EEGnet_CNN_stim_reconst3 (2).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## This is an implementation of a convnet that *tries* to reconstruct attended audio from EEG signal. \n",
    "\n",
    "This version is built from an encoder (trained in another notebook called EEGnet_CNN_autoencoder) that is supposed to *know* dense representations of the EEG features. Here we add this pre-trained encoder, freeze its weights and try to learn to reconstruct the audio based on the dense hidden representations of the EEG signal. So far, (at lease in its current form, the net fails to achieve any reconstruction at all. We get overfitting, that's it. ANY SUGGESTIONS ARE VERY WELCOME!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "# we first define the autoencoder's architecture:\n",
    "\n",
    "# make sure you don't hog all the video memory\n",
    "import tensorflow as tf\n",
    "config = tf.ConfigProto()\n",
    "config.gpu_options.allow_growth = True\n",
    "sess = tf.Session(config=config)\n",
    "from keras import backend as K\n",
    "K.set_session(sess)\n",
    "###################################\n",
    "\n",
    "from keras.layers import Input, Dense, Conv2D, MaxPooling2D, UpSampling2D, Dropout, BatchNormalization\n",
    "from keras.layers.advanced_activations import LeakyReLU\n",
    "from keras.models import Model\n",
    "from keras.models import load_model\n",
    "from keras.callbacks import TensorBoard\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# def Encoder():\n",
    "#     input_img = Input(shape=(60, 800, 1))  # adapt this if using `channels_first` image data format   \n",
    "#     e1 = Conv2D(16, (3, 3), activation='relu', padding='same')(input_img)\n",
    "#     e2 = MaxPooling2D((2, 2), padding='same')(e1)\n",
    "#     e3 = Conv2D(8, (3, 3), activation='relu', padding='same')(e2)\n",
    "#     e4 = MaxPooling2D((3, 2), padding='same')(e3)\n",
    "#     e5 = Conv2D(8, (10, 10), activation='relu', padding='same')(e4)\n",
    "#     e6 = MaxPooling2D((2, 2), padding='same')(e5)\n",
    "#     return Model(input_img, e6)\n",
    "\n",
    "def Decoder():\n",
    "    input_img = Input(shape=(5, 100, 8))  # adapt this if using `channels_first` image data format   \n",
    "    d1 = Conv2D(16, (10, 10), activation='relu', padding='same')(input_img)\n",
    "    d2 = UpSampling2D((1, 2))(d1)\n",
    "#     d3 = BatchNormalization(axis=1)(d2)\n",
    "    d4 = Dropout(0.5)(d2)\n",
    "    d5 = Conv2D(32, (5, 5), activation='relu', padding='same')(d4)\n",
    "    d6 = UpSampling2D((1, 2))(d5)\n",
    "#     d7 = BatchNormalization(axis=1)(d6)\n",
    "    d8 = Dropout(0.5)(d6)\n",
    "    d9 = Conv2D(32, (4, 4), activation='relu', padding='same')(d8)\n",
    "    d10 = UpSampling2D((1, 2))(d9)\n",
    "#     d11 = BatchNormalization(axis=1)(d10)\n",
    "    d12 = Dropout(0.5)(d10)\n",
    "    d13 = Conv2D(16, (3, 3), activation='relu', padding='same')(d12)     # !!!!!!!!!!!!!!!!relu\n",
    "    d13a = LeakyReLU(alpha=0.01)(d13)\n",
    "    d14 = MaxPooling2D((5, 1))(d13a)\n",
    "    d15 = Conv2D(1, (3, 3), activation='sigmoid', padding='same')(d14)  # !!!!!!!!!!!!!!!!relu\n",
    "    return Model(input_img, d15)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define input to the model:\n",
    "x = Input(shape=(60, 800, 1))\n",
    "\n",
    "# load the pre-trained encoder\n",
    "# enc = load_model(\"/home/amplifier/home/models/CNN_encoder_model.h5\")\n",
    "enc = load_model(\"/home/amplifier/home/NEW_DL/models/CNN_encoder_model.h5\")\n",
    "\n",
    "enc.trainable = False ############################################################################!!!!!!!!!!!!!!!!!@@@\n",
    "enc.compile(optimizer='adadelta', loss='binary_crossentropy')\n",
    "\n",
    "# make the model:\n",
    "model = Model(x, Decoder()(enc(x)))\n",
    "\n",
    "# compile the model:\n",
    "model.compile(optimizer='adadelta', loss='binary_crossentropy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         (None, 60, 800, 1)        0         \n",
      "_________________________________________________________________\n",
      "model_1 (Model)              (None, 5, 100, 8)         7728      \n",
      "=================================================================\n",
      "Total params: 7,728\n",
      "Trainable params: 0\n",
      "Non-trainable params: 7,728\n",
      "_________________________________________________________________\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_18 (InputLayer)        (None, 5, 100, 8)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_56 (Conv2D)           (None, 5, 100, 16)        12816     \n",
      "_________________________________________________________________\n",
      "up_sampling2d_34 (UpSampling (None, 5, 200, 16)        0         \n",
      "_________________________________________________________________\n",
      "dropout_34 (Dropout)         (None, 5, 200, 16)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_57 (Conv2D)           (None, 5, 200, 32)        12832     \n",
