Ensemble.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import re\n",
    "from datetime import datetime, timedelta\n",
    "import random\n",
    "import time\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import warnings\n",
    "\n",
    "# tensorflow\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Dropout, LSTM, GRU \n",
    "from tensorflow.keras.layers import BatchNormalization\n",
    "from tensorflow.keras.optimizers import Adam, RMSprop, Adagrad\n",
    "from scikeras.wrappers import KerasClassifier\n",
    "from keras.callbacks import CSVLogger, EarlyStopping, ModelCheckpoint\n",
    "\n",
    "# sklearn\n",
    "from sklearn.preprocessing import StandardScaler, MinMaxScaler\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import recall_score, precision_score, accuracy_score\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# other\n",
    "# import tqdm notebook\n",
    "from tqdm.notebook import tqdm\n",
    "\n",
    "import multiprocessing as mp\n",
    "\n",
    "#### Matplotlib settings\n",
    "%matplotlib inline\n",
    "import matplotlib as mpl\n",
    "\n",
    "# specify default rcParams so that fontsize, weight and style don't need to be set each time\n",
    "# Title in bold, fontsize 20\n",
    "mpl.rcParams['figure.titleweight'] = 'bold'\n",
    "mpl.rcParams['figure.titlesize'] = 20\n",
    "mpl.rcParams['axes.titleweight'] = 'bold'\n",
    "mpl.rcParams['axes.titlesize'] = 20\n",
    "# Plot fontsize 16 and bold\n",
    "mpl.rcParams['axes.labelweight'] = 'bold'\n",
    "mpl.rcParams['axes.labelsize'] = 16\n",
    "# set figure size\n",
    "mpl.rcParams['figure.figsize'] = (18, 8)\n",
    "# set grid on\n",
    "mpl.rcParams['axes.grid'] = True\n",
    "# set grid linestyle\n",
    "mpl.rcParams['grid.linestyle'] = '--'\n",
    "# set axis labels fontsize\n",
    "mpl.rcParams['xtick.labelsize'] = 14\n",
    "mpl.rcParams['ytick.labelsize'] = 14\n",
    "\n",
    "# race list and colors that will be used to represent them\n",
    "RACE_LIST = [\n",
    "    'Protoss',\n",
    "    'Terran',\n",
    "    'Zerg'\n",
    "    ]\n",
    "COLOR_DICT = {\n",
    "    'Protoss': 'goldenrod',\n",
    "    'Terran': 'firebrick',\n",
    "    'Zerg': 'darkviolet'\n",
    "    }\n",
    "\n",
    "#### Pandas options\n",
    "# these are set to allow better exploration of the large dataframes in this \n",
    "# notebook\n",
    "pd.options.display.max_rows = 200\n",
    "pd.options.display.max_columns = 200\n",
    "\n",
    "#IPython\n",
    "from IPython.display import display\n",
    "%load_ext autoreload\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "class BatchGenerator(keras.utils.Sequence):\n",
    "    \n",
    "    def __init__(\n",
    "        self, \n",
    "        list_filehashes, \n",
    "        labels,\n",
    "        data_dir,\n",
    "        window_size,\n",
    "        n_channels,\n",
    "        window_start=0,\n",
    "        batch_size=32,  \n",
    "        n_classes=2, \n",
    "        shuffle=True\n",
    "    ):\n",
    "        \"\"\"\n",
    "        Generate batches of data for training.\n",
    "    \n",
    "        Args:\n",
    "            data (str): the path to a directory of npy files\n",
    "            labels (dict): a dictionary of filehashes and their corresponding labels\n",
    "            list_filehashes (list): the list of filehashes to use\n",
    "            window_size (int): the number of timesteps to use as a window\n",
    "            n_channels (int): the number of features in the data\n",
    "            batch_size (int): the number of samples per batch. Defaults to 32.\n",
    "            n_classes (int): the number of classes for the target. Defaults to 2.\n",
    "            shuffle (bool): whether to shuffle the data. Defaults to True.\n",
    "        \"\"\"\n",
    "        self.data_dir = data_dir\n",
    "        self.labels = labels\n",
    "        self.batch_size = batch_size\n",
    "        self.window_size = window_size\n",
    "        self.window_start = window_start\n",
    "        self.n_channels = n_channels\n",
    "        self.list_IDs = list_filehashes\n",
    "        self.n_classes = n_classes\n",
    "        self.shuffle = shuffle\n",
    "        self.on_epoch_end()\n",
    "\n",
    "    def __len__(self):\n",
    "        'Denotes the number of batches per epoch'\n",
    "        return int(np.floor(len(self.list_IDs) / self.batch_size))\n",
    "\n",
    "    def __getitem__(self, index):\n",
    "        'Generate one batch of data'\n",
    "        # Generate indexes of the batch\n",
    "        indexes = self.indexes[index*self.batch_size:(index+1)*self.batch_size]\n",
    "\n",
    "        # Find list of IDs\n",
    "        list_IDs_temp = [self.list_IDs[k] for k in indexes]\n",
    "\n",
    "        # Generate data\n",
    "        X, y = self.__data_generation(list_IDs_temp)\n",
    "\n",
    "        return X, y\n",
    "\n",
    "    def on_epoch_end(self):\n",
    "        'Updates indexes after each epoch'\n",
    "        self.indexes = np.arange(len(self.list_IDs))\n",
    "        if self.shuffle == True:\n",
    "            np.random.shuffle(self.indexes)\n",
    "\n",
    "    def __data_generation(self, list_IDs_temp):\n",
    "        'Generates data containing batch_size samples' \n",
    "        # X : (n_samples, window_size, n_channels)\n",
    "        # Initialization\n",
    "        X = np.empty((self.batch_size, self.window_size, self.n_channels))\n",
    "        y = np.empty((self.batch_size), dtype=int)\n",
    "\n",
    "        # Generate data\n",
    "        for i, ID in enumerate(list_IDs_temp):\n",
    "\n",
    "            # get X_tmp and then cut down to the window size\n",
    "            X_tmp = np.load(os.path.join(self.data_dir, ID + '.npy'))\n",
    "\n",
    "            X[i,] = X_tmp[\n",
    "                self.window_start:self.window_start + self.window_size,\n",
    "                :\n",
    "            ]\n",
    "\n",
    "            # Store class\n",
    "            y[i] = self.labels[ID]\n",
    "\n",
    "        return X, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_labels(metadata_df):\n",
    "\n",
    "    # construct the labels dictionary\n",
    "    labels = metadata_df[['game_winner', 'filehash']].set_index('filehash').to_dict()['game_winner']\n",
    "    # drop all keys with value == 0\n",
    "    labels = {k: v for k, v in labels.items() if v != 0}\n",
    "    # convert value all keys with value == 2 to 1\n",
    "    labels = {k: 0 if v == 2 else 1 for k, v in labels.items()}\n",
    "\n",
    "    return labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def setup_data_for_modeling(\n",
    "    metadata_path='data/spawningtool_replays.csv',\n",
    "    master_columns_path='info/clean_master_columns_list.csv',\n",
    "    non_feature_columns=['winner', 'filehash', 'frame'],\n",
    "    partition_index=None,\n",
    "    window_size=60,\n",
    "    window_start=0,\n",
    "    verbose=True\n",
    "):\n",
    "    \"\"\"\n",
    "    Setup the data for modeling.\n",
    "    \n",
    "    Args:\n",
    "        metadata_path (str): the path to the metadata csv. Defaults to 'data/spawningtool_replays.csv'.\n",
    "        master_columns_path (str): the path to the master_columns_list csv. Defaults to 'info/clean_master_columns_list.csv'.\n",
    "        non_feature_columns (list): the list of non-feature columns. Defaults to ['winner', 'filehash', 'frame'].\n",
    "        window_size (int): the window size to use. Defaults to 60.\n",
    "        verbose (bool): whether to print out the progress. Defaults to True.\n",
    "\n",
    "    Returns tuple:\n",
    "        panda.DataFrame: the dataframe containing the metadata\n",
    "        list: the list of all columns\n",
    "        list: the list of feature columns\n",
    "        dict: the dictionary of train, val, and test filehashes\n",
    "        dict: the dictionary of labels\n",
    "    \"\"\"\n",
    "    # get spawningtool_df\n",
    "    spawningtool_df = pd.read_csv(metadata_path)\n",
    "\n",
    "    # get master_columns_list\n",
    "    master_columns_list = pd.read_csv(master_columns_path).values.tolist()\n",
    "    master_columns_list = [col[0] for col in master_columns_list]\n",
    "\n",
    "    # create non_feature_columns\n",
    "    non_feature_columns = ['winner', 'filehash', 'frame']\n",
    "    # create feature_columns list\n",
    "    feature_columns = [\n",
    "        col for col in master_columns_list if col not in non_feature_columns\n",
    "    ]\n",
    "\n",
    "    partition = {}\n",
    "\n",
    "    if partition_index is None:\n",
    "        # loop through files in the partitions and create a list of hashes that meet \n",
    "        # the window size\n",
    "        \n",
    "        for folder in ['train', 'val', 'test']:\n",
    "            # get the filenames in the folder\n",
    "            filehashes = [\n",
    "                os.path.splitext(f)[0] for f in os.listdir(\n",
    "                    f'data/model_data/{folder}'\n",
    "                )\n",
    "            ]\n",
    "\n",
    "            partition_list = []\n",
    "            # load the file and check its shape[0]\n",
    "            for filehash in filehashes:\n",
    "                arr = np.load(f'data/model_data/{folder}/{filehash}.npy')\n",
    "                if arr.shape[0] > (window_size + window_start):\n",
    "                    partition_list.append(filehash)\n",
    "\n",
    "            # add the filehashes to the partition\n",
    "            partition[folder] = partition_list\n",
    "\n",
    "            # calculate number of files removed\n",
    "            removed = len(filehashes) - len(partition_list)\n",
    "\n",
    "            if verbose:\n",
    "                print(f'\\t{folder} has {len(filehashes)} valid files, removed {removed} files ({round(100*removed/len(filehashes),2)}%)')\n",
    "\n",
    "    else:\n",
    "        for folder in partition_index['partition'].unique():\n",
    "            # filter by shape > window_size + window_start\n",
    "            mask = (partition_index['partition'] == folder) \\\n",
    "                & (partition_index['shape'] > (window_size + window_start)) \n",
    "\n",
    "            # get the filehashes\n",
    "            partition[folder] = partition_index.loc[mask].index.values.tolist()\n",
    "\n",
    "\n",
    "    # get labels\n",
    "    labels = create_labels(spawningtool_df)\n",
    "\n",
    "    return spawningtool_df, master_columns_list, feature_columns, partition, labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# function to create, train, evaluate, and save a model for a given window size\n",
    "def train_evaluate_model(\n",
    "    window_size,\n",
    "    partition_index,\n",
    "    window_start=0,  \n",
    "    save_models=True,\n",
    "    verbose=1\n",
    "):\n",
    "    \"\"\"\n",
    "    train_evaluate_model\n",
    "    Train, evaluate, and save a model for a given window size.\n",
    "\n",
    "    Args:\n",
    "        window_size (int): the window size to use\n",
    "        partition_index (panda.DataFrame): the dataframe containing the partition index\n",
    "        window_start (int): the window start to use. Defaults to 0.\n",
    "        save_models (bool): whether to save the models to file or not\n",
    "        verbose (int): the verbosity level for Keras. Defaults to 1.\n",
    "\n",
    "    Returns:\n",
    "        tuple: model, history, model_file, score, prediction_results\n",
    "    \"\"\"\n",
    "\n",
    "    print('\\tSetting up model data...')\n",
    "    # setup the model\n",
    "    # set up the data for modeling\n",
    "    _, _, feature_columns, partition, labels = setup_data_for_modeling(\n",
    "        window_size=window_size, \n",
    "        window_start=window_start,\n",
    "        partition_index=partition_index,\n",
    "    )\n",
    "\n",
    "    # create a dictionary of parameters to pass to both generators\n",
    "    params = {\n",
    "        'window_size': window_size, \n",
    "        'window_start': window_start,\n",
    "        'n_channels': len(feature_columns),\n",
    "        'batch_size': 32, \n",
    "        'shuffle': True\n",
    "    }\n",
    "\n",
    "    # setup generators for training and validation\n",
    "    training_generator = BatchGenerator(\n",
    "        partition['train'],\n",
    "        labels,\n",
    "        data_dir='data/model_data/train',\n",
    "        **params\n",
    "    )\n",
    "\n",
    "    validation_generator = BatchGenerator(\n",
    "        partition['val'],\n",
    "        labels,\n",
    "        data_dir='data/model_data/val',\n",
    "        **params\n",
    "    )\n",
    "\n",
    "    print('\\tCreating model architecture...')\n",
    "    # instantiate RNN model\n",
    "    model = Sequential()\n",
    "\n",
    "    # add LSTM layer\n",
    "    model.add(LSTM(\n",
    "        units=128,\n",
    "        input_shape=(window_size, len(feature_columns)),\n",
    "        return_sequences=False\n",
    "    ))\n",
    "    model.add(BatchNormalization())\n",
    "    model.add(Dropout(0.2))\n",
    "\n",
    "    # add dense layer\n",
    "    model.add(Dense(\n",
    "        1, \n",
    "        activation='sigmoid', \n",
    "        kernel_regularizer=keras.regularizers.l1(0.01)\n",
    "    ))\n",
    "\n",
    "    # compile model using binary crossentropy loss and adam optimizer\n",
    "    model.compile(\n",
    "        loss='binary_crossentropy',\n",
    "        optimizer=Adam(learning_rate=0.001),\n",
    "        metrics=['accuracy']\n",
    "    )\n",
    "\n",
    "    # generate a unique identifier for the model use LSTM nodes and window_size\n",
    "    model_file = f'models/lstm_size.{window_size}_from.{window_start}'\n",
    "    model_file_ext = '.h5'\n",
    "    # check if the model file exists and if so, append a version number\n",
    "    if os.path.exists(model_file+model_file_ext):\n",
    "        # iterate through the version number to find the next available version number\n",
    "        i = 1\n",
    "        model_file = model_file.split('_v')[0] + '_v' + str(i)\n",
    "        while os.path.exists(model_file+model_file_ext):\n",
    "            i += 1\n",
    "            model_file = model_file + '_v' + str(i)\n",
    "\n",
    "    # create a callback to log history\n",
    "    csv_logger = CSVLogger(model_file+'.csv')\n",
    "\n",
    "    # create a callback to save checkpoints\n",
    "    checkpoint = ModelCheckpoint(\n",
    "        filepath='models/checkpoints/'+model_file.split('/')[1]+'.hdf5',\n",
    "        monitor='val_loss',\n",
    "        verbose=0,\n",
    "        save_best_only=False\n",
    "    )\n",
    "\n",
    "    # create a callback which stops training when no improvement is being made\n",
    "    earlystop = EarlyStopping(\n",
    "        monitor='val_loss', # monitor validation loss to prevent overfitting\n",
    "        patience=2, # stop after 3 epochs without improvement\n",
    "        verbose=1,  # print a message when the callback is triggered\n",
    "        mode='auto' # keras infers if the monitored variable should be increasing or decreasing\n",
    "    )\n",
    "\n",
    "    # set up callbacks to be used for training\n",
    "    if save_models:\n",
    "        callbacks = [earlystop, csv_logger, checkpoint]\n",
    "    else:\n",
    "        callbacks = [earlystop]\n",
    "\n",
    "    print('\\tTraining model...')\n",
    "    # fit the model and save to history\n",
    "    history = model.fit(\n",
    "        training_generator,\n",
    "        validation_data=validation_generator,\n",
    "        epochs=50,\n",
    "        callbacks=callbacks,\n",
    "        verbose=verbose\n",
    "    )\n",
    "\n",
    "    if save_models:   \n",
    "        # pickle the model\n",
    "        model.save(model_file+model_file_ext)\n",
    "\n",
    "    # get the last value for val_accuracy\n",
    "    score = history.history['val_accuracy'][-1]\n",
    "\n",
    "    print('\\tGenerating predictions...')\n",
    "    # create a generator for predictions\n",
    "    prediction_generator = BatchGenerator(\n",
    "        partition['val'],\n",
    "        labels,\n",
    "        data_dir='data/model_data/val',\n",
    "        window_size=window_size,\n",
    "        window_start=window_start, \n",
    "        n_channels=len(feature_columns),\n",
    "        batch_size=1, \n",
    "        shuffle=False\n",
    "    )\n",
    "\n",
    "    # make predictions\n",
    "    predictions = model.predict(prediction_generator)\n",
    "\n",
    "    # get the true labels for the predictions\n",
    "    true_labels = [labels[k] for k in partition['val']]\n",
    "\n",
    "    # create the prediction_results dataframe\n",
    "    prediction_results = pd.DataFrame({\n",
    "        'filehash': partition['val'],\n",
    "        'probability': predictions.reshape(-1),\n",
    "    })\n",
    "\n",
    "    # set the filehash as the index\n",
    "    prediction_results.set_index('filehash', inplace=True)\n",
    "\n",
    "    # return the model and the score\n",
    "    return model, history, model_file, score, prediction_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_partitions():\n",
    "    \n",
    "    folder_list = [\n",
    "        'data/model_data/train',\n",
    "        'data/model_data/val',\n",
    "        'data/model_data/test'\n",
    "    ]\n",
    "\n",
    "    partition_index = pd.DataFrame(columns=['filehash', 'shape'])\n",
    "\n",
    "    for folder in folder_list:\n",
    "        for file in os.listdir(folder):\n",
    "            filehash = os.path.splitext(file)[0]\n",
    "            partition_index = partition_index.append(\n",
    "                {\n",
    "                    'filehash': filehash,\n",
    "                    'shape': int(\n",
    "                        np.load(folder+'/'+file, mmap_mode='r').shape[0]),\n",
    "                    'partition': folder.split('/')[-1]\n",
    "                },\n",
    "                ignore_index=True\n",
    "            )\n",
    "\n",
    "    partition_index.set_index('filehash', inplace=True)\n",
    "\n",
    "    print('Partitioning complete.')\n",
    "\n",
    "    return partition_index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Partitioning data...\n",
      "Partitioning complete.\n",
      "--------------------------------------------------------------------------------\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9f3f57faeea5422aa6758ddbaf93a071",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Training models:   0%|          | 0/481 [00:00<?, ?model/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00009: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00003: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00003: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00009: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00008: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00007: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00006: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00004: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n",
