implementation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Practical Deep Neural Network Performance Prediction for Hyperparameter Optimization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.10.0\n"
     ]
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "from concurrent import futures\n",
    "from functools import reduce, wraps\n",
    "from IPython.display import display\n",
    "import json\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.ticker as ticker\n",
    "import numpy as np\n",
    "import os\n",
    "import pandas as pd\n",
    "from sklearn.utils import shuffle\n",
    "import sys\n",
    "import tensorflow as tf\n",
    "os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'\n",
    "tf.logging.set_verbosity(tf.logging.WARN)\n",
    "print(tf.__version__)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "N_hidden = 16\n",
    "model_dir = 'model'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def model(n_hidden):\n",
    "    def model_loss(y, t):\n",
    "        t = tf.reshape(t, [-1])\n",
    "        mse = tf.reduce_mean(tf.square(y - t))\n",
    "        return mse\n",
    "\n",
    "    def training(loss):\n",
    "        optimizer = tf.train.AdamOptimizer()\n",
    "        train_step = optimizer.minimize(loss)\n",
    "        return train_step\n",
    "\n",
    "    x = tf.placeholder(tf.float32, shape=[None, None, 1])\n",
    "    t = tf.placeholder(tf.float32, shape=[None, 1])\n",
    "    n_batch = tf.placeholder(tf.int32, shape=[])\n",
    "    sequence_length = tf.placeholder(tf.int32, shape=[None])\n",
    "    output_keep_prob = tf.placeholder_with_default(1.0, shape=())\n",
    "    cell = tf.contrib.rnn.DropoutWrapper(\n",
    "        tf.nn.rnn_cell.LSTMCell(n_hidden),\n",
    "        output_keep_prob=output_keep_prob,\n",
    "        input_size=x.shape[-1],\n",
    "        variational_recurrent=True,\n",
    "        dtype=tf.float32)\n",
    "    zero_state = cell.zero_state(n_batch, tf.float32)\n",
    "    c_state = tf.placeholder(tf.float32, shape=[None, n_hidden])\n",
    "    h_state = tf.placeholder(tf.float32, shape=[None, n_hidden])\n",
    "    outputs, state = tf.nn.dynamic_rnn(\n",
    "        cell, x, initial_state=tf.nn.rnn_cell.LSTMStateTuple(c_state, h_state),\n",
    "        sequence_length=sequence_length, dtype=tf.float32)\n",
    "    h = tf.transpose(state.h)\n",
    "    W = tf.Variable(tf.truncated_normal([1, n_hidden], stddev=0.01))\n",
    "    b = tf.Variable(tf.zeros([1], dtype=tf.float32))\n",
    "    y = tf.sigmoid(tf.matmul(W, h) + b)\n",
    "    y = tf.reshape(y, [n_batch])\n",
    "    loss = model_loss(y, t)\n",
    "    train_step = training(loss)\n",
    "    init = tf.global_variables_initializer()\n",
    "    return x, t, n_batch, sequence_length, output_keep_prob, y, \\\n",
    "           c_state, h_state, zero_state, state, loss, train_step, init"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create model\n",
    "(x, t, n_batch, s_len,\n",
    " output_keep_prob, y, c_state,\n",
    " h_state, zero_state, lstm_state,\n",
    " loss, train_step, init) = model(n_hidden=N_hidden)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataname = 'mnist'\n",
    "batch_size = 16\n",
    "epochs = 1000\n",
    "output_keep_rate = 0.5\n",
    "N_runs = 100\n",
    "N_validation = 50\n",
    "N_train = N_runs - N_validation\n",
    "N_ensembles = 8"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "class EarlyStopping():\n",
    "    def __init__(self, sess, saver,\n",
    "                 fname, patience=30, verbose=0):\n",
    "        self._saver = saver\n",
    "        self._sess = sess\n",
    "        self._fname = fname\n",
    "        self.patience = patience\n",
    "        self.verbose = verbose\n",
    "\n",
    "        self._loss = float('inf')\n",
    "        self._step = 0\n",
    "\n",
    "    def validate(self, loss):\n",
    "        if self._loss <= loss:\n",
    "            self._step += 1\n",
    "            if self._step > self.patience:\n",
    "                if self.verbose:\n",
    "                    print('early stopping')\n",
    "                return True\n",
    "        else:\n",
    "            self._step = 0\n",
    "            self._loss = loss\n",
    "            self._saver.save(self._sess, self._fname)\n",
    "        return False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def prepare_data(df):\n",
    "    inputs = []\n",
    "    outputs = []\n",
    "    sequence_lengths = []\n",
    "    for i in range(1, df.shape[1]):\n",
    "        inputs.append(df.iloc[:, :i])\n",
    "        tmp = df.iloc[:, i:i + 1]\n",
    "        tmp.columns = [0]\n",
    "        outputs.append(tmp)\n",
    "        sequence_lengths.extend([i] * df.shape[0])\n",
    "    inputs = reduce(pd.DataFrame.append, inputs)\n",
    "    outputs = reduce(pd.DataFrame.append, outputs)\n",
    "    inputs.fillna(0, inplace=True)\n",
    "    outputs.fillna(0, inplace=True)\n",
    "    inputs.reset_index(inplace=True, drop=True)\n",
    "    outputs.reset_index(inplace=True, drop=True)\n",
    "    sequence_lengths = np.reshape(sequence_lengths, -1)\n",
    "    X = np.array(inputs).reshape(len(inputs), -1, 1)\n",
    "    Y = np.array(outputs).reshape(len(outputs), -1)\n",
    "    return X, Y, sequence_lengths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>11</th>\n",
       "      <th>12</th>\n",
       "      <th>13</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "      <th>16</th>\n",
       "      <th>17</th>\n",
       "      <th>18</th>\n",
       "      <th>19</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.963475</td>\n",
       "      <td>0.970143</td>\n",
       "      <td>0.972731</td>\n",
       "      <td>0.975119</td>\n",
       "      <td>0.977508</td>\n",
       "      <td>0.980096</td>\n",
       "      <td>0.981588</td>\n",
       "      <td>0.982584</td>\n",
       "      <td>0.982982</td>\n",
       "      <td>0.982982</td>\n",
       "      <td>0.983877</td>\n",
       "      <td>0.983877</td>\n",
       "      <td>0.983877</td>\n",
       "      <td>0.983877</td>\n",
       "      <td>0.983877</td>\n",
       "      <td>0.983877</td>\n",
       "      <td>0.983877</td>\n",
       "      <td>0.983877</td>\n",
       "      <td>0.983877</td>\n",
       "      <td>0.983877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.968451</td>\n",
       "      <td>0.968451</td>\n",
       "      <td>0.972930</td>\n",
       "      <td>0.975020</td>\n",
       "      <td>0.975020</td>\n",
       "      <td>0.977607</td>\n",
       "      <td>0.978105</td>\n",
       "      <td>0.978105</td>\n",
       "      <td>0.978901</td>\n",
       "      <td>0.979896</td>\n",
       "      <td>0.979896</td>\n",
       "      <td>0.979896</td>\n",
       "      <td>0.979896</td>\n",
       "      <td>0.980693</td>\n",
       "      <td>0.980693</td>\n",
