  test_12+21  \\\n",
       "35        1  0.850768  0.138990  0.139546  0.831239    1.682007    0.278536   \n",
       "61        1  0.788889  0.194381  0.081139  0.883951    1.672840    0.275520   \n",
       "58        1  0.829719  0.155244  0.121035  0.832751    1.662470    0.276280   \n",
       "55        1  0.854372  0.132274  0.148030  0.797729    1.652101    0.280304   \n",
       "56        1  0.862876  0.127103  0.177458  0.788410    1.651286    0.304561   \n",
       "54        1  0.787571  0.184371  0.088806  0.862547    1.650118    0.273177   \n",
       "70        1  0.788230  0.190931  0.072818  0.861572    1.649802    0.263749   \n",
       "68        1  0.794176  0.185833  0.093085  0.852272    1.646448    0.278917   \n",
       "63        1  0.798356  0.177805  0.075448  0.843776    1.642132    0.253253   \n",
       "62        1  0.840122  0.147062  0.156935  0.799764    1.639886    0.303997   \n",
       "64        1  0.839430  0.149675  0.145136  0.800236    1.639666    0.294811   \n",
       "57        1  0.795792  0.186408  0.098338  0.843807    1.639600    0.284747   \n",
       "29        1  0.810325  0.179549  0.136749  0.828797    1.639123    0.316298   \n",
       "37        1  0.841932  0.141248  0.142003  0.788619    1.630551    0.283251   \n",
       "23        1  0.769910  0.198151  0.084447  0.859092    1.629002    0.282597   \n",
       "39        1  0.847114  0.142494  0.188487  0.776054    1.623168    0.330981   \n",
       "18        1  0.751673  0.210691  0.083541  0.868307    1.619980    0.294232   \n",
       "7         1  0.809434  0.176941  0.130890  0.809682    1.619116    0.307831   \n",
       "27        1  0.801738  0.181940  0.145419  0.815258    1.616996    0.327359   \n",
       "26        1  0.800781  0.179947  0.113441  0.809683    1.610464    0.293389   \n",
       "40        1  0.766677  0.198644  0.095696  0.842744    1.609420    0.294340   \n",
       "50        1  0.732927  0.227139  0.079762  0.874268    1.607196    0.306901   \n",
       "20        1  0.805924  0.147715  0.123747  0.801041    1.606965    0.271462   \n",
       "45        1  0.812136  0.158749  0.172905  0.789296    1.601431    0.331653   \n",
       "48        1  0.732130  0.202447  0.080307  0.869290    1.601420    0.282754   \n",
       "44        1  0.862400  0.107172  0.171404  0.738618    1.601018    0.278576   \n",
       "71        1  0.713091  0.253929  0.037873  0.886563    1.599653    0.291802   \n",
       "28        1  0.751585  0.226447  0.080895  0.846833    1.598417    0.307342   \n",
       "47        1  0.752758  0.220396  0.075735  0.844798    1.597556    0.296130   \n",
       "34        1  0.695402  0.264897  0.055630  0.900725    1.596127    0.320527   \n",
       "..      ...       ...       ...       ...       ...         ...         ...   \n",
       "31        1  0.857190  0.128941  0.210557  0.731070    1.588261    0.339498   \n",
       "21        1  0.783851  0.174810  0.108014  0.803988    1.587839    0.282823   \n",
       "60        1  0.736947  0.245769  0.106176  0.848984    1.585930    0.351945   \n",
       "17        1  0.797880  0.149403  0.140641  0.786742    1.584622    0.290044   \n",
       "38        1  0.903158  0.080006  0.256125  0.680945    1.584104    0.336130   \n",
       "32        1  0.773348  0.195172  0.097723  0.808526    1.581874    0.292896   \n",
       "49        1  0.667045  0.269603  0.032461  0.912339    1.579384    0.302064   \n",