      "_________________________________________________________________\n",
      "up_sampling2d_35 (UpSampling (None, 5, 400, 32)        0         \n",
      "_________________________________________________________________\n",
      "dropout_35 (Dropout)         (None, 5, 400, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_58 (Conv2D)           (None, 5, 400, 32)        16416     \n",
      "_________________________________________________________________\n",
      "up_sampling2d_36 (UpSampling (None, 5, 800, 32)        0         \n",
      "_________________________________________________________________\n",
      "dropout_36 (Dropout)         (None, 5, 800, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_59 (Conv2D)           (None, 5, 800, 16)        4624      \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_12 (LeakyReLU)   (None, 5, 800, 16)        0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_12 (MaxPooling (None, 1, 800, 16)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_60 (Conv2D)           (None, 1, 800, 1)         145       \n",
      "=================================================================\n",
      "Total params: 46,833\n",
      "Trainable params: 46,833\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_16 (InputLayer)        (None, 60, 800, 1)        0         \n",
      "_________________________________________________________________\n",
      "model_2 (Model)              (None, 5, 100, 8)         7728      \n",
      "_________________________________________________________________\n",
      "model_16 (Model)             (None, 1, 800, 1)         46833     \n",
      "=================================================================\n",
      "Total params: 54,561\n",
      "Trainable params: 46,833\n",
      "Non-trainable params: 7,728\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# SEE INSIDE THE ENCODER AND DECODER:\n",
    "\n",
    "enc.summary()\n",
    "Decoder().summary()\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original data type: float32\n",
      "Normalized data type: float16\n",
      "test input shape (272, 60, 800, 1) Nomralized MEAN: 0.5166 min 0.0 max 1.0\n",
      "train input shape (633, 60, 800, 1) Nomralized MEAN: 0.51807 min 0.0 max 1.0\n",
      "test labels shape (272, 1, 800, 1) Nomralized MEAN: 0.27734 min 0.0 max 1.0\n",
      "train labels shape (633, 1, 800, 1) Nomralized MEAN: 0.28662 min 0.0 max 1.0\n"
     ]
    }
   ],
   "source": [
    "# get the Dataset:\n",
    "\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import preprocessing\n",
    "import scipy.io as sio\n",
    "\n",
    "# load data from the .mat file:\n",
    "# mat_contents = sio.loadmat(\"EEG_data_KOS_1+2.mat\") #('EEG_data.mat')\n",
    "mat_contents = sio.loadmat(\"EEG_big4CNN.mat\") #('EEG_data.mat')\n",
    "X = mat_contents['X']\n",
    "Y = mat_contents['Y']\n",
    "x_train, x_test, y_train, y_test = train_test_split(X, Y, test_size=0.30, shuffle=True) # !!!!!!!!!!!!!!!!!!!!!!!!\n",
    "print('Original data type:', x_train.dtype)\n",
    "\n",
    "# convert to float64 for numerical stability:\n",
    "x_train = x_train.astype('float64')\n",
    "y_train = y_train.astype('float64')\n",
    "x_test = x_test.astype('float64')\n",
    "y_test = y_test.astype('float64')\n",
    "\n",
    "# normalize to unit variance and zero mean:\n",
    "for i in range(x_train.shape[0]):\n",
    "    x_train[i,:,:] = preprocessing.scale(x_train[i,:,:], axis=1)\n",
    "    x_train[i,:,:] = preprocessing.minmax_scale(x_train[i,:,:], axis=1)\n",
    "    y_train[i,:] = preprocessing.scale(y_train[i,:], axis=0)\n",
    "    y_train[i,:] = preprocessing.minmax_scale(y_train[i,:], axis=0)\n",
    "\n",
    "for i in range(x_test.shape[0]):\n",
    "    x_test[i,:,:] = preprocessing.scale(x_test[i,:,:], axis=1)\n",
    "    x_test[i,:,:] = preprocessing.minmax_scale(x_test[i,:,:], axis=1)\n",
    "    y_test[i,:] = preprocessing.scale(y_test[i,:], axis=0)\n",
    "    y_test[i,:] = preprocessing.minmax_scale(y_test[i,:], axis=0)\n",
    "\n",
    "# ensure the tensors are sized right:\n",
    "x_train = np.expand_dims(x_train,3)\n",
    "x_test = np.expand_dims(x_test,3)\n",
    "y_test = np.expand_dims(y_test,1)\n",
    "y_test = y_test[:,:,:,None]\n",
    "y_train = np.expand_dims(y_train,1)\n",
    "y_train = y_train[:,:,:,None]\n",
    "\n",
    "# convert to float16 to save space:\n",
    "x_train = x_train.astype('float16')\n",
    "y_train = y_train.astype('float16')\n",
    "x_test = x_test.astype('float16')\n",
    "y_test = y_test.astype('float16')\n",
    "print('Normalized data type:', x_train.dtype)\n",
    "\n",
    "print('test input shape', x_test.shape, \"Nomralized MEAN:\", np.mean(x_test), \"min\", np.min(x_test),\"max\", np.max(x_test))\n",
    "print('train input shape', x_train.shape, \"Nomralized MEAN:\", np.mean(x_train), \"min\", np.min(x_train),\"max\", np.max(x_train))\n",
    "\n",
