View Predictions.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "\n",
    "np.random.seed(30) # fix random seed\n",
    "os.environ[\"CUDA_VISIBLE_DEVICES\"] = '1' # fix cuda device to be used"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Reading data... Data read. There are 96 series in total.\n",
      "Standardizing data... Data standardized.\n",
      "Separating data into train/valid... Data separated into train/valid.\n"
     ]
    }
   ],
   "source": [
    "## Read ALL data, standardize, and separate into train/valid\n",
    "\n",
    "from helpers import read_and_preprocess_data\n",
    "\n",
    "train, valid = read_and_preprocess_data()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Restore model from disk"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Restoring model from disk, path=`./results/1538524287`\n",
      "INFO:tensorflow:Restoring parameters from /home/arthurcgusmao/Projects/eeg-grasp-and-lift/results/1538524287/tf-model\n"
     ]
    }
   ],
   "source": [
    "from model import Model\n",
    "from model_defs import conv_nn_2, conv_nn_3, conv_batches_gen\n",
    "\n",
    "nn = Model('./results/1538524287')\n",
    "nn.set_data(train, valid)\n",
    "# nn.validate(4096, conv_batches_gen)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plot predictions and correct labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "valid_series = valid[3]\n",
    "valid_x = valid_series[0]\n",
    "valid_y = valid_series[1]\n",
    "preds_y = nn.predict([(valid_x, valid_y)], 4096, conv_batches_gen)\n",
    "\n",
    "preds_y = np.array(preds_y)\n",
    "valid_y = np.array(valid_y[1023:])\n",
    "# we drop the first 1023 data points because we are not able to predict them with a window size of 1024"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAzoAAAKvCAYAAACibbrwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ff5e0a25d90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "xlim_l = 80000\n",
    "xlim_r = 86000\n",
    "delta = xlim_r - xlim_l\n",
    "\n",
    "fig = plt.figure(figsize=(0.00233*delta,12))\n",
    "cols = valid[0][1].columns\n",
    "first_ax = None\n",
    "\n",
    "for idx,col in enumerate(cols):\n",
    "    if first_ax == None:\n",
    "        ax = fig.add_subplot(2*len(cols),1,2*idx+1)\n",
    "        first_ax = ax\n",
    "    else: \n",
    "        ax = fig.add_subplot(2*len(cols),1,2*idx+1, sharex=first_ax)\n",
    "    ax.plot(valid_y[:,idx], label=col+' label')\n",
    "    ax.set_xlim(xlim_l,xlim_r)\n",
    "    ax.set_ylim(0,1.2)\n",
    "    ax.legend()\n",
    "    \n",
    "    ax = fig.add_subplot(2*len(cols),1,2*idx+2, sharex=first_ax)\n",
    "    ax.plot(preds_y[:,idx], label=col+' preds')\n",
    "    ax.set_xlim(xlim_l,xlim_r)\n",
    "    ax.set_ylim(0,1.2)\n",
    "    \n",
    "    ax.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Calculate AUC separately"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Run\n",
      "Run\n",
      "Run\n",
      "Run\n",
      "Run\n",
      "Run\n",
      "Run\n",
      "Run\n",
      "Run\n",
      "Run\n",
      "Run\n",
      "Run\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import roc_auc_score\n",
    "\n",
    "partial_results = [] # indexed by subject id\n",
    "\n",
    "for x,y in valid:\n",
    "    preds = nn.predict([(x, y)], 4096, conv_batches_gen)\n",
    "    mean_auc = roc_auc_score(y[1023:], preds, average=None)\n",
    "    partial_results.append(mean_auc)\n",
    "    print(\"Run\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "df = pd.DataFrame(partial_results, columns=['HandStart', 'FirstDigitTouch', 'BothStartLoadPhase', 'LiftOff', 'Replace', 'BothReleased'])\n",
    "df.index = ['subject_1', 'subject_2', 'subject_3', 'subject_4', 'subject_5', 'subject_6', 'subject_7', 'subject_8', 'subject_9', 'subject_10', 'subject_11', 'subject_12']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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>HandStart</th>\n",
       "      <th>FirstDigitTouch</th>\n",
       "      <th>BothStartLoadPhase</th>\n",
       "      <th>LiftOff</th>\n",
       "      <th>Replace</th>\n",
       "      <th>BothReleased</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>subject_1</th>\n",
       "      <td>0.864008</td>\n",
       "      <td>0.846042</td>\n",
       "      <td>0.867201</td>\n",
       "      <td>0.794682</td>\n",
       "      <td>0.879466</td>\n",
       "      <td>0.862761</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_2</th>\n",
       "      <td>0.901226</td>\n",
       "      <td>0.952016</td>\n",
       "      <td>0.968780</td>\n",
       "      <td>0.941687</td>\n",
       "      <td>0.896037</td>\n",
       "      <td>0.916459</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_3</th>\n",
       "      <td>0.957713</td>\n",
       "      <td>0.962371</td>\n",
       "      <td>0.955505</td>\n",
       "      <td>0.920559</td>\n",
       "      <td>0.958759</td>\n",
       "      <td>0.963216</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_4</th>\n",
       "      <td>0.897950</td>\n",
       "      <td>0.973719</td>\n",
       "      <td>0.979499</td>\n",
       "      <td>0.968707</td>\n",
       "      <td>0.927034</td>\n",
       "      <td>0.893260</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_5</th>\n",
       "      <td>0.952513</td>\n",
       "      <td>0.970425</td>\n",
       "      <td>0.964235</td>\n",
       "      <td>0.920775</td>\n",
       "      <td>0.937080</td>\n",