       "19        1  0.776913  0.212756  0.153442  0.801818    1.578732    0.366198   \n",
       "9         1  0.713201  0.271862  0.050931  0.864683    1.577885    0.322792   \n",
       "2         1  0.926572  0.058502  0.308779  0.649881    1.576453    0.367281   \n",
       "25        1  0.727369  0.244711  0.052268  0.846252    1.573621    0.296979   \n",
       "42        1  0.867145  0.109188  0.208108  0.705932    1.573077    0.317296   \n",
       "69        1  0.858918  0.123538  0.207265  0.712497    1.571415    0.330803   \n",
       "0         1  0.918915  0.067399  0.295044  0.649303    1.568218    0.362443   \n",
       "15        1  0.882287  0.095098  0.265452  0.683058    1.565344    0.360550   \n",
       "3         1  0.825539  0.124662  0.174211  0.737446    1.562985    0.298872   \n",
       "43        1  0.687646  0.279961  0.035215  0.873425    1.561071    0.315177   \n",
       "65        1  0.688659  0.278057  0.038709  0.861727    1.550385    0.316765   \n",
       "22        1  0.856205  0.096011  0.211923  0.685681    1.541885    0.307934   \n",
       "8         1  0.737849  0.235875  0.069771  0.803042    1.540891    0.305646   \n",
       "46        1  0.658901  0.298491  0.048867  0.873030    1.531931    0.347359   \n",
       "41        1  0.612606  0.369583  0.025029  0.916149    1.528755    0.394613   \n",
       "10        1  0.712039  0.238079  0.062735  0.804453    1.516491    0.300814   \n",
       "24        1  0.717924  0.221304  0.074288  0.791369    1.509293    0.295592   \n",
       "36        1  0.578580  0.393927  0.014724  0.908180    1.486760    0.408650   \n",
       "66        1  0.650680  0.270954  0.061796  0.815973    1.466653    0.332750   \n",
       "53        1  0.563482  0.411848  0.021130  0.900106    1.463589    0.432978   \n",
       "51        1  0.522508  0.421381  0.012391  0.933210    1.455717    0.433772   \n",
       "52        1  0.587643  0.369888  0.033761  0.866562    1.454205    0.403649   \n",
       "33        1  0.532673  0.418215  0.014700  0.865243    1.397916    0.432915   \n",
       "\n",
       "       Dice0     Dice1     Dice2     Dice3  \n",
       "35  0.992319  0.825083  0.831247  0.809081  \n",
       "61  0.994809  0.827094  0.851354  0.843518  \n",
       "58  0.995374  0.823699  0.834863  0.831108  \n",
       "55  0.995481  0.820110  0.825959  0.833727  \n",
       "56  0.995355  0.807689  0.810778  0.819124  \n",
       "54  0.992172  0.821132  0.846978  0.825889  \n",
       "70  0.995917  0.831604  0.848871  0.843631  \n",
       "68  0.995408  0.821614  0.839846  0.833980  \n",
       "63  0.994820  0.835899  0.845961  0.837082  \n",
       "62  0.991060  0.809171  0.816202  0.808539  \n",
       "64  0.995972  0.815201  0.816830  0.823648  \n",
       "57  0.996155  0.818509  0.840425  0.852010  \n",
       "29  0.995893  0.802237  0.815528  0.807952  \n",
       "37  0.996288  0.818332  0.814715  0.815823  \n",
       "23  0.995134  0.813365  0.843567  0.837692  \n",
       "39  0.994722  0.796587  0.780807  0.751619  \n",
       "18  0.993344  0.802652  0.835804  0.798312  \n",
       "7   0.992859  0.807523  0.819547  0.813614  \n",
       "27  0.993349  0.794069  0.807127  0.788786  \n",
       "26  0.991585  0.811223  0.825079  0.821793  \n",
       "40  0.995408  0.804334  0.833286  0.827636  \n",
       "50  0.995883  0.792475  0.832953  0.808690  \n",
       "20  0.995025  0.811626  0.828520  0.822260  \n",
       "45  0.991836  0.784101  0.790127  0.740799  \n",