    "print('test labels shape', y_test.shape, \"Nomralized MEAN:\", np.mean(y_test), \"min\", np.min(y_test),\"max\", np.max(y_test))\n",
    "print('train labels shape', y_train.shape, \"Nomralized MEAN:\", np.mean(y_train), \"min\", np.min(y_train),\"max\", np.max(y_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'Channels 11 in test samples 6 and 16')"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEICAYAAACktLTqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvAOZPmwAAIABJREFUeJzsXXmcHEX1/77umdkj2c19BxICCSHcGBAwQLiPyKEIgqLiT0UURPFAUFRUEFRADjlVUOQWVCAJtwlHCEcASQhJyH3fyebaa2a6fn9UV3dVdfUxs7Ob3dhfP5idPqurq7/16vtevSLGGFKkSJEixa4Fa2cXIEWKFClSVB4puadIkSLFLoiU3FOkSJFiF0RK7ilSpEixCyIl9xQpUqTYBZGSe4oUKVLsgkjJvYNBRNcQ0YM7uxw6iGgqEX19J5dhOxGN2Jll6CroDO+rkiCiJUR0ws4ux66ElNzbAUT0BSKa4ZLVaiJ6lojG7exytQeI6F4imkdEDhFdqO3bj4ieJ6INRBQ7oYIx1p0xtqiMMgwnIkZEmVLPNVzrr0R0bVuv01VBRCcQ0XtEtIOIlhPRuZ2gTMcS0RQi2kJES0KO+S4RLXbLPYeIRnVwMTsdUnKvMIjo+wBuAfAbAAMA7A7gTgBn7sxytSM+APBtAO8Z9uUBPA7gax1aohRlgYjGAHgYwE8B9ABwEIB3d2qhOHYAuA/Aj0w73RHM1wBMANAdwKcBbOiw0nVWMMbS/yr0H/gHsR3AORHHXANOeA8A2AZgNoCx0v4rASx0930E4DPSvgsBvA7gRgCbASwGcKq0fyqAXwOY5p7/AoC+0v7DAbwBoAGclMdr537d/XsvAK8A2AL+kTyW4NlfB3BhyL69eFOLvQYDsJf7918B3AFgkvssbwHYM+S8Ze65293/jnC3/x+AOW5dPQ9gmLudAPwBwDr3GWcC2A/AReAdUqt7nWcM9zKe6+6bAOB9AFsBLAdwjXTecLeMX3X3bQZwMYBD3Ws0APij9q6nAbjdvc9cAMeb3lc5zxpSjw8D+HXCtt4LwEQA6917TgQwtIS2+CUASwFsBO9MlgA4IeaeJwBYom2z3Po8Pkm5/5f+2+kF2JX+A3AKgAKATMQx1wBoBnAaABvA9QDelPafA2Cw22g/D261DHL3XeiSzzfcc78FYBUAcvdPBe8YRgGocX/f4O4b4n5Ip7nXPtH93U86V5D7I+4HZwGoBjAuwbO3B7lvAnAYgAyAhwA8GnLecPfcjLTtLAALAOzjnn81gDfcfSeDW6Q9XfLbR6rjvwK4NqKMUeeOB7C/W28HAFgL4CytjHe7dXqS2w7+DaC/+37WAThGetcFAJcDyLptYQuA3ob3VdazGp5tETghzwKwGsCD4n6GY/sAOBtALYA6AP8A8G9p/1SEt8Ux4J3n0QCqANzsPms55L67W6/fBSf5xQB+CcDa2Xyws/9LZZnKog+ADYyxQsxxrzPGJjPGigD+DuBAsYMx9g/G2CrGmMMYewzAfHCCE1jKGPuTe+7fAAwCl38E7meMfcwYawIfIRzkbr8AwGT3vg5j7EUAM8DJXkcewDAAgxljzYyx15NWQIXxT8bY2259PgT/WZLgmwCuZ4zNcc//DYCDiGgY+PPVARgN3jHOYYytTnjd0HMZY1MZY7Pc+p0J3kkeo53/a7dOXwDvuB9hjK1jjK0E8BqAg6Vj1wG4hTGWd9vCPPDRQXs961Bwi/psACPBSfl204GMsY2MsScZY42MsW0ArjM8a1hb/ByAiYyxVxljLQB+BsAJKVMchrr/ngTesR4L4HykUmBK7hXGRgB9Ezj21kh/NwKoFucQ0ZeJ6L9E1EBEDeByQV/TuYyxRvfP7hHXFvuGAThHXNe99jjwzkHHFeBW3ttENJuI/i/medoLYc+SBMMA3Co96ybwZxrCGPsPgD+Cyz5rXadwfZKLRp1LRJ90HX/riWgLuOzSV7vEWunvJsNv+RlXMtc8dbEUfFTXXs/aBJ+Qt4N3EqbOH0RUS0T3ENFSItoK4FUAPYnIlg4Le3+Dwa1sAABjbAf4t1MOmtx/f8cYa2CMLQFwT1i5/5eQkntlMR18qH1WOSe7ltafAFwKoA9jrCeAD8E/1LZiOYC/M8Z6Sv91Y4zdoB/IGFvDGPsGY2wwuFV4JxHtVYEytBdMkTjLAXxTe94axtgbAMAYu40x9gkA+4JLBz+KuJZ6s/BzHwbwNIDdGGM9wCWYtry7IUQkn787uAyno9xn1TETCZ7fxQ8A7A3gk4yxenCJBUj2vKsB7CZ+EFEt+Ki3HMwD95Gk6W01pOReQTDGtgD4OYA7iOgs17rJEtGpRPS7BJfoBt5I1wMAEX0V3HKvBB4EcDoRnUxENhFVE9F4IhqqH0hE50jbN7tlKpouSkQ5IqoG/6iz7nUtdx+5+3Lu72oiqqrQ88hYDz6sl2Pk7wZwFRHt6967BxGd4/59qGtlZ8GlkWbp+dZq11EQc24dgE2MsWYiOgzAF9r4XP0BXOa2oXPA9fLJhuPKfVYd9wP4KhGNcAn3x+COUhPqwK3mBiLqDeAXJTzXEwA+TUTjiCgH4FeI4CIistx2lOU/qdo9T4xeHwNwBRHVue32GxHl/p9BSu4VBmPsZgDfB3dqrQe3qi4Fd5zFnfsRgJvARwBrwTXEaRUq13LwcMyfSOX6Ecxt4FAAbxHRdnBL9LuMscUhl34B/CM/EsC97t/Cihvm/p7t/m4Ct7QqCvcDvw7ANFeaOJwx9i8AvwXwqCsbfAjgVPeUevAR0mb4ERs3uvv+AmCMex3TO4s699sAfkVE28A7+cfb+GhvgWvfG9zn+xxjLCBftOFZ9evcBx7F9ZZ7bAuAy0LKdgu4Jr8BwJsAnkv6UIyx2QAuAR/prHbLtiLilKPB285k8NFLE3i7E7gU3EG7CvzbeRg8dPJ/GiLKIkWKFJ0IxCeEfZ0xtktOfkvR/kgt9xQpUqTYBZGSe4oUKVLsgkhlmRQpUqTYBZFa7ilSpEixC6LNWfTKRd++fdnw4cN31u1TpEiRokvi3Xff3cAY6xd33E4j9+HDh2PGjBk76/YpUqRI0SVBREuTHJfKMilSpEixCyIl9xQpUqTYBZGSe4oUKVLsgkjJPUWKFCl2QaTkniJFihS7IGLJnYjuI6J1RPRhyH4iotuIaAERzSSiQypfzBQpUqRIUQqSWO5/BV8+LgyngmeuGwm+BuVdbS9WihQpUqRoC2LJnTH2KvjKLmE4E8ADjONN8NVYTKv7pEiKNbOA5W/v7FKkSAEAWL2lCS99tDb+wBSdCpXQ3IdAWjILPC/zENOBRHQREc0gohnr16+vwK13Udw9DvjLiTu7FClSAADOvvMNfP2BdMJhV0MlyN20rJYxGxlj7F7G2FjG2Nh+/WJnz6ZIkaITYNWWZgBAoVjuGtYpdgYqQe4rIK2HCL4auWmdxxQpUnRh5ItpBtmuhEqQ+9MAvuxGzRwOYAtjbHUFrpsiRYpOhNbUcu9SiE0cRkSPABgPoC8RrQBfCDcLAIyxu8HXNTwNwAIAjQC+2l6FTZEixc5DayEl966EWHJnjJ0fs5+BL3abIkWKXRj51HLvUkhnqKb4n8S8NdvQ0Ni6s4vRJUBuyERK7l0LKbmn+J/Eybe8is/c+cbOLkaXgOWyeyrLdC2k5J7ifxaLN+zY2UXotFi3tRnvLt0MALBcyz11qHYtpOSeIkWKACbc/jrOvusNzF61xQuBzBcZHnl7GYZfOQmNrYWdXMIUcUjJvTPDSS2l9kDRSeO147B+WwsAYMJtr3vbWgsOHp/BJ6O/s2TzTilXiuRIyb0zw8nv7BLskkgdg+UhX3QwemAdAGDZxlTS6uxIyb0zo5iSe3sg1Y7LA683LsAX0tFPp0dK7p0ZxTRUrz3w6sdp0rpy0FpwwKe1pNJWV0BK7p0ZTuq0ag9c+vD7O7sInR5kSAe4pSnvWeyp5d75ETtDNcVORCrLVByp3p4MOdtCixbXfsUTM72/0wyRnR+p5d6ZkTpUK47N6azURMhloqkhtdw7P1Jy78woprJMpZGmrU2GqhhyTzX3zo+U3DsQf5++BFf9cxY27WjFuq3N8SekDtWK45RbXt3ZRegSqMnZkftTy73zI9XcOxA/e2o2AOCRt5cBAJbcMCH6hNShWnFsa07rNAm65aKpIbXcOz9Sy72DUJ4jL/2AUuwc1MZZ7qm81emRknsHYUtT0Dn6wfKG6JNYGpGQYucgzjAvpqkxOj1Sci8X854FNi5MfHhDY5Dcz7xjmjcpxIiU3FPsJMTJLnPXbDMaLCk6D1JyLxePnAfcfkjiw//82iLjdhPpe4gi/hQlY/mmRuP2d5duxvWT53RwaTo34mTEtxZvwjl3VyYf/tw1W/GX1xdX5FopfKTk3kF49J3lxu1roqJmUsu9ojjqd1MC26Yv3Iiz73oD97y6KHUSSig6DKftPxC/OnPf0GM+Xru9Ivc65ZbX8OuJH1XkWil8pOS+k5GSe8dAJ+6+3avQt3sOL89Z621Lc5T7KDgMtmWl0UVdGCm5dxDGDKpHddbC6QcOVrZvjdItU3KvGAqaA3B4n1ps2N6KP0tyQGNrsaOL1WlRcBxkLUJzvuPqJNL/lKJkpOReKRRagAUvh+4mAsbt1Rc5W63y5nwRv31uLv71/gqceutrePLdFf7OlNwrBj10zzRJZ0eLZKXmm4EdG9u7WJ0WhSKDbRG+Pm4Ezjt0N/Tplmv3e6ayWGWRknulMHci8OBngS0rjLvFx6JxO5pai7hr6kJc/tgHmLN6K37wjw/8nSm5Vwz6jMqabJDcFcv9wc8Cvx/R3sXqtCg4DBnbQo/aLG44+wD0q6tq93um3F5ZpOReDkzDxxbXuZQ3a+gFx0HGsrxTLzqaE0dTPoLAU3KvGPQshqZJOorlvnRaexep02L2qi1Yv60Fq7c0edsG1Fe3+32dCsoyja0FXPLQe1ibJM3HLoqU3MuBiXRFBkdm1iiLDkPGJjS7aVR3710LAGiK0jRTci8JC9Ztw/ArJxknh+lDfqMskzpUAQCTZ60GAEyd5y9qcvO5B7b7fStJ7hM/WI1Js1bjxufnVeyaXQ3/0+R+/bNzcMeUBaWfaCJdkcHRMZM1jz4grN3CLYlhfWpRnbVUazHJfToJ1mxpxpuLOpcm/fKcdQCAiTNXBfblNXKvNsgyO1pShyoA9Krl+vodX/DncfTpXoXjRvdv1/tWUnMvuh2FZVp1ZCegteDgvWUdu6j4LkfuzfkiFq5PFn97zyuL8PtyenYTgSex3C3Cum2c3If2qkVN1o6e5deJowdOu+01nHfvmzu7GArEd2yqNl2WMWnu+uIU/6vY2lwAEXDqfgOV7bbVvkRZSc1djAKsEhjuV898hGNvnFq5Qkj4zeQ5+Oydb2D+2m3tcn0Tdjlyv+KJmTj+plewPcoibitMBC4yOIZY2yJu+A+fPwifOXgIdu/NyT16hmrnJZtNO3g64s4UvkYIJx/doZrRPduorCzQlbG1KY/uVRlYGpln7fD6nb92GxasaxtxORVkd3EpKsFyv2/aYizesKNiZZAxe9UWAMDmqO+9wtjlyP2NhRsA8CiUdkM5skzRQcYiHLx7L/zh8wfBtgjVORsNUSsDdUJyv/fVhRh+5STvd2snXG7NRBFyKGTGImNon+MwPPHuCuz/i+fbsXSlo6GxFcOvnIRpCzZ0yP22NuVRX50NbM8YzGBByCf+4VWccHPbcuUXK9i5Mk+Wqdgluxx2OXIXPXW7WpRrDVOlPVkm3HLPaJZPTdaOXvatncl9yrx1eHvxppLOuWOKmiytubXzkHuUkSZPYrrs+JGBdwFwa+/nT32Ibe056isDM1dwq++uqckT1bUFW5vzqK8xkTuvs9ED6zBmUD2Ayi7aUcmRk+h07BjLfWtzHr946sN2n6zFvJFEu95GwS63WIfoqdtjpZj/Lm/AsqWLcMZLJwV3isWsQyx3obnLqK/OYtlGczIrAO1O7l+9/x0ACRYNiUBTvogeCBJBR2LK3HXYrXeN91vniCnz1uHqf33o/bYIyBqs0CJjnXIBbfE4HUUMLQUH1dlg/YgOcfze/VFfk8FHq7dWmJArdikUE8oyt700H3+bvhQj+nWv3M0N8N5hu95FRSLLnYhOIaJ5RLSAiK407N+diKYQ0ftENJOITqt8UZNB6K7tsZjAWXdMw22T3jHvjHGoCs1dRu/uuWgrsRNqwPqIKDKUsww0thYw4qpJeNYNxwN4x/jb5+Z6zmgdX/3rOzjh5lf9UZsmzHz1/newssGP2bYsUiz3G8/hYX6MsU65xmpH+zXyRcfY+Qk/hW35FnElI1wq2VGIOotzAgsjMMlzvL14Ex57Z1nbC9dBiCV3IrIB3AHgVABjAJxPRGO0w64G8Dhj7GAA5wG4s9IFTQrxLvPttJiAFbY6UozmbrLcY6d0d0LNXX/6Svs2lm9qgsOAm1/82Ns2Y8km3DV1Ia58clbkuSKzYBxHWESKQ/V4N8Svs05/ZwmsUMYYWisU7VMoMmQzwXuJ9msTeaRZSZ28PTqKpJZyklHRufdMx49j2mAYdkbgQRLL/TAACxhjixhjrQAeBXCmdgwDUO/+3QNAMNC4gyA+gFKG10krvn9dFSiM3CM0d8YYim6cu4zeXZDc9cdfG2JNtxXyxya++cg5ASXAJkJWehciKsTILZ1o9BTFP4+8vRyjrn4Wq6QRSrnIFx2j81TEjNuW5ZO7NtJpi3ZdyaoW71KP+BFoai0q4bEd9Zo7UnNPQu5DAMjJyFe422RcA+ACIloBYDKA75guREQXEdEMIpqxfv160yFthmiT+YL5bbUUiliihTsljW/O2hYyCDm2GC7LiKGfbrmLySKh6ITkrteq0O0rd/3gexPVVu4HqHeqRGoopNhtDMXrBOQu6iSKGJ7+YCUABNp2OcgXmTHs0Sd3KJa7bByN/tlzZd/3gxUNGH7lJMxYUpqT3wQvzj2k0vb5+XP4ziPvd6pQ3kojCbmbakevkfMB/JUxNhTAaQD+TkSBazPG7mWMjWWMje3Xr1/ppU0A8TLDZJkrnpiJ8TdOVeLgk0oLDmPIIsR6FHHuhvuK4aYeW22aAq9csgPXqVyzpRnXPD27rKHxXVPV8MhKQI5ZF6OxpBKA/sHqERNclvG3CaIytplO0MGKYkUZfaK+KkFVBcdB1jAPQFSjZZH3nTkOq9jkrynz+Azjl+eua/O1REcdJbk/++Ea4/ZdhfCTkPsKALtJv4ciKLt8DcDjAMAYmw6gGkDfShSwVHjkHtLgXpvPY4VbpOFjUgnHYQw5hExCKMNyr8r41V9XHQxcemB6xy09dsWTM/HXN5bEphQQDX9APc8S2Ld7Dr99bi6AykxCMX1XQksuV5PVFQbbIsVhKNrM754zzVbe+R+6aJ9RmnvU7NxSUSgy4yQvcXebyGvLBYdVzO8i2k8llAtPlkmog5hkwEpiZ7SiJOT+DoCRRLQHEeXAHaZPa8csA3A8ABDRPuDk3j66SwzEOwqLejD1ykktQocBOQqz3MNDIYUuqcsDcn4T03T4mcvaPjxNCrGafVzEgthbleHlHTWgztvXXk7slkIxUdkE9KOCljsUyz2SADqB5f6th94DwNtuS6Fo7EQ9cgfD4zOWh67ZmwStRUfxSQgI/ZpJfxcdhuZCZci9nEClD1duwfArJwXkKF+WCZ4jc4Cpw9QXdqkkOjLSNpbcGWMFAJcCeB7AHPComNlE9CsiOsM97AcAvkFEHwB4BMCFbCeNbcS7irPG5Zea1CJkDOGWe0T6AdFY9IkzCrkbJBqLOq4KvWF9zC3FfmFNy4SbOIxw21qgwbymrFceqapaSrTc9WfQnWqWRYqmHBkt516sUHTw4cotie7fVjz34Wr8Y0awfqbMW4+9r34O594zHUWH4ZaXPsa2Zt4e5fd3xRMzce2k8hf85pZ7sFLEFsb8DtNhDJNnmeWNUiEMjFKcjk+4i9v8R5NyotIPyO3IRFMmbp+1ojLvviNTXCSaxMQYmwzuKJW3/Vz6+yMAn6ps0cqDlTBaRq7kpB01Ywy5MM1dhEJGyDIBy12SZaozQXIPjcxpB/iWXzKItAOOw8nRYcHkXAKOw1SCvWkU//ea4AeTRJZ56r8rsa25gAsOH2b8OHWnrF7vFpESDRIZC+121r9/fh7ueXURXrj8aGW00h64+EFuqZ8zdjfj/hlLN2PSrNW45aX52LC9BdeetX/J7y8K4Zq7P49AkH/RYRVb3LqiWSHdj9rUPuImOHJjzP8eZ6/agtP/+HqbyiOK8dR/V+LQ4b3bPQkbsAumH/DJPUSWcf+VyT25LBNB7hGyjGhM+sQQ2XKvNlnunUDv1SGI0yNc5od4mvLMMMYw4ieT8bN/fxjYF3V9GUKWEa/pu4/+F1e710tCCLqvwyZSyEu27vSZmY2t/L1+sILniN+wrSX2fh0BUf/6uq+VGDC3FszkLkctie+sXEKeNet93PviB8q2ouc4Jnz+nul47J1leO7DNdi4Pb7OdQNdfP/zDFkY48hdN/bWbKlcuO8jby/HfdK6ve2JXY7ck8oy8gtM2kC55h7nUDVEyyTQ3Af3ECvd+GWx4CRKXzxzRQNufqFtixLoOXmWb2rE8Csn4fX5arIqT5Yp+tZ0VIe6cD3XQv/+5tLYMrQUirjphY+VbQvXb/cmjpiGtKYPNSDLaF8+UVAGE++mZ40annrcjVP4OR06cTweSeLJn/twdewxJhQME+4A2Wnrd+imd57Esb7/k+Mx7rUvKVkYRcCCzQp4a/Em/PjJWbj4wXfxf3+bUfIziO//+dlrsUHrHPTYfEDlAF1zr/TktsUb2yfzpI5djtzjZBnx4cvWukwaU+auw8wVwZV8AN5owy13vv1v04LJncI0dxEtY1uEG84+AIAqxVhgOP6mV5RzpsxdhxFXTcIUSWM844/TcNt/FrTJavP0VPffd5fyhQUeN2i/gKq5R0UoLd/Ec+f0qg3mn3n4rWUYfuUkbHV148feWe5pp6KzmbfGt7xMQ1kT4QccqrrlbhG6aeQuDumplXN7s9qZd/RYKuyd6uGHvmTiQ8g7pYLPUDVFy/i6vnjnV/1zZvD8hGQ4xlqKOau3er+LDsPH1V/BF2b9n3LcsggyDKsfmZAbtUVYZPIWR8mdVJExOA7Dn19bhIkzV3nprSuFbc0dk5iuSycOY4yh4DBtiM3/jXPuydaF3BC++lc1mdYzH6zCgnXbcfmJoxKFQn5giHAphmnuruVuW4QeNVlM2H8Qnp210n8WA5W8uXgjHAbMWbMVx2or4zgMWN3QiNpcJn72qwbS2D3MqaWXSJ55a4oyEBa+KdfP/dP48HR1QzPqB2aV6fPi9nK4qEkqMFlVuuWoW+4WEWqrMoZjWCAbopDGdtaCPmHtWE814BWvjb0PYwx5JyRaRkz2Yr7U9YHB0Rhl6TLGUCg6Xqo52QgTHfWAHXMTX09AL63cwejvznQ92V/kODwGvi1OaR3yHXWDob3QpS33P7+2GCN/+iw2Sz1rOQ7VqMbznUfex60vzwfAX5CJ3AtFBys2cgvEpuB9xQeqD3WFvis+JNsiWNIMWMswG9Ym33rSUXQYxv12Cg697qXQ59nSmMdr84NRqj63qxcWv4oO4xkstfsWHeZ9PK3urOAF67Z7kSWChFoM70MO3+O/oyMbTD4o44eqbdM71aLDAqGn4ozuVbq9o9VHB5vuYWF5LVr4IXnEq5eXlTRrtegwMGZezETcxGHBVBoyokJif/f8PIy5eqL3Wx6BhH2H5agicgevv7OCYZ+8DGPBcSJlr1JGyA+9tTTgcG7XhYQkdGlyF5LBum0tuH/aYjQ0tnqNPKyhiBcj704SnrSjpQCHMdQgOER7e8kmbN3Bc3qYnKC+5W52qJ66/yB3Pynnm64V5cgSzxHVWX39gXfwpb+8HdrAwqriyfdW4OjfTwk4TR1Ff+X7Trj5FXz69teVbabOVg+/NPGF/Ci641vk7NGhR+3oRJQvOoFtggx00i/Xch9+5ST8/KlkTmQgnDDC0miEJQnT6+OlOesw/sapeGF2snBFz/kfMYmJITxnCwC89vGG0Od54I0lioEkP0eYnFNO+KB8Lb2zWS85xUXnqVvuUcvzJelsVjY04Y0FG/DTf32Iv2gO1I6SZbo0uYs6/u6j7+OXz3yEq/7pZ2yLG8rJ+5MM+1Zs5tkKa8jguWdAxtXibYO1HRXn/uZVx+P6z+7Pz7UooLnnDNonYA47lBt0c75o1Apnr+IjDP2DEVaz6QNbvaUJj79j1t4d5o8mdALf0VLwPl7T9ynPqmSMYZ300emW6O69awMTQFoKjjHSSX8GnYdMHY04p1bT4tsSjvrA9KUKkYRhZUMT9rhqMv753orAvjAr+E6xcIeQ0dztelsW9588K5lzVdRNVG4ZMLPDVeCSh9/DX99YYtznMCgpPBRZpoLkrjhINWnrzDumSffn+wqa5R41sS1Jecb/fgq+8Oe3/A0hPr72RNcmd7eS5rpON3mx6dhwp4SVLSy8hsZWMMZQC/VjXbetGYs37kAGfBhnklK83DKGD2Jgj2rPSuqWsxUyITgBshFkljc8n+I7uP8dHPLrFwPHiI9JD+oJIwfGGM67903MWGpeud1xmNcx6Fb98s2NiVI7MDDc/cqiwEpDC9dv90IoszbBcdQkVc35YuQIRkDPcNga4Y/R69sCQ3O+iNVuOJwpVFOH3PEeet1LGHX1s7jiiQ9Cj//I7XAnzlQJePGGHYlTY4R1zuJ5VmxOli2yECIh8nvwfx0W9GPoIaTh5K7mZ5Itd7mjrpa+s6h5KGFvQ24XUXVoGlnKQQImJCFn3VeiBnDEnl4RdHFyV3/LQ+1iSIsQp6iaO9eiTVO2xRULDoPDgFrNcj/supfx0399iKxL7jacwJC04DCcbL2NfiteiHye4/cZELDcu+UybhkZ5q/d5lk3psgU2fKZHpIjRjQ63eIVbVmQgyCLiTNXY2nEalE8zp3//cwHKjkVisGkUndOXWAoNwJ+ACLgkofew0Z39JG1LXelJJncnRCHqvpblxCiPnZ9vgEBuPjBd7HIDekUt9M7GhnN2jO3Fhw8PoMF11z3AAAgAElEQVRb5Z++/TVc/W81J7jQz3WCPOkPryRedCascxaLqSRd61a8r5xhUp0ljaZ0WevLo4q4LXu7R9w7WsyaNWNAFZnJXW6/veCHACchU9LkyqQjc1G/cj0XnBhyLyOFwHZJikkt9wTQq4iIPMIvOAwrG5pCV++RX/jyTY34xLUvRnrHW4uctGtgvp5NvuXuXdpxgMcuQO2qN3FP7haMfuXbkc+zW+/agOYujM7bXp6PE//wqjdKMRFUKQsnBBu8a/mVmPyi6DBPlnnkbXWVGkcj4807Wo3JufKOY5Rt5I8gl7GwYN12vPCRrx035YNx8UCwHnQrVHSMf/ryWDz8jU8q+6oyOrk7mDrP73iWbNiBjdtbMOInk/HDfwTDAIHoGPQPV27Fg2+q9dScd4z3zhdZLCmv2crbo945C4gc+EkX8mhs5cfrIxh+Dz/cUpbkjxvdH+eu/j3OsKdjrDVPuY4O3XLPF/K4xP43llR/QSl7b5JCJKX3OWnmary7NBiRtnpLM95Zsgl7/mQyZizZpDiid7QU0JwvYuq8YLZJ8R21ao7dKAJ2GMO6bc0l5a6XVYWOcsp3bXLXaolPg/d77k/d8B8cdt3LxnNlcrviyZnxU5KLruUOs4aalWQZr2G1bAXmPIMxr1yU6Hm6V2UUcs9avkPtvWVcFlmxmVvRJmlB6OlJIDdexlgoOcReR08toNxD7YSOu2mqsl+3tpR97v8EhHR16cPve9ua80X86/2VgXP1dqHrx6JMJ44ZgCP3VJOXBpKMaSbEL56ejXG/nQKAO5kBHosvO81KzZIoSMK0bmmc5f7Gwo3uJB1RlyqJC8d5UnlHzHg15TqSfSRycEC+6MBxqUSMPPWZswDwy2dmo+Co4cSFQit+lH3cLbtM7v78Bvl1XvLwezj7runKMowAcPcrC3HO3dMBAJ+7ezqen73W2/eFP7+Fz975Bi68/x3MXaN+I6K9y9FHE257PZK4Cw7DYde9jG89+G7oMTpkck8t9wTQq8gi8hqITFL6DDV9v4lchl85SVnVRrx8XZYRpchIsoz3QVruB5Iws2D3qowiyxwwpM7rKGRN1YKDvCET31fuezvRfQD1mU+77XVffnLLnjQ4hC/WEH4P2SLa3KiGkYp75ItOQMsmUiNUTA6+MMLS36ce+RE1wtGd3iaHqr5u7Kdvfw2/nviR16noYYpxEESiW+5AMlLetKM1tHMWcoAYHcRBPJvJcv/0/oNBBJx9yBClE2xsLYK5bzMsZUah6OD+aUsAqA7VK945xj/I8bf3QjBtgAyRKTMpPnInS+mRKqJ+9XcaterXStd/MWVefOJbMWosNTqvEujS5K5XkkVmzc20WlASLXPG0s3eRyOssRrNcrcC5M6kD4yfTCHrquqozloKmWTIJ0dhHOfy27Co+gIctf6RRNcM04Xl55+zeqsUQlqqLBMei81lmfDriXtOmrk6dqhqCs0Lkxr0vlqOOLrg8N3xzWP2DL3PPoPqcNr+A73fSfL7COlJSChNraXVodC5qw1pn5No5UWHhWruwnLfmnDijLC4TeS+e59aLL5+AkYOqEN9jT8foOgwOO7aPKaAAkD1Q4TN8rYc/9vqQcli80vlSd1EkNMUyGiK6Aznr4vueGSY3mkqyySAXklEhAZ3+COT18em5EEJPpqqjOVJA6Jn1+PcRWMW1ogFR7q2J77H3kuUXyZ3m3zNWjh4alp5rpfDG5KtfOQwNSRRQA+xE89ZqizDGFOyJMqdSdExrdLj7xfk/tBby/DW4qCOKju1cgZyl63RQ3bv6f0d7PT961x71v6orw6mQhCoyti4eoK//nspoZDiWUvNb94WWQaIDvvb5pL7tuZgGzChydXKa7LRk9eH9emGbx49wrt/nOUuyxxhayJYBd+fJWTOSkP3k4bVb1OIzwBASekITO80JfcI/GfuWgy/clIgvIsx5kkwsgWqEJb750tz4pfyki0+YdHYWqMTce1yKKT3sblyDJXwNi3FcuedypuLNnqW+7YW8VzJOoyiw7DvL57HefdOV7brFrU3rC/y0L8nDTHXxutrE4lkrVW33PelxTjDeiPRdeUyAWbLXXaWy+GOOrmXMgzOZSzFAUsxOfXl2PSWvLDczcQUFscdZbnHyTKH0MfotujZ0Ml7cpSGSZ7UEWW56zhsj97ePZmmueuQ6yRsqUqr2Bx7TFuhz4L+cJU5T7su08iQ8x3FwSS1pbJMBCZ+YJ6QsWF7qxItI1B0OMn8d7mfEOy+aYv10wOQdUXROC0wvFI8wNvOGzNDxk07YJPjx6C7hSFDjvcwnLSPv7as7RLLefe+6TVKR0g9CXV80ZDeW6YmQ9MtFlmz/cNLHysRImHI2tzHIROKrFU6jiqdTKr6KW7L3aHsDwNBI3fDZC7ZJxKVn7+UbylrW0qYX5zl/v3H/fh1IaGEOePCJBax5KMptvyL8kQYA/5ZdQ32ePmbIBB6YSu671ii7Jffx9qt8alrd5RA7j1aVqE/NnNZxrPc49MlhOVnooLf+SQh98mzVhtTaQj07R7Mr6TXcJgvwuQQFng0ZEKfCVWGdpta7mVg4w6/cegWzK0vzcdZd0zzhqlJkC86XmsQPbkFBxtQj+vyX3B/M2UIacGRUooK52Tyt3mDO1sV8MkdkGKM3QIlXTx7c6N5CBmw3KVoi80Jh501WRuOFjYmfxQX/OUtNEZYQJEWDOnRMkHiWyXl2ZadpLrDtCTLXSP3UnLqt8TElIdZ4bItEOYjqTessSuDCJhS9QOc/9ZnlO3bSyR3T5aJI/dremDsv8fj7epLUHAcT5aRZ2jLOfxlEg2TXGzHb3fZsOUsJXz7ofewJGIOhukZkr7NKMtduV5M2zKtaJVa7mWgJR8ixYAnsyoVX/vbDGlRBFdTJz4EdaTGnJGsDFsOhSznJSoWeVBLFcPfQrGIuWu2xqZOOOL6/xi3h2UbzBfDUx7oqM1lArKMrrG/GjECiNP3ZcvdpLkvWCtPdPG36x9cKfm4c5nklrtuacctBxjuAGbev2F1EpflkwjoaXBCNrYWPQt2e8jEIhktITH3UXAYAqGQgJ/Df9aKLbhRWm/AlKID0GUZtazlLHFoeoakbSFpOGvRYVi/rSX0eH12NJDOUC0LsmWkv8TuMZZPHEQEBIHBYSQ1ZkdpiEq0TBvJfaOkKevDRALDKbe8Fogdj4Ks+eoOZVFfG3e0IGeHf9gPfs2f9FOTs+E46sLGuiQRNVKKWtSBoA6hTZr720skJ2xEls9SXkPOtpQPMspy12dpxpJ7jOXusPBzTeReJTn3wwZyja1Fz4JNEgklJEXTSCkMXJYR0TKaQ7c5j9P/+Lon8+07uD5Uusk44bKMSERXCkzOzHwxmNJD4Ovj9vD+jpJlZKxsaMKh172E8+6djuNumopTbnlVkQtNb7M9F+CW0TXJPaTdtYRkmMvZliGVa2loyvuJwRz45M4td1WWKWiyTGnwz9mrf610/6KyX1hIUakBdMhEK4imKmOhOmt5UsbMFVuMlvvwPrV47KLDMaRXjbetJmt7CxsEyxmPKMudNNHdRO4yonJ3lDIM7l6dUTICRlnuepmunzwHk2etLtlyZ5LlHibd3Hrewd7fYhRTJWnXYbH7Ta0F1GZFCgvjIQryRQdZm4zpl6UCKz/tYrNRlgGAo343Rfl90