       "      <td>0.980792</td>\n",
       "      <td>0.983977</td>\n",
       "      <td>0.983977</td>\n",
       "      <td>0.984674</td>\n",
       "      <td>0.986465</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.971935</td>\n",
       "      <td>0.971935</td>\n",
       "      <td>0.976015</td>\n",
       "      <td>0.978802</td>\n",
       "      <td>0.978802</td>\n",
       "      <td>0.980892</td>\n",
       "      <td>0.981887</td>\n",
       "      <td>0.985669</td>\n",
       "      <td>0.985669</td>\n",
       "      <td>0.985868</td>\n",
       "      <td>0.985868</td>\n",
       "      <td>0.985868</td>\n",
       "      <td>0.985868</td>\n",
       "      <td>0.985868</td>\n",
       "      <td>0.985868</td>\n",
       "      <td>0.985868</td>\n",
       "      <td>0.985868</td>\n",
       "      <td>0.985868</td>\n",
       "      <td>0.985868</td>\n",
       "      <td>0.985868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.970442</td>\n",
       "      <td>0.970442</td>\n",
       "      <td>0.975717</td>\n",
       "      <td>0.977807</td>\n",
       "      <td>0.978802</td>\n",
       "      <td>0.980792</td>\n",
       "      <td>0.980792</td>\n",
       "      <td>0.980892</td>\n",
       "      <td>0.980892</td>\n",
       "      <td>0.981887</td>\n",
       "      <td>0.984574</td>\n",
       "      <td>0.984674</td>\n",
       "      <td>0.984674</td>\n",
       "      <td>0.984674</td>\n",
       "      <td>0.984674</td>\n",
       "      <td>0.984674</td>\n",
       "      <td>0.984674</td>\n",
       "      <td>0.984674</td>\n",
       "      <td>0.984674</td>\n",
       "      <td>0.984773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.964570</td>\n",
       "      <td>0.972432</td>\n",
       "      <td>0.975617</td>\n",
       "      <td>0.978105</td>\n",
       "      <td>0.979100</td>\n",
       "      <td>0.981290</td>\n",
       "      <td>0.981290</td>\n",
       "      <td>0.981290</td>\n",
       "      <td>0.982683</td>\n",
       "      <td>0.982783</td>\n",
       "      <td>0.982783</td>\n",
       "      <td>0.982783</td>\n",
       "      <td>0.982783</td>\n",
       "      <td>0.982882</td>\n",
       "      <td>0.982882</td>\n",
       "      <td>0.983181</td>\n",
       "      <td>0.983181</td>\n",
       "      <td>0.983678</td>\n",
       "      <td>0.983678</td>\n",
       "      <td>0.983678</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         0         1         2         3         4         5         6   \\\n",
       "0  0.963475  0.970143  0.972731  0.975119  0.977508  0.980096  0.981588   \n",
       "1  0.968451  0.968451  0.972930  0.975020  0.975020  0.977607  0.978105   \n",
       "2  0.971935  0.971935  0.976015  0.978802  0.978802  0.980892  0.981887   \n",
       "3  0.970442  0.970442  0.975717  0.977807  0.978802  0.980792  0.980792   \n",
       "4  0.964570  0.972432  0.975617  0.978105  0.979100  0.981290  0.981290   \n",
       "\n",
       "         7         8         9         10        11        12        13  \\\n",
       "0  0.982584  0.982982  0.982982  0.983877  0.983877  0.983877  0.983877   \n",
       "1  0.978105  0.978901  0.979896  0.979896  0.979896  0.979896  0.980693   \n",
       "2  0.985669  0.985669  0.985868  0.985868  0.985868  0.985868  0.985868   \n",
       "3  0.980892  0.980892  0.981887  0.984574  0.984674  0.984674  0.984674   \n",
       "4  0.981290  0.982683  0.982783  0.982783  0.982783  0.982783  0.982882   \n",
       "\n",
       "         14        15        16        17        18        19  \n",
       "0  0.983877  0.983877  0.983877  0.983877  0.983877  0.983877  \n",
       "1  0.980693  0.980792  0.983977  0.983977  0.984674  0.986465  \n",
       "2  0.985868  0.985868  0.985868  0.985868  0.985868  0.985868  \n",
       "3  0.984674  0.984674  0.984674  0.984674  0.984674  0.984773  \n",
       "4  0.982882  0.983181  0.983181  0.983678  0.983678  0.983678  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Train data\n",
    "df = pd.read_json('%s.json' % (dataname), orient='split')\n",
    "display(df.head())\n",
    "df.head().T.plot(title='Previous learning curves')\n",
    "dlen = df.shape[1]\n",
    "tmp = df.copy()\n",
    "for i in range(1, df.shape[1]):\n",
    "    tmp.iloc[:, i] = (df.iloc[:, i] - df.iloc[:, i - 1]) / (1 - df.iloc[:, i - 1])\n",
    "    tmp.fillna(0, inplace=True)\n",
    "df = tmp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ensemble: 7\tepoch: 242\tvalidation loss:0.0029982498"
     ]
    }
   ],
   "source": [
    "# Training\n",
    "with tf.Session() as sess:\n",
    "    for e in list(range(N_ensembles)):\n",
    "        shuffled_idx = np.arange(N_runs)\n",
    "        np.random.shuffle(shuffled_idx)\n",
    "        sess.run(init)\n",
    "        saver = tf.train.Saver()\n",
    "        early_stopping = EarlyStopping(\n",
    "                sess, saver, \"%s/%s/%d\" % (model_dir, dataname, e))\n",
    "        df_t = df.iloc[shuffled_idx][:N_train]\n",
    "        tmp = np.array(df_t).reshape(-1)\n",
    "        X_train, Y_train, SL_train = prepare_data(df_t)\n",
    "        df_v = df.iloc[shuffled_idx][N_train:]\n",
    "        X_validation, Y_validation, SL_validation = prepare_data(df_v)\n",
    "        for epoch in range(epochs):\n",
    "            X_, Y_, SL_ = shuffle(X_train, Y_train, SL_train)\n",
    "            N_batches = X_.shape[0] // batch_size\n",
    "            for i in range(N_batches):\n",
    "                z = sess.run(zero_state, feed_dict={n_batch: batch_size})\n",
    "                start = i * batch_size\n",
    "                end = start + batch_size\n",
    "                sess.run([train_step, loss], feed_dict={\n",
    "                    x: X_[start:end],\n",
    "                    t: Y_[start:end],\n",
    "                    s_len: SL_[start:end],\n",
    "                    n_batch: batch_size,\n",
    "                    output_keep_prob: output_keep_rate,\n",
    "                    c_state: z[0],\n",
    "                    h_state: z[1]\n",
    "                })\n",
    "            z = sess.run(zero_state, feed_dict={n_batch: len(X_validation)})\n",
    "            val_loss = loss.eval(session=sess, feed_dict={\n",
    "                x: X_validation,\n",
    "                t: Y_validation,\n",
    "                s_len: SL_validation,\n",
    "                n_batch: len(X_validation),\n",
    "                c_state: z[0],\n",
    "                h_state: z[1]\n",
    "            })\n",
    "            print('\\rensemble: %s\\tepoch: %s\\tvalidation loss:%s' % (\n",
    "                e, epoch, val_loss), end='')\n",
    "            if early_stopping.validate(val_loss):\n",
    "                break"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataname = 'mnist'\n",
    "validation_dataname = 'mnist_test'\n",
    "ylim = [0.95, 1.0]  # [0, 1]\n",
    "dlen = 20  # 300\n",