      "\tSetting up model data...\n",
      "\tCreating model architecture...\n",
      "\tTraining model...\n",
      "Epoch 00005: early stopping\n",
      "\tGenerating predictions...\n",
      "Done!\n",
      "--------------------------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "model_dict = {}\n",
    "i = 0\n",
    "print('Partitioning data...')\n",
    "partition_index = create_partitions()\n",
    "print('-'*80)\n",
    "\n",
    "# create a list of window sizes and start tuples\n",
    "model_params = []\n",
    "for window_size in np.arange(48, 121, 6):\n",
    "    for window_start in np.arange(0, 121 - window_size, step=1):\n",
    "        model_params.append((window_size, window_start))\n",
    "\n",
    "for i in tqdm(range(len(model_params)), \n",
    "    desc='Training models',\n",
    "    unit='model'\n",
    "):\n",
    "    # get the window size and start tuple\n",
    "    window_size, window_start = model_params[i]\n",
    "\n",
    "    # set up the model\n",
    "    model, history, model_file, score, new_pred_results = train_evaluate_model(\n",
    "        window_size=window_size,\n",
    "        window_start=window_start,\n",
    "        save_models=False,\n",
    "        partition_index=partition_index,\n",
    "        verbose=0\n",
    "    )\n",
    "\n",
    "    model_index = str(window_start)+'_'+str(window_size)\n",
    "\n",
    "    new_pred_results.rename(\n",
    "        columns={'probability': model_index},\n",
    "        inplace=True\n",
    "    )\n",
    "\n",
    "    if i == 0:\n",
    "        prediction_results = new_pred_results\n",
    "    else:\n",
    "        prediction_results = pd.concat(\n",
    "            [prediction_results, new_pred_results], \n",
    "            axis=1\n",
    "        )\n",
    "\n",
    "    # save models in a dict keyed by window_start + window_size\n",
    "    model_dict[model_index] = model\n",
    "    print('Done!')\n",
    "    print('-'*80)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# loop through each model and save the model\n",
    "count = 0\n",
    "for model_index in model_dict.keys():\n",
    "    model = model_dict[model_index]\n",
    "    model.save('models/ensemble/'+model_index+'_'+str(count)+'.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>0_48</th>\n",
       "      <th>1_48</th>\n",
       "      <th>2_48</th>\n",
       "      <th>3_48</th>\n",
       "      <th>4_48</th>\n",
       "      <th>5_48</th>\n",
       "      <th>6_48</th>\n",
       "      <th>7_48</th>\n",
       "      <th>8_48</th>\n",
       "      <th>9_48</th>\n",
       "      <th>10_48</th>\n",
       "      <th>11_48</th>\n",
       "      <th>12_48</th>\n",
       "      <th>13_48</th>\n",
       "      <th>14_48</th>\n",
       "      <th>15_48</th>\n",
       "      <th>16_48</th>\n",
       "      <th>17_48</th>\n",
       "      <th>18_48</th>\n",
       "      <th>19_48</th>\n",
       "      <th>20_48</th>\n",
       "      <th>21_48</th>\n",
       "      <th>22_48</th>\n",
       "      <th>23_48</th>\n",
       "      <th>24_48</th>\n",
       "      <th>25_48</th>\n",
       "      <th>26_48</th>\n",
       "      <th>27_48</th>\n",
       "      <th>28_48</th>\n",
       "      <th>29_48</th>\n",
       "      <th>30_48</th>\n",
       "      <th>31_48</th>\n",
       "      <th>32_48</th>\n",
       "      <th>33_48</th>\n",
       "      <th>34_48</th>\n",
       "      <th>35_48</th>\n",
       "      <th>36_48</th>\n",
       "      <th>37_48</th>\n",
       "      <th>38_48</th>\n",
       "      <th>39_48</th>\n",
       "      <th>40_48</th>\n",
       "      <th>41_48</th>\n",
       "      <th>42_48</th>\n",
       "      <th>43_48</th>\n",
       "      <th>44_48</th>\n",
       "      <th>45_48</th>\n",
       "      <th>46_48</th>\n",
       "      <th>47_48</th>\n",
       "      <th>48_48</th>\n",
       "      <th>49_48</th>\n",
       "      <th>50_48</th>\n",
       "      <th>51_48</th>\n",
       "      <th>52_48</th>\n",
       "      <th>53_48</th>\n",
       "      <th>54_48</th>\n",
       "      <th>55_48</th>\n",
       "      <th>56_48</th>\n",
       "      <th>57_48</th>\n",
       "      <th>58_48</th>\n",
       "      <th>59_48</th>\n",
       "      <th>60_48</th>\n",
       "      <th>61_48</th>\n",
       "      <th>62_48</th>\n",
       "      <th>63_48</th>\n",
       "      <th>64_48</th>\n",
       "      <th>65_48</th>\n",
       "      <th>66_48</th>\n",
       "      <th>67_48</th>\n",
       "      <th>68_48</th>\n",
       "      <th>69_48</th>\n",
       "      <th>70_48</th>\n",
       "      <th>71_48</th>\n",
       "      <th>72_48</th>\n",
       "      <th>0_54</th>\n",
       "      <th>1_54</th>\n",
       "      <th>2_54</th>\n",
       "      <th>3_54</th>\n",
       "      <th>4_54</th>\n",
       "      <th>5_54</th>\n",
       "      <th>6_54</th>\n",
       "      <th>7_54</th>\n",
       "      <th>8_54</th>\n",
       "      <th>9_54</th>\n",
       "      <th>10_54</th>\n",
       "      <th>11_54</th>\n",
       "      <th>12_54</th>\n",
       "      <th>13_54</th>\n",
       "      <th>14_54</th>\n",
       "      <th>15_54</th>\n",
       "      <th>16_54</th>\n",
       "      <th>17_54</th>\n",
       "      <th>18_54</th>\n",
       "      <th>19_54</th>\n",
       "      <th>20_54</th>\n",
       "      <th>21_54</th>\n",
       "      <th>22_54</th>\n",
       "      <th>23_54</th>\n",
       "      <th>24_54</th>\n",
       "      <th>25_54</th>\n",
       "      <th>26_54</th>\n",
       "      <th>...</th>\n",
       "      <th>33_84</th>\n",
       "      <th>34_84</th>\n",
       "      <th>35_84</th>\n",
       "      <th>36_84</th>\n",
       "      <th>0_90</th>\n",
       "      <th>1_90</th>\n",
       "      <th>2_90</th>\n",
       "      <th>3_90</th>\n",
       "      <th>4_90</th>\n",
       "      <th>5_90</th>\n",
       "      <th>6_90</th>\n",
       "      <th>7_90</th>\n",
       "      <th>8_90</th>\n",
       "      <th>9_90</th>\n",
       "      <th>10_90</th>\n",
       "      <th>11_90</th>\n",
       "      <th>12_90</th>\n",
       "      <th>13_90</th>\n",
       "      <th>14_90</th>\n",
       "      <th>15_90</th>\n",
       "      <th>16_90</th>\n",
       "      <th>17_90</th>\n",
       "      <th>18_90</th>\n",
       "      <th>19_90</th>\n",
       "      <th>20_90</th>\n",
       "      <th>21_90</th>\n",
       "      <th>22_90</th>\n",
       "      <th>23_90</th>\n",
       "      <th>24_90</th>\n",
       "      <th>25_90</th>\n",
       "      <th>26_90</th>\n",
       "      <th>27_90</th>\n",
       "      <th>28_90</th>\n",
       "      <th>29_90</th>\n",
       "      <th>30_90</th>\n",
       "      <th>0_96</th>\n",
       "      <th>1_96</th>\n",
       "      <th>2_96</th>\n",
       "      <th>3_96</th>\n",
       "      <th>4_96</th>\n",
       "      <th>5_96</th>\n",
       "      <th>6_96</th>\n",
       "      <th>7_96</th>\n",
       "      <th>8_96</th>\n",
       "      <th>9_96</th>\n",
       "      <th>10_96</th>\n",
       "      <th>11_96</th>\n",
       "      <th>12_96</th>\n",
       "      <th>13_96</th>\n",
       "      <th>14_96</th>\n",
       "      <th>15_96</th>\n",
       "      <th>16_96</th>\n",
       "      <th>17_96</th>\n",
       "      <th>18_96</th>\n",
       "      <th>19_96</th>\n",
       "      <th>20_96</th>\n",
       "      <th>21_96</th>\n",
       "      <th>22_96</th>\n",
       "      <th>23_96</th>\n",
       "      <th>24_96</th>\n",
       "      <th>0_102</th>\n",
       "      <th>1_102</th>\n",
       "      <th>2_102</th>\n",
       "      <th>3_102</th>\n",
       "      <th>4_102</th>\n",
       "      <th>5_102</th>\n",
       "      <th>6_102</th>\n",
       "      <th>7_102</th>\n",
       "      <th>8_102</th>\n",
       "      <th>9_102</th>\n",
       "      <th>10_102</th>\n",
       "      <th>11_102</th>\n",
       "      <th>12_102</th>\n",
       "      <th>13_102</th>\n",
       "      <th>14_102</th>\n",
       "      <th>15_102</th>\n",
       "      <th>16_102</th>\n",
       "      <th>17_102</th>\n",
       "      <th>18_102</th>\n",
       "      <th>0_108</th>\n",
       "      <th>1_108</th>\n",
       "      <th>2_108</th>\n",
       "      <th>3_108</th>\n",
       "      <th>4_108</th>\n",
       "      <th>5_108</th>\n",
       "      <th>6_108</th>\n",
       "      <th>7_108</th>\n",
       "      <th>8_108</th>\n",
       "      <th>9_108</th>\n",
       "      <th>10_108</th>\n",
       "      <th>11_108</th>\n",
       "      <th>12_108</th>\n",
       "      <th>0_114</th>\n",
       "      <th>1_114</th>\n",
       "      <th>2_114</th>\n",
       "      <th>3_114</th>\n",
       "      <th>4_114</th>\n",
       "      <th>5_114</th>\n",
       "      <th>6_114</th>\n",
       "      <th>0_120</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>filehash</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0016ddd5b36473f259ba8630c6c0047540a6858071737e8f890358d3f7ad355d</th>\n",
       "      <td>0.029698</td>\n",
       "      <td>0.080832</td>\n",
       "      <td>0.049923</td>\n",
       "      <td>0.118166</td>\n",
       "      <td>0.223824</td>\n",
       "      <td>0.177816</td>\n",
       "      <td>0.045554</td>\n",
       "      <td>0.106425</td>\n",
       "      <td>0.194602</td>\n",
       "      <td>0.048929</td>\n",
       "      <td>0.061413</td>\n",
       "      <td>0.047669</td>\n",
       "      <td>0.088884</td>\n",
       "      <td>0.206801</td>\n",
       "      <td>0.064651</td>\n",
       "      <td>0.275370</td>\n",
       "      <td>0.151547</td>\n",
       "      <td>0.077658</td>\n",
       "      <td>0.029862</td>\n",
       "      <td>0.041826</td>\n",
       "      <td>0.051233</td>\n",
       "      <td>0.011541</td>\n",
       "      <td>0.049218</td>\n",
       "      <td>0.089420</td>\n",
       "      <td>0.022117</td>\n",
       "      <td>0.013123</td>\n",
       "      <td>0.030460</td>\n",
       "      <td>0.013854</td>\n",
       "      <td>0.183592</td>\n",
       "      <td>0.064270</td>\n",
       "      <td>0.283902</td>\n",
       "      <td>0.078608</td>\n",
       "      <td>0.084389</td>\n",
       "      <td>0.008901</td>\n",
       "      <td>0.031846</td>\n",
       "      <td>0.214726</td>\n",
       "      <td>0.013098</td>\n",
       "      <td>0.327583</td>\n",
       "      <td>0.740116</td>\n",
       "      <td>0.506495</td>\n",
       "      <td>0.331147</td>\n",
       "      <td>0.346144</td>\n",
       "      <td>0.167521</td>\n",
       "      <td>0.276274</td>\n",
       "      <td>0.395539</td>\n",
       "      <td>0.860666</td>\n",
       "      <td>0.751504</td>\n",
       "      <td>0.311051</td>\n",
       "      <td>0.804861</td>\n",
       "      <td>0.598100</td>\n",
       "      <td>0.753049</td>\n",
       "      <td>0.142164</td>\n",
       "      <td>0.060092</td>\n",
       "      <td>0.031496</td>\n",
       "      <td>0.560822</td>\n",
       "      <td>0.885624</td>\n",
       "      <td>0.319505</td>\n",
       "      <td>0.479512</td>\n",
       "      <td>0.493804</td>\n",
       "      <td>0.150953</td>\n",
       "      <td>0.022642</td>\n",
       "      <td>0.618027</td>\n",
       "      <td>0.570352</td>\n",
       "      <td>0.170242</td>\n",
       "      <td>0.271364</td>\n",
       "      <td>0.845575</td>\n",
       "      <td>0.094429</td>\n",
       "      <td>0.559305</td>\n",
       "      <td>0.352102</td>\n",
       "      <td>0.923105</td>\n",
       "      <td>0.621745</td>\n",
       "      <td>0.295658</td>\n",
       "      <td>0.145317</td>\n",
       "      <td>0.052655</td>\n",
       "      <td>0.125262</td>\n",
       "      <td>0.296246</td>\n",
       "      <td>0.158113</td>\n",
       "      <td>0.260551</td>\n",
       "      <td>0.172064</td>\n",
       "      <td>0.261298</td>\n",
       "      <td>0.182585</td>\n",
       "      <td>0.107669</td>\n",
       "      <td>0.137240</td>\n",
       "      <td>0.360731</td>\n",
       "      <td>0.112294</td>\n",
       "      <td>0.012958</td>\n",
       "      <td>0.013463</td>\n",
       "      <td>0.055620</td>\n",
       "      <td>0.005191</td>\n",
       "      <td>0.103964</td>\n",
       "      <td>0.058726</td>\n",
       "      <td>0.020764</td>\n",
       "      <td>0.009628</td>\n",
       "      <td>0.024620</td>\n",
       "      <td>0.109718</td>\n",
       "      <td>0.467515</td>\n",
       "      <td>0.022748</td>\n",
       "      <td>0.094592</td>\n",
       "      <td>0.035569</td>\n",
       "      <td>0.520413</td>\n",
       "      <td>...</td>\n",
       "      <td>0.494885</td>\n",
       "      <td>0.551332</td>\n",
       "      <td>0.894536</td>\n",
       "      <td>0.101025</td>\n",
       "      <td>0.085483</td>\n",
       "      <td>0.204991</td>\n",
       "      <td>0.740870</td>\n",
       "      <td>0.647179</td>\n",
       "      <td>0.394482</td>\n",
       "      <td>0.303274</td>\n",
       "      <td>0.893054</td>\n",
       "      <td>0.264133</td>\n",
       "      <td>0.273751</td>\n",
       "      <td>0.057169</td>\n",
       "      <td>0.138298</td>\n",
       "      <td>0.220179</td>\n",
       "      <td>0.605863</td>\n",
       "      <td>0.527785</td>\n",
       "      <td>0.706976</td>\n",
       "      <td>0.273590</td>\n",
       "      <td>0.363384</td>\n",
       "      <td>0.744208</td>\n",
       "      <td>0.128260</td>\n",
       "      <td>0.798777</td>\n",
       "      <td>0.098031</td>\n",
       "      <td>0.288894</td>\n",
       "      <td>0.065271</td>\n",
       "      <td>0.651618</td>\n",
       "      <td>0.943892</td>\n",
       "      <td>0.758336</td>\n",
       "      <td>0.699642</td>\n",
       "      <td>0.156586</td>\n",
       "      <td>0.336604</td>\n",
       "      <td>0.060895</td>\n",
       "      <td>0.469109</td>\n",
       "      <td>0.481473</td>\n",
       "      <td>0.316594</td>\n",
       "      <td>0.207696</td>\n",
       "      <td>0.245382</td>\n",
       "      <td>0.449606</td>\n",
       "      <td>0.220301</td>\n",
       "      <td>0.559647</td>\n",
       "      <td>0.556434</td>\n",
       "      <td>0.724366</td>\n",
       "      <td>0.327392</td>\n",
       "      <td>0.737083</td>\n",
       "      <td>0.367128</td>\n",
       "      <td>0.221972</td>\n",
       "      <td>0.841491</td>\n",
       "      <td>0.041223</td>\n",
       "      <td>0.837291</td>\n",
       "      <td>0.008387</td>\n",
       "      <td>0.633813</td>\n",
       "      <td>0.318139</td>\n",
       "      <td>0.024553</td>\n",
       "      <td>0.163532</td>\n",
       "      <td>0.089204</td>\n",
       "      <td>0.026484</td>\n",
       "      <td>0.137825</td>\n",
       "      <td>0.041519</td>\n",
       "      <td>0.250772</td>\n",
       "      <td>0.436267</td>\n",
       "      <td>0.308587</td>\n",
       "      <td>0.481186</td>\n",
       "      <td>0.482252</td>\n",
       "      <td>0.368783</td>\n",
       "      <td>0.330872</td>\n",
       "      <td>0.637380</td>\n",
       "      <td>0.630386</td>\n",
       "      <td>0.452975</td>\n",
       "      <td>0.618075</td>\n",
       "      <td>0.540244</td>\n",
       "      <td>0.727182</td>\n",
       "      <td>0.165879</td>\n",
       "      <td>0.576970</td>\n",
       "      <td>0.032162</td>\n",
       "      <td>0.121999</td>\n",
       "      <td>0.176663</td>\n",
       "      <td>0.097464</td>\n",
       "      <td>0.130436</td>\n",
       "      <td>0.714333</td>\n",
       "      <td>0.175717</td>\n",
       "      <td>0.739042</td>\n",
       "      <td>0.020250</td>\n",
       "      <td>0.485943</td>\n",
       "      <td>0.401508</td>\n",
       "      <td>0.636606</td>\n",
       "      <td>0.870755</td>\n",
       "      <td>0.282143</td>\n",
       "      <td>0.194977</td>\n",
       "      <td>0.077225</td>\n",
       "      <td>0.023264</td>\n",
       "      <td>0.804053</td>\n",
       "      <td>0.802499</td>\n",
       "      <td>0.432499</td>\n",
       "      <td>0.559297</td>\n",
       "      <td>0.016355</td>\n",
       "      <td>0.094903</td>\n",
       "      <td>0.114713</td>\n",
       "      <td>0.206466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e7827e10660ad88bf5bee</th>\n",
       "      <td>0.295901</td>\n",
       "      <td>0.616100</td>\n",
       "      <td>0.585490</td>\n",
       "      <td>0.763552</td>\n",
       "      <td>0.874480</td>\n",
       "      <td>0.942619</td>\n",
       "      <td>0.703923</td>\n",
       "      <td>0.629009</td>\n",
       "      <td>0.812236</td>\n",
       "      <td>0.865754</td>\n",
       "      <td>0.784470</td>\n",
       "      <td>0.449863</td>\n",
       "      <td>0.337121</td>\n",
       "      <td>0.661061</td>\n",
       "      <td>0.956659</td>\n",
       "      <td>0.866433</td>\n",
       "      <td>0.686306</td>\n",
       "      <td>0.927653</td>\n",
       "      <td>0.934034</td>\n",
       "      <td>0.947765</td>\n",
       "      <td>0.808483</td>\n",
       "      <td>0.833280</td>\n",
       "      <td>0.770048</td>\n",
       "      <td>0.862283</td>\n",
       "      <td>0.960639</td>\n",
       "      <td>0.920410</td>\n",
       "      <td>0.881784</td>\n",
       "      <td>0.942431</td>\n",
       "      <td>0.725764</td>\n",
       "      <td>0.811090</td>\n",
       "      <td>0.913210</td>\n",
       "      <td>0.927040</td>\n",
       "      <td>0.346670</td>\n",
       "      <td>0.804140</td>\n",
       "      <td>0.554704</td>\n",
       "      <td>0.771561</td>\n",
       "      <td>0.360767</td>\n",
       "      <td>0.902478</td>\n",
       "      <td>0.600150</td>\n",
       "      <td>0.942712</td>\n",
       "      <td>0.617142</td>\n",
       "      <td>0.657846</td>\n",
       "      <td>0.777856</td>\n",
       "      <td>0.892649</td>\n",
       "      <td>0.845431</td>\n",
       "      <td>0.937457</td>\n",
       "      <td>0.516100</td>\n",
       "      <td>0.339713</td>\n",
       "      <td>0.418890</td>\n",
       "      <td>0.245681</td>\n",
       "      <td>0.449455</td>\n",
       "      <td>0.308119</td>\n",
       "      <td>0.552498</td>\n",
       "      <td>0.709674</td>\n",
       "      <td>0.107609</td>\n",
       "      <td>0.854450</td>\n",
       "      <td>0.504910</td>\n",
       "      <td>0.444506</td>\n",
       "      <td>0.328661</td>\n",
       "      <td>0.489310</td>\n",
       "      <td>0.256028</td>\n",
       "      <td>0.322243</td>\n",
       "      <td>0.130332</td>\n",
       "      <td>0.021160</td>\n",
       "      <td>0.318908</td>\n",
       "      <td>0.374743</td>\n",