       "48  0.993884  0.791720  0.840746  0.794139  \n",
       "44  0.990123  0.816217  0.805659  0.824301  \n",
       "71  0.995389  0.806570  0.845419  0.829573  \n",
       "28  0.994971  0.803225  0.828606  0.824020  \n",
       "47  0.995467  0.808363  0.829791  0.823269  \n",
       "34  0.995051  0.783159  0.840469  0.826030  \n",
       "..       ...       ...       ...       ...  \n",
       "31  0.994264  0.790629  0.776794  0.798461  \n",
       "21  0.995970  0.807650  0.824122  0.823838  \n",
       "60  0.991282  0.778757  0.799489  0.742067  \n",
       "17  0.996218  0.797222  0.815309  0.810516  \n",
       "38  0.996178  0.789520  0.764866  0.822953  \n",
       "32  0.995460  0.807557  0.819218  0.813403  \n",
       "49  0.995689  0.777983  0.844907  0.795348  \n",
       "19  0.993348  0.776299  0.787699  0.784175  \n",
       "9   0.995328  0.797492  0.828315  0.831142  \n",
       "2   0.994202  0.777533  0.752149  0.844704  \n",
       "25  0.996687  0.806246  0.830782  0.827865  \n",
       "42  0.993086  0.797152  0.782825  0.825374  \n",
       "69  0.993320  0.792416  0.782156  0.833675  \n",
       "0   0.993768  0.780229  0.744064  0.805368  \n",
       "15  0.995164  0.777106  0.762458  0.836195  \n",
       "3   0.992718  0.794706  0.796743  0.786574  \n",
       "43  0.996453  0.791053  0.833452  0.826392  \n",
       "65  0.995060  0.789468  0.826526  0.808840  \n",
       "22  0.994825  0.791488  0.773357  0.810508  \n",
       "8   0.996067  0.802167  0.808498  0.805656  \n",
       "46  0.992602  0.761938  0.822676  0.801886  \n",
       "41  0.995426  0.742947  0.825055  0.837594  \n",
       "10  0.993799  0.789505  0.808477  0.781285  \n",
       "24  0.990993  0.786643  0.805821  0.768787  \n",
       "36  0.995960  0.723195  0.817109  0.825188  \n",
       "66  0.993133  0.748033  0.802822  0.755055  \n",
       "53  0.993331  0.707024  0.806870  0.806477  \n",
       "51  0.995112  0.678361  0.816866  0.802167  \n",
       "52  0.987312  0.717716  0.798132  0.743359  \n",
       "33  0.995076  0.685615  0.792096  0.784187  \n",
       "\n",
       "[66 rows x 13 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "eval_df_sort"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "top_num = 1\n",
      "[[9.99249704e-01 0.00000000e+00 2.85438750e-04 4.64857393e-04]\n",
      " [9.68858131e-04 8.50768166e-01 1.38989619e-01 9.27335640e-03]\n",
      " [2.47586994e-03 1.39546219e-01 8.31238515e-01 2.67393953e-02]\n",
      " [6.24177624e-02 2.07506896e-02 1.79085566e-01 7.37745982e-01]]\n",
      "[0.99231927 0.82508269 0.83124655 0.80908148]\n",
      "top_num = 3\n",
      "[[9.99133490e-01 0.00000000e+00 1.38641679e-04 7.27868812e-04]\n",
      " [7.91695502e-04 8.28844291e-01 1.59031142e-01 1.13328720e-02]\n",
      " [2.41784174e-03 1.07479835e-01 8.59769048e-01 3.03332753e-02]\n",
      " [4.26027544e-02 1.65498689e-02 1.48646673e-01 7.92200704e-01]]\n",
      "[0.99435505 0.83214754 0.84699829 0.83739691]\n",
      "top_num = 5\n",
      "[[9.98939799e-01 0.00000000e+00 1.73302098e-04 8.86898973e-04]\n",
      " [1.77162630e-04 8.45115571e-01 1.44348789e-01 1.03584775e-02]\n",
      " [1.66734366e-03 1.23236426e-01 8.41722277e-01 3.33739531e-02]\n",
      " [3.52147682e-02 2.22419322e-02 1.41921461e-01 8.00621838e-01]]\n",
      "[0.99533165 0.83094183 0.84203988 0.83958605]\n",
      "top_num = 7\n",
      "[[9.98968343e-01 0.00000000e+00 1.54952464e-04 8.76704732e-04]\n",
      " [2.26989619e-04 8.31490657e-01 1.55022837e-01 1.32595156e-02]\n",