    "inputlen = 1\n",
    "plot_ticks = np.array(range(dlen))\n",
    "N_test_cases = 1  # 20\n",
    "N_sigma = 2\n",
    "N_ensembles = 8"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "models = list(range(N_ensembles))\n",
    "max_workers = len(models)\n",
    "saver = tf.train.Saver()\n",
    "config = tf.ConfigProto(device_count={\"GPU\": 0})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Predictor():\n",
    "    def __init__(self,\n",
    "                 dataname,\n",
    "                 validation_dataname,\n",
    "                 N_sigma,\n",
    "                 dlen,\n",
    "                 inputlen,\n",
    "                 input_df,\n",
    "                 modelpath):\n",
    "        self.dataname = dataname\n",
    "        self.validation_dataname = validation_dataname\n",
    "        self.N_sigma = N_sigma\n",
    "        self.dlen = dlen\n",
    "        self.inputlen = inputlen\n",
    "        self.input_df = input_df\n",
    "        self.modelpath = modelpath\n",
    "\n",
    "    def __call__(self):\n",
    "        with tf.Session(config=config) as sess:\n",
    "            sess.run(init)\n",
    "            saver.restore(sess, self.modelpath)\n",
    "            sess.graph.finalize()\n",
    "            predicted = self.input_df.values.tolist()\n",
    "            z = sess.run(zero_state, feed_dict={n_batch: 1})\n",
    "            y_, z = sess.run([y, lstm_state], feed_dict={\n",
    "                x: np.array(predicted).reshape(1, -1, 1),\n",
    "                n_batch: 1,\n",
    "                s_len: [len(predicted)],\n",
    "                c_state: z[0],\n",
    "                h_state: z[1]\n",
    "            })\n",
    "            predicted.append(y_.reshape(-1)[0])\n",
    "            for _ in range(self.dlen - len(predicted)):\n",
    "                y_, z = sess.run([y, lstm_state], feed_dict={\n",
    "                    x: np.array(predicted)[-1:].reshape(1, -1, 1),\n",
    "                    n_batch: 1,\n",
    "                    s_len: [1],\n",
    "                    c_state: z[0],\n",
    "                    h_state: z[1]\n",
    "                })\n",
    "                predicted.append(y_.reshape(-1)[0])\n",
    "        for i in range(1, len(predicted)):\n",
    "            predicted[i] = predicted[i - 1] + (1 - predicted[i - 1]) * predicted[i]\n",
    "        predicted = np.array(predicted)\n",
    "        return predicted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "class MultiPredictor():\n",
    "    def __init__(self,\n",
    "                 dataname,\n",
    "                 validation_dataname,\n",
    "                 N_sigma,\n",
    "                 dlen,\n",
    "                 inputlen,\n",
    "                 N_ensembles,\n",
    "                 max_workers):\n",
    "        self.dataname = dataname\n",
    "        self.validation_dataname = validation_dataname\n",
    "        self.N_sigma = N_sigma\n",
    "        self.dlen = dlen\n",
    "        self.inputlen = inputlen\n",
    "        self.N_ensembles = N_ensembles\n",
    "        self.max_workers = max_workers\n",
    "        self.models = ['%s/%s/%d' % (model_dir, self.dataname, e) for e in models]\n",
    "        self.executor = futures.ProcessPoolExecutor(max_workers=self.max_workers)\n",
    "\n",
    "    def predict(self, input_df):\n",
    "        predictions = []\n",
    "        fs = [self.executor.submit(\n",
    "            Predictor(self.dataname,\n",
    "                      self.validation_dataname,\n",
    "                      self.N_sigma,\n",
    "                      self.dlen,\n",
    "                      self.inputlen,\n",
    "                      input_df,\n",
    "                      m)) for m in self.models]\n",
    "        for future in futures.as_completed(fs):\n",
    "            predictions.append(future.result())\n",
    "        predictions = pd.DataFrame(predictions).iloc[:, input_df.shape[0]:]\n",
    "        return predictions\n",
    "\n",
    "    def __del__(self):\n",
    "        self.executor.shutdown()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot(mean, original):\n",
    "    plt.figure()\n",
    "    plt.xlabel('epoch')\n",
    "    plt.ylabel('accuracy')\n",
    "    plt.xticks(plot_ticks, plot_ticks + 1)\n",
    "    plt.ylim(ylim)\n",
    "    ax = plt.gca()\n",
    "    plt.plot(mean, color='red', label='Prediction')\n",
    "    original.T.plot(ax=ax, linestyle='dashed', color='gray', label='Ground truth')\n",
    "    plt.legend()\n",
    "    plt.grid()\n",
    "    plt.show()\n",
    "    plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 1\n",
      "predictive mean: 0.9828084261242099\n",
      "predictive std: 0.0007846972472526188\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 2\n",
      "predictive mean: 0.9845828571729797\n",
      "predictive std: 0.0005189400841878884\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEKCAYAAADjDHn2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzt3Xl8VeW59//PRRJIIMgokyAgRuYxyCCCDII4VCtqxdbxFH16KnY6pbXHHm3tY8Whgz1a69E6Ha2oiP7Uaq0iiNZHZDBIIGEUMKDIJBCSCEmu3x9rJeyMe++QTZB836/Xeu211n1fa917Z2dde033MndHRESkNk0augEiInLsU7IQEZGolCxERCQqJQsREYlKyUJERKJSshARkagSlizM7FEz+8LMsmsoNzP7k5mtN7OPzWxYRNk1ZrYuHK5JVBtFRCQ2idyzeByYWkv5uUBGONwAPAhgZm2B24CRwAjgNjNrk8B2iohIFAlLFu6+CNhdS5WLgCc98AHQ2sw6A+cAb7r7bnffA7xJ7UlHREQSLLkB130S8GnEdF44r6b5VZjZDQR7JaSlpWV269atzo0pLS2lSZO6507FK17xiv86xq9du3anu58YtaK7J2wAegDZNZS9CpwZMT0fGA78FPhlxPz/An4abV2ZmZl+JBYsWKB4xSte8Y0uHljqMWzPG/JqqK1A5K5A13BeTfNFRKSBNGSyeBm4OrwqahSw190/A94ApphZm/DE9pRwnoiINJCEnbMws2eA8UB7M8sjuMIpBcDd/wK8BpwHrAcKgOvCst1m9htgSbio2929thPlIiKSYAlLFu5+RZRyB26soexR4NFEtEtEjp5Dhw6Rl5dHUVFRrfVatWpFTk5Ondej+OjxqampdO3alZSUlDqtoyGvhhKR41xeXh4tW7akR48emFmN9fbv30/Lli3rvB7F1x7v7uzatYu8vDx69uxZp3Wouw8RSZiioiLatWtXa6KQxDMz2rVrF3UPrzZKFiKSUEoUx4Yj/TsoWYiISFRKFiJy3EpKSmLIkCEMGDCAyy67jIKCgjova+HChVxwwQUAvPzyy8yePbvGul9++SV//vOfy6e3bdvGpZdeWud1HwuULETkuJWWlkZWVhbZ2dk0bdqUv/zlLxXK3Z3S0tK4l3vhhRdy880311heOVl06dKFuXPnxr2eY4mShYg0CmPHjmX9+vVs2rSJ3r17c/XVVzNgwAA+/fRT5s+fz+jRoxk2bBiXXXYZ+fn5APzjH/+gT58+DBs2jHnz5pUv6/HHH2fmzJkAbN++nW9/+9sMHjyYwYMH8/7773PzzTezYcMGhgwZwqxZs9i0aRMDBgwAgpP+1113HQMHDmTo0KEsWLCgfJnTpk1j6tSpZGRk8LOf/ewof0K106WzInJ0/OhHkJVVbVFaSQkkJcW/zCFD4I9/jFqtuLiY119/nalTgw6s161bxxNPPMGoUaPYuXMn99xzD2+99RYtWrTgrrvu4ve//z0/+9nPuP7663n77bc59dRTufzyy6td9g9+8APGjBnDK6+8QklJCfn5+cyePZvs7Gyywve7adOm8voPPPAAZsbKlSvJzc1lypQpLFu2DICsrCw++ugjmjVrRu/evbnppps4kg5S65P2LETkuFVYWMiQIUMYPnw4J598Mt/97ncB6N69O6NGjQLggw8+IDc3lzFjxjBkyBCeeOIJNm/eTG5uLj179iQjIwMz48orr6x2HW+//TYzZswAgnMkrVq1qrVN7733Xvmy+vTpQ/fu3Vm/fj0AkyZNolWrVqSmptKvXz82b95cL59DfdCehYgcHbXsARQe4U1pNSk7Z1FZixYtysfdnQkTJlQ5p1BdXKI1a9asfDwpKYni4uKj3oaaaM9CRBq1UaNGsXjx4vJf9wcOHGDt2rX06dOHTZs2sWHDBgCeeeaZauMnTZrEI488AkBJSQl79+6lZcuW7N+/v9r6Y8eO5emnnwZg7dq1bNmyhYyMjPp+W/VOyUJEGrUTTzyRBx98kCuuuIJBgwYxevRocnNzSU1N5X/+5384//zzGTZsGB06dKg2/r777uPdd99l4MCBZGZmsnr1atq1a8eYMWMYMGAAs2bNqlD/+9//PqWlpQwcOJDLL7+cxx9/vMIexbFKh6FE5LhVdlVTpB49epCdnV1h3llnncWSJUuq1J06dSq5ublV5l977bVce+21AHTs2JE5c+ZUOYz2t7/9rcJ02TpTU1N57LHHKpTt37+/wjIBXn311ZrfWAPQnoWIiESlZCEiIlEpWYiISFRKFiIiEpWShYiIRKVkISIiUSlZiMhxrayjv1NOOYXMzExGjx7Niy++eNTb0aNHD3bu3Fll/m9/+9s6Le+ll15i9erV5dPjx49n6dKldW5fNEoWInLccne++c1vMm7cODZu3MiyZcuYM2cOeXl5Veo2VNcaNSWLaN2nV04WiaZkISLHrbfffpumTZvyve99r3xe9+7duemmm4CgW/ALL7yQCy64gEmTJuHuzJo1iwEDBjBw4ECeffZZoOKDjwBmzpzJ448/DgR7DHfccQfDhg1j4MCB5Tfx7dq1iylTptC/f39mzJiBu1dp380330xhYSFjxozhO9/5TrXdp6enp5fXnzt3Ltdeey3vv/8+L7/8MrNmzWLIkCFs3LgRgOeff54RI0Zw2mmn8e6779brZ6k7uEXkqCnbwEbq378/ffr04dChQ+V9JkUaMmQIQ4YMoaCggOeee65CWeQdz9VZtWoVw4YNq7XO8uXL+de//kX37t154YUXyMrKYsWKFezcuZPTTz+dcePGRX1f7dq1Y/ny5fz5z3/m3nvv5ZFHHuHXv/41Z555Jrfeeit///vf+etf/1olbvbs2dx///3861//omXLlmzatKlC9+k1OeOMM8qT3KWXXlreD1VxcTEffvghr732Gr/+9a956623orY9VtqzEJFG48Ybb2Tw4MGcfvrp5fMmT55M27ZtgaD78CuuuIKkpCQ6duxYYzcglV144YUAZGZmlj+7YtGiReVdkZ9//vm0adMmpjZGdp8er2nTplVpR33RnoWIHDU17Qns37+flJSUWvcUmjdvHnVPorL+/fvzwgsvlE8/8MAD7Ny5k+HDh5fPi+yuvCbJyckVzh8UFRVVKC/rCLA+uhWv3B4zq3G9ldVnOyrTnoWIHLcmTpxIUVERDz74YPm8goKCGuuPHTuWZ599lpKSEnbs2MGiRYsYMWIE3bt3Z/Xq1Xz11Vd8+eWXzJ8/P+q6x40bV96Z4Ouvv86ePXuqrZeSksKhQ4dqXE7Hjh3JycmhtLS0wlVctXWDnghKFiJy3DIzXnrpJd555x169uzJiBEjuOaaa7jrrruqrX/xxRczaNAgBg8ezMSJE7n77rvp1KkT3bp141vf+hYDBgzgW9/6FkOHDo267ttuu41FixbRv39/5s2bx8knn1xtvRtuuIHRo0fzne98p9ry2bNnc8EFF3DGGWfQuXPn8vnTp0/nnnvuYejQoeUnuBPK3Y+LITMz04/EggULFK94xddz/OrVq2OK37dv3xGtX/GxxVf39wCWegzbWO1ZiIhIVEoWIiISlZKFiCSUV3Mzmhx9R/p3ULIQkYRJTU1l165dShgNzN3ZtWsXqampdV6G7rMQkYTp2rUreXl57Nixo9Z6RUVFR7QhU3z0+NTUVLp27VrndShZiEjCpKSk0LNnz6j1Fi5cGNPlqIpPTHwsdBhKRESiSmiyMLOpZrbGzNab2c3VlHc3s/lm9rGZLTSzrhFld5lZdjhcnsh2iohI7RKWLMwsCXgAOBfoB1xhZv0qVbsXeNLdBwG3A3eGsecDw4AhwEjgp2Z2QqLaKiIitUvknsUIYL27b3T3g8Ac4KJKdfoBb4fjCyLK+wGL3L3Y3Q8AHwNTE9hWERGphSXqkjYzuxSY6u4zwumrgJHuPjOizt+Axe5+n5lNA14A2gOZwG3AZKA58CHwgLv/rtI6bgBuAOjYsWPmnDlz6tze/Pz8Cg8ZUbziFa/4xhA/YcKEZe4+PGrFWPoEqcsAXAo8EjF9FXB/pTpdgHnAR8B9QB7QOiy7BcgC3gSeBn5U2/rUN5TiFa94xcePY6BvqK1At4jpruG8cu6+zd2nufvQMDng7l+Gr3e4+xB3nwwYsDaBbRURkVokMlksATLMrKeZNQWmAy9HVjCz9mZW1oZfAI+G85PMrF04PggYBPwzgW0VEZFaJOymPHcvNrOZwBtAEvCou68ys9sJdnteBsYDd5qZA4uAG8PwFODd8AlR+4Ar3b1+H/skIiIxS+gd3O7+GvBapXm3RozPBeZWE1dEcEWUiIgcA3QHt4iIRKVkISIiUSlZiIhIVEoWIiISlZKFiIhEpWQhIiJRKVmIiEhUShYiIhKVHqsq0kiUlJRQUFBQPhw4cICCggJOP/10ANauXcuWLVuqxJ199tkA5OTksHVrhe7dSE5OZvz48QCsXLmS7du3Vyhv1qwZY8eOBSArK4udO3dWKG/evDlnnHEGAMuWLWPPnj0Vyk844QRGjBgBwOLFi9m/f3+F8jZt2pCZmQkEjxbdv38/JSUlFBcXU1JSQufOnRk3bhwATz31FPn5+eVlJSUl9O7dm/PPPx+Au+++m0OHDlVY/rBhwzj33HNxd377299W+WxGjRrFpEmTKC4u5o477qhSPm7cOMaOHcu+ffv47//+7yrlZ599NiNHjqSgoKDa+PPOO4+hQ4eybds2HnvssSrlF198Mf36HZ37l5UsRBrQV199RVJSEgCFhYXs2LGjSp0OHTqQmprKgQMH2LVrV5XyTp06AbB9+3Y2bNhQnggKCws5cOAA06dPp0WLFixatIhFixZViR84cCAAmzZt4sMPP6xSXpYsNm7cyEcffVShrGnTpuXJYv369axatapCeXp6enmyyM3NZf369RXK27ZtW54sVq1aVSVZde7cuTxZrFy5ks8//7xCeffu3cuTxfr169m7dy9JSUkkJyeTlJREq1atyus2b96c5OTk8rKkpCS6dOlSXp6ZmUlpaWmF5XftWv7wzvJ2ROrWLegr1cyqLT/ppJPKP6fqysv+dsnJydWWd+jQAYAWLVpUW962bdsq8xJFyUIkgdyd0tJSkpKS2Lt3b/mv5z179rB7924KCwu54oorANiyZQvVPZPlmmuuoUePHmzcuJF58+ZVKb/++usByMvL480336RJkyY0b96cFi1a0Lx5c0pKSgA47bTTSE9Pr1DWvHlzUlNTAZgyZQpTpkyp8b2cf/755b/Cq3PxxRdz8cUX11g+ffr0GssArr766lrLZ8yYcUTl06ZNq7V80qRJNZaZGZMnT66xPCkpqdb41NTUWuMjk251WrVqVWv80aBkIRKhsLCwyrykpCSaNm2Ku1NUVFSlPDk5+DcqKioiKyuL3bt38+WXX5a/Tp06leHDh1NUVMR7771Hq1ataNu2LX379qVt27a0a9eObdu20bVrV6688soqy+/YsSMAPXr0qLa8Xbt2QLCH0L9/f5o1a0bYCWcFJ510UvkvXZF4KVlIo/XZZ5+xbNkyWrRoUb5x/cMf/lDluHVmZiYXXHABEBzXrmz06NE0bdoUgDfeeIOmTZvStm1bOnToQO/evcs39h06dOCWW24pP+xUWYsWLejVq1eN7W3ZsiUtW7assbysDSKJoGQhjcqhQ4fIzs5m2bJlbN26leTkZDIzM8sPxZx99tlVjluXbewBzjnnnCrL7Ny5M5988gmpqanMmjWLtLS0an/Zm1mNiULkWKdkIY3KG2+8wbJly2jfvj3nnHMOgwcPJi0tjYULFwLVn8QsY2aMGjWq2rJPPvkECE6iihyPlCzkuFVcXExOTg5Lly5l6tSpdO7cmVGjRjFw4EBOPvnkan/9i0j1lCzkuLNr1y6WLVvGihUrKCgooE2bNhQUFADQvn172rdv38AtFKlBaSkcPAhffVX9aw1lbTZvhlqupqoPShZyXCkuLubhhx/m4MGD9OnTh8zMTE455RTtRchhpaXBhjZiSNu6FVatOrwRrmljHfkaMd5rwwaYOzd6TA2vY4uKoLg4GOqgZ58+MGtWPX9QFSlZyBHZuXMn2dnZrFu3jiFDhtC6dWs+/fRTVq5cWaXuuHHjSE9P55NPPiEnJ6dC2datWxk5ciRpaWmsX7+etWvXVomfPHkyKSkp5ObmsnHjxvL5JSUlbNy4kbPOOovk5GQuueQSOnXqVOuVQ9JASkqgsPDwUFAAhYWcsGoVuENRUbARLSo6PMQwPXDrVkhPr5IEqh2q2SCPPJL31LQpnZOSoHlzaNoUmjWr+Fo23qpVjWXbtm+n26mnVh9X22s4vio7m9FH8h5ioGQhcTt48CArV64kKyuLvLy88qt8CgsLad26Nbt37yY7O7tK3MiRI0lPT2fXrl1Vyg8dOkRx+E/8xRdfVBs/ceJEUlJS+Pzzz6uUp6Wl8dVXX5GamkpGRkY9vttGyB0OHID9+2Me+m7cCPfdVyUJVBk/eLDaVQ6LtW1Nm0JqarCxTE0tH085dAiSk4P56enBaxxDzsaN9B08uOLGuLoNdOXxlBQw472FC2u9qS6aDQsX0u0I4r+q5s7/+qZkITFxdwoLC2nevDkHDhzg1Vdf5cQTT2Ty5MkMGjSIpUuX0rlzZwAGDx7M4MGDa1zW8OHDGT58eIV5CxcuLN8TOOOMM8q7gKjO+PHjq/xjLly4sPzyV4lQWgq7d8P27fDFF1Vfw/ER27YFv/r374f8/CBhxKJFC2jZkhOaNIF27SAtLRhOPDF4bd788LyaxtPSWLFuHYNHjDicAConhNTUYCPdpPq+T5cf4cZ6+8KF9E3wMf+vOyULqdWePXvIyspixYoVdOrUienTp9OmTRu+//3v0759e50LaEiFhbBmDaxeTbd33oFXXqmaDHbsCJJAZUlJwQa9Y0fo0IH8tDSa9+oFLVtGH9LTD7+GG+/FR7ix3rNwIZx1Vp3jJfGULKRaubm5fPDBB2zevBmAXr16MWjQoPLyE088saGa1vgcOAC5ubB6dTCsWhW8btxYvgfQC4Jf+R06BAmgRw8YMaI8GVR5bdu2wq/01QsX0kG/rKUWShYCBIeZtmzZQpcuXUhOTmb79u3s37+fCRMmMHjw4Aq9d0qC5OdDTs7hZFA2bNp0+LBQSgqcdhoMGwZXXQX9+kG/fizasoVx557boM2X45uShbB06VKWLFnCokWLuOyyy+jXrx9jxoxh3LhxOsyUKHv2wJIlsHgxfPgho5YsCQ4dlWnaFHr3DvYOrruuPClw6qlBwqik9Cic4JTGTcmikduzZw9///vfadmyJd/85jc59dRTgcM9qUo9OHgQVqwoTwwsXgyRlwb37cve/v1Jvemmw0mhV6/g6h6RY4S+jY1c2f0Offv2rfUKJomRe3AuYfHiw8nho4+C6/shOGcwciRcc02w13D66dCqFTkLF9JR5wzkGKZk0cht27aNTp06kZaW1tBN+fpxh7w8WLmS7s8/D3ffHSSHsqfZpaXB8OEwc2aQIEaOhG7dQIf25GtIyaKRu+SSSygsLKz2cZoSYe9eyM6Gjz+GlSuDITsbvvwSgB5mweGjiy4KksKIETBggA4lyXFD3+RGzszUrXakgweDexfKEkLZEPls6BNOgIEDYfr04HXgQN7bt4+xtTxyVOTrTsmiEZs7dy5t27Zl4sSJDd2Uo+/QIdiwIbh/ISeHvvPnww9+EEyXPSkvORn69IExY+B73wsSw6BB1R5KKgmfhyFyvFKyaKQKCwvJyclh9OhEdz/WwPbuDfYUwqRAbm4wrF9foUO5Vh06BOcXzjsvSAgDBwaXrupRpSKAkkWjtXbtWkpLS+nbt29DN+XIudNsxw54662KCSEnBz777HC95GTIyIC+fWHatGCvoU8f6N2bD5YvP6LuKkSOd0oWjVRubi4tW7akS5cuDd2U+OTnByeWV6wITjaHr6P37z9cp1WrICGcc87hhNC3L/TsWe0NbSISXUKThZlNBe4DkoBH3H12pfLuwKPAicBu4Ep3zwvL7gbOB5oAbwI/dI+1K0ypzcGDB1m/fj1Dhw49du/Qdg+6uShLCGXJYcOGw11fnHACDB4MV1/N2pQUTrvooiAxdOyoy1NF6lnCkoWZJQEPAJOBPGCJmb3s7qsjqt0LPOnuT5jZROBO4CozOwMYA5T1XPcecBawMFHtbUwOHTpEZmYm/fv3b+imBA4c4ITVq4O7msuSwscfw759QblZ0M3FkCFw9dVBghg0CLp3L08K2xYu5DQdRhJJmETuWYwA1rv7RgAzmwNcBEQmi37AT8LxBcBL4bgDqUBTwIAUIKLjHDkSLVq0YOrUqQ3XAPfgfMLrr8Nrr8G77zKs7AqkE04IEsGVVwZJYfBg6N8/6A5bRBqMJerIjpldCkx19xnh9FXASHefGVHnb8Bid7/PzKYBLwDt3X2Xmd0LzCBIFve7+y3VrOMG4AaAjh07Zs6ZM6fO7c3Pzyf9CDZIX5f40tJS9u3bR6tWrSocgkr0+pMKC2m9fDltP/yQdosXkxp2mpffsye7R4xg+6mnUtK/P0WdOtXpENLX5fNXvOKPtfgJEyYsc/fhUSu6e0IG4FKC8xRl01cRbPQj63QB5gEfEZzbyANaA6cCfwfSw+H/AWNrW19mZqYfiQULFjSK+HXr1vmvfvUrX7NmTWLXX1rqvnq1++9+53722e5Nm7qDe3q6+0UXuT/0kPvmzYlbv+IVr/iYAEs9hm16TIehzGwe8FfgdXcvjTFhbQW6RUx3DedFJqptwLRwHenAJe7+pZldD3zg7vlh2evAaODdGNctNcjJyaFp06accsop9b/wAwfg7beDw0uvvx6coIagG4ybbgruYTjzTN27IPI1FOs5iz8D1wF/MrPngcfcfU2UmCVAhpn1JEgS04FvR1Yws/bA7jAB/YLgyiiALcD1ZnYnwWGos4A/xthWqUFpaSlr1qwhIyOj/rogLyiAxx9n0GOPBSelDx4Mntg2aRL8/Odw7rnBiWgR+VqLaYvh7m8Bb5lZK+CKcPxT