       "      <td>0.091339</td>\n",
       "      <td>0.308014</td>\n",
       "      <td>0.311482</td>\n",
       "      <td>0.105572</td>\n",
       "      <td>0.245726</td>\n",
       "      <td>0.194622</td>\n",
       "      <td>0.853658</td>\n",
       "      <td>0.634948</td>\n",
       "      <td>0.893021</td>\n",
       "      <td>0.749243</td>\n",
       "      <td>0.796109</td>\n",
       "      <td>0.669738</td>\n",
       "      <td>0.554182</td>\n",
       "      <td>0.465155</td>\n",
       "      <td>0.070758</td>\n",
       "      <td>0.905468</td>\n",
       "      <td>0.877355</td>\n",
       "      <td>0.868039</td>\n",
       "      <td>0.876720</td>\n",
       "      <td>0.749644</td>\n",
       "      <td>0.920808</td>\n",
       "      <td>0.781602</td>\n",
       "      <td>0.775674</td>\n",
       "      <td>0.880736</td>\n",
       "      <td>0.807706</td>\n",
       "      <td>0.838220</td>\n",
       "      <td>0.957286</td>\n",
       "      <td>0.863977</td>\n",
       "      <td>0.944742</td>\n",
       "      <td>0.857148</td>\n",
       "      <td>0.972580</td>\n",
       "      <td>0.507750</td>\n",
       "      <td>0.855397</td>\n",
       "      <td>0.791277</td>\n",
       "      <td>...</td>\n",
       "      <td>0.172838</td>\n",
       "      <td>0.239157</td>\n",
       "      <td>0.183369</td>\n",
       "      <td>0.112225</td>\n",
       "      <td>0.291436</td>\n",
       "      <td>0.819209</td>\n",
       "      <td>0.447267</td>\n",
       "      <td>0.907690</td>\n",
       "      <td>0.720795</td>\n",
       "      <td>0.661324</td>\n",
       "      <td>0.468694</td>\n",
       "      <td>0.626997</td>\n",
       "      <td>0.557113</td>\n",
       "      <td>0.429908</td>\n",
       "      <td>0.326878</td>\n",
       "      <td>0.585498</td>\n",
       "      <td>0.180201</td>\n",
       "      <td>0.446671</td>\n",
       "      <td>0.340976</td>\n",
       "      <td>0.566979</td>\n",
       "      <td>0.423755</td>\n",
       "      <td>0.598135</td>\n",
       "      <td>0.845344</td>\n",
       "      <td>0.179377</td>\n",
       "      <td>0.158628</td>\n",
       "      <td>0.141114</td>\n",
       "      <td>0.310970</td>\n",
       "      <td>0.445229</td>\n",
       "      <td>0.293999</td>\n",
       "      <td>0.151553</td>\n",
       "      <td>0.257232</td>\n",
       "      <td>0.205178</td>\n",
       "      <td>0.743767</td>\n",
       "      <td>0.550082</td>\n",
       "      <td>0.193352</td>\n",
       "      <td>0.356057</td>\n",
       "      <td>0.942356</td>\n",
       "      <td>0.357973</td>\n",
       "      <td>0.245077</td>\n",
       "      <td>0.311749</td>\n",
       "      <td>0.817325</td>\n",
       "      <td>0.082411</td>\n",
       "      <td>0.252692</td>\n",
       "      <td>0.696436</td>\n",
       "      <td>0.352933</td>\n",
       "      <td>0.553481</td>\n",
       "      <td>0.315569</td>\n",
       "      <td>0.297744</td>\n",
       "      <td>0.395014</td>\n",
       "      <td>0.430080</td>\n",
       "      <td>0.601526</td>\n",
       "      <td>0.671068</td>\n",
       "      <td>0.114176</td>\n",
       "      <td>0.102735</td>\n",
       "      <td>0.056013</td>\n",
       "      <td>0.205072</td>\n",
       "      <td>0.061710</td>\n",
       "      <td>0.180499</td>\n",
       "      <td>0.352650</td>\n",
       "      <td>0.282756</td>\n",
       "      <td>0.259089</td>\n",
       "      <td>0.313544</td>\n",
       "      <td>0.612125</td>\n",
       "      <td>0.344980</td>\n",
       "      <td>0.502287</td>\n",
       "      <td>0.382664</td>\n",
       "      <td>0.372557</td>\n",
       "      <td>0.361522</td>\n",
       "      <td>0.090581</td>\n",
       "      <td>0.099027</td>\n",
       "      <td>0.159275</td>\n",
       "      <td>0.329697</td>\n",
       "      <td>0.055109</td>\n",
       "      <td>0.214870</td>\n",
       "      <td>0.089066</td>\n",
       "      <td>0.072910</td>\n",
       "      <td>0.149949</td>\n",
       "      <td>0.259793</td>\n",
       "      <td>0.205103</td>\n",
       "      <td>0.318784</td>\n",
       "      <td>0.522218</td>\n",
       "      <td>0.133060</td>\n",
       "      <td>0.282103</td>\n",
       "      <td>0.193713</td>\n",
       "      <td>0.164036</td>\n",
       "      <td>0.089202</td>\n",
       "      <td>0.466361</td>\n",
       "      <td>0.384753</td>\n",
       "      <td>0.141033</td>\n",
       "      <td>0.186505</td>\n",
       "      <td>0.388967</td>\n",
       "      <td>0.302577</td>\n",
       "      <td>0.588224</td>\n",
       "      <td>0.137935</td>\n",
       "      <td>0.356914</td>\n",
       "      <td>0.291190</td>\n",
       "      <td>0.082553</td>\n",
       "      <td>0.467127</td>\n",
       "      <td>0.473013</td>\n",
       "      <td>0.528720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0033334e86cb352b0131027e58effa23adb23cf1cdd90d5760cc62254611c99a</th>\n",
       "      <td>0.688192</td>\n",
       "      <td>0.856472</td>\n",
       "      <td>0.740203</td>\n",
       "      <td>0.928254</td>\n",
       "      <td>0.940908</td>\n",
       "      <td>0.941489</td>\n",
       "      <td>0.899931</td>\n",
       "      <td>0.877026</td>\n",
       "      <td>0.915597</td>\n",
       "      <td>0.972010</td>\n",
       "      <td>0.915853</td>\n",
       "      <td>0.928046</td>\n",
       "      <td>0.983680</td>\n",
       "      <td>0.964526</td>\n",
       "      <td>0.906938</td>\n",
       "      <td>0.924819</td>\n",
       "      <td>0.962019</td>\n",
       "      <td>0.911155</td>\n",
       "      <td>0.922102</td>\n",
       "      <td>0.940164</td>\n",
       "      <td>0.843983</td>\n",
       "      <td>0.893247</td>\n",
       "      <td>0.902425</td>\n",
       "      <td>0.894499</td>\n",
       "      <td>0.951999</td>\n",
       "      <td>0.704206</td>\n",
       "      <td>0.960727</td>\n",
       "      <td>0.959247</td>\n",
       "      <td>0.960867</td>\n",
       "      <td>0.914251</td>\n",
       "      <td>0.911883</td>\n",
       "      <td>0.548480</td>\n",
       "      <td>0.959223</td>\n",
       "      <td>0.900190</td>\n",
       "      <td>0.726230</td>\n",
       "      <td>0.941350</td>\n",
       "      <td>0.972824</td>\n",
       "      <td>0.990653</td>\n",
       "      <td>0.575059</td>\n",
       "      <td>0.843176</td>\n",
       "      <td>0.180662</td>\n",
       "      <td>0.611569</td>\n",
       "      <td>0.907423</td>\n",
       "      <td>0.855664</td>\n",
       "      <td>0.989813</td>\n",
       "      <td>0.713413</td>\n",
       "      <td>0.967392</td>\n",
       "      <td>0.867970</td>\n",
       "      <td>0.645537</td>\n",
       "      <td>0.710680</td>\n",
       "      <td>0.781099</td>\n",
       "      <td>0.939137</td>\n",
       "      <td>0.635680</td>\n",
       "      <td>0.882481</td>\n",
       "      <td>0.632616</td>\n",
       "      <td>0.338599</td>\n",
       "      <td>0.837811</td>\n",
       "      <td>0.793955</td>\n",
       "      <td>0.876178</td>\n",
       "      <td>0.844483</td>\n",
       "      <td>0.873082</td>\n",
       "      <td>0.710679</td>\n",
       "      <td>0.921149</td>\n",
       "      <td>0.423949</td>\n",
       "      <td>0.316755</td>\n",
       "      <td>0.898936</td>\n",
       "      <td>0.942313</td>\n",
       "      <td>0.519842</td>\n",
       "      <td>0.678256</td>\n",
       "      <td>0.756128</td>\n",
       "      <td>0.559718</td>\n",
       "      <td>0.576955</td>\n",
       "      <td>0.801891</td>\n",
       "      <td>0.930794</td>\n",
       "      <td>0.973386</td>\n",
       "      <td>0.953991</td>\n",
       "      <td>0.964543</td>\n",
       "      <td>0.884326</td>\n",
       "      <td>0.972037</td>\n",
       "      <td>0.979429</td>\n",
       "      <td>0.949053</td>\n",
       "      <td>0.948033</td>\n",
       "      <td>0.874371</td>\n",
       "      <td>0.950503</td>\n",
       "      <td>0.955456</td>\n",
       "      <td>0.959813</td>\n",
       "      <td>0.945319</td>\n",
       "      <td>0.873793</td>\n",
       "      <td>0.915594</td>\n",
       "      <td>0.902709</td>\n",
       "      <td>0.915216</td>\n",
       "      <td>0.885734</td>\n",
       "      <td>0.860223</td>\n",
       "      <td>0.805894</td>\n",
       "      <td>0.796016</td>\n",
       "      <td>0.927726</td>\n",
       "      <td>0.550648</td>\n",
       "      <td>0.751867</td>\n",
       "      <td>0.884138</td>\n",
       "      <td>0.865604</td>\n",
       "      <td>...</td>\n",
       "      <td>0.622803</td>\n",
       "      <td>0.776738</td>\n",
       "      <td>0.636121</td>\n",
       "      <td>0.633256</td>\n",
       "      <td>0.919222</td>\n",
       "      <td>0.917695</td>\n",
       "      <td>0.983487</td>\n",
       "      <td>0.893850</td>\n",
       "      <td>0.980378</td>\n",
       "      <td>0.855946</td>\n",
       "      <td>0.713308</td>\n",
       "      <td>0.876186</td>\n",
       "      <td>0.699076</td>\n",
       "      <td>0.660392</td>\n",
       "      <td>0.950420</td>\n",
       "      <td>0.823178</td>\n",
       "      <td>0.908768</td>\n",
       "      <td>0.858369</td>\n",
       "      <td>0.837563</td>\n",
       "      <td>0.788934</td>\n",
       "      <td>0.785083</td>\n",
       "      <td>0.928731</td>\n",
       "      <td>0.875839</td>\n",
       "      <td>0.717817</td>\n",
       "      <td>0.815259</td>\n",
       "      <td>0.906099</td>\n",
       "      <td>0.483409</td>\n",
       "      <td>0.886357</td>\n",
       "      <td>0.916344</td>\n",
       "      <td>0.721530</td>\n",
       "      <td>0.602255</td>\n",
       "      <td>0.671580</td>\n",
       "      <td>0.561276</td>\n",
       "      <td>0.156611</td>\n",
       "      <td>0.949887</td>\n",
       "      <td>0.419144</td>\n",
       "      <td>0.749896</td>\n",
       "      <td>0.341601</td>\n",
       "      <td>0.608766</td>\n",
       "      <td>0.949398</td>\n",
       "      <td>0.936689</td>\n",
       "      <td>0.775339</td>\n",
       "      <td>0.436182</td>\n",
       "      <td>0.747801</td>\n",
       "      <td>0.815390</td>\n",
       "      <td>0.953011</td>\n",
       "      <td>0.917669</td>\n",
       "      <td>0.764154</td>\n",
       "      <td>0.603275</td>\n",
       "      <td>0.744248</td>\n",
       "      <td>0.885763</td>\n",
       "      <td>0.423015</td>\n",
       "      <td>0.408318</td>\n",
       "      <td>0.587098</td>\n",
       "      <td>0.770504</td>\n",
       "      <td>0.585670</td>\n",
       "      <td>0.741818</td>\n",
       "      <td>0.312818</td>\n",
       "      <td>0.503492</td>\n",
       "      <td>0.777177</td>\n",
       "      <td>0.806311</td>\n",
       "      <td>0.328509</td>\n",
       "      <td>0.797078</td>\n",
       "      <td>0.825528</td>\n",
       "      <td>0.867705</td>\n",
       "      <td>0.931840</td>\n",
       "      <td>0.860384</td>\n",
       "      <td>0.769309</td>\n",
       "      <td>0.661027</td>\n",
       "      <td>0.706649</td>\n",
       "      <td>0.546251</td>\n",
       "      <td>0.621745</td>\n",
       "      <td>0.643203</td>\n",
       "      <td>0.393842</td>\n",
       "      <td>0.848718</td>\n",
       "      <td>0.933219</td>\n",
       "      <td>0.503421</td>\n",
       "      <td>0.680889</td>\n",
       "      <td>0.860788</td>\n",
       "      <td>0.817465</td>\n",
       "      <td>0.660918</td>\n",
       "      <td>0.731410</td>\n",
       "      <td>0.589714</td>\n",
       "      <td>0.430669</td>\n",
       "      <td>0.896000</td>\n",
       "      <td>0.572216</td>\n",
       "      <td>0.821138</td>\n",
       "      <td>0.283826</td>\n",
       "      <td>0.893737</td>\n",
       "      <td>0.516957</td>\n",
       "      <td>0.839548</td>\n",
       "      <td>0.964948</td>\n",
       "      <td>0.837720</td>\n",
       "      <td>0.782127</td>\n",
       "      <td>0.697565</td>\n",
       "      <td>0.918200</td>\n",
       "      <td>0.192754</td>\n",
       "      <td>0.734890</td>\n",
       "      <td>0.940787</td>\n",
       "      <td>0.874146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0055f4541fe30a72e964fc168db7abef35822d27e4cdce8ef0adb8b3cba51b7d</th>\n",
       "      <td>0.910836</td>\n",
       "      <td>0.928538</td>\n",
       "      <td>0.527132</td>\n",
       "      <td>0.888624</td>\n",
       "      <td>0.915498</td>\n",
       "      <td>0.909505</td>\n",
       "      <td>0.964951</td>\n",
       "      <td>0.962932</td>\n",
       "      <td>0.948339</td>\n",
       "      <td>0.975626</td>\n",
       "      <td>0.925406</td>\n",
       "      <td>0.976905</td>\n",
       "      <td>0.986096</td>\n",
       "      <td>0.977565</td>\n",
       "      <td>0.980430</td>\n",
       "      <td>0.956631</td>\n",
       "      <td>0.976478</td>\n",
       "      <td>0.949210</td>\n",
       "      <td>0.906159</td>\n",
       "      <td>0.983026</td>\n",
       "      <td>0.974964</td>\n",
       "      <td>0.961030</td>\n",
       "      <td>0.980169</td>\n",
       "      <td>0.964771</td>\n",
       "      <td>0.931328</td>\n",
       "      <td>0.992986</td>\n",
       "      <td>0.991289</td>\n",
       "      <td>0.989885</td>\n",
       "      <td>0.994662</td>\n",
       "      <td>0.992780</td>\n",
       "      <td>0.954595</td>\n",
       "      <td>0.970848</td>\n",
       "      <td>0.911303</td>\n",
       "      <td>0.979076</td>\n",
       "      <td>0.985443</td>\n",
       "      <td>0.973002</td>\n",
       "      <td>0.991811</td>\n",
       "      <td>0.988492</td>\n",
       "      <td>0.960801</td>\n",
       "      <td>0.997184</td>\n",
       "      <td>0.943886</td>\n",
       "      <td>0.994764</td>\n",
       "      <td>0.992135</td>\n",
       "      <td>0.991517</td>\n",
       "      <td>0.981689</td>\n",
       "      <td>0.994041</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.890511</td>\n",
       "      <td>0.971366</td>\n",
       "      <td>0.922571</td>\n",
       "      <td>0.992779</td>\n",
       "      <td>0.988530</td>\n",
       "      <td>0.984489</td>\n",
       "      <td>0.989918</td>\n",
       "      <td>0.981094</td>\n",
       "      <td>0.917718</td>\n",
       "      <td>0.929050</td>\n",
       "      <td>0.913255</td>\n",
       "      <td>0.955568</td>\n",
       "      <td>0.987435</td>\n",
       "      <td>0.983431</td>\n",
       "      <td>0.917882</td>\n",
       "      <td>0.955393</td>\n",
       "      <td>0.931522</td>\n",
       "      <td>0.823321</td>\n",
       "      <td>0.954594</td>\n",
       "      <td>0.981330</td>\n",
       "      <td>0.928382</td>\n",
       "      <td>0.980354</td>\n",
       "      <td>0.995421</td>\n",
       "      <td>0.994731</td>\n",
       "      <td>0.943641</td>\n",
       "      <td>0.967961</td>\n",
       "      <td>0.924548</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.995000</td>\n",
       "      <td>0.990181</td>\n",
       "      <td>0.981080</td>\n",
       "      <td>0.994067</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e46cd9f2cbf3fa3a99</th>\n",
       "      <td>0.847571</td>\n",
       "      <td>0.963856</td>\n",
       "      <td>0.809928</td>\n",
       "      <td>0.791543</td>\n",
       "      <td>0.949490</td>\n",
       "      <td>0.851008</td>\n",
       "      <td>0.970709</td>\n",
       "      <td>0.963803</td>\n",
       "      <td>0.949462</td>\n",
       "      <td>0.952987</td>\n",
       "      <td>0.916027</td>\n",
       "      <td>0.980209</td>\n",
       "      <td>0.975572</td>\n",
       "      <td>0.982690</td>\n",
       "      <td>0.973211</td>\n",
       "      <td>0.980387</td>\n",
       "      <td>0.992644</td>\n",
       "      <td>0.959100</td>\n",
       "      <td>0.954591</td>\n",
       "      <td>0.971579</td>\n",
       "      <td>0.983339</td>\n",
       "      <td>0.992214</td>\n",
       "      <td>0.985897</td>\n",
       "      <td>0.980644</td>\n",
       "      <td>0.988835</td>\n",
       "      <td>0.994517</td>\n",
       "      <td>0.994570</td>\n",
       "      <td>0.995637</td>\n",
       "      <td>0.985682</td>\n",
       "      <td>0.982689</td>\n",
       "      <td>0.906536</td>\n",
       "      <td>0.978538</td>\n",
       "      <td>0.937355</td>\n",
       "      <td>0.970239</td>\n",
       "      <td>0.944574</td>\n",
       "      <td>0.923227</td>\n",
       "      <td>0.980998</td>\n",
       "      <td>0.985862</td>\n",
       "      <td>0.992878</td>\n",
       "      <td>0.983027</td>\n",
       "      <td>0.873222</td>\n",
       "      <td>0.996068</td>\n",
       "      <td>0.981925</td>\n",
       "      <td>0.958488</td>\n",
       "      <td>0.996247</td>\n",
       "      <td>0.996054</td>\n",
       "      <td>0.970371</td>\n",
       "      <td>0.973013</td>\n",
       "      <td>0.973622</td>\n",
       "      <td>0.965161</td>\n",
       "      <td>0.988407</td>\n",
       "      <td>0.970234</td>\n",
       "      <td>0.985868</td>\n",
       "      <td>0.943630</td>\n",
       "      <td>0.819314</td>\n",
       "      <td>0.938482</td>\n",
       "      <td>0.952506</td>\n",
       "      <td>0.980911</td>\n",
       "      <td>0.988051</td>\n",
       "      <td>0.975935</td>\n",
       "      <td>0.921501</td>\n",
       "      <td>0.971955</td>\n",
       "      <td>0.976553</td>\n",
       "      <td>0.961308</td>\n",
       "      <td>0.962721</td>\n",
       "      <td>0.996436</td>\n",
       "      <td>0.990649</td>\n",
       "      <td>0.986348</td>\n",
       "      <td>0.954686</td>\n",
       "      <td>0.984010</td>\n",
       "      <td>0.998333</td>\n",
       "      <td>0.995046</td>\n",
       "      <td>0.996945</td>\n",
       "      <td>0.935542</td>\n",
       "      <td>0.971322</td>\n",
       "      <td>0.963798</td>\n",
       "      <td>0.989304</td>\n",
       "      <td>0.961676</td>\n",
       "      <td>0.909218</td>\n",
       "      <td>0.972137</td>\n",
       "      <td>0.984640</td>\n",
       "      <td>0.941276</td>\n",
       "      <td>0.914055</td>\n",
       "      <td>0.996700</td>\n",
       "      <td>0.973780</td>\n",
       "      <td>0.984575</td>\n",
       "      <td>0.983629</td>\n",
       "      <td>0.979572</td>\n",
       "      <td>0.989709</td>\n",
       "      <td>0.988069</td>\n",
       "      <td>0.879544</td>\n",
       "      <td>0.990320</td>\n",
       "      <td>0.977579</td>\n",
       "      <td>0.969606</td>\n",
       "      <td>0.985269</td>\n",
       "      <td>0.807804</td>\n",
       "      <td>0.976838</td>\n",
       "      <td>0.851307</td>\n",
       "      <td>0.988131</td>\n",
       "      <td>0.943015</td>\n",
       "      <td>...</td>\n",
       "      <td>0.982313</td>\n",
       "      <td>0.985683</td>\n",
       "      <td>0.992984</td>\n",
       "      <td>0.988352</td>\n",
       "      <td>0.834004</td>\n",
       "      <td>0.991213</td>\n",
       "      <td>0.999283</td>\n",