       "      <td>0.905560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_6</th>\n",
       "      <td>0.794479</td>\n",
       "      <td>0.905650</td>\n",
       "      <td>0.925436</td>\n",
       "      <td>0.916274</td>\n",
       "      <td>0.772466</td>\n",
       "      <td>0.825394</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_7</th>\n",
       "      <td>0.916921</td>\n",
       "      <td>0.958898</td>\n",
       "      <td>0.977873</td>\n",
       "      <td>0.975841</td>\n",
       "      <td>0.943662</td>\n",
       "      <td>0.945678</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_8</th>\n",
       "      <td>0.854580</td>\n",
       "      <td>0.953829</td>\n",
       "      <td>0.972916</td>\n",
       "      <td>0.972331</td>\n",
       "      <td>0.972535</td>\n",
       "      <td>0.957155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_9</th>\n",
       "      <td>0.796382</td>\n",
       "      <td>0.928739</td>\n",
       "      <td>0.920870</td>\n",
       "      <td>0.876802</td>\n",
       "      <td>0.950319</td>\n",
       "      <td>0.928125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_10</th>\n",
       "      <td>0.825524</td>\n",
       "      <td>0.942940</td>\n",
       "      <td>0.938562</td>\n",
       "      <td>0.937935</td>\n",
       "      <td>0.927681</td>\n",
       "      <td>0.864203</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_11</th>\n",
       "      <td>0.804732</td>\n",
       "      <td>0.842774</td>\n",
       "      <td>0.853059</td>\n",
       "      <td>0.831441</td>\n",
       "      <td>0.749318</td>\n",
       "      <td>0.792366</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_12</th>\n",
       "      <td>0.715163</td>\n",
       "      <td>0.806805</td>\n",
       "      <td>0.779347</td>\n",
       "      <td>0.788293</td>\n",
       "      <td>0.863270</td>\n",
       "      <td>0.851891</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            HandStart  FirstDigitTouch  BothStartLoadPhase   LiftOff  \\\n",
       "subject_1    0.864008         0.846042            0.867201  0.794682   \n",
       "subject_2    0.901226         0.952016            0.968780  0.941687   \n",
       "subject_3    0.957713         0.962371            0.955505  0.920559   \n",
       "subject_4    0.897950         0.973719            0.979499  0.968707   \n",
       "subject_5    0.952513         0.970425            0.964235  0.920775   \n",
       "subject_6    0.794479         0.905650            0.925436  0.916274   \n",
       "subject_7    0.916921         0.958898            0.977873  0.975841   \n",
       "subject_8    0.854580         0.953829            0.972916  0.972331   \n",
       "subject_9    0.796382         0.928739            0.920870  0.876802   \n",
       "subject_10   0.825524         0.942940            0.938562  0.937935   \n",
       "subject_11   0.804732         0.842774            0.853059  0.831441   \n",
       "subject_12   0.715163         0.806805            0.779347  0.788293   \n",
       "\n",
       "             Replace  BothReleased  \n",
       "subject_1   0.879466      0.862761  \n",
       "subject_2   0.896037      0.916459  \n",
       "subject_3   0.958759      0.963216  \n",
       "subject_4   0.927034      0.893260  \n",
       "subject_5   0.937080      0.905560  \n",
       "subject_6   0.772466      0.825394  \n",
       "subject_7   0.943662      0.945678  \n",
       "subject_8   0.972535      0.957155  \n",
       "subject_9   0.950319      0.928125  \n",
       "subject_10  0.927681      0.864203  \n",
       "subject_11  0.749318      0.792366  \n",
       "subject_12  0.863270      0.851891  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>HandStart</th>\n",
       "      <th>FirstDigitTouch</th>\n",
       "      <th>BothStartLoadPhase</th>\n",
       "      <th>LiftOff</th>\n",
       "      <th>Replace</th>\n",
       "      <th>BothReleased</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>12.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>12.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.856766</td>\n",
       "      <td>0.920351</td>\n",
       "      <td>0.925273</td>\n",
       "      <td>0.903777</td>\n",
       "      <td>0.898135</td>\n",
       "      <td>0.892172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.072593</td>\n",
       "      <td>0.057176</td>\n",
       "      <td>0.062070</td>\n",
       "      <td>0.066495</td>\n",
       "      <td>0.071820</td>\n",
       "      <td>0.053840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.715163</td>\n",
       "      <td>0.806805</td>\n",
       "      <td>0.779347</td>\n",
       "      <td>0.788293</td>\n",
       "      <td>0.749318</td>\n",
       "      <td>0.792366</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.802645</td>\n",
       "      <td>0.890748</td>\n",
       "      <td>0.907453</td>\n",
       "      <td>0.865462</td>\n",
       "      <td>0.875417</td>\n",
       "      <td>0.860043</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.859294</td>\n",
       "      <td>0.947478</td>\n",
       "      <td>0.947033</td>\n",
       "      <td>0.920667</td>\n",
       "      <td>0.927357</td>\n",
       "      <td>0.899410</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.905150</td>\n",
       "      <td>0.959766</td>\n",
       "      <td>0.969814</td>\n",
       "      <td>0.948442</td>\n",
       "      <td>0.945326</td>\n",
       "      <td>0.932513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>0.957713</td>\n",