dEjQutUfp5KOFRN6xKLlMbG47M2bj73QADhM2x1zHTz2X+wYgsWrd+BuWu24cgb/NGyqUNNOlu4reia5G5A3+45pbctOg56uAsv1FVn0K0q+TDThEbJoerAQlGyVALkXiFZZq++tfj2eB6TLZzBwjIKG9pGYYs0iUh0QDnbUpyiq7c0G8n9qJH98MkRfRQnc23ODjhUt5eQzjRuiKxY7ploS3JY727e3zqZl5KWoXtVRrHco8hdt9zfWrwJ337ovcQLbAgMbpyHS+1/wXFYaGjebr1r8eNTRgPwrWq53b340VrjeTtai6itSm65F4qOUUpQUFQdotXODjhkDoVs0Cau1eYyoaOhamkkErriWQmoMoWWOizSchZhshu3J/M7BUN9tfsZ3mdjazF0hbNKokuvxCSjZ20OG9wXUpO1eboAlzxaiw66tdVyd8mdXMudSdEBGUWWcSRZpoweWiYixgLWoSCbchbPlqdACws7l7GwvbXgz+wtOkYPv9gmf/diuN8gOW23tSRfZSaO3OW9Js0dAM47dDectv8gFB2GSe6U9GAoJP/32rP2iy1TLmOBSR9/VJcSlvZ2SciycGGyzGULvw5kgVvZjxOFoGYzFtBaVNpdFIQMkWQOQ77I4iUZR33H/bvZcAr8HhRjdNhWeERNpS13kXRPRqHIQjvfbc15zwG6PkHYKBA/GzmsznvVlrZSWjnYZSx3WV+rERalSI9bdELJISn8aBmD5U4VlGXkDoE5ASva0mSZUiBPrvnOIzxHS9a2wJjvrygUmbGuhBUkTwgSQ/dVW5o9QqjkQgSypRm28k911sbRo/opnbfOoYwxHLt3P1xw+LBE95XDTMOICDCvHAUAC0Oc93HDcYcxY/733519gPJbWNZhk4F01GT9zKJx4LJMaZb7jWePiZ3EJDC0Vy2skLkDVeQbCfI3FYaIlPIAeKd29wWHKNui3sG25oJXt0nbcdw7DZssOWpA90TXbwu6JLmbhjryULI2Z7vDL+Ydn9RLPvWH45XfvzpzX3TL2ZosQ160zC+zf1OsjIpGy7BgpyTIpi9txZLqL+BIK/lqP8Z6c0lZNNJWwypFgB95oIQJSoeJ+t/alNxyLxQdHLhbT3zvhJGBfYs37FCGxnpCLwEx8UeW3fRombj83AGwZJZ7mMwhj5BG0zIcYc0GECSClkIRs6VJNMwxL7CiJ2bzZZlwAhITjIDSLPdCMThaDMBR79uvNuNHjxmWmZTRszaLk/bpb9wnyzJJLPfe3aoi99uWheF9uynbhKVtGp0WHKaELsppKMJgIvdB2Oh9p6Y6/9b4PbF779rA9kqjS5L7sk1BJ6I8lKzN2Xjl4/VexctEH4fdetfic58Y6v3ed3A96muynixTZfOp1sKheor9ToUdqhIMlrtusZ9rT018OVMHJ3jQW8g6JN2pKIdMkj2kxaTLsdzzhQLqqjJG4m5ozHsLZACAHSIViFO7V6m5TmQUnXAr2whp0llVxBcSRoINErk/V3UlHsldB0CVZfJFB5c98j4m3OZHgTDmYHNjHqMH1mHCAYO87fptxH2jZBmZvGq9aJlklrspNlvB0mnq72IeDoQsExP3bVkIi7Q92PJz/Zs0dz0csmctb39hr9a2gouKiHh7eTbwgPoqfPPoEbj60/soPNK/rlpp4ybomntfbMH06u/gisyjAMx1/uNTRpfWHstElyP3TTtalZmmAnJMa42ktQlLsyVhFIdtEW4850DpN4979np8m1CEBYf5L0cldynioSzNXbXcg5p7+TA5FgXhtEphfKaOUJCF+KAANexRlHOrgdzDJJV8ocDXjY0bXyPcchfD3lrpnZsWiY4zRj/3iaE466DB7gX854oK9zY57ADVcpchR8Jc9MCMQLIqx3HQ0NiK3t1yip4vCGpQj2oAwLA+3OqTQ3B1rVsld14389duw/ArJ+Hrfwsm0vPK6MRY7qveBx7/srpt85LEsoxtEYRvfFL1p5V9BxNfiH4962G03H+mrUnbuzaHH58yGlN+MD7kXlZAujHl8SEQrjptH/Svq1ZSGVdnbXRztuMXmb8pYacydMu9hrhB8mn7TQCl52qqJLqcQ1UM+z+5R28l2ZQcu5qTet/qjIUdrcXA6jhJkbEIGYv8SQrMUTR3AFqcu+P35mXJMrJD1aS5lx9GZZIRBOHIFsgNz84NHCc6z6xt4ddn7ottLQVMX+iv9uSTe5DYcraFJkNmzGKxgIxFiRaf1juIb43fE3dNXeg5x2TL3RQKGSfLyB26YrlHkHuYH8e8bqqa/tiULnb+6gasbGjFwbv3VGRG8exnHjQY9TUZFB1g9vxFeLbqKv8YrV3I7aY6a4EI+Pd/VwGIzquULzjRDtX5waUb8cjncVRIOXTw74m/oJerjseE5ol+OSkPp+cwLN1UZSR3PYKFiLcDGafsOxDPuYuB87alPkuTR+7+i5UJWO5UszbhEjyKL2aexxy2Ox4vHhsok+5QFZMM68HVBVlzP3pUP/zlK2MD12gvdDnLXUyn1r3NwqGWy1gKUYmIjqbWYskr2APcarIs8joHYiJaxq86Wfu0yJFeeNs1d/lD61GTLStKRsCkuYtVieIcQzJZfOmI4fj2+L2Uqe1iKG/Kgx1m4dagBaPycwLE+0lJLxbQjxEha2LkIVtipmiZkjR3qSPqHjENX75L3OVr0RIbNvfa/HU8FNVWk5eJUQsR4bjRA5CxyVvxyDtGJ3ep48llrNCRj46CExMK2RKdNEtPrBfYbxEyrkO1iamSRxVaAbKRZxmMs2fjm/Yzyv6NWgSL6Z3e/aVPeKkabIsCx3hJ2qQh2U8njPb+liWprG2hF+NSUDMz6/v6dyOev55ccpc6Dpvi52tUEl2O3IVjU5YHvj1+T89yr8pYilwgLM6WQjG0gV/kpi41IWNzy93XzrjmrljupFnu+bZY7uEO1fsuPLRN5G7S/4TlGxfSZdLh9+jjO6uEJWRKmmVyXgHAb7N/wo9Xfgfdm9co200frW6517hkLjonIsKcX52Csw4aHHhOvoi3sQhmSO+ge4TpLssspgkzMuqxI7YDFQRtW5ZCMopsVcxj/9cvRX9SpUldf5ct06qMHai/8ORmzJikzUNMRcaFZ8qyTJOjCgc5FEBWBnlXv78qq65ZIJKa1bmGXFgfJJ7VIgrIMkKeFQbHeYfuhs8c7PvY5I4tYxPqXAt8K8wOUL3D1p9fbosdLdB0OXIXlqEYhk/YfxCuOGW0Z7lXZSyltxQWXXPegWURnvzWkcr1Pn3AIFx16miEwbZI0eGIFZVoGQCBOHdflilHQjHLMr275bDv4PrYOOIoRGnu5UzGuPYz+3l51Gs8cg9eJyxXzQHWIgBAdVF1kJv4Q9fl9xvSg19jaA9vW03ORsYN7ZRhWtA5EpLlXhfhURVLIALmqe4y6qkxduENb4KaBaPlDgBYNwd9lz+Pa7P3K+fqFrOcNEtPYwyEv+9C0UE2sq6i6zFOlrGJYFvCclfJvQp5wLLRGqMW93ANu7DRmNiezQQtd09zD+mM5XrK2RbqwOcsNCa23NXfBcccWtwRSHRXIjqFiOYR0QIiujLkmHOJ6CMimk1ED1e2mD6EFFArNFbt/Y4aUKd4uGWLMmMRdpOmzwPA8D7dIj3XQnP3wBzuUFVkGdmxxbgVvHom0LgRJUOx3JlHmvmig+qs3TaHqsFyH783TzEcS+4Gvb42l8G4kfx8QW7hzu7gvT3/gfZQpgGPPuoaO7w3Xvr+MYHYdXk1LoGSZRnpHQgnpgmK5e6+p5H9zfHL3dHkjTLCQB65W4pBoUQKOWZnbUYhFaZM4NGToell17dHSgdttNwtSZZpZKq0mqUiYNkoInoUJMgy7LsV26tsK2AUCONDzHjW25o8YspYvuUe9tRxljtQWp6eSiKW3InIBnAHgFMBjAFwPhGN0Y4ZCeAqAJ9ijO0L4HvtUFYAsuWuNgAxtX7CAYPQ00TuBQc2kZ9r3cV3DTHWMrjlzl9On245EHOUrJAAAnHurQUHuOco4M/Hl/p4AVlmRD9OFiLEcHB9+TPbdOv7oa9/Eme6ESJhksE5blho2H5h5UXptDnbMobICXLXiVdfT5VfP/iB7NW/e+ADty0yLtZRmizjf6DnHzrEqwOBZ797FPbs102xxIWlbFpwA+DPGme5D6EN2I8WcYNCIgSlYyuaQ02Hku+gtcC8lAMAH81mNMIOe5/5IosOhYyRGpOkxRCXb3UsIKMaW2TFx3iIkWB4KXkZs3YwWuYxd9p/WAcmb89mLNRRk3tP8510OdP0/JEyVzsiyV0PA7CAMbaIMdYK4FEAZ2rHfAPAHYyxzQDAGItf5qhMCMtdTycgws961uRwhghpg29RvvrxemxrKWBgfbUSQxzn4LAly32oa/XzrJDSx+e+0CIjVZYpB1q0zIB6PhwU4W/3aDPuBI4b3R+XHBu+NigQtGj711WpkpPha7n0uL2we+9anH7g4OBOCVGE0JQvGhu92KYTtIk/YmOvXRAZyN0pcRKT7FCtyuArRw5Xdg/qUY3de9cq1u/Aem7hhy2sbGnRMiZMrvoJJlZdrRgUgOZvCLHcn676mXQvR7HcqwyWe1hZCk7MbO6YhWeS5GEXlnvBISCrjYysaKsdkOdcmPcLWTaXsYzvfSitQ84tg25IyIEaWdvy5K4fnTjKOPEomeXeecl9CAA5y80Kd5uMUQBGEdE0InqTiE4xXYiILiKiGUQ0Y/36+JV+TBjWpxtO239gIG3n2a519YlhvfCZg4d6Fqn+cm2L8MfzD4aOy44fiUuP3Yv/YAyjaLl3vBja1bv6q6M7VN0X2oIcLDhoybdhCr5muRMRXrz8aDxx8ZHB/RKOGNEnMlUvELQyuldnFIvY5Pgc1qcbXr3iWAyoN8sTor8Iy7MC8ElnJkdwb+LT9C1L3WfiYTGCiYNNZAiFDI+TN4LJPpTgpC4RhSH7Fwb35B2/Kc4f4OsAJM0GaGtSoNKGi/EzgC0wRXOvMkTLmGSZJRt24MOVW6M70pjkY2I09sLlR