       "      <td>0.992114</td>\n",
       "      <td>0.992033</td>\n",
       "      <td>0.936641</td>\n",
       "      <td>0.958363</td>\n",
       "      <td>0.981761</td>\n",
       "      <td>0.963807</td>\n",
       "      <td>0.980325</td>\n",
       "      <td>0.989877</td>\n",
       "      <td>0.916012</td>\n",
       "      <td>0.959801</td>\n",
       "      <td>0.957176</td>\n",
       "      <td>0.937554</td>\n",
       "      <td>0.973596</td>\n",
       "      <td>0.951490</td>\n",
       "      <td>0.973167</td>\n",
       "      <td>0.959685</td>\n",
       "      <td>0.987583</td>\n",
       "      <td>0.994991</td>\n",
       "      <td>0.986589</td>\n",
       "      <td>0.990358</td>\n",
       "      <td>0.984640</td>\n",
       "      <td>0.983943</td>\n",
       "      <td>0.981917</td>\n",
       "      <td>0.985673</td>\n",
       "      <td>0.979008</td>\n",
       "      <td>0.995468</td>\n",
       "      <td>0.992109</td>\n",
       "      <td>0.995718</td>\n",
       "      <td>0.905527</td>\n",
       "      <td>0.978246</td>\n",
       "      <td>0.949338</td>\n",
       "      <td>0.961583</td>\n",
       "      <td>0.991132</td>\n",
       "      <td>0.698479</td>\n",
       "      <td>0.878103</td>\n",
       "      <td>0.938664</td>\n",
       "      <td>0.941828</td>\n",
       "      <td>0.910275</td>\n",
       "      <td>0.993352</td>\n",
       "      <td>0.973185</td>\n",
       "      <td>0.974632</td>\n",
       "      <td>0.879820</td>\n",
       "      <td>0.971996</td>\n",
       "      <td>0.991848</td>\n",
       "      <td>0.946107</td>\n",
       "      <td>0.996981</td>\n",
       "      <td>0.953958</td>\n",
       "      <td>0.969832</td>\n",
       "      <td>0.987080</td>\n",
       "      <td>0.954606</td>\n",
       "      <td>0.960255</td>\n",
       "      <td>0.999015</td>\n",
       "      <td>0.995428</td>\n",
       "      <td>0.927151</td>\n",
       "      <td>0.961038</td>\n",
       "      <td>0.962591</td>\n",
       "      <td>0.933289</td>\n",
       "      <td>0.916554</td>\n",
       "      <td>0.985206</td>\n",
       "      <td>0.978720</td>\n",
       "      <td>0.978963</td>\n",
       "      <td>0.986775</td>\n",
       "      <td>0.915564</td>\n",
       "      <td>0.968877</td>\n",
       "      <td>0.993506</td>\n",
       "      <td>0.995939</td>\n",
       "      <td>0.982688</td>\n",
       "      <td>0.960686</td>\n",
       "      <td>0.987055</td>\n",
       "      <td>0.985477</td>\n",
       "      <td>0.980632</td>\n",
       "      <td>0.996868</td>\n",
       "      <td>0.975097</td>\n",
       "      <td>0.978805</td>\n",
       "      <td>0.972199</td>\n",
       "      <td>0.980035</td>\n",
       "      <td>0.979632</td>\n",
       "      <td>0.981614</td>\n",
       "      <td>0.946957</td>\n",
       "      <td>0.983215</td>\n",
       "      <td>0.993577</td>\n",
       "      <td>0.989796</td>\n",
       "      <td>0.985959</td>\n",
       "      <td>0.998428</td>\n",
       "      <td>0.995931</td>\n",
       "      <td>0.994079</td>\n",
       "      <td>0.984256</td>\n",
       "      <td>0.995169</td>\n",
       "      <td>0.982770</td>\n",
       "      <td>0.970700</td>\n",
       "      <td>0.997255</td>\n",
       "      <td>0.989502</td>\n",
       "      <td>0.989360</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 481 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                        0_48      1_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.029698  0.080832   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.295901  0.616100   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.688192  0.856472   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.910836  0.928538   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.847571  0.963856   \n",
       "\n",
       "                                                        2_48      3_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.049923  0.118166   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.585490  0.763552   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.740203  0.928254   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.527132  0.888624   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.809928  0.791543   \n",
       "\n",
       "                                                        4_48      5_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.223824  0.177816   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.874480  0.942619   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.940908  0.941489   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.915498  0.909505   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.949490  0.851008   \n",
       "\n",
       "                                                        6_48      7_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.045554  0.106425   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.703923  0.629009   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.899931  0.877026   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.964951  0.962932   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.970709  0.963803   \n",
       "\n",
       "                                                        8_48      9_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.194602  0.048929   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.812236  0.865754   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.915597  0.972010   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.948339  0.975626   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.949462  0.952987   \n",
       "\n",
       "                                                       10_48     11_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.061413  0.047669   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.784470  0.449863   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.915853  0.928046   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.925406  0.976905   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.916027  0.980209   \n",
       "\n",
       "                                                       12_48     13_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.088884  0.206801   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.337121  0.661061   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.983680  0.964526   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.986096  0.977565   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.975572  0.982690   \n",
       "\n",
       "                                                       14_48     15_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.064651  0.275370   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.956659  0.866433   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.906938  0.924819   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.980430  0.956631   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.973211  0.980387   \n",
       "\n",
       "                                                       16_48     17_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.151547  0.077658   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.686306  0.927653   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.962019  0.911155   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.976478  0.949210   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.992644  0.959100   \n",
       "\n",
       "                                                       18_48     19_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.029862  0.041826   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.934034  0.947765   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.922102  0.940164   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.906159  0.983026   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.954591  0.971579   \n",
       "\n",
       "                                                       20_48     21_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.051233  0.011541   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.808483  0.833280   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.843983  0.893247   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.974964  0.961030   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.983339  0.992214   \n",
       "\n",
       "                                                       22_48     23_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.049218  0.089420   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.770048  0.862283   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.902425  0.894499   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.980169  0.964771   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.985897  0.980644   \n",
       "\n",
       "                                                       24_48     25_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.022117  0.013123   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.960639  0.920410   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.951999  0.704206   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.931328  0.992986   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.988835  0.994517   \n",
       "\n",
       "                                                       26_48     27_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.030460  0.013854   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.881784  0.942431   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.960727  0.959247   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.991289  0.989885   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.994570  0.995637   \n",
       "\n",
       "                                                       28_48     29_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.183592  0.064270   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.725764  0.811090   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.960867  0.914251   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.994662  0.992780   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.985682  0.982689   \n",
       "\n",
       "                                                       30_48     31_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.283902  0.078608   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.913210  0.927040   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.911883  0.548480   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.954595  0.970848   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.906536  0.978538   \n",
       "\n",
       "                                                       32_48     33_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.084389  0.008901   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.346670  0.804140   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.959223  0.900190   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.911303  0.979076   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.937355  0.970239   \n",
       "\n",
       "                                                       34_48     35_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.031846  0.214726   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.554704  0.771561   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.726230  0.941350   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.985443  0.973002   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.944574  0.923227   \n",
       "\n",
       "                                                       36_48     37_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.013098  0.327583   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.360767  0.902478   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.972824  0.990653   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.991811  0.988492   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.980998  0.985862   \n",
       "\n",
       "                                                       38_48     39_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.740116  0.506495   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.600150  0.942712   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.575059  0.843176   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.960801  0.997184   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.992878  0.983027   \n",
       "\n",
       "                                                       40_48     41_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.331147  0.346144   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.617142  0.657846   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.180662  0.611569   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.943886  0.994764   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.873222  0.996068   \n",
       "\n",
       "                                                       42_48     43_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.167521  0.276274   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.777856  0.892649   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.907423  0.855664   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.992135  0.991517   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.981925  0.958488   \n",
       "\n",
       "                                                       44_48     45_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.395539  0.860666   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.845431  0.937457   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.989813  0.713413   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.981689  0.994041   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.996247  0.996054   \n",
       "\n",
       "                                                       46_48     47_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.751504  0.311051   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.516100  0.339713   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.967392  0.867970   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.970371  0.973013   \n",
       "\n",
       "                                                       48_48     49_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.804861  0.598100   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.418890  0.245681   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.645537  0.710680   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.973622  0.965161   \n",
       "\n",
       "                                                       50_48     51_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.753049  0.142164   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.449455  0.308119   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.781099  0.939137   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.988407  0.970234   \n",
       "\n",
       "                                                       52_48     53_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.060092  0.031496   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.552498  0.709674   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.635680  0.882481   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.985868  0.943630   \n",
       "\n",
       "                                                       54_48     55_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.560822  0.885624   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.107609  0.854450   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.632616  0.338599   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.819314  0.938482   \n",
       "\n",
       "                                                       56_48     57_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.319505  0.479512   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.504910  0.444506   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.837811  0.793955   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.952506  0.980911   \n",
       "\n",
       "                                                       58_48     59_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.493804  0.150953   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.328661  0.489310   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.876178  0.844483   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.988051  0.975935   \n",
       "\n",
       "                                                       60_48     61_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.022642  0.618027   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.256028  0.322243   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.873082  0.710679   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.921501  0.971955   \n",
       "\n",
       "                                                       62_48     63_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.570352  0.170242   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.130332  0.021160   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.921149  0.423949   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.976553  0.961308   \n",
       "\n",
       "                                                       64_48     65_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.271364  0.845575   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.318908  0.374743   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.316755  0.898936   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.962721  0.996436   \n",
       "\n",
       "                                                       66_48     67_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.094429  0.559305   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.091339  0.308014   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.942313  0.519842   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.990649  0.986348   \n",
       "\n",
       "                                                       68_48     69_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.352102  0.923105   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.311482  0.105572   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.678256  0.756128   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.954686  0.984010   \n",
       "\n",
       "                                                       70_48     71_48  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.621745  0.295658   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.245726  0.194622   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.559718  0.576955   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.998333  0.995046   \n",
       "\n",