       "      <td>0.973719</td>\n",
       "      <td>0.979499</td>\n",
       "      <td>0.975841</td>\n",
       "      <td>0.972535</td>\n",
       "      <td>0.963216</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       HandStart  FirstDigitTouch  BothStartLoadPhase    LiftOff    Replace  \\\n",
       "count  12.000000        12.000000           12.000000  12.000000  12.000000   \n",
       "mean    0.856766         0.920351            0.925273   0.903777   0.898135   \n",
       "std     0.072593         0.057176            0.062070   0.066495   0.071820   \n",
       "min     0.715163         0.806805            0.779347   0.788293   0.749318   \n",
       "25%     0.802645         0.890748            0.907453   0.865462   0.875417   \n",
       "50%     0.859294         0.947478            0.947033   0.920667   0.927357   \n",
       "75%     0.905150         0.959766            0.969814   0.948442   0.945326   \n",
       "max     0.957713         0.973719            0.979499   0.975841   0.972535   \n",
       "\n",
       "       BothReleased  \n",
       "count     12.000000  \n",
       "mean       0.892172  \n",
       "std        0.053840  \n",
       "min        0.792366  \n",
       "25%        0.860043  \n",
       "50%        0.899410  \n",
       "75%        0.932513  \n",
       "max        0.963216  "
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>subject_1</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.852360</td>\n",
       "      <td>0.030218</td>\n",
       "      <td>0.794682</td>\n",
       "      <td>0.850222</td>\n",
       "      <td>0.863384</td>\n",
       "      <td>0.866403</td>\n",
       "      <td>0.879466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_2</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.929368</td>\n",
       "      <td>0.029284</td>\n",
       "      <td>0.896037</td>\n",
       "      <td>0.905034</td>\n",
       "      <td>0.929073</td>\n",
       "      <td>0.949434</td>\n",
       "      <td>0.968780</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_3</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.953020</td>\n",
       "      <td>0.016163</td>\n",
       "      <td>0.920559</td>\n",
       "      <td>0.956057</td>\n",
       "      <td>0.958236</td>\n",
       "      <td>0.961468</td>\n",
       "      <td>0.963216</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_4</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.940028</td>\n",
       "      <td>0.039095</td>\n",
       "      <td>0.893260</td>\n",
       "      <td>0.905221</td>\n",
       "      <td>0.947871</td>\n",
       "      <td>0.972466</td>\n",
       "      <td>0.979499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_5</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.941765</td>\n",
       "      <td>0.025358</td>\n",
       "      <td>0.905560</td>\n",
       "      <td>0.924851</td>\n",
       "      <td>0.944796</td>\n",
       "      <td>0.961304</td>\n",
       "      <td>0.970425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_6</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.856616</td>\n",
       "      <td>0.067256</td>\n",
       "      <td>0.772466</td>\n",
       "      <td>0.802208</td>\n",
       "      <td>0.865522</td>\n",
       "      <td>0.913618</td>\n",
       "      <td>0.925436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_7</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.953145</td>\n",
       "      <td>0.022880</td>\n",
       "      <td>0.916921</td>\n",
       "      <td>0.944166</td>\n",
       "      <td>0.952288</td>\n",
       "      <td>0.971605</td>\n",
       "      <td>0.977873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_8</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.947224</td>\n",
       "      <td>0.046165</td>\n",
       "      <td>0.854580</td>\n",
       "      <td>0.954661</td>\n",
       "      <td>0.964743</td>\n",
       "      <td>0.972484</td>\n",
       "      <td>0.972916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_9</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.900206</td>\n",
       "      <td>0.056319</td>\n",
       "      <td>0.796382</td>\n",
       "      <td>0.887819</td>\n",
       "      <td>0.924497</td>\n",
       "      <td>0.928585</td>\n",
       "      <td>0.950319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_10</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.906141</td>\n",
       "      <td>0.049271</td>\n",
       "      <td>0.825524</td>\n",
       "      <td>0.880073</td>\n",
       "      <td>0.932808</td>\n",
       "      <td>0.938406</td>\n",
       "      <td>0.942940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_11</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.812282</td>\n",
       "      <td>0.038413</td>\n",
       "      <td>0.749318</td>\n",
       "      <td>0.795458</td>\n",
       "      <td>0.818087</td>\n",
       "      <td>0.839941</td>\n",
       "      <td>0.853059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subject_12</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.800795</td>\n",
       "      <td>0.053850</td>\n",
       "      <td>0.715163</td>\n",
       "      <td>0.781584</td>\n",
       "      <td>0.797549</td>\n",
       "      <td>0.840620</td>\n",
       "      <td>0.863270</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            count      mean       std       min       25%       50%       75%  \\\n",