4GHgKXc/VE1MsZnNBN4guHT2UXdfZWa3E+z2vAyMB+40MwcWATeG4XOBicBKgpPd/3D3V47gfQqQl5fHgQMH6NOnz5EvrKAA/vKXoKfV7dtpdvLJQe+q554LY8dCs2ZHvg4ROWbE/PPSzNoBVxKce/gIeBo4E7iGYKNfhbu/BrxWad6tEeNzCRJD5bgS4P/E2jaJzZo1a0hKSiIjI6PuCzlwAB58EO65B774ItiDeP55lpSU6A5okeNYrOcsXgR6A/8LfMPdy/pQeNbMliaqcVK/Jk6cyIABA2hWl1/9Bw7An/8cJIkdO2DyZLjttqCTPQB1pCdyXIt1z+JP7r6gugKP5ZIrOSYkJSXRuXPn+ILy8+GBB+Dee2HnTpgyJUgSZ5yRmEaKyDGpSYz1+plZ67IJM2tjZt9PUJskAT788EPmz59fdslydPv3w+zZ0KMH3Hxz0CPr++/DG28oUYg0QrEmi+vd/cuyCXffA1yfmCZJIixdupS8vLzofUHt3w933hl0uveLXwRPfPvgg+BS2OO9O3MRqVGsySLJIrYyYb9Pulj+a2Lnzp3s2LGj1qugkg4cgDvuCPYk/vM/YdQoWLw46I5j5Mij11gROSbFes7iHwQnsx8Kp/9POE++BnJzcwGqTxYFBfDHPzJq9uxgr+KCC+DWW+H0049yK0XkWBZrsvg5QYL493D6TeCRhLRI6l1OTg5dunShVatWh2e6w3PPwaxZ8Omn7D3jDNr/6U+QmdlwDRWRY1asN+WVAg+Gg3yNlJaW0rZtW7p27Xp45rJl8KMfwXvvwdCh8NRTZJeWMl6JQkRqEOt9FhkEz5roR9B1OADunoAOhqQ+NWnShEsuuSSY2L4dbrkFHn0U2reHhx+G666DpCTdJyEitYr1BPdjBHsVxcAE4EngqUQ1SurPvn374KuvgpvpMjLgySfhP/4D1q2DGTOCRCEiEkWsySLN3ecTPP9is7v/iuCRp3IMO5Cfzx//8Ac+uOQS+NnPYPx4WLUqSByR5y9ERKKI9QT3V2bWBFgXdg64leA5E3KsWrWKtXfeiWdk0H33bvjHP+Cccxq6VSLyNRXrnsUPgebAD4BMgg4Fr0lUo+QI7N4dPDti8GBykpJo1aQJnRYuVKIQkSMSdc8ivAHvcnf/KZBP8FwLOcZYSQncf39wj8TevXz1/e+zsUMHTh8xAtPDhkTkCEVNFu5eYmZnHo3GSB29+SbDb7gheDLdxInwxz+yzoySF16gb9++Dd06ETkOxHrO4iMzexl4HjhQNtPd5yWkVRKb99+HX/4SFiygSZcu8OKLcNFFYEavwkKmTZtW8f4KEZE6ijVZpAK7CJ5eV8YBJYuGsGwZ/Nd/BZ37dewI993Hh336cNaUKeVV0tLSGDhwYAM2UkSOJ7Hewa3zFMeC7OzgWRLz5kHbtnDXXXDjjdCiBR5xU92nn37K1q1byczMJCUlpeHaKyLHjVjv4H6MYE+iAnf/t3pvkVS1bh386lfwzDPQsmUw/uMfwwknVFt9+fLl5OTkcLo6AxSRehLrYahXI8ZTgYuBbfXfHKlg0yb4zW/giSegWTP4+c/hpz+Fdu1qDCktLWXNmjWcdtppJOnubBGpJ7EehnohctrMngHeS0iLBLZtC54t8fDD0KRJcN/EzTcH5yei2Lx5M4WFhboKSkTqVax7FpVlAB3qsyEC7NgRnId44AEoLobvfjfo+K9bt5gXkZubS3JyMr169UpgQ0WksYn1nMV+Kp6z+JzgGRdSH/bsoedf/xpc+lpYCFddFdxcd0r8nfoWFBSQkZFBU92IJyL1KNbDUC0T3ZBGa8sWGDSI7nv3wuWXByeva3n8aTSXXHIJpaWl9dc+ERFi37O4GHjb3feG062B8e7+UiIb1yh06wYzZ7KkRw9OnzHjiBZVUlJCUlISTZrE2uWXiEhsYt2q3FaWKADc/UvgtsQ0qZExg//7fzlw6qlHtBh356GHHuKtt96qp4aJiBwWa7Korl5dT45LAhQUFLBjxw5at27d0E0RkeNQrMliqZn93sx6hcPvgWXEu/NgAAATF0lEQVSJbJjEZ+fOnQD07t27gVsiIsejWJPFTcBB4FlgDlAE3JioRkn8du7cSbdu3WjZUtciiEj9i/VqqAPAzQlui9TRnj17yM/PZ/To0Q3dFBE5TsW0Z2Fmb4ZXQJVNtzGzNxLXLIlHs2bNOOWUU+jXr19DN0VEjlOxnqRuH14BBYC77zEz3cFdjbVr11a5z6F169Z06tQJCO6wrqxt27ZAcOnrunXrqpS3b9+e9u3bc+jQITZs2FClvEOHDnTr1k0nt0UkYWJNFqVmdrK7bwEwsx5U0wttY7R7927ef/99kpODj/KFF17g4MGDFeoMGzaMb3zjGwA8++yzVZYxatQomjVrRnFxcbXlZ511FuPHj6eoqKja8smTJ9fHWxERqVGsyeIW4D0zewcwYCxwQ8Ja9TWSlZXF8uXLGTlyJADXXXcd7hXzaPPmzcvHb7ih6sfWokULli9fTkpKSrXl6enp5cuprrxly5YsXbr0iN6HiEhtYj3B/Q8zG06QID4CXgIKE9mwr4PS0lJWrFhBr169aNasGUD54aaadO7cucayJk2a1FqelJRUa7mISKLEeoJ7BjAf+A/gp8D/Ar+KIW6qma0xs/VmVuVqKjPrbmbzzexjM1toZl3D+RPMLCtiKDKzb8bzxo6GTz75hH379jFkyJCGboqISELFep/FD4HTgc3uPgEYCnxZW4CZJQEPAOcC/YArzKzy5Tr3Ak+6+yDgduBOAHdf4O5D3H0IwXO/C4B/xtjWoyYrK4vU1FTdCCcix71Yk0WRuxcBmFkzd88Fom0hRwDr3X2jux8kuJnvokp1+gFvh+MLqikHuBR43d0LYmzrUZOens7w4cPLT26LiByvrPLJ2Gormb0IXAf8iOCX/h4gxd3PqyXmUmCqu88Ip68CRrr7zIg6fwMWu/t9ZjYNeIHgMt1dEXXeBn7v7q9SiZndQHiivWPHjplz5syJ4S1XLz8/v/xEsuIVr3jFN5b4CRMmLHP34VEruntcA3AWcCHQNEq9S4FHIqavAu6vVKcLMI/gpPl9QB7QOqK8M7CDIDHV2q7MzEw/EgsWLIir/meffealpaV1jj/S9Ste8YpXfH3EA0s9hm1/3A8+cPd33P1lDw4t1WYrEPk80K7hvMhlbXP3ae4+lODy3LLuz8t8C3jR3Q/F285E2rFjBw899BDLly9v6KaIiBwViXxKzhIgw8x6mllTYDrwcmQFM2tvZmVt+AXwaKVlXAE8k8A21klWVhZNmjShzxE80U5E5OskYcnC3YuBmcAbQA7wnLuvMrPbzezCsNp4YI2ZrQU6AneUxYd3iXcD3klUG+uitLSUjz/+mIyMDFq0aNHQzREROSoSehmPu78GvFZp3q0R43OBuTXEbgJOSmT76mL9+vXk5+fr3goRaVT0sOY4rVq1iubNm5ORkdHQTREROWp0g0CcvvGNb7Bz506SkpIauikiIkeN9izilJycHLX/JxGR442SRRyee+45srKyGroZIiJHnZJFjD7//HNycnKqPKtCRKQxULKIUVZWFklJSQwYMKChmyIictQpWcSgpKSElStX0rt37woPMhIRaSyULGKwdu1aCgoKdG+FiDRaShYxSE9PZ9CgQfTq1auhmyIi0iB0n0UMunXrRrdu3aJXFBE5TmnPIopPP/2UPXv2NHQzREQalJJFLdydV199lXnz5jV0U0REGpSSRS0+++wzvvjiCwYPHtzQTRERaVBKFrX46KOPSE5O1r0VItLoKVnUoLi4mOzsbPr06UNqampDN0dEpEEpWdRg27ZtHDx4UPdWiIigS2drdPLJJ/OTn/yEtLS0hm6KiEiDU7KohrtjZnpsqohISIehqvGvf/2Lxx9/nOLi4oZuiojIMUHJohJ3JysrC3cnOVk7XiIioGRRRV5eHrt27dKJbRGRCEoWlWRlZZGSkkK/fv0auikiIscMJYsIhw4dIjs7m379+tGsWbOGbo6IyDFDB+UrmThxIl27dm3oZoiIHFOULCKkpKQwcuTIhm6GiMgxR4ehQkVFRSxfvpyDBw82dFNERI45Shahzz//nFdeeYWCgoKGboqIyDFHyYLg3ort27fTs2dPWrdu3dDNERE55ihZAJs3b6aoqEj3VoiI1EDJguDeiqSkJPr27dvQTREROSY1+mTh7uzdu5cTTzyRlJSUhm6OiMgxqdEnCzPjmmuuISMjo6GbIiJyzGr0yaJMkyb6KEREaqItpIiIRJXQZGFmU81sjZmtN7ObqynvbmbzzexjM1toZl0jyk42s3+aWY6ZrTazHolsq4iI1CxhycLMkoAHgHOBfsAVZla5K9d7gSfdfRBwO3BnRNmTwD3u3hcYAXyRqLaKiEjtErlnMQJY7+4b3f0gMAe4qFKdfsDb4fiCsvIwqSS7+5sA7p7v7rq1WkSkgZi7J2bBZpcCU919Rjh9FTDS3WdG1PkbsNjd7zOzacALQHtgLDADOAj0BN4Cbnb3kkrruAG4AaBjx46Zc+bMqXN78/PzSU9PV7ziFa/4RhU/YcKEZe4+PGpFd0/IAFwKPBIxfRVwf6U6XYB5wEfAfUAe0DqM3QucQtAz7gvAd2tbX2Zmph+JBQsWKF7xild8o4sHlnoM2/REHobaCnSLmO4azivn7tvcfZq7DwVuCed9GSaNLA8OYRUDLwHDEthWERGpRSKTxRIgw8x6mllTYDrwcmQFM2tvZmVt+AXwaERsazM7MZyeCKxOYFtFRKQWCUsW4R7BTOANIAd4zt1XmdntZnZhWG08sMbM1gIdgTvC2BLgp8B8M1sJGPBwotoqIiK1S+iT8tz9NeC1SvNujRifC8ytIfZNYFAi2yciIrHRHdwiIhKVkoWIiESlZCEiIlEpWYiISFRKFiIiEpWShYiIRKVkISIiUSlZiIhIVEoWIiISlZKFiIhEpWQhIiJRKVmIiEhUShYiIhKVkoWIiESlZCEiIlEpWYiISFRKFiIiEpWShYiIRKVkISIiUSlZiIhIVEoWIiISlZKFiIhEpWQhIiJRKVmIiEhUShYiIhKVkoWIiESlZCEiIlEpWYiISFRKFiIiEpWShYiIRKVkISIiUSlZiIhIVEoWIiISlZKFiIhEpWQhIiJRJTRZmNlUM1tjZuvN7OZqyrub2Xwz+9jMFppZ14iyEjPLCoeXE9lOERGpXXKiFmxmScADwGQgD1hiZi+7++qIavcCT7r7E2Y2EbgTuCosK3T3IYlqn4iIxC6RexYjgPXuvtHdDwJzgIsq1ekHvB2OL6imXEREjgEJ27MATgI+jZjOA0ZWqrMCmAbcB1wMtDSzdu6+C0g1s6VAMTDb3V+qvAIzuwG4IZzMN7M1R9De9sBOxSte8YpvZPHdY6rl7gkZgEuBRyKmrwLur1SnCzAP+IggYeQBrcOyk8LXU4BNQK9EtTVcz1LFK17xim+M8bEMidyz2Ap0i5juGs4r5+7bCPYsMLN04BJ3/zIs2xq+bjSzhcBQYEMC2ysiIjVI5DmLJUCGmfU0s6bAdKDCVU1m1t7MytrwC+DRcH4bM2tWVgcYA0SeGBcRkaMoYcnC3YuBmcAbQA7wnLuvMrPbzezCsNp4YI2ZrQU6AneE8/sCS81sBcGJ79le8SqqRPgfxSte8YpvpPFRWXi8S0REpEa6g1tERKJSshARkagafbIws0fN7Aszy65DbDczW2Bmq81slZn9MM74VDP70MxWhPG/jrcN4XKSzOwjM3u1DrGbzGxl2K3K0jrEtzazuWaWa2Y5ZjY6jtjeEV26ZJnZPjP7UZzr/3H42WWb2TNmlhpn/A/D2FWxrru674yZtTWzN81sXfjaJs74y8I2lJrZ8Dqs/57wb/Cxmb1oZq3jjP9NGJtlZv80sy7xxEeU/YeZeXhhSjzr/5WZbY34LpwX7/rN7KbwM1hlZnfHuf5nI9a9ycyy4owfYmYflP0fmdmIOOMHm9n/C/8XXzGzE2qIrXabE8/3r84SfW3usT4A44BhQHYdYjsDw8LxlsBaoF8c8Qakh+MpwGJgVB3a8RPgb8CrdYjdBLQ/gs/vCWBGON6U8D6ZOiwnCfgc6B5HzEnAJ0BaOP0ccG0c8QOAbKA5wQ2qbwGn1uU7A9wN3ByO3wzcFWd8X6A3sBAYXof1TwGSw/G76rD+EyLGfwD8JZ74cH43ggtaNtf2naph/b8Cfhrj3626+Anh369ZON0h3vZHlP8OuDXO9f8TODccPw9YGGf8EuCscPzfgN/UEFvtNiee719dh0a/Z+Hui4DddYz9zN2Xh+P7Ca76OimOeHf3/HAyJRziuuLAgs4XzwceiSeuPphZK4Iv/l8B3P2gh/fJ1MEkYIO7b44zLhlIM7Nkgo3+tjhi+wKL3b3Ag6v33iG876c2NXxnLiJInISv34wn3t1z3D2mHghqiP9n+B4APiC4ryme+H0Rky2o5XtYy//MH4Cf1RYbJT4mNcT/O8FVk1+Fdb6oy/rNzIBvAc/EGe9A2d5AK2r5HtYQfxqwKBx/E7ikhtiatjkxf//qqtEni/piZj0IbhxcHGdcUrjL+wXwprvHFQ/8keAftDTOuDIO/NPMllnQfUo8egI7gMfCw2CPmFmLOrZjOrX8g1bHgxs37wW2AJ8Be939n3EsIhsYa2btzKw5wS/CblFiatLR3T8Lxz8nuBS8ofwb8Hq8QWZ2h5l9CnwHuDXO2IuAre6+It71RpgZHgp7tA6HUU4j+FsuNrN3zOz0OrZhLLDd3dfFGfcj4J7w87uX4L6xeKzicN94lxHD97DSNifh3z8li3pgwd3nLwA/qvQLLSp3L/Ggd92uwAgzGxDHei8AvnD3ZXE1uKIz3X0YcC5wo5mNiyM2mWB3+kF3HwocINgFjosFN21eCDwfZ1wbgn+wngRdx7QwsytjjXf3HIJDNv8E/gFkASXxtKGG5Tpx7iHWFzO7haA/tafjjXX3W9y9Wxg7M451Ngf+kzgTTCUPAr2AIQSJ/3dxxicDbYFRwCzguXAvIV5XEOePltC/Az8OP78fE+5tx+HfgO+b2TKCw0sHa6tc2zYnUd8/JYsjZGYpBH+0p919Xl2XEx6+WQBMjSNsDHChmW0i6NV3opk9Fed6y7pV+QJ4kaC34FjlAXkRe0NzCZJHvM4Flrv79jjjzgY+cfcd7n6IoJ+xM+JZgLv/1d0z3X0csIfgGHBdbDezzgDha42HQRLFzK4FLgC+E24w6uppajgMUoNeBAl7Rfhd7AosN7NOsS7A3beHP5xKgYeJ73sIwXdxXnho90OCPe0aT7JXJzyUOQ14Ns51A1xD8P2D4EdPXO1391x3n+LumQTJqsaujWrY5iT8+6dkcQTCXy5/BXLc/fd1iD+x7KoVM0sjePZHbqzx7v4Ld+/q7j0IDuO87e4x/7I2sxZm1rJsnOAkacxXhbn758CnZtY7nDWJunXLUtdfc1uAUWbWPPxbTCI4hhszM+sQvp5MsKH4Wx3aAUFXNteE49cA/18dl1MnZjaV4HDkhe5eUIf4jIjJi4jve7jS3Tu4e4/wu5hHcBL28zjW3zli8mLi+B6GXiI4yY2ZnUZwsUW8vbCeDeS6e16ccRCcozgrHJ8IxHUYK+J72AT4JfCXGurVtM1J/Pevvs+Yf90Ggo3UZ8Ahgi/5d+OIPZNgd+9jgkMYWcB5ccQPIuhx92OCf44ar8CIYVnjifNqKIIefVeEwyrgljqsdwiwNHwPLwFt4oxvAewCWtXxff+aYMOWDfwv4dUwccS/S5DgVgCT6vqdAdoB8wk2Em8BbeOMvzgc/wrYDrwRZ/x6gkcClH0Pa7uaqbr4F8LP8GPgFcJen+vyP0OUK+xqWP//AivD9b8MdI4zvinwVPgelgMT420/8DjwvTr+/c8EloXfo8VAZpzxPyTYq10LzCbsXaOa2Gq3OfF8/+o6qLsPERGJSoehREQkKiULERGJSslCRESiUrIQEZGolCxERCQqJQuRY4CZjbc69BoscrQoWYiISFRKFiJxMLMrLXgGSZaZPRR2BJlvZn8Iny8w38xODOuWPeOg7BkTbcL5p5rZWxY8x2S5mfUKF59uh58N8nQd+zYSSQglC5EYmVlf4HJgjAedP5YQ9NDaAljq7v0Jujm/LQx5Evi5uw8iuDu5bP7TwAPuPpigL6uy3kKHEvRe2o/g7voxCX9TIjFKbugGiHyNTAIygSXhj/40gg7bSjnc+dxTwLzwWR+t3f2dcP4TwPNhX1wnufuLAO5eBBAu70MP+yUKu63vAbyX+LclEp2ShUjsDHjC3Ss8q8DM/qtSvbr2ofNVxHgJ+v+UY4gOQ4nEbj5waUQPoW3NrDvB/9GlYZ1vA++5+15gj5mNDedfBbzjwdPN8szsm+EymoXPgxA5pumXi0iM3H21mf2S4MmCTQh6Db2R4KFPI8KyLwjOa0DQVfRfwmSwEbgunH8V8JCZ3R4u47Kj+DZE6kS9zoocITPLd/f0hm6HSCLpMJSIiESlPQsREYlKexYiIhKVkoWIiESlZCEiIlEpWYiISFRKFiIiEtX/DzChInLT8D03AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 3\n",
      "predictive mean: 0.9838942958645751\n",
      "predictive std: 0.0004802426372715161\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 4\n",
      "predictive mean: 0.9832994417619269\n",
      "predictive std: 0.0005039129118147263\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 5\n",
      "predictive mean: 0.9862716747832435\n",
      "predictive std: 0.0010705891687547956\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 6\n",
      "predictive mean: 0.9866320667622309\n",
      "predictive std: 0.0010411638343490463\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 7\n",
      "predictive mean: 0.9863178105595968\n",
      "predictive std: 0.0009177139146090559\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 8\n",
      "predictive mean: 0.9860348232150422\n",
      "predictive std: 0.0007843878014747506\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 9\n",
      "predictive mean: 0.9869562162422549\n",
      "predictive std: 0.0006433944708444138\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 10\n",
      "predictive mean: 0.9867520915091901\n",
      "predictive std: 0.0005704487950481779\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 11\n",
      "predictive mean: 0.9870073881027711\n",
      "predictive std: 0.0004910045015181523\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 12\n",
      "predictive mean: 0.9868619034099901\n",
      "predictive std: 0.00043518316936959394\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 13\n",
      "predictive mean: 0.9870285796231452\n",
      "predictive std: 0.0003671023946043446\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 14\n",
      "predictive mean: 0.9869363556049985\n",
      "predictive std: 0.00031001065315402164\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 15\n",
      "predictive mean: 0.9868525205619156\n",
      "predictive std: 0.0002540351281448753\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 16\n",
      "predictive mean: 0.9867734250151068\n",
      "predictive std: 0.0002000601524658402\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 17\n",
      "predictive mean: 0.9867929765207396\n",
      "predictive std: 0.00014677457319661475\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEKCAYAAADjDHn2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzt3Xl8VfWd//HXhyQQQlJW2QQBMbLIEgirCLKIxWpREStWrTpVZ6aly3S0Y3/OqHUetm7T1vnVWltr1akWLaI/tVoXFnEZF8AgSwJEBAkoEkAkJBGSfH5/nJP0Zr33hlwukPfz8biP3HO+3885n5vc3M892/eYuyMiItKUNslOQEREjn4qFiIiEpWKhYiIRKViISIiUalYiIhIVCoWIiISVcKKhZk9ZGafmdnaRtrNzP7bzArN7AMzGx3RdqWZbQofVyYqRxERiU0ityweBmY10X4OkB0+rgPuBzCzLsAtwHhgHHCLmXVOYJ4iIhJFwoqFuy8H9jTR5XzgUQ+8DXQys17AV4FX3H2Pu+8FXqHpoiMiIgmWmsR1nwhsi5guCuc1Nr8eM7uOYKuE9u3b5/bt27fZyVRVVdGmTfNrp+IVr3jFH4vxGzduLHb3E6J2dPeEPYD+wNpG2p4HzoiYXgyMAa4H/j1i/n8A10dbV25urh+OpUuXKl7xild8q4sHVngMn+fJPBtqOxC5KdAnnNfYfBERSZJkFotngW+FZ0VNAPa5+yfAS8DZZtY5PLB9djhPRESSJGHHLMzsz8BUoJuZFRGc4ZQG4O6/BV4AvgYUAqXA1WHbHjP7T+C9cFG3uXtTB8pFRCTBElYs3P3SKO0OfLeRtoeAhxKRl4gcOYcOHaKoqIjy8vIm+3Xs2JH8/Pxmr0fx0ePT09Pp06cPaWlpzVpHMs+GEpHjXFFREVlZWfTv3x8za7Tf/v37ycrKavZ6FN90vLuze/duioqKGDBgQLPWoeE+RCRhysvL6dq1a5OFQhLPzOjatWvULbymqFiISEKpUBwdDvfvoGIhIiJRqViIyHErJSWFnJwchg0bxsUXX0xpaWmzl7Vs2TLOO+88AJ599lnuuOOORvt+/vnn/OY3v6mZ3rFjB3Pnzm32uo8GKhYictxq3749eXl5rF27lrZt2/Lb3/62Vru7U1VVFfdyZ8+ezY033thoe91i0bt3bxYuXBj3eo4mKhYi0ipMnjyZwsJCtmzZwqBBg/jWt77FsGHD2LZtG4sXL2bixImMHj2aiy++mJKSEgD+9re/MXjwYEaPHs2iRYtqlvXwww8zf/58AHbu3Mk3v/lNRo4cyciRI3nrrbe48cYb+fDDD8nJyeGGG25gy5YtDBs2DAgO+l999dUMHz6cUaNGsXTp0pplzpkzh1mzZpGdnc2Pf/zjI/wbappOnRWRI+OHP4S8vAab2ldWQkpK/MvMyYFf/Spqt4qKCl588UVmzQoGsN60aROPPPIIEyZMoLi4mLvvvptXX32VDh06cOedd/KLX/yCH//4x1x77bUsWbKEU045hUsuuaTBZX//+99n0qRJPPfcc1RWVlJSUsIdd9zB2rVryQtf75YtW2r633fffZgZa9asoaCggLPPPpuVK1cCkJeXx/vvv0+7du0YNGgQ3/ve9zicAVJbkrYsROS4VVZWRk5ODmPGjOGkk07i29/+NgD9+vVjwoQJALz99tsUFBQwadIkcnJyeOSRR9i6dSsFBQUMGDCA7OxszIzLL7+8wXUsWbKEa665BgiOkXTs2LHJnN54442aZQ0ePJh+/fpRWFgIwIwZM+jYsSPp6ekMHTqUrVu3tsjvoSVoy0JEjowmtgDKDvOitMZUH7Ooq0OHDjXP3Z1p06bVO6bQUFyitWvXruZ5SkoKFRUVRzyHxmjLQkRatQkTJvDOO+/UfLs/cOAAGzduZPDgwWzZsoUPP/wQgD//+c8Nxs+YMYMHH3wQgMrKSvbt20dWVhb79+9vsP/kyZN57LHHANi4cSMff/wx2dnZLf2yWpyKhYi0aieccAL3338/l156KSNGjGDixIkUFBSQnp7O7373O84991xGjx5N9+7dG4y/9957ef311xk+fDi5ubmsX7+erl27MmnSJIYNG8YNN9xQq/93vvMdqqqqGD58OJdccgkPP/xwrS2Ko5V2Q4nIcav6rKZI/fv3Z+3atbXmnXnmmbz33nv1+s6aNYuCgoJ686+66iquuuoqAHr06MGCBQvq7UZ7/PHHa01XrzM9PZ0//vGPtdr2799fa5kAzz//fOMvLAm0ZSEiIlGpWIiISFQqFiIiEpWKhYiIRKViISIiUalYiIhIVCoWInJcqx7o7+STTyY3N5eJEyfy9NNPH/E8+vfvT3Fxcb35P/vZz5q1vGeeeYb169fXTE+dOpUVK1Y0O79oVCxE5Ljl7lxwwQVMmTKFzZs3s3LlShYsWEBRUVG9vskaWqOxYhFt+PS6xSLRVCxE5Li1ZMkS2rZtyz/90z/VzOvXrx/f+973gGBY8NmzZ3PeeecxY8YM3J0bbriBYcOGMXz4cJ544gmg9o2PAObPn8/DDz8MBFsMt99+O6NHj2b48OE1F/Ht3r2bs88+m9NOO41rrrkGd6+X34033khZWRmTJk3isssua3D49MzMzJr+Cxcu5KqrruKtt97i2Wef5YYbbiAnJ4fNmzcD8Je//IVx48Zx6qmn8vrrr7fo71JXcIvIEVP9ARvptNNOY/DgwRw6dKhmzKRIOTk55OTkUFpaypNPPlmrLfKK54asW7eO0aNHN9ln1apVvPnmm/Tr14+nnnqKvLw8Vq9eTXFxMWPHjmXKlClRX1fXrl1ZtWoVv/nNb7jnnnt48MEH+elPf8oZZ5zBzTffzF//+lf+8Ic/1Iu74447+PWvf82bb75JVlYWW7ZsqTV8emNOP/30miI3d+7cmnGoKioqePfdd3nhhRf46U9/yquvvho191hpy0JEWo3vfve7jBw5krFjx9bMmzlzJl26dAGC4cMvvfRSUlJS6NGjR6PDgNQ1e/ZsAHJzc2vuXbF8+fKaocjPPfdcOnfuHFOOkcOnx2vOnDn18mgp2rIQkSOmsS2B/fv3k5aW1uSWQkZGRtQtibpOO+00nnrqqZrp++