       "                                                       72_48      0_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.145317  0.052655   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.853658  0.634948   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.801891  0.930794   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN  0.890511   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.996945  0.935542   \n",
       "\n",
       "                                                        1_54      2_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.125262  0.296246   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.893021  0.749243   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.973386  0.953991   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.971366  0.922571   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.971322  0.963798   \n",
       "\n",
       "                                                        3_54      4_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.158113  0.260551   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.796109  0.669738   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.964543  0.884326   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.992779  0.988530   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.989304  0.961676   \n",
       "\n",
       "                                                        5_54      6_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.172064  0.261298   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.554182  0.465155   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.972037  0.979429   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.984489  0.989918   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.909218  0.972137   \n",
       "\n",
       "                                                        7_54      8_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.182585  0.107669   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.070758  0.905468   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.949053  0.948033   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.981094  0.917718   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.984640  0.941276   \n",
       "\n",
       "                                                        9_54     10_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.137240  0.360731   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.877355  0.868039   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.874371  0.950503   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.929050  0.913255   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.914055  0.996700   \n",
       "\n",
       "                                                       11_54     12_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.112294  0.012958   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.876720  0.749644   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.955456  0.959813   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.955568  0.987435   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.973780  0.984575   \n",
       "\n",
       "                                                       13_54     14_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.013463  0.055620   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.920808  0.781602   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.945319  0.873793   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.983431  0.917882   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.983629  0.979572   \n",
       "\n",
       "                                                       15_54     16_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.005191  0.103964   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.775674  0.880736   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.915594  0.902709   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.955393  0.931522   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.989709  0.988069   \n",
       "\n",
       "                                                       17_54     18_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.058726  0.020764   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.807706  0.838220   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.915216  0.885734   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.823321  0.954594   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.879544  0.990320   \n",
       "\n",
       "                                                       19_54     20_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.009628  0.024620   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.957286  0.863977   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.860223  0.805894   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.981330  0.928382   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.977579  0.969606   \n",
       "\n",
       "                                                       21_54     22_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.109718  0.467515   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.944742  0.857148   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.796016  0.927726   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.980354  0.995421   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.985269  0.807804   \n",
       "\n",
       "                                                       23_54     24_54  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.022748  0.094592   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.972580  0.507750   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.550648  0.751867   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.994731  0.943641   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.976838  0.851307   \n",
       "\n",
       "                                                       25_54     26_54  ...  \\\n",
       "filehash                                                                ...   \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.035569  0.520413  ...   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.855397  0.791277  ...   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.884138  0.865604  ...   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.967961  0.924548  ...   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.988131  0.943015  ...   \n",
       "\n",
       "                                                       33_84     34_84  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.494885  0.551332   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.172838  0.239157   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.622803  0.776738   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.982313  0.985683   \n",
       "\n",
       "                                                       35_84     36_84  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.894536  0.101025   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.183369  0.112225   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.636121  0.633256   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.992984  0.988352   \n",
       "\n",
       "                                                        0_90      1_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.085483  0.204991   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.291436  0.819209   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.919222  0.917695   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.995000  0.990181   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.834004  0.991213   \n",
       "\n",
       "                                                        2_90      3_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.740870  0.647179   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.447267  0.907690   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.983487  0.893850   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.981080  0.994067   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.999283  0.992114   \n",
       "\n",
       "                                                        4_90      5_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.394482  0.303274   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.720795  0.661324   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.980378  0.855946   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.992033  0.936641   \n",
       "\n",
       "                                                        6_90      7_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.893054  0.264133   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.468694  0.626997   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.713308  0.876186   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.958363  0.981761   \n",
       "\n",
       "                                                        8_90      9_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.273751  0.057169   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.557113  0.429908   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.699076  0.660392   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.963807  0.980325   \n",
       "\n",
       "                                                       10_90     11_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.138298  0.220179   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.326878  0.585498   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.950420  0.823178   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.989877  0.916012   \n",
       "\n",
       "                                                       12_90     13_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.605863  0.527785   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.180201  0.446671   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.908768  0.858369   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.959801  0.957176   \n",
       "\n",
       "                                                       14_90     15_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.706976  0.273590   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.340976  0.566979   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.837563  0.788934   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.937554  0.973596   \n",
       "\n",
       "                                                       16_90     17_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.363384  0.744208   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.423755  0.598135   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.785083  0.928731   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.951490  0.973167   \n",
       "\n",
       "                                                       18_90     19_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.128260  0.798777   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.845344  0.179377   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.875839  0.717817   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.959685  0.987583   \n",
       "\n",
       "                                                       20_90     21_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.098031  0.288894   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.158628  0.141114   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.815259  0.906099   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.994991  0.986589   \n",
       "\n",
       "                                                       22_90     23_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.065271  0.651618   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.310970  0.445229   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.483409  0.886357   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.990358  0.984640   \n",
       "\n",
       "                                                       24_90     25_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.943892  0.758336   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.293999  0.151553   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.916344  0.721530   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.983943  0.981917   \n",
       "\n",
       "                                                       26_90     27_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.699642  0.156586   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.257232  0.205178   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.602255  0.671580   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.985673  0.979008   \n",
       "\n",
       "                                                       28_90     29_90  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.336604  0.060895   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.743767  0.550082   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.561276  0.156611   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.995468  0.992109   \n",
       "\n",
       "                                                       30_90      0_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.469109  0.481473   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.193352  0.356057   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.949887  0.419144   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.995718  0.905527   \n",
       "\n",
       "                                                        1_96      2_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.316594  0.207696   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.942356  0.357973   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.749896  0.341601   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.978246  0.949338   \n",
       "\n",
       "                                                        3_96      4_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.245382  0.449606   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.245077  0.311749   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.608766  0.949398   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.961583  0.991132   \n",
       "\n",
       "                                                        5_96      6_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.220301  0.559647   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.817325  0.082411   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.936689  0.775339   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.698479  0.878103   \n",
       "\n",
       "                                                        7_96      8_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.556434  0.724366   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.252692  0.696436   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.436182  0.747801   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.938664  0.941828   \n",
       "\n",
       "                                                        9_96     10_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.327392  0.737083   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.352933  0.553481   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.815390  0.953011   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.910275  0.993352   \n",
       "\n",
       "                                                       11_96     12_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.367128  0.221972   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.315569  0.297744   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.917669  0.764154   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.973185  0.974632   \n",
       "\n",
       "                                                       13_96     14_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.841491  0.041223   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.395014  0.430080   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.603275  0.744248   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.879820  0.971996   \n",
       "\n",
       "                                                       15_96     16_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.837291  0.008387   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.601526  0.671068   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.885763  0.423015   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.991848  0.946107   \n",
       "\n",
       "                                                       17_96     18_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.633813  0.318139   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.114176  0.102735   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.408318  0.587098   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.996981  0.953958   \n",
       "\n",
       "                                                       19_96     20_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.024553  0.163532   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.056013  0.205072   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.770504  0.585670   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.969832  0.987080   \n",
       "\n",
       "                                                       21_96     22_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.089204  0.026484   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.061710  0.180499   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.741818  0.312818   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.954606  0.960255   \n",
       "\n",
       "                                                       23_96     24_96  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.137825  0.041519   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.352650  0.282756   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.503492  0.777177   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.999015  0.995428   \n",
       "\n",
       "                                                       0_102     1_102  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.250772  0.436267   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.259089  0.313544   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.806311  0.328509   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.927151  0.961038   \n",
       "\n",
       "                                                       2_102     3_102  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.308587  0.481186   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.612125  0.344980   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.797078  0.825528   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.962591  0.933289   \n",
       "\n",