       "subject_1     6.0  0.852360  0.030218  0.794682  0.850222  0.863384  0.866403   \n",
       "subject_2     6.0  0.929368  0.029284  0.896037  0.905034  0.929073  0.949434   \n",
       "subject_3     6.0  0.953020  0.016163  0.920559  0.956057  0.958236  0.961468   \n",
       "subject_4     6.0  0.940028  0.039095  0.893260  0.905221  0.947871  0.972466   \n",
       "subject_5     6.0  0.941765  0.025358  0.905560  0.924851  0.944796  0.961304   \n",
       "subject_6     6.0  0.856616  0.067256  0.772466  0.802208  0.865522  0.913618   \n",
       "subject_7     6.0  0.953145  0.022880  0.916921  0.944166  0.952288  0.971605   \n",
       "subject_8     6.0  0.947224  0.046165  0.854580  0.954661  0.964743  0.972484   \n",
       "subject_9     6.0  0.900206  0.056319  0.796382  0.887819  0.924497  0.928585   \n",
       "subject_10    6.0  0.906141  0.049271  0.825524  0.880073  0.932808  0.938406   \n",
       "subject_11    6.0  0.812282  0.038413  0.749318  0.795458  0.818087  0.839941   \n",
       "subject_12    6.0  0.800795  0.053850  0.715163  0.781584  0.797549  0.840620   \n",
       "\n",
       "                 max  \n",
       "subject_1   0.879466  \n",
       "subject_2   0.968780  \n",
       "subject_3   0.963216  \n",
       "subject_4   0.979499  \n",
       "subject_5   0.970425  \n",
       "subject_6   0.925436  \n",
       "subject_7   0.977873  \n",
       "subject_8   0.972916  \n",
       "subject_9   0.950319  \n",
       "subject_10  0.942940  \n",
       "subject_11  0.853059  \n",
       "subject_12  0.863270  "
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.apply(pd.Series.describe, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now calculating statistics for all cells:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.stack().describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Debug"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(valid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>HandStart</th>\n",
       "      <th>FirstDigitTouch</th>\n",
       "      <th>BothStartLoadPhase</th>\n",
       "      <th>LiftOff</th>\n",
       "      <th>Replace</th>\n",
       "      <th>BothReleased</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1023</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1024</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1025</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1026</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1027</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1028</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1029</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1030</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1031</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1032</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1033</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1034</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1035</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1036</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1037</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1038</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1039</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1040</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1041</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1042</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1043</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1044</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1045</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1046</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1047</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1048</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1049</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1050</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1051</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_1052</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134259</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134260</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134261</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134262</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134263</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134264</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134265</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134266</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134267</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134268</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134269</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134270</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134271</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134272</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134273</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134274</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134275</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134276</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134277</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134278</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134279</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134280</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134281</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134282</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134283</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134284</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134285</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134286</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134287</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>subj11_series8_134288</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>133266 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                       HandStart  FirstDigitTouch  BothStartLoadPhase  \\\n",