4ceIxahaXUAZDXCtDJgMeKjCIeuC1kLV4xQs7YVaEsDsAmvV30P5zT8yXhuLmNhb3dN4Kztr2k8sn8tHrno8MDxOrnLvg8xou/MmruppvUvNQNgJIDxAM4H8Gci6hk4ibF7GWNjGWNj+/Xrp+9OhBPHDMCdX/yE1+BJ2r7khgkY6GqkwpIyOQKFpXj0KL8M3z9xFH548t78x7v344WqH+MIa7Y7o44fL1LC962rQdb2LSMRCtmCLGw4KLbGr3gTDtVyB4CRA+rQr64quL9E6Dmq66qzikWnSwpxTkLAj6KxI2SZosMwhDaE7rel6ez11RmjtTV6YF1gmwkmzb3IWGhkhREKgQUTTZkmXglyD8s+asFBS5Kcu0CM5R5vOBDU5F+y5i7asml0Of7GqQBiLM0Yy12MxnpGzOy0PcsdQFaVZUB2bAsfPbAeAPA1Kf/6recdhNtdo034h0yWez83ymh003/Dy+fWkVwP1RnLuGyfcNAeuBunu4yUfuGe7M0AwqN62htJbrsCwG7S76EAVhmOeYoxlmeMLQYwD5zs2w2iN6wJ0TgFUYVZIR/84iT8+cshEwo28SiOA2mh8qFVuf9++VMjcO9XDvUOF0mOBLk7+TaQOwuSu7rfTBAM8bqyTu61WVtZQUaPIJj761OjLwg/j0uU5T4qswZTqn4Qul+emPXMd8YZn6M6a+OVH42PLY9l0txLlWWevtT/mzmB5E+2RQFLuCZr43efOwB/++ph5nKBKdE1UbAtUohF4dpElrujPG/Otr3vQMiZUfp/JLkbJqPJENQctZKTIPe8Q0FyTyDLjBvZF/OvO9UjVAA486AhnnQoOvecrZL7n7881nM8Mwq/j5e7x7I845HgGJ9pgzth7YmLj+DFl9rycTbvQEQZOmrtVIEk5P4OgJFEtAcR5QCcB+Bp7Zh/AzgWAIioL7hMs6iSBdVx4pgBuPTYvXD1hDHG/UK2CRsS9ajJhi8n130AAKA/NSjRMl6ADqkOQiHLtLIMX6mpLZa7JstE7kdpDUdfSMOyCKu2+KvGhE02ioK4bVjuFwD4VO+tkdewXWtwn0H1GNanm7J2qoxhfboZt8uwiAIrPBUZi+x8Alg23f+bsYCVbhMFrDEi4NyxuymJqsTwHhAO1WQpMDKa5a50TCGauwwLauhnVda33EUIcVTkTmRdxVjugtwooikJiiwwAJZm4Scgd73z0yE0dx4K6W8/elQ/r3yC3I2RWe5JuQz5eXBCPrB1LrmLzl7X3I+13i/NsKggYr9mxlgBwKUAngcwB8DjjLHZRPQrIjrDPex5ABuJ6CMAUwD8iDFWRhxgcmRsCz88eW8v5lWHN6mpDMJCTS8AfAgnx7l7I26ylJedkTT33Wkdiq3JV0cKoGTLnQx/mWFa5/HYvf0MfXETcUxwEljuViE6N7Yeoy1/C3df8Al88POTEpeHR8uo24rFIEEnBnMwdlgvb6EMcY8kH6zsrLPkyW0xsEjV3BVZphC/iIQFpows5Bmv3SXLPWxx58jojhjLPUl6DNvix/DINa0MEdEyp1lv8oyLMVUvHitn28p7z2UsTzZiEZ2IiCzKyiu4uRd99rtHecfVV2e8jkTcRw8cuD/3+7JmxlcCiZiPMTaZMTaKMbYnY+w6d9vPGWNPu38zxtj3GWNjGGP7M8Yebc9CJ4GQZcryVLsEWo28S+58c5WYWkeWQrJCcx9prcQwax0OaXgu4toxJrYW5x53/tCe3Lfwqb36lizLANxaPnU/ntq0nI7Qd6iGn8uK0YQkyN1U/IE9qkM7cBOIzCsxleRQVcBAREoOEyIKRJ8YHVNSMXphO/67bHOiO0Za7oX4USHBCZKa+36ET+DG5+dhj6smG9exjZzEFGO55yxxDQtfCsmf71nuDgU7C7JDHKoMd+Zuw1NVP0tsCWftYCcsJNQksgzBd6iK71LudPt2534w+V2ZomXKb3ttQ5eboZoUJ+83EIcN743LTxhV+sme04qTgiCuKhHVoZG7vlL7Xs2zwq8tf/FbddcFSpZl+nSvwpIbJmC/IT1i04iKVX5uPOdAZVUiYZ30LGN1mESWuxNN7hZTnYTyc9SXuLC5RYSWgmqVDnVW4ZCNupKYECE+Dp00TFUv0+Yfcndh3KYnvN++g9xwbSvKclfJ/YyWXwfP1yx3HgrJ/+7uRph84K79OUtLowsgEBOvICZa5oTRfTDpsnGoydmhaT18zRZOy2QAACAASURBVJ0F9ZsQy11EW/WnhsTkzh2q+nVCNPfmrV7eHPG95x3Hf4luO5A7zU/t1ReA6sA3hfx2Wlmmq6K+OovHLz4Cw/t2w35D6ks72RGWJB/Oi4+rWujKlq189GOspcrpVpTTS5y3cSFw8z7AqzfqBwSPNZ3vopy80GcfMkSZ2SmcbMP7lL6AgODQqPA5irPcmYh357/lD7Jei7r43gkj8YfPH4gwiPkO901b4m170v4JTlt8fSwxGREy0ErywZ62n7rYw0nWu97fe/br5jnhdGSIqTNUFXJXJa4VrB+uy39BLZumucuJw/TJf6samqAjMlWDbrnn6oDjf+79HFSXw76De0RexyZ3rggjoFtfdaelBs2KGaKy3JM0+kR3qALSTF69E7lhN+C3PPpGWO5qpF3QiBlpWE3JZLl3almmS6PQgomHvB+5ak0AGjn7DhZ3g2a5n2FPV46nqHA18XGI7HpTr9f2x1nuekx4uMxz/Gjzijd6hyAs9/qQuOEoVGfEcD/cwp4wpk/kNSz33YhiyaXTy/S9E0YpiZ50LHXXL5040x8ViVmGiOlkjAix3HXj1uQE/p42arSkMDnGwmezZkhdMEPhyKJK7gVttjTAyd2yyJsAZ1vkxd8P6qFGp5ScblyXUY65wgtAANS88uG5X5hf5lqd3NV2dGWGJw+TKb8Uy10/1PPvCIeqvNN1VovvnafIVmUZ+d7GBHdGck8t9/bBtNuAF3+G8+wp6vYZ9wF/Odl8jkvO+isReiKIIpOCWSyB5S4ajd4RaAtkh5/v3ktZ01Q99OxPhJOgDDHhxWEM15xujj4Kw7fG74XLjtsLF3zSrK8CwEGDakL3Ab7lLiB/DKERTSEQaRpEp6Boygn06iDM7JeEYHQnrgUHR+7Zx7tqmGWbJQf7DAwZbWqjDz73V1/QhafaePgbh+PCI4djQF21t4i5vuycKZlcJGTL/WsvAkdc6hElL5+/vyYkXbItk/u47wFZqUyaXFJDLe4zlU7uWZPl7nawNdVcgtx3cLCev+i25QOG9vS/QSE/ymkhDO8vY5Bl+rsSXNI1fCuFXZ/cm7gTq1pfSWXi5cDyN83neOQuJlu4sbtKtIz6Eu8u+KvK2EnIPUwiiLPcNbLpEaGTJ/1uxTA0X3Rw4af2SHaSi5qcje+ftLfijH3p+9rsxJhoGQtqfYlP5vzDdi+pLACw3SX37q5Wr5BXTDkAAB8/r/4O09x1h6r0UzjadPSusXHWwf7k7jA/RZYc7D+0h9nfoMkiBdjKbGlAWO7Anv2645oz9oVlEYa4qTP20MJJTQ7VSMiW+26HcY1E1s2l+upRk8XFxwRTYthg/noI/fcBfuQvr6db7p89aJD7TP51k2b4NJF7zr1Or+61mHzZUbhQSy0B8JDJJTdM8Camyc9V6y42fvqBg42OUtuwsHePmiyW3DABJ44ZENjXntg1yd0pAgtdS90lav0DiD0fwJF78kUjxApGuQhyb4FPsnZURIE4L8zyL9GhGrWYRFJ4DiRXY7z1vIPwa8nhmgTyR7RX/zrM/qU0KioEdV0Znubu0rq41LF7lz6LeVuLsNxdcpfJa4FhFSEdD5+r/g55T/pnLcfgv/7jYzH318EMHDYcn9BZsIPwjnOtS6PT1QmSu2656w5VALj9/ENw9iFDcegevRRyNGUKjaROU31YZssdAI7YMyjJWXBUKUnuHPT0A4zhV2fuqzgqk0a1dquyA8eKtRfItjFmcD0fJeaj2qcqy/SozeLlHxyDWz5/kPH9mRyq+vySjsKuSe5TrgP+fhaw7K0yyZ1bkkJP9me8ufvJBgaqTr1mKX1p5CLBHrmHvXBzRE7gfAN03deUXdEEEdcs8o2cedCQ0DC2MOjtXNHgY2Kjiek+Cn6xcib0nbwvt45EAiglLPKpS0q/YMjwR+bEm889ULHKqrO2UU+3UPSIlSF8YlWWwt+x3m709NP8Pk7Aut2jbzfcdO6BqMrYqJNGBCZZJrLeTe9S7ki09mlKd0tgEeRuYQ1TV+L68hHD8dmD/WR/SWWZuupsQO/OuYnNlOyTTcE01Si08NF963Z3g18re/brroRIyzCSuyEEuSOwa5L7gpfdP5j3MXxz/Cg8c+m44LEmeUTo4O6/IuFTjRzn3ncv4LirvVOa4Tv+ItOeerJMyAsXHxvZCck9XHNnDHjp+8fgpnPCo0sAf83ZsBzfSVBShIV+rpcdkv/2jNsy2P0Xp+8LwNfqf/n0R6VfRIZU35MuG4e7vsizcsp1/YlhvRJdymI+6TIWTlK6DyKsPHlmA1oSO1626ElbcsbIkmUZ07uUdXJtv2mGODGmdkiy5W9l8PvC5+Ub8v+Xypmc3IOyVpVnuUv7PAKXMOcZ7pfzihGsJ1M5TNEyqeVeSWxewv+1Mh5BD5n2U+wvrdrjwdRYBbm7ubM3uAvz9uvmEriwNEb7OvseA32vvx1leW1ZATRuipBlmF92vUH940LgsQv0E0JvxcBznh+5V3S0ipjopSfIKgWREQExlnv/NVMBaTBOHrmXXh7xLIK0RP7usiG9p30H98Cp+3MLUs6gGZYuQQcxx5NLohyqzG1/e7vJ0mpy5lGQIHU9WsaGEzlxRnZ0miz3yKcxWu5W6H55EuF9F7q5nJgDBvLTSMvnk61InKaRbtJWYZrAeNLevCO2ZXI3RVHZmi/L8L2a3p8pWia13CuJVh4OB6egWuZRVroMsSCCK8+s28ajLPrXuS+cJGfQ3hP4LclvDDkRdmlKsHH3OODmMeHWrGhEVibYoGb/y3yOC72pCXKUradbzzsocJ6