       "                                                       4_102     5_102  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.482252  0.368783   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.502287  0.382664   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.867705  0.931840   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.916554  0.985206   \n",
       "\n",
       "                                                       6_102     7_102  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.330872  0.637380   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.372557  0.361522   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.860384  0.769309   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.978720  0.978963   \n",
       "\n",
       "                                                       8_102     9_102  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.630386  0.452975   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.090581  0.099027   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.661027  0.706649   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.986775  0.915564   \n",
       "\n",
       "                                                      10_102    11_102  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.618075  0.540244   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.159275  0.329697   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.546251  0.621745   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.968877  0.993506   \n",
       "\n",
       "                                                      12_102    13_102  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.727182  0.165879   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.055109  0.214870   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.643203  0.393842   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.995939  0.982688   \n",
       "\n",
       "                                                      14_102    15_102  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.576970  0.032162   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.089066  0.072910   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.848718  0.933219   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.960686  0.987055   \n",
       "\n",
       "                                                      16_102    17_102  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.121999  0.176663   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.149949  0.259793   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.503421  0.680889   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.985477  0.980632   \n",
       "\n",
       "                                                      18_102     0_108  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.097464  0.130436   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.205103  0.318784   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.860788  0.817465   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.996868  0.975097   \n",
       "\n",
       "                                                       1_108     2_108  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.714333  0.175717   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.522218  0.133060   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.660918  0.731410   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.978805  0.972199   \n",
       "\n",
       "                                                       3_108     4_108  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.739042  0.020250   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.282103  0.193713   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.589714  0.430669   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.980035  0.979632   \n",
       "\n",
       "                                                       5_108     6_108  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.485943  0.401508   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.164036  0.089202   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.896000  0.572216   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.981614  0.946957   \n",
       "\n",
       "                                                       7_108     8_108  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.636606  0.870755   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.466361  0.384753   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.821138  0.283826   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.983215  0.993577   \n",
       "\n",
       "                                                       9_108    10_108  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.282143  0.194977   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.141033  0.186505   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.893737  0.516957   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.989796  0.985959   \n",
       "\n",
       "                                                      11_108    12_108  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.077225  0.023264   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.388967  0.302577   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.839548  0.964948   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.998428  0.995931   \n",
       "\n",
       "                                                       0_114     1_114  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.804053  0.802499   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.588224  0.137935   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.837720  0.782127   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.994079  0.984256   \n",
       "\n",
       "                                                       2_114     3_114  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.432499  0.559297   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.356914  0.291190   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.697565  0.918200   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.995169  0.982770   \n",
       "\n",
       "                                                       4_114     5_114  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.016355  0.094903   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.082553  0.467127   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.192754  0.734890   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.970700  0.997255   \n",
       "\n",
       "                                                       6_114     0_120  \n",
       "filehash                                                                \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.114713  0.206466  \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.473013  0.528720  \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.940787  0.874146  \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN  \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.989502  0.989360  \n",
       "\n",
       "[5 rows x 481 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sort columns \n",
    "prediction_results.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# read in the metadata\n",
    "spawningtool_df = pd.read_csv('data/spawningtool_replays.csv')\n",
    "\n",
    "# get the labels\n",
    "labels = create_labels(spawningtool_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>48</th>\n",
       "      <th>49</th>\n",
       "      <th>50</th>\n",
       "      <th>51</th>\n",
       "      <th>52</th>\n",
       "      <th>53</th>\n",
       "      <th>54</th>\n",
       "      <th>55</th>\n",
       "      <th>56</th>\n",
       "      <th>57</th>\n",
       "      <th>58</th>\n",
       "      <th>59</th>\n",
       "      <th>60</th>\n",
       "      <th>61</th>\n",
       "      <th>62</th>\n",
       "      <th>63</th>\n",
       "      <th>64</th>\n",
       "      <th>65</th>\n",
       "      <th>66</th>\n",
       "      <th>67</th>\n",
       "      <th>68</th>\n",
       "      <th>69</th>\n",
       "      <th>70</th>\n",
       "      <th>71</th>\n",
       "      <th>72</th>\n",
       "      <th>73</th>\n",
       "      <th>74</th>\n",
       "      <th>75</th>\n",
       "      <th>76</th>\n",
       "      <th>77</th>\n",
       "      <th>78</th>\n",
       "      <th>79</th>\n",
       "      <th>80</th>\n",
       "      <th>81</th>\n",
       "      <th>82</th>\n",
       "      <th>83</th>\n",
       "      <th>84</th>\n",
       "      <th>85</th>\n",
       "      <th>86</th>\n",
       "      <th>87</th>\n",
       "      <th>88</th>\n",
       "      <th>89</th>\n",
       "      <th>90</th>\n",
       "      <th>91</th>\n",
       "      <th>92</th>\n",
       "      <th>93</th>\n",
       "      <th>94</th>\n",
       "      <th>95</th>\n",
       "      <th>96</th>\n",
       "      <th>97</th>\n",
       "      <th>98</th>\n",
       "      <th>99</th>\n",
       "      <th>100</th>\n",
       "      <th>101</th>\n",
       "      <th>102</th>\n",
       "      <th>103</th>\n",
       "      <th>104</th>\n",
       "      <th>105</th>\n",
       "      <th>106</th>\n",
       "      <th>107</th>\n",
       "      <th>108</th>\n",
       "      <th>109</th>\n",
       "      <th>110</th>\n",
       "      <th>111</th>\n",
       "      <th>112</th>\n",
       "      <th>113</th>\n",
       "      <th>114</th>\n",
       "      <th>115</th>\n",
       "      <th>116</th>\n",
       "      <th>117</th>\n",
       "      <th>118</th>\n",
       "      <th>119</th>\n",
       "      <th>120</th>\n",
       "      <th>winner</th>\n",
       "      <th>length</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>filehash</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0016ddd5b36473f259ba8630c6c0047540a6858071737e8f890358d3f7ad355d</th>\n",
       "      <td>0.029698</td>\n",
       "      <td>0.080832</td>\n",
       "      <td>0.049923</td>\n",
       "      <td>0.118166</td>\n",
       "      <td>0.223824</td>\n",
       "      <td>0.177816</td>\n",
       "      <td>0.049105</td>\n",
       "      <td>0.115844</td>\n",
       "      <td>0.245424</td>\n",
       "      <td>0.103521</td>\n",
       "      <td>0.160982</td>\n",
       "      <td>0.109867</td>\n",
       "      <td>0.189026</td>\n",
       "      <td>0.170682</td>\n",
       "      <td>0.089482</td>\n",
       "      <td>0.184983</td>\n",
       "      <td>0.259445</td>\n",
       "      <td>0.117815</td>\n",
       "      <td>0.027187</td>\n",
       "      <td>0.032933</td>\n",
       "      <td>0.044005</td>\n",
       "      <td>0.011349</td>\n",
       "      <td>0.046533</td>\n",
       "      <td>0.070703</td>\n",
       "      <td>0.024941</td>\n",
       "      <td>0.022455</td>\n",
       "      <td>0.029916</td>\n",
       "      <td>0.040424</td>\n",
       "      <td>0.224837</td>\n",
       "      <td>0.062168</td>\n",
       "      <td>0.141769</td>\n",
       "      <td>0.212538</td>\n",
       "      <td>0.139552</td>\n",
       "      <td>0.019185</td>\n",
       "      <td>0.147976</td>\n",
       "      <td>0.067908</td>\n",
       "      <td>0.075190</td>\n",
       "      <td>0.331753</td>\n",
       "      <td>0.708202</td>\n",
       "      <td>0.202371</td>\n",
       "      <td>0.288190</td>\n",
       "      <td>0.394197</td>\n",
       "      <td>0.117126</td>\n",
       "      <td>0.412181</td>\n",
       "      <td>0.543048</td>\n",
       "      <td>0.634000</td>\n",
       "      <td>0.587407</td>\n",
       "      <td>0.256679</td>\n",
       "      <td>0.734369</td>\n",
       "      <td>0.640322</td>\n",
       "      <td>0.439475</td>\n",
       "      <td>0.184596</td>\n",
       "      <td>0.140991</td>\n",
       "      <td>0.148540</td>\n",
       "      <td>0.543569</td>\n",
       "      <td>0.567708</td>\n",
       "      <td>0.661769</td>\n",
       "      <td>0.381437</td>\n",
       "      <td>0.415634</td>\n",
       "      <td>0.366015</td>\n",
       "      <td>0.301227</td>\n",
       "      <td>0.725185</td>\n",
       "      <td>0.409948</td>\n",
       "      <td>0.529443</td>\n",
       "      <td>0.245262</td>\n",
       "      <td>0.498787</td>\n",
       "      <td>0.577625</td>\n",
       "      <td>0.498656</td>\n",
       "      <td>0.491801</td>\n",
       "      <td>0.317357</td>\n",
       "      <td>0.196038</td>\n",
       "      <td>0.296253</td>\n",
       "      <td>0.211617</td>\n",
       "      <td>0</td>\n",
       "      <td>161.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e7827e10660ad88bf5bee</th>\n",
       "      <td>0.295901</td>\n",
       "      <td>0.616100</td>\n",
       "      <td>0.585490</td>\n",
       "      <td>0.763552</td>\n",
       "      <td>0.874480</td>\n",
       "      <td>0.942619</td>\n",
       "      <td>0.669436</td>\n",
       "      <td>0.761015</td>\n",
       "      <td>0.780740</td>\n",
       "      <td>0.830931</td>\n",
       "      <td>0.727104</td>\n",
       "      <td>0.502022</td>\n",
       "      <td>0.452369</td>\n",
       "      <td>0.415986</td>\n",
       "      <td>0.937942</td>\n",
       "      <td>0.897614</td>\n",
       "      <td>0.818862</td>\n",
       "      <td>0.907613</td>\n",
       "      <td>0.867234</td>\n",
       "      <td>0.850644</td>\n",
       "      <td>0.780323</td>\n",
       "      <td>0.793507</td>\n",
       "      <td>0.883239</td>\n",
       "      <td>0.890958</td>\n",
       "      <td>0.809717</td>\n",
       "      <td>0.923362</td>\n",
       "      <td>0.632340</td>\n",
       "      <td>0.882474</td>\n",
       "      <td>0.871142</td>\n",
       "      <td>0.917342</td>\n",
       "      <td>0.860490</td>\n",
       "      <td>0.872128</td>\n",
       "      <td>0.701284</td>\n",
       "      <td>0.727705</td>\n",
       "      <td>0.736100</td>\n",
       "      <td>0.632429</td>\n",
       "      <td>0.410252</td>\n",
       "      <td>0.742076</td>\n",
       "      <td>0.761465</td>\n",
       "      <td>0.723962</td>\n",
       "      <td>0.601294</td>\n",
       "      <td>0.647466</td>\n",
       "      <td>0.433692</td>\n",
       "      <td>0.549625</td>\n",
       "      <td>0.630394</td>\n",
       "      <td>0.847661</td>\n",
       "      <td>0.676614</td>\n",
       "      <td>0.515724</td>\n",
       "      <td>0.395499</td>\n",
       "      <td>0.573178</td>\n",
       "      <td>0.545287</td>\n",
       "      <td>0.363934</td>\n",
       "      <td>0.389985</td>\n",
       "      <td>0.591466</td>\n",
       "      <td>0.226051</td>\n",
       "      <td>0.464468</td>\n",
       "      <td>0.534707</td>\n",
       "      <td>0.401697</td>\n",
       "      <td>0.424739</td>\n",
       "      <td>0.504880</td>\n",
       "      <td>0.414498</td>\n",
       "      <td>0.333180</td>\n",
       "      <td>0.191836</td>\n",
       "      <td>0.216881</td>\n",
       "      <td>0.304967</td>\n",
       "      <td>0.250442</td>\n",
       "      <td>0.154725</td>\n",
       "      <td>0.221425</td>\n",
       "      <td>0.303488</td>\n",
       "      <td>0.252059</td>\n",
       "      <td>0.258118</td>\n",
       "      <td>0.325752</td>\n",
       "      <td>0.351423</td>\n",
       "      <td>0</td>\n",
       "      <td>178.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0033334e86cb352b0131027e58effa23adb23cf1cdd90d5760cc62254611c99a</th>\n",
       "      <td>0.688192</td>\n",
       "      <td>0.856472</td>\n",
       "      <td>0.740203</td>\n",
       "      <td>0.928254</td>\n",
       "      <td>0.940908</td>\n",
       "      <td>0.941489</td>\n",
       "      <td>0.915362</td>\n",
       "      <td>0.925206</td>\n",
       "      <td>0.934794</td>\n",
       "      <td>0.968276</td>\n",
       "      <td>0.900090</td>\n",
       "      <td>0.950041</td>\n",
       "      <td>0.979137</td>\n",
       "      <td>0.937121</td>\n",
       "      <td>0.930140</td>\n",
       "      <td>0.918943</td>\n",
       "      <td>0.949939</td>\n",
       "      <td>0.938874</td>\n",
       "      <td>0.943798</td>\n",
       "      <td>0.951096</td>\n",
       "      <td>0.826471</td>\n",
       "      <td>0.914995</td>\n",
       "      <td>0.889222</td>\n",
       "      <td>0.913643</td>\n",
       "      <td>0.910013</td>\n",
       "      <td>0.822716</td>\n",
       "      <td>0.902883</td>\n",
       "      <td>0.926848</td>\n",
       "      <td>0.795082</td>\n",
       "      <td>0.777135</td>\n",
       "      <td>0.872945</td>\n",
       "      <td>0.733354</td>\n",
       "      <td>0.857619</td>\n",
       "      <td>0.900528</td>\n",
       "      <td>0.869856</td>\n",
       "      <td>0.940228</td>\n",
       "      <td>0.932453</td>\n",
       "      <td>0.961672</td>\n",
       "      <td>0.549144</td>\n",
       "      <td>0.761709</td>\n",
       "      <td>0.613517</td>\n",
       "      <td>0.657306</td>\n",
       "      <td>0.775918</td>\n",
       "      <td>0.829999</td>\n",
       "      <td>0.944010</td>\n",
       "      <td>0.686689</td>\n",
       "      <td>0.959113</td>\n",
       "      <td>0.884538</td>\n",
       "      <td>0.642170</td>\n",
       "      <td>0.843438</td>\n",
       "      <td>0.711400</td>\n",
       "      <td>0.754922</td>\n",
       "      <td>0.828797</td>\n",
       "      <td>0.842460</td>\n",
       "      <td>0.787398</td>\n",
       "      <td>0.600437</td>\n",
       "      <td>0.844334</td>\n",
       "      <td>0.792091</td>\n",
       "      <td>0.816067</td>\n",
       "      <td>0.901142</td>\n",
       "      <td>0.837865</td>\n",
       "      <td>0.705645</td>\n",
       "      <td>0.779353</td>\n",
       "      <td>0.681035</td>\n",
       "      <td>0.570873</td>\n",
       "      <td>0.747662</td>\n",
       "      <td>0.742645</td>\n",
       "      <td>0.679350</td>\n",
       "      <td>0.696317</td>\n",
       "      <td>0.789129</td>\n",
       "      <td>0.540936</td>\n",
       "      <td>0.661435</td>\n",
       "      <td>0.849607</td>\n",
       "      <td>1</td>\n",
       "      <td>214.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0055f4541fe30a72e964fc168db7abef35822d27e4cdce8ef0adb8b3cba51b7d</th>\n",
       "      <td>0.910836</td>\n",
       "      <td>0.928538</td>\n",
       "      <td>0.527132</td>\n",
       "      <td>0.888624</td>\n",
       "      <td>0.915498</td>\n",
       "      <td>0.909505</td>\n",
       "      <td>0.927731</td>\n",
       "      <td>0.967149</td>\n",
       "      <td>0.935455</td>\n",
       "      <td>0.984202</td>\n",
       "      <td>0.956968</td>\n",
       "      <td>0.980697</td>\n",
       "      <td>0.987332</td>\n",
       "      <td>0.939530</td>\n",
       "      <td>0.952741</td>\n",
       "      <td>0.953666</td>\n",
       "      <td>0.946691</td>\n",
       "      <td>0.956564</td>\n",
       "      <td>0.917624</td>\n",
       "      <td>0.982332</td>\n",
       "      <td>0.945112</td>\n",
       "      <td>0.966749</td>\n",
       "      <td>0.957718</td>\n",
       "      <td>0.879283</td>\n",