       "id                                                                      \n",
       "subj11_series8_1023            0                0                   0   \n",
       "subj11_series8_1024            0                0                   0   \n",
       "subj11_series8_1025            0                0                   0   \n",
       "subj11_series8_1026            0                0                   0   \n",
       "subj11_series8_1027            0                0                   0   \n",
       "subj11_series8_1028            0                0                   0   \n",
       "subj11_series8_1029            0                0                   0   \n",
       "subj11_series8_1030            0                0                   0   \n",
       "subj11_series8_1031            0                0                   0   \n",
       "subj11_series8_1032            0                0                   0   \n",
       "subj11_series8_1033            0                0                   0   \n",
       "subj11_series8_1034            0                0                   0   \n",
       "subj11_series8_1035            0                0                   0   \n",
       "subj11_series8_1036            0                0                   0   \n",
       "subj11_series8_1037            0                0                   0   \n",
       "subj11_series8_1038            0                0                   0   \n",
       "subj11_series8_1039            0                0                   0   \n",
       "subj11_series8_1040            0                0                   0   \n",
       "subj11_series8_1041            0                0                   0   \n",
       "subj11_series8_1042            0                0                   0   \n",
       "subj11_series8_1043            0                0                   0   \n",
       "subj11_series8_1044            0                0                   0   \n",
       "subj11_series8_1045            0                0                   0   \n",
       "subj11_series8_1046            0                0                   0   \n",
       "subj11_series8_1047            0                0                   0   \n",
       "subj11_series8_1048            0                0                   0   \n",
       "subj11_series8_1049            0                0                   0   \n",
       "subj11_series8_1050            0                0                   0   \n",
       "subj11_series8_1051            0                0                   0   \n",
       "subj11_series8_1052            0                0                   0   \n",
       "...                          ...              ...                 ...   \n",
       "subj11_series8_134259          0                0                   0   \n",
       "subj11_series8_134260          0                0                   0   \n",
       "subj11_series8_134261          0                0                   0   \n",
       "subj11_series8_134262          0                0                   0   \n",
       "subj11_series8_134263          0                0                   0   \n",
       "subj11_series8_134264          0                0                   0   \n",
       "subj11_series8_134265          0                0                   0   \n",
       "subj11_series8_134266          0                0                   0   \n",
       "subj11_series8_134267          0                0                   0   \n",
       "subj11_series8_134268          0                0                   0   \n",
       "subj11_series8_134269          0                0                   0   \n",
       "subj11_series8_134270          0                0                   0   \n",
       "subj11_series8_134271          0                0                   0   \n",
       "subj11_series8_134272          0                0                   0   \n",
       "subj11_series8_134273          0                0                   0   \n",
       "subj11_series8_134274          0                0                   0   \n",
       "subj11_series8_134275          0                0                   0   \n",
       "subj11_series8_134276          0                0                   0   \n",
       "subj11_series8_134277          0                0                   0   \n",
       "subj11_series8_134278          0                0                   0   \n",
       "subj11_series8_134279          0                0                   0   \n",
       "subj11_series8_134280          0                0                   0   \n",
       "subj11_series8_134281          0                0                   0   \n",
       "subj11_series8_134282          0                0                   0   \n",
       "subj11_series8_134283          0                0                   0   \n",
       "subj11_series8_134284          0                0                   0   \n",
       "subj11_series8_134285          0                0                   0   \n",
       "subj11_series8_134286          0                0                   0   \n",