      " [1.52807598e-03 1.03943983e-01 8.57772878e-01 3.67550630e-02]\n",
      " [3.53414751e-02 1.38597842e-02 1.29533426e-01 8.21265314e-01]]\n",
      "[0.99536009 0.83645963 0.85020476 0.84605344]\n",
      "top_num = 9\n",
      "[[9.98774652e-01 0.00000000e+00 1.12136652e-04 1.11321112e-03]\n",
      " [1.16262976e-04 8.22305882e-01 1.62651903e-01 1.49259516e-02]\n",
      " [1.50486470e-03 9.17077700e-02 8.65742084e-01 4.10452813e-02]\n",
      " [3.12868546e-02 1.14523533e-02 1.19309155e-01 8.37951637e-01]]\n",
      "[0.99571021 0.83906393 0.85422551 0.84943485]\n",
      "top_num = 11\n",
      "[[9.98829701e-01 0.00000000e+00 8.97093214e-05 1.08058955e-03]\n",
      " [7.75086505e-05 8.27847751e-01 1.58289273e-01 1.37854671e-02]\n",
      " [1.58610418e-03 9.99284319e-02 8.57846380e-01 4.06390839e-02]\n",
      " [3.22517763e-02 1.18422207e-02 1.23061628e-01 8.32844375e-01]]\n",
      "[0.99562343 0.8372317  0.85041946 0.84782188]\n",
      "top_num = 13\n",
      "[[9.98727759e-01 0.00000000e+00 8.97093214e-05 1.18253196e-03]\n",
      " [6.64359862e-05 8.25799308e-01 1.60675433e-01 1.34588235e-02]\n",
      " [1.28822608e-03 9.76692006e-02 8.60689762e-01 4.03528115e-02]\n",
      " [2.85090498e-02 1.16570337e-02 1.23490482e-01 8.36343434e-01]]\n",
      "[0.99604104 0.83744119 0.85126312 0.85021972]\n",
      "top_num = 15\n",
      "[[9.98619700e-01 0.00000000e+00 7.33985357e-05 1.30690171e-03]\n",
      " [3.87543253e-05 8.23800692e-01 1.61948789e-01 1.42117647e-02]\n",
      " [1.14508985e-03 9.44815180e-02 8.62419002e-01 4.19543898e-02]\n",
      " [2.57507383e-02 1.09552725e-02 1.19133715e-01 8.44160274e-01]]\n",
      "[0.99631616 0.83837271 0.85261131 0.85220061]\n"
     ]
    }
   ],
   "source": [
    "## Majority vote for the top ones\n",
    "\n",
    "for top_num in [1,3,5,7,9,11,13,15]:\n",
    "    im_arr = np.zeros([96,112,96,top_num])\n",
    "    for i in range(top_num):\n",
    "        im_arr[:,:,:,i] = sitk.GetArrayFromImage(\n",
    "                sitk.ReadImage(\n",
    "                    os.path.join(eval_df_sort.iloc[i,1],'1_niftynet_out.nii.gz')))\n",
    "\n",
    "    true_arr = sitk.GetArrayFromImage(\n",
    "        sitk.ReadImage('/proj/NIRAL/users/siyangj/myData/min_normal/seg_inf_1/1.nrrd'))\n",
    "\n",
    "    label_vote_arr = np.zeros([96,112,96,4])\n",
    "    for i in range(4):\n",
    "        label_vote_arr[:,:,:,i] = (im_arr==i).sum(axis=3)\n",
    "\n",
    "    res_arr = label_vote_arr.argmax(axis=3)\n",
    "\n",
    "    res_arr.shape\n",
    "\n",
    "    c_mat = skl.metrics.confusion_matrix(true_arr.flatten(),res_arr.flatten())\n",
    "    p_mat = c_mat.T / c_mat.astype(np.float).sum(axis=1)\n",
    "    p_mat = p_mat.T\n",
    "    print('top_num = {:d}'.format(top_num))\n",
    "    print(p_mat)\n",
    "    print(Dice(res_arr,true_arr))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "arr1 = sitk.GetArrayFromImage(\n",
    "        sitk.ReadImage('/proj/NIRAL/users/siyangj/myData/min_normal/seg_inf_1/1.nrrd'))\n",
    "arr2 = sitk.GetArrayFromImage(\n",
    "        sitk.ReadImage('/proj/NIRAL/users/siyangj/myData/min_normal/seg/1.nrrd'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4282"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(arr1 != arr2).sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.004148453000992063"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "4282/(96*96*112)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}