       "      <td>0.942988</td>\n",
       "      <td>0.981035</td>\n",
       "      <td>0.961570</td>\n",
       "      <td>0.974043</td>\n",
       "      <td>0.988248</td>\n",
       "      <td>0.992702</td>\n",
       "      <td>0.977239</td>\n",
       "      <td>0.951895</td>\n",
       "      <td>0.964291</td>\n",
       "      <td>0.984865</td>\n",
       "      <td>0.988028</td>\n",
       "      <td>0.986600</td>\n",
       "      <td>0.984741</td>\n",
       "      <td>0.990989</td>\n",
       "      <td>0.964949</td>\n",
       "      <td>0.992553</td>\n",
       "      <td>0.981271</td>\n",
       "      <td>0.984607</td>\n",
       "      <td>0.987333</td>\n",
       "      <td>0.986069</td>\n",
       "      <td>0.981104</td>\n",
       "      <td>0.987983</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>94.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e46cd9f2cbf3fa3a99</th>\n",
       "      <td>0.847571</td>\n",
       "      <td>0.963856</td>\n",
       "      <td>0.809928</td>\n",
       "      <td>0.791543</td>\n",
       "      <td>0.949490</td>\n",
       "      <td>0.851008</td>\n",
       "      <td>0.953126</td>\n",
       "      <td>0.967563</td>\n",
       "      <td>0.956630</td>\n",
       "      <td>0.971145</td>\n",
       "      <td>0.938852</td>\n",
       "      <td>0.944714</td>\n",
       "      <td>0.977624</td>\n",
       "      <td>0.977813</td>\n",
       "      <td>0.957068</td>\n",
       "      <td>0.957609</td>\n",
       "      <td>0.987875</td>\n",
       "      <td>0.970643</td>\n",
       "      <td>0.975519</td>\n",
       "      <td>0.968132</td>\n",
       "      <td>0.940672</td>\n",
       "      <td>0.966959</td>\n",
       "      <td>0.988862</td>\n",
       "      <td>0.956328</td>\n",
       "      <td>0.983833</td>\n",
       "      <td>0.986910</td>\n",
       "      <td>0.983948</td>\n",
       "      <td>0.980048</td>\n",
       "      <td>0.924302</td>\n",
       "      <td>0.977331</td>\n",
       "      <td>0.920068</td>\n",
       "      <td>0.951111</td>\n",
       "      <td>0.968223</td>\n",
       "      <td>0.956626</td>\n",
       "      <td>0.960833</td>\n",
       "      <td>0.968167</td>\n",
       "      <td>0.980781</td>\n",
       "      <td>0.985699</td>\n",
       "      <td>0.990506</td>\n",
       "      <td>0.989244</td>\n",
       "      <td>0.972191</td>\n",
       "      <td>0.989290</td>\n",
       "      <td>0.958299</td>\n",
       "      <td>0.984030</td>\n",
       "      <td>0.995166</td>\n",
       "      <td>0.991015</td>\n",
       "      <td>0.985913</td>\n",
       "      <td>0.951528</td>\n",
       "      <td>0.966056</td>\n",
       "      <td>0.964792</td>\n",
       "      <td>0.977254</td>\n",
       "      <td>0.976929</td>\n",
       "      <td>0.982447</td>\n",
       "      <td>0.713689</td>\n",
       "      <td>0.861239</td>\n",
       "      <td>0.932462</td>\n",
       "      <td>0.941604</td>\n",
       "      <td>0.952223</td>\n",
       "      <td>0.964997</td>\n",
       "      <td>0.968056</td>\n",
       "      <td>0.968781</td>\n",
       "      <td>0.961734</td>\n",
       "      <td>0.974157</td>\n",
       "      <td>0.968409</td>\n",
       "      <td>0.978671</td>\n",
       "      <td>0.987025</td>\n",
       "      <td>0.977041</td>\n",
       "      <td>0.980725</td>\n",
       "      <td>0.979807</td>\n",
       "      <td>0.981203</td>\n",
       "      <td>0.982252</td>\n",
       "      <td>0.993341</td>\n",
       "      <td>0.993303</td>\n",
       "      <td>1</td>\n",
       "      <td>126.4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                          48        49  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.029698  0.080832   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.295901  0.616100   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.688192  0.856472   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.910836  0.928538   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.847571  0.963856   \n",
       "\n",
       "                                                          50        51  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.049923  0.118166   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.585490  0.763552   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.740203  0.928254   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.527132  0.888624   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.809928  0.791543   \n",
       "\n",
       "                                                          52        53  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.223824  0.177816   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.874480  0.942619   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.940908  0.941489   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.915498  0.909505   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.949490  0.851008   \n",
       "\n",
       "                                                          54        55  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.049105  0.115844   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.669436  0.761015   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.915362  0.925206   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.927731  0.967149   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.953126  0.967563   \n",
       "\n",
       "                                                          56        57  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.245424  0.103521   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.780740  0.830931   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.934794  0.968276   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.935455  0.984202   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.956630  0.971145   \n",
       "\n",
       "                                                          58        59  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.160982  0.109867   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.727104  0.502022   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.900090  0.950041   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.956968  0.980697   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.938852  0.944714   \n",
       "\n",
       "                                                          60        61  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.189026  0.170682   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.452369  0.415986   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.979137  0.937121   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.987332  0.939530   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.977624  0.977813   \n",
       "\n",
       "                                                          62        63  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.089482  0.184983   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.937942  0.897614   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.930140  0.918943   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.952741  0.953666   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.957068  0.957609   \n",
       "\n",
       "                                                          64        65  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.259445  0.117815   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.818862  0.907613   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.949939  0.938874   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.946691  0.956564   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.987875  0.970643   \n",
       "\n",
       "                                                          66        67  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.027187  0.032933   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.867234  0.850644   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.943798  0.951096   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.917624  0.982332   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.975519  0.968132   \n",
       "\n",
       "                                                          68        69  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.044005  0.011349   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.780323  0.793507   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.826471  0.914995   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.945112  0.966749   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.940672  0.966959   \n",
       "\n",
       "                                                          70        71  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.046533  0.070703   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.883239  0.890958   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.889222  0.913643   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.957718  0.879283   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.988862  0.956328   \n",
       "\n",
       "                                                          72        73  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.024941  0.022455   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.809717  0.923362   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.910013  0.822716   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.942988  0.981035   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.983833  0.986910   \n",
       "\n",
       "                                                          74        75  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.029916  0.040424   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.632340  0.882474   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.902883  0.926848   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.961570  0.974043   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.983948  0.980048   \n",
       "\n",
       "                                                          76        77  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.224837  0.062168   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.871142  0.917342   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.795082  0.777135   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.988248  0.992702   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.924302  0.977331   \n",
       "\n",
       "                                                          78        79  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.141769  0.212538   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.860490  0.872128   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.872945  0.733354   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.977239  0.951895   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.920068  0.951111   \n",
       "\n",
       "                                                          80        81  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.139552  0.019185   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.701284  0.727705   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.857619  0.900528   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.964291  0.984865   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.968223  0.956626   \n",
       "\n",
       "                                                          82        83  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.147976  0.067908   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.736100  0.632429   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.869856  0.940228   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.988028  0.986600   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.960833  0.968167   \n",
       "\n",
       "                                                          84        85  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.075190  0.331753   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.410252  0.742076   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.932453  0.961672   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.984741  0.990989   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.980781  0.985699   \n",
       "\n",
       "                                                          86        87  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.708202  0.202371   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.761465  0.723962   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.549144  0.761709   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.964949  0.992553   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.990506  0.989244   \n",
       "\n",
       "                                                          88        89  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.288190  0.394197   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.601294  0.647466   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.613517  0.657306   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.981271  0.984607   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.972191  0.989290   \n",
       "\n",
       "                                                          90        91  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.117126  0.412181   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.433692  0.549625   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.775918  0.829999   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.987333  0.986069   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.958299  0.984030   \n",
       "\n",
       "                                                          92        93  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.543048  0.634000   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.630394  0.847661   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.944010  0.686689   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...  0.981104  0.987983   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.995166  0.991015   \n",
       "\n",
       "                                                          94        95  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.587407  0.256679   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.676614  0.515724   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.959113  0.884538   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.985913  0.951528   \n",
       "\n",
       "                                                          96        97  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.734369  0.640322   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.395499  0.573178   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.642170  0.843438   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.966056  0.964792   \n",
       "\n",
       "                                                          98        99  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.439475  0.184596   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.545287  0.363934   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.711400  0.754922   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.977254  0.976929   \n",
       "\n",
       "                                                         100       101  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.140991  0.148540   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.389985  0.591466   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.828797  0.842460   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.982447  0.713689   \n",
       "\n",
       "                                                         102       103  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.543569  0.567708   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.226051  0.464468   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.787398  0.600437   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.861239  0.932462   \n",
       "\n",
       "                                                         104       105  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.661769  0.381437   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.534707  0.401697   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.844334  0.792091   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.941604  0.952223   \n",
       "\n",
       "                                                         106       107  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.415634  0.366015   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.424739  0.504880   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.816067  0.901142   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.964997  0.968056   \n",
       "\n",
       "                                                         108       109  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.301227  0.725185   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.414498  0.333180   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.837865  0.705645   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.968781  0.961734   \n",
       "\n",
       "                                                         110       111  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.409948  0.529443   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.191836  0.216881   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.779353  0.681035   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.974157  0.968409   \n",
       "\n",
       "                                                         112       113  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.245262  0.498787   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.304967  0.250442   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.570873  0.747662   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.978671  0.987025   \n",
       "\n",
       "                                                         114       115  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.577625  0.498656   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.154725  0.221425   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.742645  0.679350   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.977041  0.980725   \n",
       "\n",
       "                                                         116       117  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.491801  0.317357   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.303488  0.252059   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.696317  0.789129   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.979807  0.981203   \n",
       "\n",
       "                                                         118       119  \\\n",