67j+LiYsaMGVMzL3K48sakpqbWOn5QXl5eq716IMCWGFa8bj5m1uh662rJPOrSloWIHLemT59OeXk5999/f8280tLSRvtPnjyZJ554gsrKSnbt2sXy5csZN24c/fr1Y/369Xz55Zd8/vnnLF68OOq6p0yZUjOY4IsvvsjevXsb7JeWlsahQ4caXU6PHj3Iz8+nqqqq1llcTQ2DnggqFiJy3DIznnnmGV577TUGDBjAuHHjuPLKK7nzzjsb7H/hhRcyYsQIRo4cyfTp07nrrrvo2bMnffv25Rvf+AbDhg3jG9/4BqNGjYq67ltuuYXly5dz2mmnsWjRIk466aQG+1133XVMnDiRyy67rMH2O+64g/POO4/TTz+dXr161cyfN28ed999N6NGjao5wJ1Q7n5cPHJzc/1wLF26VPGKV3wLx69fvz6m+C+++OKw1q/42OIb+nsAKzyGz1htWYiISFQqFiIiEpWKhYgklDdwMZoceYf7d1CxEJGESU9PZ/fu3SoYSebu7N69m/T09GYvQ9dZiEjC9OnTh6KiInbt2tVkv/Ly8sP6IFN89Pj09HT69OnT7HWoWIhIwqSlpTFgwICo/ZYtWxbT6aiKT0x8LLQbSkREokposTCzWWa2wcwKzezGBtr7mdliM/vAzJaZWZ+ItjvNbG34uCSReYqISNMSVizMLAW4DzgHGApcamZD63S7B3jU3UcAtwE/D2PPBUYDOcB44Hoz+0qichURkaYlcstiHFDo7pvd/SCwADi/Tp+hwJLw+dKI9qHAcnevcPcDwAfArATmKiIiTbBEndJmZnOBWe5+TTh9BTDe3edH9HkceMfd7zWzOcBTQDcgF7gFmAlkAO8C97n7f9VZx3XAdQA9evTIXbBgQbPzLSkpqXWTEcUrXvGKbw3x06ZNW+nuY6J2jGVMkOY8gLnAgxHTVwC/rtOnN7AIeB+4FygCOoVtNwF5wCvAY8APm1qfxoZSvOIVr/j4cRSMDbUd6Bsx3SecV8Pdd7j7HHcfFRYH3P3z8Oft7p7j7jMBAzYmMFcREWlCIovFe0C2mQ0ws7bAPODZyA5m1s3MqnP4CfBQOD/FzLqGz0cAI4CXE5iriIg0IWEX5bl7hZnNB14CUoCH3H2dmd1GsNnzLDAV+LmZObAc+G4Ynga8Ht4h6gvgcndv2ds+iYhIzBJ6Bbe7vwC8UGfezRHPFwILG4grJzgjSkREjgK6gltERKJSsRARkahULEREJCoVCxERiUrFQkREolKxEBGRqFQsREQkKt0pT0TkcPzv/9L1jTeguBgOHYKDB4Of1Y/I6UaeD9m2De6/v+F+DU3XmTdy4EB4//2EvkwVC5FWorKyktLS0prHgQMHKC0tZezYsQBs3LiRjz/+uF7cWWedBUB+fj7bt9ca3o3U1FSmTp0KwJo1a9i5c2et9nbt2jF58mQA8vLyKC4urtWekZHB6aefDsDKlSvZu3dvrfavfOUrjBs3DoB33nmH/fv312rv3Lkzubm5QHBr0f3791NZWUlFRQWVlZX06tWLKVOmAPCnP/2JkpKSmrbKykoGDRrEueeeC8Bdd93FoUOHai1/9OjRnHPOObg7P/vZz+r9biZMmMCMW28l+733uP366+u1T1m+nMmvv84XWVn83+9/P5iZlhY8MjI46+23Gb9hA4eysrh9woS/BwajV/C1wkJG7dnDjo4d+ePgwbUXbsaF+/cz1J3iigo611t7y1KxEEmiL7/8kpSUFADKysrYtWtXvT7du3cnPT2dAwcOsHv37nrtPXv2BGDnzp18+OGHNYWgrKyMAwcOMG/ePDp06MDy5ctZvnx5vfjhw4cDsGXLFt5999167dXFYvPmzbxf59tr27Zta4pFYWEh69atq9WemZlZUywKCgooLCys1d6lS5eaYrFu3bp6xapXr141xWLNmjV8+umntdr79etXUywKCwvZt28fKSkppKamkpKSQseOHWv6ZmRkkJqaWtOWkpJC7969a9pzc3Opqqqqtfw+fWpu3lmTR6S+ffvC737H6tdfZ1xWFrRpEzxSUqBNG06cOxcGDqRtVRXj3nmnpghU63nttdCvH+tffplxddoAuv/jP8KJJ9Jh3z7GNfC36TJ8OPTsyfZly8iu19qyVCxEEsjdqaqqIiUlhX379tV8e967dy979uyhrKyMSy+9FICPP/6Yhu7JcuWVV9K/f382b97MokWL6rVfe+21ABQVFfHKK6/Qpk0bMjIy6NChAxkZGVRWVgJw6qmnkpmZWastIyOD9PR0AM4++2zOPvvsRl/LueeeW/MtvCEXXnghF154YaPt8+bNa7QN4Fvf+laT7ddcc81htc+ZM6fJ9hkzZjTaZmbMnDmz0faynTuZGRbNhqQDM5v43UYW3YZ07NixyfUfCSoWIhHKysrqzUtJSaFt27a4O+Xl5fXaU1ODf6Py8nLy8vLYs2cPn3/+ec3PWbNmMWbMGMrLy3njjTfo2LEjXbp0YciQIXTp0oWuXbuyY8cO+vTpw+WXX15v+T169ACgf//+DbZ37doVCLYQTjvtNNq1a4c18C31xBNP5MQTT4zvFyISUrGQVuuTTz5h5cqVdOjQoebD9Ze//GW9/da5ubmcd955QLBfu66JEyfStm1bAF566SXatm1Lly5d6N69O4MGDar5sO/evTs33XRTzW6nujp06MDAgQMbzTcrK4usrKxG26tzEEkEFQtpVQ4dOsTatWtZuXIl27dvJzU1ldzc3JpdMWeddVa9/dbVH/YAX/3qV+sts1evXnz00Uekp6dzww030L59+wa/2ZtZo4VC5GinYiGtyksvvcTKlSvp1q0bX/3qVxk5ciTt27dn2bJlQMMHMauZGRMiz1iJ8NFHHwHBQVSR45GKhRy3KioqyM/PZ8WKFcyaNYtevXoxYcIEhg8fzkknndTgt38RaZiKhRx3du/ezcqVK1m9ejWlpaV07tyZ0tJSALp160a3bt2SnKHIsUfFQo4rFRUV/P73v+fgwYMMHjyY3NxcTj75ZG1FiBwmFQs5LMXFxaxdu5ZNmzaRk5NDp06d2LZtG2vWrKnXd8qUKWRmZvLRRx+Rn59fq2379u2MHz+e9u3bU1hYyMaNG+vFz5w5k7S0NAoKCti8eXPN/MrKSjZv3syZZ55JamoqF110ET179mzyzCERiY+KhcTt4MGDrFmzhry8PIqKimrO8ikrK6NTp07s2bOHtWvX1osbP348mZmZ7N69u177oUOHqKioAOCzzz5rMH769OmkpaXx6aef1mtv3749X375Jenp6WRnJ/paVpHWR8VCYuLulJWVkZGRwYEDB3j++ec54YQTmDlzJiNGjGDFihX06tULgJEjRzJy5MhGlzVmzBjGjBlTa96yZctqtgROP/30miEgGjJ16tR6V7suW7as5vRXEWl5KhbSpL1795KXl8fq1avp2bMn8+bNo3PnznznO9+hW7duOhYg0kqoWEiDCgoKePvtt9m6dSsAAwcOZMSIETXtJ5xwQrJSE5EkULEQINjN9PHHH9O7d29SU1PZuXMn+/fvZ9q0aYwcObLW6J0i0vqoWAgrVqzgvffeY/ny5Vx88cUMHTqUSZMmMWXKFO1mEhFAt1Vt9fbu3ctf//pXUlNTueCCCzjllFOAYCRVFQoRqaYti1au+nqHIUOGNHkGk4i0btqyaOV27NhBz549ad++fbJTEZGjmLYsWrmLLrqIsrKyBm+nKSJSTVsWrZyZaVhtEYlKxaIVW7hwIUuWLEl2GiJyDFCxaKXKysrIz8+vd1c4EZGGqFi0Uhs3bqSqqoohQ4YkOxUROQaoWLRSBQUFZGVl0bt372SnIiLHgIQWCzObZWYbzKzQzG5soL2fmS02sw/MbJmZ9Ylou8vM1plZvpn9t+kKsRZz8OBBCgsLGTx4sC68E5GYJKxYmFkKcB9wDjAUuNTMhtbpdg/wqLuPAG4Dfh7Gng5MAkYAw4CxwJmJyrW1OXToELm5uQwfPjzZqYjIMSKR11mMAwrdfTOAmS0AzgfWR/QZCvwofL4UeCZ87kA60BYwIA3YmcBcW5UOHTowa9asZKchIscQc/fELNhsLjDL3a8Jp68Axrv7/Ig+jwPvuPu9ZjYHeAro5u67zewe4BqCYvFrd7+pgXVcB1wH0KNHj9wFCxY0O9+SkhIyMzOP+/iqqiq++OILOnbsWGsX1LGSv+IVr/iWjZ82bdpKdx8TtaO7J+QBzAUejJi+guBDP7JPb2AR8D5wL1AEdAJOAf4KZIaP/wUmN7W+3NxcPxxLly5tFfGbNm3yW2+91Tds2JCU9Ste8Yo/uuKBFR7DZ3pMxyzMbJGZnWtm8Rzj2A70jZjuE86LLFQ73H2Ou48CbgrnfQ5cCLzt7iXuXgK8CEyMY93SiPz8fNq2bcvJJ5+c7FRE5BgS64f/b4BvApvM7A4zGxRDzHtAtpkNMLO2wDzg2cgOZtYtogD9BHgofP4xcKaZpZpZGsHB7fwYc5VGVFVVsWHDBrKzs0lN1bBgIhK7mIqFu7/q7pcBo4EtwKtm9paZXR1+mDcUUwHMB14i+KB/0t3XmdltZjY77DYV2GBmG4EewO3h/IXAh8AaYDWw2t2fa84LlL8rKiriwIEDDB48ONmpiMgxJuavl2bWFbic4NjD+8BjwBnAlQQf+vW4+wvAC3Xm3RzxfCFBYagbVwn8Y6y5SWw2bNhASkoK2dnZyU5FRI4xMRULM3saGAT8D/B1d/8kbHrCzFYkKjlpWdOnT2fYsGG0a9cu2amIyDEm1i2L/3b3pQ01eCynXMlRISUlhV69eiU7DRE5BsV6gHuomXWqnjCzzmb2nQTlJAnw7rvvsnjx4upTlkVE4hJrsbg2PKUVAHffC1ybmJQkEVasWEFRUZHGghKRZom1WKREDuQXjvvUNjEpSUsrLi5m165dOgtKRJot1mMWfyM4mP1AOP2P4Tw5BhQUFACoWIhIs8VaLP6NoED8czj9CvBgQjKSFpefn0/v3r3p2LFjslMRkWNUTMXC3auA+8OHHEOqqqro0qULffr0id5ZRKQRsV5nkU1wr4mhBEOHA+DuGmDoKNemTRsuuuiiZKchIse4WA9w/5Fgq6ICmAY8CvwpUUlJy/niiy+SnYKIHAdiLRbt3X0xwf0vtrr7rcC5iUtLWsKBAwf41a9+xdtvv53sVETkGBfrAe4vw9FhN5nZfIKhxpt/pw45IjZu3Ii7069fv2SnIiLHuFi3LH4AZADfB3IJBhS8MlFJScvIz8+nY8eO9OzZM9mpiMgxLuqWRXgB3iXufj1QAlyd8KzksH355Zds3ryZsWPH6qptETlsUbcswuHCzzgCuUgL2rRpE5WVlQwZMiTZqYjIcSDWYxbvm9mzwF+AA9Uz3X1RQrKSwzZw4EDmzJmj6ytEpEXEWizSgd3A9Ih5DqhYHKXat2/P8OHDk52GiBwnYr2CW8cpjiHbtm1j+/bt5ObmkpbW4F1vRUTiEusV3H8k2JKoxd3/ocUzksO2atUq8vPzGTt2bLJTEZHjRKy7oZ6PeJ4OXAjsaPl05HBVVVWxYcMGTj31VFJSUpKdjogcJ2LdDfVU5LSZ/Rl4IyEZyWHZunUrZWVlOgtKRFpUrBfl1ZUNdG/JRKRlFBQUkJqaysCBA5OdiogcR2I9ZrGf2scsPiW4x4UcZUpLS8nOzqZtW93IUERaTqy7obISnYi0jIsuuoiqqqpkpyEix5mYdkOZ2YVm1jFiupOZXZC4tKQ5KisrgeAeFiIiLSnWT5Vb3H1f9YS7fw7ckpiUpDncnQceeIBXX3012amIyHEo1mLRUL9YT7uVI6C0tJRdu3bRqVOnZKciIsehWIvFCjP7hZkNDB+/AFYmMjGJT3FxMQCDBg1KciYicjyKtVh8DzgIPAEsAMqB7yYqKYlfcXExffv2JStL5yKISMuL9WyoA8CNCc5Fmmnv3r2UlJQwceLEZKciIsepWM+GesXMOkVMdzazlxKXlsSjXbt2nHzyyQwdOjTZqYjIcSrWg9TdwjOgAHD3vWamK7gbsHHjxnrXOXTq1Knm1qYFBQX1Yrp06QIEp75u2rSpXnu3bt3o1q0bhw4d4sMPP6zX3r17d/r27auD2yKSMLEWiyozO8ndPwYws/40MApta7Rnzx7eeustUlODX+VTTz3FwYMHa/UZPXo0X//61wF44okn6i1jwoQJtGvXjoqKigbbzzzzTKZOnUp5eXmD7TNnzmyJlyIi0qhYi8VNwBtm9hpgwGTguoRldQzJy8tj1apVjB8/HoCrr74a99p1NCMjo+b5ddfV/7V16NCBVatWkZaW1mB7ZmZmzXIaas/KymLFihWH9TpERJoS6wHuv5nZGIIC8T7wDFCWyMSOBVVVVaxevZqBAwfSrl07gJrdTY3p1atXo21t2rRpsj0lJaXJdhGRRIn1APc1wGLgX4Hrgf8Bbo0hbpaZbTCzQjOrdzaVmfUzs8Vm9oGZLTOzPuH8aWaWF/EoPxqHF/noo4/44osvyMnJSXYqIiIJFet1Fj8AxgJb3X0aMAr4vKkAM0sB7gPOAYYCl5pZ3dN17gEedfcRwG3AzwHcfam757h7DsF9v0uBl2PM9YjJy8sjPT1dF8KJyHEv1mJR7u7lAGbWzt0LgGifkOOAQnff7O4HCS7mO79On6HAkvD50gbaAeYCL7p7aYy5HjGZmZmMGTOm5uC2iMjxyuoejG2wk9nTwNXADwm+6e8F0tz9a03EzAVmufs14fQVwHh3nx/R53HgHXe/18zmAE8RnKa7O6LPEuAX7v48dZjZdYQH2nv06JG7YMGCGF5yw0pKSmoOJCte8YpXfGuJnzZt2kp3HxO1o7vH9QDOBGYDbaP0mws8GDF9BfDrOn16A4sIDprfCxQBnSLaewG7CApTk3nl5ub64Vi6dGlc/T/55BOvqqpqdvzhrl/xile84lsiHljhMXz2x33jA3d/zd2f9WDXUlO