       "subj11_series8_134287          0                0                   0   \n",
       "subj11_series8_134288          0                0                   0   \n",
       "\n",
       "                       LiftOff  Replace  BothReleased  \n",
       "id                                                     \n",
       "subj11_series8_1023          0        0             0  \n",
       "subj11_series8_1024          0        0             0  \n",
       "subj11_series8_1025          0        0             0  \n",
       "subj11_series8_1026          0        0             0  \n",
       "subj11_series8_1027          0        0             0  \n",
       "subj11_series8_1028          0        0             0  \n",
       "subj11_series8_1029          0        0             0  \n",
       "subj11_series8_1030          0        0             0  \n",
       "subj11_series8_1031          0        0             0  \n",
       "subj11_series8_1032          0        0             0  \n",
       "subj11_series8_1033          0        0             0  \n",
       "subj11_series8_1034          0        0             0  \n",
       "subj11_series8_1035          0        0             0  \n",
       "subj11_series8_1036          0        0             0  \n",
       "subj11_series8_1037          0        0             0  \n",
       "subj11_series8_1038          0        0             0  \n",
       "subj11_series8_1039          0        0             0  \n",
       "subj11_series8_1040          0        0             0  \n",
       "subj11_series8_1041          0        0             0  \n",
       "subj11_series8_1042          0        0             0  \n",
       "subj11_series8_1043          0        0             0  \n",
       "subj11_series8_1044          0        0             0  \n",
       "subj11_series8_1045          0        0             0  \n",
       "subj11_series8_1046          0        0             0  \n",
       "subj11_series8_1047          0        0             0  \n",
       "subj11_series8_1048          0        0             0  \n",
       "subj11_series8_1049          0        0             0  \n",
       "subj11_series8_1050          0        0             0  \n",
       "subj11_series8_1051          0        0             0  \n",
       "subj11_series8_1052          0        0             0  \n",
       "...                        ...      ...           ...  \n",
       "subj11_series8_134259        0        0             0  \n",
       "subj11_series8_134260        0        0             0  \n",
       "subj11_series8_134261        0        0             0  \n",
       "subj11_series8_134262        0        0             0  \n",
       "subj11_series8_134263        0        0             0  \n",
       "subj11_series8_134264        0        0             0  \n",
       "subj11_series8_134265        0        0             0  \n",
       "subj11_series8_134266        0        0             0  \n",
       "subj11_series8_134267        0        0             0  \n",
       "subj11_series8_134268        0        0             0  \n",
       "subj11_series8_134269        0        0             0  \n",
       "subj11_series8_134270        0        0             0  \n",
       "subj11_series8_134271        0        0             0  \n",
       "subj11_series8_134272        0        0             0  \n",
       "subj11_series8_134273        0        0             0  \n",
       "subj11_series8_134274        0        0             0  \n",
       "subj11_series8_134275        0        0             0  \n",
       "subj11_series8_134276        0        0             0  \n",
       "subj11_series8_134277        0        0             0  \n",
       "subj11_series8_134278        0        0             0  \n",
       "subj11_series8_134279        0        0             0  \n",
       "subj11_series8_134280        0        0             0  \n",
       "subj11_series8_134281        0        0             0  \n",
       "subj11_series8_134282        0        0             0  \n",
       "subj11_series8_134283        0        0             0  \n",
       "subj11_series8_134284        0        0             0  \n",
       "subj11_series8_134285        0        0             0  \n",
       "subj11_series8_134286        0        0             0  \n",
       "subj11_series8_134287        0        0             0  \n",
       "subj11_series8_134288        0        0             0  \n",
       "\n",
       "[133266 rows x 6 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "valid[0][1][1023:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.5278880e-03, 1.9892847e-02, 1.3149024e-02, 1.3340203e-03,\n",
       "        3.2382278e-04, 7.5215957e-04],\n",
       "       [1.4925188e-03, 1.9966707e-02, 1.4492883e-02, 1.3846192e-03,\n",
       "        2.5933399e-04, 7.2086893e-04],\n",
       "       [1.6619033e-03, 2.1948349e-02, 1.6930113e-02, 1.6569407e-03,\n",
       "        2.9008201e-04, 7.9023757e-04],\n",
       "       ...,\n",
       "       [5.7007857e-03, 1.1026986e-04, 2.1352037e-04, 6.0135481e-04,\n",
       "        3.2178956e-05, 1.1429013e-05],\n",
       "       [8.5503999e-03, 8.0567683e-05, 1.7943539e-04, 6.2627235e-04,\n",
       "        4.6632220e-05, 1.3431395e-05],\n",
       "       [1.1009806e-02, 7.9406476e-05, 1.8507462e-04, 6.4887857e-04,\n",
       "        5.2054475e-05, 1.5220159e-05]], dtype=float32)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preds_y"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}