       "filehash                                                                 \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.196038  0.296253   \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.258118  0.325752   \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.540936  0.661435   \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       NaN   \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.982252  0.993341   \n",
       "\n",
       "                                                         120  winner  length  \n",
       "filehash                                                                      \n",
       "0016ddd5b36473f259ba8630c6c0047540a6858071737e8...  0.211617       0   161.2  \n",
       "00283dcee1b0bec45ff654a0df06248d9f69a5c3a66e782...  0.351423       0   178.2  \n",
       "0033334e86cb352b0131027e58effa23adb23cf1cdd90d5...  0.849607       1   214.6  \n",
       "0055f4541fe30a72e964fc168db7abef35822d27e4cdce8...       NaN       1    94.2  \n",
       "005ac860a39ec7ae855a22b68f13ea17c1d7006bafe381e...  0.993303       1   126.4  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def aggregate_predictions(predictions, metadata):\n",
    "    # loop through each col in prediction results and get a dict of index to sum\n",
    "    prob_index_dict = {}\n",
    "    for col in predictions.columns:\n",
    "        # calculate the row at which the window ends\n",
    "        end_frame = int(col.split('_')[0]) + int(col.split('_')[1])\n",
    "        \n",
    "        # if the end frame is not in the dict, add it\n",
    "        if end_frame not in prob_index_dict.keys():\n",
    "            prob_index_dict[end_frame] = [col]\n",
    "        else:\n",
    "            prob_index_dict[end_frame].append(col)\n",
    "\n",
    "\n",
    "    agg_predictions = pd.DataFrame()\n",
    "    # loop through each key in the dict and get the mean of the values\n",
    "    for end_frame in prob_index_dict.keys():\n",
    "        # calculate the row-wise mean of the values in these columns and add to the dataframe\n",
    "        agg_prob = predictions.loc[:, prob_index_dict[end_frame]].mean(axis=1)\n",
    "        \n",
    "        # add the mean as a column to the dataframe\n",
    "        agg_predictions[end_frame] = agg_prob\n",
    "\n",
    "    # sort columns in ascending order\n",
    "    agg_predictions = agg_predictions.reindex(\n",
    "        sorted(agg_predictions.columns),\n",
    "        axis=1\n",
    "    )\n",
    "\n",
    "    # append game_winner to the dataframe\n",
    "    agg_predictions['winner'] = pd.DataFrame().from_dict(labels, orient='index')\n",
    "\n",
    "    # append game length to the dataframe\n",
    "    agg_predictions['length'] = metadata.set_index('filehash')['game_length']/5\n",
    "\n",
    "    return agg_predictions\n",
    "\n",
    "agg_predictions = aggregate_predictions(prediction_results, spawningtool_df)\n",
    "\n",
    "agg_predictions.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.7633699 , 0.7372874 , 0.37481397, 0.28510046, 0.61181474,\n",
       "       0.825375  , 0.31476364, 0.31972808, 0.22428766, 0.3435039 ,\n",
       "       0.37225485, 0.6057095 , 0.6381684 , 0.42742038, 0.45988026,\n",
       "       0.36916852, 0.49774107, 0.34848297, 0.6333327 , 0.793322  ,\n",
       "       0.7194455 , 0.75525856, 0.821038  , 0.6036366 , 0.4731819 ,\n",
       "       0.4419209 , 0.49901134, 0.33664095, 0.6101788 , 0.46319962,\n",
       "       0.18749762, 0.13039787, 0.1610338 , 0.11424755, 0.06664199,\n",
       "       0.06557623, 0.24670045, 0.25098208, 0.0790699 , 0.1621588 ,\n",
       "       0.2866229 , 0.29493645, 0.38711393, 0.17443529, 0.05348754,\n",
       "       0.1294706 , 0.17898266, 0.2165153 , 0.1295036 , 0.14634909,\n",
       "       0.02838423, 0.1397459 , 0.16743451, 0.12015149, 0.03842833,\n",
       "       0.01219733, 0.00991571, 0.03444169, 0.0140866 , 0.00928502,\n",
       "       0.01122712, 0.02539749, 0.03498638, 0.02168916, 0.03397271,\n",
       "       0.01436423, 0.05187571, 0.03283773, 0.13080224, 0.0625976 ,\n",
       "       0.03246713, 0.03219239, 0.05170704], dtype=float32)"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agg_predictions.drop(columns=['winner', 'length']).loc[game_hash].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index([ 48,  49,  50,  51,  52,  53,  54,  55,  56,  57,  58,  59,  60,  61,\n",
       "        62,  63,  64,  65,  66,  67,  68,  69,  70,  71,  72,  73,  74,  75,\n",
       "        76,  77,  78,  79,  80,  81,  82,  83,  84,  85,  86,  87,  88,  89,\n",
       "        90,  91,  92,  93,  94,  95,  96,  97,  98,  99, 100, 101, 102, 103,\n",
       "       104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117,\n",
       "       118, 119, 120],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agg_predictions.drop(columns=['winner', 'length']).loc[game_hash].index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2dde24972fa14f75942f12473e4a57cc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Predicting:   0%|          | 0/481 [00:00<?, ?model/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1296x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# function to load clean_events, generate predictions and return the predictions\n",
    "def predict_game(game_hash, model_dict):\n",
    "    \n",
    "    # data directory\n",
    "    data_dir = 'data/model_data/'\n",
    "\n",
    "    # locate the filename containing game_hash in data_dir\n",
    "    for root, dirs, files in os.walk(data_dir):\n",
    "        for file in files:\n",
    "            if game_hash in file:\n",
    "                filename = root+'/'+file\n",
    "                break\n",
    "\n",
    "    # load the data\n",
    "    data = np.load(filename)\n",
    "\n",
    "    new_predictions = {}\n",
    "    # loop through each model and get the prediction\n",
    "    for key, model in tqdm(\n",
    "        model_dict.items(), \n",
    "        desc='Predicting',\n",
    "        unit='model'\n",
    "    ):\n",
    "\n",
    "        # create a window from the data\n",
    "        window_start = int(key.split('_')[0])\n",
    "        window_size = int(key.split('_')[1])\n",
    "        windowed_data = data[window_start:window_start+window_size]        \n",
    "\n",
    "        # reshape the windowed data\n",
    "        windowed_data = windowed_data.reshape(1, window_size, windowed_data.shape[1])\n",
    "\n",
    "        # add the prediction to the prediction_results\n",
    "        new_predictions[key] = model.predict(windowed_data)[0]\n",
    "        \n",
    "        # convert the dict to a df with index of game_hash\n",
    "        results_df = pd.DataFrame.from_dict(new_predictions, orient='columns')\n",
    "        results_df.index = [game_hash]\n",
    "\n",
    "        # aggregate_predictions\n",
    "        agg_results = aggregate_predictions(results_df, spawningtool_df)\n",
    "\n",
    "    return agg_results\n",
    "\n",
    "game_hash = '010f527ac4796137aeb19378d3d0901067d96553409022395eba9a6745e3c595'\n",
    "results = predict_game(game_hash, model_dict)\n",
    "\n",
    "# get the values for the game_hash\n",
    "y_values = results.drop(columns=['winner', 'length']).loc[game_hash].values\n",
    "# subtract 0.5 from each value and then multiply by 200\n",
    "# this gives the probability of player 1 winning as 100% and player 2 winning as -100%\n",
    "y_values = (y_values - 0.5) * 200\n",
    "\n",
    "# get the x values from the index\n",
    "x_values = results.drop(columns=['winner', 'length']).loc[game_hash].index\n",
    "\n",
    "# convert x values to minutes\n",
    "x_values = x_values * 5 / 60\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import savgol_filter\n",
    "from scipy.signal import savgol_filter\n",
    "\n",
    "# smooth the y values\n",
    "smoothed_y_values = savgol_filter(y_values, 5, 3)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot\n",
    "fig, ax = plt.subplots(figsize=(12,4))\n",
    "plt.plot(x_values, smoothed_y_values, color='orange', linewidth=3)\n",
    "plt.axhline(y=0, color='white', linestyle='--')\n",
    "\n",
    "# make x ticks every 1 minute\n",
    "plt.xticks(np.arange(0, 11, step=1))\n",
    "\n",
    "start_point_x = 4 + 54/60\n",
    "# interpolate the y values at the start point\n",
    "start_point_y = np.interp(start_point_x, x_values, smoothed_y_values)\n",
    "\n",
    "end_point_x = 5 + 24/60\n",
    "# interpolate the y values at the end point\n",
    "end_point_y = np.interp(end_point_x, x_values, smoothed_y_values)\n",
    "\n",
    "# plot shaded area between start and end points of x\n",
    "plt.axvspan(start_point_x, end_point_x, alpha=0.5, color='gray')\n",
    "\n",
    "plt.ylim(-100, 100)\n",
    "plt.xlim(4, 10)\n",
    "plt.xlabel('Game Time (minutes)')\n",
    "plt.ylabel('Player 1 Win Probability (%)')\n",
    "# set all background to black\n",
    "plt.gca().set_facecolor('black')\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAABDCAYAAACBQ/NEAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8/fFQqAAAACXBIWXMAAAsTAAALEwEAmpwYAAAS/UlEQVR4nO3debgU1ZnH8W+zCQgqCkgU8aoZI4xjEjM4TqLORYO4b5NgNDGjiVuMJs+MGaNi4jVG0RmjqIk6biHGKDqaKOCwimgmbhFcxjVuF4yyurApIHDmj/fUU6frVi/30re7b9fv8zz1dHXVqerT1ef2fev0W6dyzjlERERERMR0q3UFRERERETqiQJkEREREZGAAmQRERERkYACZBERERGRgAJkEREREZGAAmQRERERkUA5AXIL4BLTRmAZMAXYP1G+1ZdprVAda+ksYCr2XqP3/kxNayQiIiJSnhayGcM1AVcCTwDvAeuABcCDwD7l7KCjPcjdgIHAEcBc4NgO7qfenQ4cjr1XERERka4uCzHcvsB5/vEzQC9gGHAUFjQfU2oH7Q2QLwFywNbATcE+ftHO/dSL3iXWPwCcAxzXzv0mz9Y0adKkSZMmTZo6c9pEcVmL4R4HvgZsAwwB7vbLuwE/K7XzjvYgrwTGBc93oXgv6xeA3wNv+G0/BRb7ZX/vy2wPrMU+5GmJ7Y8kbgAXBMsPAWYAHwDrsZ8Erk+pS6vfthX7OeFx4BPiBlJIC/BL4NkS5fKtWQO5nCZNmjRp0qSpE6Y6qELdTT6uKUcWYriHgK8A9wMrgCXAD4L1uxfZFti8i/Tas+0eWBf+bkB/oAd2MI8FHgWGY5Wf5MsfjH1gkW/4xw3ARD9/LvYhHAwMAHoCOwNnA08Dg1PqMQiYCfwjpc88RERERBpRo8dwq1KWhdu8U2L7DgfIWwGXBs/fApYXKT8fGIPlgWzht/+eX9cXOMPPXxPU63Q/3wfLGQE7I1gE7ASM98umYwe1N/GHsAtwUUo9+gKPAbsC/YDLitRZREREpNFkNYa7PJgvlUHQ7gD5YqybewVwpl/msEToYhYDBwFzgI+wLvobg/Wf84/PYwnjAN/BziiOwA4EwK3+8RC/LppfgHXtR2cvYGclaU4B3gbWAK+XqLeIiIhII8hqDJfDUjdO8s8fIA7mC+pR5s6THPAh8CRwFfBIifL3YmcfhfQJ5icAzVj3+nHAWL/8XeK8lrSu96TtUpYtxYb7EBEREcmiLMVwPbG0jhP98weA4yl9QWOHR7HohlX+cEof2AHEB3YJ8LdAd2CvAuWnAG/6+X8DDvPzv8bG7gM7SJFxvk7JKe0DKDeBXURERKSRZC2G64MFxFFwfBs2qsX6cjauxp30NmBnK9H8SuwKxUsLlN+EdYWDDebc229/e1BmBnYVJcCPsC76vlh+zCgst+THFaj71r6uA4JlPfyygVgujoiIiEgj6qox3DbALOIA/XLgVOIgvaRqBMirgIf9/I7YlYNLgBFFtrkd+xAiD2M5J5GFxAncA7Bu+zVY1/scLGG8EqNUPIjdbWZ+sOzzftky4IQKvIaIiIhIPeqqMdwx2DBvkQtpO250U7EdVCNABvgWcA+W87ICuBPLASlkFdYVHrklpcx/YGcG04D3sTObxdgdUlqAOza30iIiIiIZl8kYLuecK12qNu7CemgXY0OAlJUzUifq9qCKiIhIQ/oY2LLWlfC6cgwHdHwUi870MJYEvr1/fjld78A6dEFgqAd2dihGxyOmY5FPxyOfjkc+HY98Oh75+pQu0ukaIYYD6jNA3g07sIuAm4mTvbuS+cS3XxR4Bh2PkI5HTMcin45HPh2PfDoe+XQ88j1T6wrQGDEcUJ8BclOtKyAiIiIi7dZU6wpUSrUu0hMRERER6RIUIHeOm2tdgTqj45FPxyOmY5FPxyOfjkc+HY98Oh75dDwqqJ5HsRARERERqTr1IIuIiIiIBBQgi4iIiIgEFCCX5wBgMvAuNsbxyYn1OezOL+9h4x/PxcYBDG2BDXeyHLul4mRgaCfVt9paaXsLRwc85NdPTFn3ZLUrWUUttH2/i4P15bSXRnIB8Gfs1qPLgCnAnokyE8lWG0lzFnY71rXAPGD/2lanKtQ28rWg745QK9n+36LYo4YUIJenH/Ai8EPSbwByHnAucA4wElgKzAL6B2UmAP+M3Vlmf2ArYCrQvbMqXUUjgc8E097YH/O9QZnZiTKHVbmO1fYa+e/374J15bSXRtIM3AB8GTgQG9h/NrBtolzW2kjoeOBabFD9LwKPY7dgHVbLSlVBM2obSfruiGX9f4tij1pyzmlq37TaOXdy8DznnFvknBsXLOvjnFvlnDvDP9/aObfeOffNoMxOzrlNzrkxdfCeKj2Nc8595Jzr659PdM5NrYN6VWtqcc69WGBdOe2l0ad+zrmNzrkjg2VZayPJ6Snn3C2JZa8758bXQd3UNqo3tTh9dxSbsvy/RbFHlSf1IG++XYAhwMxg2SfAY1ivCMCXgJ6JMu8ArwRlGkUO+C5wJ3Zf+Mh+2NntX4BbgMHVr1pV7Yr9LPY2MMk/h/LaS6Prj/169WFiedbaSKQX9h0xM7F8JtlpExG1DX13FKL/LfkUe3QyBcibb4h/XJJYviRYNwTYiOUAFSrTKEZjf7i3BsumA98GDsJ+DtoHmIPlRjWip7BcsUOB07DP+HFgO8prL43uWuA54IlgWdbaSGgg9nNnlttEJOttQ98dhel/Sz7FHp2sHm813VUlB5TOpSxLKqdMV3MadtHNc8GyScH8/2EXIC0ADgd+X7WaVc+0xPMngbeAfyG+gKQj7aURXI31+OyHfXFHstZG0mS1TUTUNvTdUYz+t6RT7NFJ1IO8+aIrjJNnY4OJz+wWYz1EA4uUaQSDgaOxn7mKeQ/4K/A3nV6j+rAaeAl7v+W0l0Z1DXahyIHYP/1istRGlmMBYRbbRERtI52+O4z+t7Sl2KOTKUDefG9jjXB0sKw3drXo4/75PODTRJmhwPCgTCM4BVhH/ll9moHAjsCiTq9RfegN7IG933LaSyO6FjgRC4BeLaN8ltrIeuw7YnRi+Wgau01E1DYK03eH0f+WthR7dDKlWJSnH/BZP98NG3rpC8AHwEJsGJVx2Jf7X4CLsDP/u/w2K4DbgP/ELiZ4H/s58QVsiJpGkANOxb7AVgXL+2HjNN6PfWk1AeOx4/CHqtaweq7CxnNdiJ2p/wTYEvgN9rPWBIq3l0bzK+Ak4Bjs4quox2O1n7LYRpKuBn4LPA38CTgT2AG4qZaVqgK1jXz67mgry/9bFHvUUq2H0egiU7NLN9GvzzkbnmeRc26tc+5R59yeiX30ds5d75x73zn3sXNuirPhVmr93io1jfLHZJ/E8j7OuRnOuaXOhptZ4I9bI7335DTJOfeef7/vOufud86NCNaX014aaSqkxWW3jaRNZznnWp1z65xz85xzB9RBndQ2qjvpu6PtlOX/Lc0u3URXfnto9Nij06acc8rTFhERERGJKAdZRERERCSgAFlEREREJKAAWUREREQkoABZRERERCSgAFlEREREJKAAWUREREQkoABZJJvmYDcecMAhKetzwDtBmV2xQfmj583tfL0mv31LB7attonE7zOaNmA3I7gH2KsTXzt6vbkV3u/cYN+lNAdlW0rso1DZZuLPu6mddS1Xf+x2uQ6701rkG8CLwBr/ODZl29v9+p1T1t3h9/k/laysiHQtCpBFsunOYP6ElPX7YbckBbsl6Vub+XpNwMV+at7MfdVCd+wub2OBp4B/qG116l4z8efd1EmvcT52t7mFxO15X+wuYp8CX/WPd5P/eX0ZOBn4ObAgZb+XYwHyoX4fIpJBCpBFsuk+YK2fPxbonVgfBs2/9Y8tWM9yjsr3cFZC8j1UwinY+x0MTA5eZ3yN6lMNc4k/55YKlq2kvsD3/fyvsUAY7JbVOexXgCf8Yze/HKAHcCPwGvCLAvt+Ffijn/9RxWosIl2KAmSRbFpJHPD1B44I1vUAvubn1wP3+vkW0lMscsB3gf8FVgDrgFbgd379ROCRoPzFpP8sfxQwG/jQv+5C4Dba9kDODbbfE5iJ/Vw+HXiAOCVix2Cb7sBiv+5V2m8ZcEnwfJ9gPkyLOAZ4ztf/fL++B/CvwHxfz7XAy8DPgC2LvOZo4BlffgFwXmL9QcBU7Fiv9q/5Dtab+tki+93Nb7ca+AC4GdgqWN9M+ueTJq1sK/YZRx4Jyowh/hxeTuxrz6Dcf5V43eOBrf38pGB5L//4aeIxWn4OliLzfex4FRLt82DS0zBEpMEpQBbJrkJpFl8FBvn5aVgQVUgOC6BvBb6CBVq9sKDixHbU5QLgQSzo2wboCewEfAcLLEcU2G4uFkj29c8n+MfuwKlBuQOB7f38be2oV6jU9+VewP3A57H6R/WYDFwNfNHXcwtgOPAT4DHSg+QRWA7sl3z5YcCVWFAdGQkcjh3rLf1rDgW+iaXFDCxQz8f8dlsCA4DTsGNfjf8H64Cb/Pxw4J+CdWEbvLXEfsb4x+Xkn/DM8Y9HYyd+R/vnDwM7YCc5dwXlCol6kHNYkCwiGaMAWSS7pmMBBljAFPXIpaVXFPJ14t7md7Fgoh/Wgxn1uJ4MjAq2uYT8n+V3Jg78PsKCpq2xHFGwIG5CgddfgPU89gXOxALm5/y6U7EAFezCLbAexd+UeE9pBmEBbeSplDIDsF7zHbAgf6J/3UP9+mex3tshwAy/bG/ghwVe76fYcTgY+MQv/zHxycss4AAs8O8JbAtcFmz/rQLv5WksZWQPLNUArCf4iALl26uJ/N72UeSn5tyABcpgn1nkeP/4PPDnEq8x0j++kFg+FbjUv+ZK7H1dgp1sXIP1Tp8blO9JupexXyFA+eYimaQAWSS7PsVGZQDrpYxykY/xyz7CAo5ijgrmx2FB2xrgTcrPSR2DpSGAjSDwGBbctBAH8KNIz+k9B3gJCyCjnsQJ/nEoFvT1xN4bWG/u0jLrBZbf6vw20XtdB1yUUnYFFvAt8vOt2IlH5FLsYsclWKAbOSxlX+8CV2DHYRbwB7+8FxYUR2VOwHqLo3SJccE+PlfgPZ2HpYwk83BHFyhfaUuxC+cAjsOC+ZHYyQOU7j0GO8mAuH2EfoqdpO2G9SK3YL+KjMU+t4+AX2Gf0VosIN8/sY9NxL+cDEFEMkcBski2hWkWJ2IBXZSPeh9xT18h2wfzyZzScg0K5hcG8xuxIBAsgN42ZdtnU5bdjQWhYAHrIVjvLpQXfKXZ5Pd5HzZSwhMpZV4DPk4sK/TewtETBqfsKxpiL23bgdh398PA97BAcIuUffRJWZbcV3K/1TLBP/bC0miiHv615LfJUgoNW7ceOxlZ71/jl1hbuQE7iTgLO/E4Cfu140Fgu8Q+cu2oh4g0GAXIItn2JPCGnz8Q+EGwrlR6BcSBKFhOaSHFxt9dFswPC+a7E19ot4H0XOhPUpatx0YqAEtP+Hc/vxC7oK89olEsomHevk6cwlFOXQq9t3A+rUd7KPkBWlh+OZbvHOVlv4SlNXQjv0e/kEL1SOuN7ahS4y0/TzwSyunEYxXfh/XwlhK1u0FFS5nzgN2xk4mNxCkvP8fykWdjJ1D7Btt0Iz6pCtu4iGSEAmQRiXrsuhP/fL+Q+EKlYiYH85dhF9n1xQK2MGf3/WB+OPGoAmC50FG+50nYGMxbYT+VR72ac4iHpSvHjVjvdzfin89vx3qCqylMURkH7IL1ul8RLH8oZbuhWGDXH0t9iFJE1mMpKBuCsuuwFIth2MWOpVyBHdfdyc/HnVXGtuUKP++9SP9fM8E/7ko85na5PfxRjnKpm7Y0ARcCtxDnjUdtIDnSxcZguxHEaT+l8qFFpAEpQBaRtJ+0f0d5d137b6zXDyzImY3lIL9N/ogLbxD3UI7FgrpouLiFWDAM1mv3Ryw/NFr2ITZMWnssxXoHI5uwALna7sFGAgEbkeItbJizqBdzPnBdynbLsB7OlVivd5QucaVf9yrwil+2N3ZsWykvTWJf4hzkKE95LqXzzdvjyWD+Wiz4TLanKcS/XgC8Djxa5v6jixwHYhcbFnId1h7DE4con/tsLLf9IKyXOEybCXOSK3niICJdhAJkEXmTtjm15eaBOizgPQ34ExbQRWMYh+PTrsVGKZhH2zxdsBtvHI31FK/Aekj/igW1e9Ox/OYJwfwMLK+32jZiaQ/nYjmwH2MnB69gF+0dgAVwSS9j+eDzfPl3sAv7ovGFN/j9TgNWYQHydeSnyBRyANZrvQZLZ7gVO/aV7F1/xtflTeIe2qRNwPXB8/bkh9+DtTWI85eTjgKOxHriw/Scq7C75R2B/QLyAnah5IqgTLTPOdh7EJGMyTlXTieRiEiXMwZL3wALACcXKSu1MR67ocpabLi/9owwEm27ELtQcUPx4mXbAztByWEXeM4oXlxEGpF6kEWk0ZyN/XQf5fbOw37Ol/pxB9YrHt1t8AbaFxyDBchLsdzrkypXNS7EguPpKDgWyawepYuIiHQpA7EexdXYBW1nUF4+tVTPMCxnfTl2J8bzixdPtZL8YQYr5dt+EpEMU4qFiIiIiEhAKRYiIiIiIgEFyCIiIiIiAQXIIiIiIiIBBcgiIiIiIgEFyCIiIiIiAQXIIiIiIiKB/wcvgD0Hgza9OAAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x72 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# create array with values every second between start and end points\n",
    "x_values_array = np.arange(start_point_x, end_point_x + 1/60, step=1/60)\n",
    "\n",
    "params = {\"ytick.color\" : \"w\",\n",
    "          \"xtick.color\" : \"w\",\n",
    "          \"axes.labelcolor\" : \"w\",\n",
    "          \"axes.edgecolor\" : \"w\"}\n",
    "plt.rcParams.update(params)\n",
    "# loop through each value in x_values_array and interpolate the y values\n",
    "for x_val in x_values_array:\n",
    "    y_val = np.interp(x_val, x_values, smoothed_y_values)\n",
    "    \n",
    "    # plot stacked barh\n",
    "    # player 1 starts at 100% and player 2 starts where player 1 ends\n",
    "    fig, ax = plt.subplots(figsize=(10, 1))\n",
    "\n",
    "    plt.barh(\n",
    "        y=0,\n",
    "        width=100+y_val,\n",
    "        height=1,\n",
    "        color='red',\n",
    "        label='Player 1',\n",
    "        left=-100\n",
    "    )\n",
    "    ax.barh(\n",
    "        y=0,\n",
    "        width=200 - y_val,\n",
    "        height=1,\n",
    "        color='blue',\n",
    "        label='Player 2',\n",
    "        left=y_val\n",
    "    )\n",
    "    plt.xlim(-100, 100)\n",
    "    plt.grid(False)\n",
    "    # plt.grid(axis='x', color='white', linestyle='--')\n",
    "\n",
    "    # make x ticks absolute values of themselves\n",
    "    new_xticks = np.arange(100, 0, -25)\n",
    "    new_xticks = np.concatenate((new_xticks, np.arange(0, 101, 25)))\n",
    "    # convert x ticks to strings\n",
    "    new_labels = [str(x) for x in new_xticks]\n",
    "    # set x ticks and labels\n",
    "    plt.xticks(np.arange(-100, 101, 25), new_xticks)\n",
    "\n",
    "    # set y label and rotate it\n",
    "    h = plt.ylabel('Player 1')\n",
    "    h.set_rotation(0)\n",
    "    h.set_horizontalalignment('right')\n",
    "    h.set_verticalalignment('center')\n",
    "    # remove the y ticks\n",
    "    ax.set_yticks([])\n",
    "\n",
    "    # plot secondary y label\n",
    "    ax2 = ax.twinx()\n",
    "    h = ax2.set_ylabel('Player 2')\n",
    "    h.set_rotation(0)\n",
    "    h.set_horizontalalignment('left')\n",
    "    h.set_verticalalignment('center')\n",
    "    ax2.set_yticks([])\n",
    "\n",
    "    ax.set_xlabel('Victory Probability (%)')\n",
    "\n",
    "    # make background transparent\n",
    "    fig.patch.set_visible(False)\n",
    "\n",
    "    plt.tight_layout()\n",
    "    # save as png\n",
    "    save_path = 'C:/Users/jared/Desktop/predictions/'+str(x_val)+'.png'\n",
    "    plt.savefig(save_path, transparent=True)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([100,  75,  50,  25,   0,  25,  50,  75])"
      ]
     },
     "execution_count": 175,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_xticks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "# save data to csv\n",
    "agg_predictions.to_csv('data/new_agg_predictions.csv')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "prediction_results.to_csv('data/ensemble_prediction_results.csv')"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "e547aa29629eb5c25aeeb45870ab448aed1803ed873d9ecef66cb983aa54d8de"
  },
  "kernelspec": {
   "display_name": "Python 3.9.7 64-bit ('tflow2': conda)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}