2A30jpvuE8yKXtcPd57j7KILTc6uHP6/2DeBpdz8Ub56JtGvXLh544AFWrVqV7FRERI6IRN4l5z0g28wGmFlbYB7wbGQHM+tmZtU5/AR4qM4yLgX+nMAcmyUvL482bdowePDgZKciInJEJKxYuHsFMB94CcgHnnT3dWZ2m5nNDrtNBTaY2UagB3B7dXx4lXhf4LVE5dgcVVVVfPDBB2RnZ9OhQ4dkpyMickQk9DQed38BeKHOvJsjni8EFjYSuwU4MZH5NUdhYSElJSW6tkJEWhXdrDlO69atIyMjg+zs7GSnIiJyxOgCgTh9/etfp7i4mJSUlGSnIiJyxGjLIk6pqalRx38SETneqFjE4cknnyQvLy/ZaYiIHHEqFjH69NNPyc/Pr3evChGR1kDFIkZ5eXmkpKQwbNiwZKciInLEqVjEoLKykjVr1jBo0KBaNzISEWktVCxisHHjRkpLS3VthYi0WioWMcjMzGTEiBEMHDgw2amIiCSFrrOIQd++fenbt2/0jiIixyltWUSxbds29u7dm+w0RESSSsWiCe7O888/z6JFi5KdiohIUqlYNOGTTz7hs88+Y+TIkclORUQkqVQsmvD++++TmpqqaytEpNVTsWhERUUFa9euZfDgwaSnpyc7HRGRpFKxaMSOHTs4ePCgrq0QEUGnzjbqpJNO4kc/+hHt27dPdioiIkmnYtEAd8fMdNtUEZGQdkM14M033+Thhx+moqIi2amIiBwVVCzqcHfy8vJwd1JTteElIgIqFvUUFRWxe/duHdgWEYmgYlFHXl4eaWlpDB06NNmpiIgcNVQsIhw6dIi1a9cydOhQ2rVrl+x0RESOGtopX8f06dPp06dPstMQETmqqFhESEtLY/z48clOQ0TkqKPdUKHy8nJWrVrFwYMHk52KiMhRR8Ui9Omnn/Lcc89RWlqa7FRERI46KhYE11bs3LmTAQMG0KlTp2SnIyJy1FGxALZu3Up5ebmurRARaYSKBcG1FSkpKQyQ48SCAAAM2klEQVQZMiTZqYiIHJVafbFwd/bt28cJJ5xAWlpastMRETkqtfpiYWZceeWVZGdnJzsVEZGjVqsvFtXatNGvQkSkMfqEFBGRqBJaLMxslpltMLNCM7uxgfZ+ZrbYzD4ws2Vm1iei7SQze9nM8s1svZn1T2SuIiLSuIQVCzNLAe4DzgGGApeaWd2hXO8BHnX3EcBtwM8j2h4F7nb3IcA44LNE5SoiIk1L5JbFOKDQ3Te7+0FgAXB+nT5DgSXh86XV7WFRSXX3VwDcvcTddWm1iEiSmLsnZsFmc4FZ7n5NOH0FMN7d50f0eRx4x93vNbM5wFNAN2AycA1wEBgAvArc6O6VddZxHXAdQI8ePXIXLFjQ7HxLSkrIzMxUvOIVr/hWFT9t2rSV7j4makd3T8gDmAs8GDF9BfDrOn16A4uA94F7gSKgUxi7DziZYGTcp4BvN7W+3NxcPxxLly5VvOIVr/hWFw+s8Bg+0xO5G2o70Ddiuk84r4a773D3Oe4+CrgpnPd5WDTyPNiFVQE8A4xOYK4iItKERBaL94BsMxtgZm2BecCzkR3MrJuZVefwE+ChiNhOZnZCOD0dWJ/AXEVEpAkJKxbhFsF84CUgH3jS3deZ2W1mNjvsNhXYYGYbgR7A7WFsJXA9sNjM1gAG/D5RuYqISNMSeqc8d38BeKHOvJsjni8EFjYS+wowIpH5iYhIbHQFt4iIRKViISIiUalYiIhIVCoWIiISlYqFiIhEpWIhIiJRqViIiEhUKhYiIhKVioWIiESlYiEiIlGpWIiISFQqFiIiEpWKhYiIRKViISIiUalYiIhIVCoWIiISlYqFiIhEpWIhIiJRqViIiEhUKhYiIhKVioWIiESlYiEiIlGpWIiISFQqFiIiEpWKhYiIRKViISIiUalYiIhIVCoWIiISlYqFiIhEpWIhIiJRqViIiEhUKhYiIhKVioWIiESlYiEiIlGpWIiISFQJLRZmNsvMNphZoZnd2EB7PzNbbGYfmNkyM+sT0VZpZnnh49lE5ikiIk1LTdSCzSwFuA+YCRQB75nZs+6+PqLbPcCj7v6ImU0Hfg5cEbaVuXtOovITEZHYJXLLYhxQ6O6b3f0gsAA4v06focCS8PnSBtpFROQokLAtC+BEYFvEdBEwvk6f1cAc4F7gQiDLzLq6+24g3cxWABXAHe7+TN0VmNl1wHXhZImZbTiMfLsBxYpXvOIV38ri+8XUy90T8gDmAg9GTF8B/LpOn97AIuB9goJRBHQK204Mf54MbAEGJirXcD0rFK94xSu+NcbH8kjklsV2oG/EdJ9wXg1330GwZYGZZQIXufvnYdv28OdmM1sGjAI+TGC+IiLSiEQes3gPyDazAWbWFpgH1Dqrycy6mVl1Dj8BHgrndzazdtV9gElA5IFxERE5ghJWLNy9ApgPvATkA0+6+zozu83MZofdpgIbzGwj0AO4PZw/BFhhZqsJDnzf4bXPokqE3yle8YpXfCuNj8rC/V0iIiKN0hXcIiISlYqFiIhE1eqLhZk9ZGafmdnaZsT2NbOlZrbezNaZ2Q/ijE83s3fNbHUY/9N4cwiXk2Jm75vZ882I3WJma8JhVVY0I76TmS00swIzyzeziXHEDooY0iXPzL4wsx/Guf5/CX93a83sz2aWHmf8D8LYdbGuu6H3jJl1MbNXzGxT+LNznPEXhzlUmdmYZqz/7vBv8IGZPW1mneKM/88wNs/MXjaz3vHER7T9q5l5eGJKPOu/1cy2R7wXvhbv+s3se+HvYJ2Z3RXn+p+IWPcWM8uLMz7HzN6u/j8ys3Fxxo80s/8N/xefM7OvNBLb4GdOPO+/Zkv0ublH+wOYAowG1jYjthcwOnyeBWwEhsYRb0Bm+DwNeAeY0Iw8fgQ8DjzfjNgtQLfD+P09AlwTPm9LeJ1MM5aTAnwK9Isj5kTgI6B9OP0kcFUc8cOAtUAGwQWqrwKnNOc9A9wF3Bg+vxG4M874IcAgYBkwphnrPxtIDZ/f2Yz1fyXi+feB38YTH87vS3BCy9am3lONrP9W4PoY/24NxU8L/37twunu8eYf0f5fwM1xrv9l4Jzw+deAZXHGvwecGT7/B+A/G4lt8DMnnvdfcx+tfsvC3ZcDe5oZ+4m7rwqf7yc46+vEOOLd3UvCybTwEdcZBxYMvngu8GA8cS3BzDoSvPH/AODuBz28TqYZZgAfuvvWOONSgfZmlkrwob8jjtghwDvuXurB2XuvEV7305RG3jPnExROwp8XxBPv7vnuHtMIBI3Evxy+BoC3Ca5riif+i4jJDjTxPmzif+aXwI+bio0SH5NG4v+Z4KzJL8M+nzVn/WZmwDeAP8cZ70D11kBHmngfNhJ/KrA8fP4KcFEjsY195sT8/muuVl8sWoqZ9Se4cPCdOONSwk3ez4BX3D2ueOBXBP+gVXHGVXPgZTNbacHwKfEYAOwC/hjuBnvQzDo0M495NPEP2hAPLty8B/gY+ATY5+4vx7GItcBkM+tqZhkE3wj7RolpTA93/yR8/inBqeDJ8g/Ai/EGmdntZrYNuAy4Oc7Y84Ht7r463vVGmB/uCnuoGbtRTiX4W75jZq+Z2dhm5jAZ2Onum+KM+yFwd/j7u4fgurF4rOPvY+NdTAzvwzqfOQl//6lYtAALrj5/CvhhnW9oUbl7pQej6/YBxpnZsDjWex7wmbuvjCvh2s5w99HAOcB3zWxKHLGpBJvT97v7KOAAwSZwXCy4aHM28Jc44zoT/IMNIBg6poOZXR5rvLvnE+yyeRn4G5AHVMaTQyPLdeLcQmwpZnYTwXhqj8Ub6+43uXvfMHZ+HOvMAP4PcRaYOu4HBgI5BIX/v+KMTwW6ABOAG4Anw62EeF1KnF9aQv8M/Ev4+/sXwq3tOPwD8B0zW0mwe+lgU52b+sxJ1PtPxeIwmVkawR/tMXdf1NzlhLtvlgKz4gibBMw2sy0Eo/pON7M/xbne6mFVPgOeJhgtOFZFQFHE1tBCguIRr3OAVe6+M864s4CP3H2Xux8iGGfs9HgW4O5/cPdcd58C7CXYB9wcO82sF0D4s9HdIIliZlcB5wGXhR8YzfUYjewGacRAgoK9Onwv9gFWmVnPWBfg7jvDL05VwO+J730IwXtxUbhr912CLe1GD7I3JNyVOQd4Is51A1xJ8P6D4EtPXPm7e4G7n+3uuQTFqtGhjRr5zEn4+0/F4jCE31z+AOS7+y+aEX9C9VkrZtae4N4fBbHGu/tP3L2Pu/cn2I2zxN1j/mZtZh3MLKv6OcFB0pjPCnP3T4FtZjYonDWD5g3L0txvcx8DE8wsI/xbzCDYhxszM+se/jyJ4IPi8WbkAcFQNleGz68E/l8zl9MsZjaLYHfkbHcvbUZ8dsTk+cT3Plzj7t3dvX/4XiwiOAj7aRzr7xUxeSFxvA9DzxAc5MbMTiU42SLeUVjPAgrcvSjOOAiOUZwZPp8OxLUbK+J92Ab4d+C3jfRr7DMn8e+/lj5ifqw9CD6kPgEOEbzJvx1H7BkEm3sfEOzCyAO+Fkf8CIIRdz8g+Odo9AyMGJY1lTjPhiIY0Xd1+FgH3NSM9eYAK8LX8AzQOc74DsBuoGMzX/dPCT7Y1gL/Q3g2TBzxrxMUuNXAjOa+Z4CuwGKCD4lXgS5xxl8YPv8S2Am8FGd8IcEtAarfh02dzdRQ/FPh7/AD4DnCUZ+b8z9DlDPsGln//wBrwvU/C/SKM74t8KfwNawCpsebP/Aw8E/N/PufAawM30fvALlxxv+AYKt2I3AH4egaDcQ2+JkTz/uvuQ8N9yEiIlFpN5SIiESlYiEiIlGpWIiISFQqFiIiEpWKhYiIRKViIXIUMLOp1oxRg0WOFBULERGJSsVCJA5mdrkF9yDJM7MHwoEgS8zsl+H9BRab2Qlh3+p7HFTfY6JzOP8UM3vVgvuYrDKzgeHiM+3v9wZ5rJljG4kkhIqFSIzMbAhwCTDJg8EfKwlGaO0ArHD30wiGOb8lDHkU+Dd3H0FwdXL1/MeA+9x9JMFYVtWjhY4iGL10KMHV9ZMS/qJEYpSa7AREjiEzgFzgvfBLf3uCAduq+Pvgc38CFoX3+ujk7q+F8x8B/hKOxXWiuz8N4O7lAOHy3vVwXKJw2Pr+wBuJf1ki0alYiMTOgEfcvda9CszsP+r0a+4YOl9GPK9E/59yFNFuKJHYLQbmRowQ2sXM+hH8H80N+3wTeMPd9wF7zWxyOP8K4DUP7m5WZGYXhMtoF94PQuSopm8uIjFy9/Vm9u8EdxZsQzBq6HcJbvo0Lmz7jOC4BgRDRf82LAabgavD+VcAD5jZbeEyLj6CL0OkWTTqrMhhMrMSd89Mdh4iiaTdUCIiEpW2LEREJCptWYiISFQqFiIiEpWKhYiIRKViISIiUalYiIhIVP8fSNpdSVusVMwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 18\n",
      "predictive mean: 0.9867177358196214\n",
      "predictive std: 9.692467576388345e-05\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test case: 0\n",
      "number of inputs: 19\n",
      "predictive mean: 0.9866416431183805\n",
      "predictive std: 4.818862190307515e-05\n",
      "ground truth: 0.986564\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "predictor = MultiPredictor(dataname, validation_dataname, N_sigma, dlen,\n",
    "                           inputlen, N_ensembles, max_workers)\n",
    "pred_df = pd.read_json('%s.json' % (validation_dataname), orient='split')\n",
    "for target_num in range(N_test_cases):\n",
    "    for i in range(dlen - inputlen):\n",
    "        input_df = pred_df.iloc[target_num, :inputlen + i]\n",
    "        tmp = input_df.copy()\n",
    "        for i in range(1, input_df.shape[0]):\n",
    "            tmp[i] = (input_df[i] - input_df[i - 1]) / (1 - input_df[i - 1])\n",
    "        input_df = tmp\n",
    "        predictions = predictor.predict(input_df)\n",
    "        mean = predictions.mean()\n",
    "        std = predictions.std()\n",
    "        original = pred_df.iloc[target_num]\n",
    "        print('test case: %s\\nnumber of inputs: %s\\npredictive mean: %s\\npredictive std: %s\\nground truth: %s' % (\n",
    "            target_num, inputlen + i, mean.values[-1], std.values[-1], original.values[-1]))\n",
    "        plot(mean, original)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}