YtBK3mEQUTA39HxcfgYnfGRdL7gNXvogTpRhwaoMsA3BHZclRIGEI6YTlJFKm9WVNkOUSFpHvpljg7e73nzsQD339kxhicOwBQN5djs5hBs090nKXy56o6NL9De/SCrfcZYI9brQrXTEHdTU53Pp5ty2SG/MOBOPP3bol6bk3N5YR0upi3IgebpFlcjcEQOS0OR+Gd2x6f6ZombD1atsbpU3/62pwCmpjcwrAjs1ApkrdZjoP8F5696oM1qIF/buLzGHSx6RntQNQDTcqI0xaKTQBHz5pLrNH7iGaewDhsoyAHE545kF6Kn5fltGzH5YCk6V46HBXO/1vfOOukaKZqA2yDMAdlYWSWSsEIe9AToectCORNXcg3KFaKPgL0ohZkB6kjnILuBM3GC3jREoXMuG2KVpGQImWUa+XyxjKwRyefE96fgfEY9kDOV9UhyYAbNxePrl7VrosJZnIXVshytQOksoySReaqTR2fXKXydspADeNAqp7qNtM5wGe5f63/zsM7y1rQJXtTjyyguSeZX6YnUdUUVLFew+E7JBlmdLIdqA2QcVLoJSxcKT1IWY4e6snbF0N1A0MRMuUg0gZNEEO8g2o9y7ikXuZtrttyAxZNkKIW3aY1kVM4JJBsuaO8FDIYshSevxEnzw+dHjYqi7LEBiy+S3A8rnA0EMDL0cmJVO0TCREm+wnLSofEucOhOR2Yk5gVOs9g07u7v2YNOo+aqTW4WkY0a8bVmwOiYARdassgGLqLPR6CdaTaXSUM6xE9ecvHxrY1hHYNWUZeYKQQu5uD90s5dMwfUia5j60Vy3OOHCw37AVy50P37JOC7Yw928R6xqRnMiIj54CNi/1zy2R3E8/YBDuv/DQwJJ4ufUf4uHcb/DTzIP+xqVvADePBmb/KxAtUw4iHaoma6/fPuohCGrY5VrumQ6w3IXmfviI3ugfskqVDlWWCbfcnULEPAmpLnffi9ehKRRyyIvfBv5yIvB2cMUhJYOwqZIjO+oisMfRwCVvScfLlnu55O7+1r8Zt3xMeg/7DTH4ziS8ePkx+OiXIRMURfkUP4GhvgOTCw2Wu4ncEbzW7mWk9agEdlFyd+EUVWIxkUwCyx2zngBWvmsm9098BQDwfvXhOL7lpraV9/EvA89fxf8uw3InIhw7ur+3ALGwfGkHXwVpBK32D571D/5vwzKMHdYLYwbV44qTR6NcREYwmJ5jnLqGOpkcquWWxfIt96yediJG//cwYrxbCPPxQnMf1js+x7xXLs1yD1trIKnlPmYonwdgmsRUvWYG/zH/hWA5Yiz3HhGrKIEVgwQsj2TLttzd3yGJw6iEb8G2KDz5mSif3KmZZBm9nZgcqoY2n0MBLFMDNtS8aEtHYtckdyZb7prmriOB5o4nvwb86TgzuQ/YF7hmC9ZmB2MDeqDJlpaDiwkBVO+pHSvI/b+PAFtWJr8O4K132r3K/UiF9SObZO5qU+jWF92qMpj83aPM0UQJETmxxFTH2sdN0kpSIvKmnGgZQHWo1rj+j6Yqd0JUXAqCXHdg2DjgNDehWwipjB3Ooy6ODcnfY4KFok8IjKE6axtXAzM55TzI5bG57yiYW8YBxDUMz6tMYpLqWCwj942jzNkc+c2KQelEt4InXg5smA8gpANjTkAq8koRkGXEngo5JcU3Kdejidz19/70d4CZjyubTKPVLApAtgY04pi2lrTN+N/T3E3HhG1zCsBrN0vbDUM6F4KIHEo+eUeB3sAsC2huAP59MZcwLglJlWDA908ahZED6nDCPi7xuA1VsfAa3bVUkkzJTwBByMeMMswqjXTCEQCmJIYSn0xbHKrCIhXk3prriZqW9UC+GchFWNtOERhysO8oDyH3UQPqMP+6U0taL4CBFMsdAE4wTEn/7EER09TliC87i8tPGIXqRQuVhS8JDCQqz/B+FVlGstwZA774yd3j11DVLXf59/p5wOJXuex3yVvIGh2qwVS/VRmL87dO7luW8QVKKiWziW9bIXdJcxedl4kXZv0DOMCfwWwareaQB2Wqgk7mnZA8bNcndz1aJnBMVJx7Hnj5l/52z3IP19IdeVhZkuWulc3K+CGd29cmvw64ZPA5OWmYWw5Fm91RWXIH+LR7Y14VuR5O/g2wz+lc5gLcuQh5JXfI1RP2QcYmnKKlzE0KmyRydxNPteTc9TbzOwBExP0L8hIfZ4QcUOpCMAyWormrf/iojnLVyHVp5/DdE0YC8xYCj/ibLWnyXpzlLnNmouUIHSfachftP89T9ZrrKEjuGZHdVP+21swCJn4PzDlRK0OZooMny0j1KBtWxbxL7jFRQYhwqNo5zclcAOzSM662FV1PlmEM+OAxwNWRQw7i/wQ094SWu1gmT3e0eLNHg9Um8o47VukLXhjvZWV8iyLBupKRcAlKGb67CdX0FLJtwdBeteY0tvI7GHgA0HN3v/G7ZGBJolH/+mrcfO5BoSlx42BbBOEbrnUt9409D+Abpt0WfbKw3BKQeyzWqIu2OPBXBmKGED+lDFHlExAhvVp7tOD41zV03mGae7Go5pE3ghWD7V9un1tXKvc1dhYGWUaevBdIWTB3opJKuNR3Ul+d8dNpeJOipF5N/vY8yz5kJm7jJu+nqd/KocDfi1yPCRY1bw90PXJfMQP410XAiz+PP9YpBHtl0zEyWhuBxa+4x+secyHLBBvsb87aHz89bR9UVYWvsBMJ/V5kcwlB/N0WiHAyRZsVw/Y2xAwnhUxIghjEvy65R+bjKRGc3Pkzi2yga3uPBWr7AHMnRp8snH2VIPd/flPbQMGmk9Tg8MonW+6uNagR8jBa5zsgDZa7zLc1Lev5wjHg2U9jY7LjNHcBt+0aZy4bHKpyCPCMq09UdxULqkO1lBExgJnXnOwnwouVZVyOMHWwK94GfrcHMHcyL2qILBO03FNyTwZBvFG6qexQlRt33rBwtV7xrdv5y85UBxuReOEGS7pHbRbfOHoEqFwiDsgytp9NMWop+SQwae6ijuIsd8bKF7+9a0j1KOpO+9eCU7FFDWyLIEL2hSzTalUD+5wRQ5wMXDKQLPdyVm8SyKohko67Kqd3K6B0cpdJxxaWu9rmbsv90f8RY7lfPOM04HaeL6fosPi8LUbN3dA+oxzXJnL3LHc7GK1TbFVCIUvyZekQ5zphskwhuF9gx3r+r2sgCFlGrrIcXFlG5oio6Kd2RNcj92Gf4v/WJdBjnYLauE2pPcNCnjLVhljXBPHr5WqBO9Zp15Hi3CPJPQHxOkJzlz1p7rPFae73nwr8smf8PRLcH4BkuYfLMm2FbLmLUMgC5cK1VL2clZJlsmp8M5Msd5/cTT6fBCmjAd9yj2ofRss9ZPJUIs29EAxXNN2/2MIT+C1/O7gv0nI3yXp59blLtNzV28REywhjz3QP8T638ZBiUY9yfeaQ57KMMgO2ctJnKeh65L6bGz8a2Ru6DWXi5cC62f5mI7mHxEHbufB9ERo4laOPb14C3HO0uk3+gKI6jCRWjCfLCMKSHW4xDW/Z9Oj9SSDXo0fu6oxECwz7t7wPbNc6uTIgO1TF4uVF2G54aQLiLEWWee4nwH2nmvdVa6GlJlI1WelJygj4mntUmzO8X9si7E3L0AN+NkQ29bcoOtE5aQBwItSTaoV1Lg9+lk+kAnCgHGYbZbmHGU5tsdzzTcDDnwc2LDBr6rIsU4yQZUTZtvEAB89ylw7Jkus8ld9JWSuAtR1dL1rG4mlOy9KxkpC7Z7lXBT9s0SBCJlr45SsRIoulch3pHlEjhSSWpfvhfHKEO21bGZJ2gFXBDJa7FnlEYPhFw0+A+/4CXPZem24nO1RFro8CbH6vSKtYCnVNSu5v3hG+T2snjmy5e/dMmMzO26dGywAIbx9WhhOLForXu1sOz1ddiUWOP/qlqb9BFg/EW+7F1mDkR0yb/+DnJykrdcVp7iaomnvC0VS+md9n8avAx8/x88SaraGWe4RD1TOIOI8Iclct9wKXy+R3UsGItFLQ9Sx3gDeucjzQichdWO6G0CUnXpYpS3M3TgGXrrN5MbBkWsi5CRq6+4z96muU3wA6yKEqlVEnTTc7n5cDf9PCNt9OlmUynuVuRcsyq973R0+VkmVatwODD/F+MkVzl/xCOhI7VLUspTqytQBY4Fu56jQ+E3mEtUbZ3hPbI/PAA3Atd+3biPEJ9ajNqpFPRstdJMxzj/vWG+b9+t9RuG4AcG0/f3UtK2uWZRxD0IVxROUo+zxSVzT3fDBaRnSw1/QAXmvjLPYS0DXJ3cqGfwBRlpnRoRpiuetDT/naETIJRVn1YYhbugwA/npa8nMB3pjee4B3aHomPIXcO2DIKN9PPJec2hhARk8T0AZYrkPVcZhH7mNH9A+fnAIAL/wM2PAx/1uJc2+DvtuyDajyZyz36ha/bF7oNtM+0UbDLGcxEUsbndXmzG20F21LYLmbZJkSDRrDJCYP4vsZsC/Qa7h0TgUcqvL7lw0OMZ8EiI6WEdtcSdi33P1DuOWeVZ+v0OKX/+VflVf2MtA1yd3OmjO5bV0NPHBmcLuwnkzkHgh3ZP49dCRyqFbIck96nTBy//h5PmX65V/5DdaLAJGe2ZgRr8JIIMtkDQsLl4uMm1sm7zjIuNcd3q9HtOYuvwPLDnZCcTBF1eQblZTQNRkLtW4u9YE93EiapBPrvHIWg397FqTWZsS9dVkgJPqpF7bHa+6OwXJPkPVTvb8hzl0gjPQr4VC1s9IkJvd6a2YB06XoImG5G1dBE5E2LrkbHaomWabZv25bI99KQBcmd4Msc9vBwJLXgtu7uTpbEstddqiGHRthnZflUDU1pKTWUBj5tLjpibev9etKjDhk8hBWS/NWYMb9bQ97NEG+n7i+ZrkHEny1ATYRCo4Dx5Hya9tZeJk29WeUHcxAeXHupkXAHUdtK8zB8L7dcNv5B+PWzx/sHlOqLCOVp3mru839XaNFNYlQSf1bCbl+rOXOGDcGLI3c5SyrHqTrTL5ClUSNmruLJJp72Za7QZZxY9b9a0eEQnqyDK9P8TnJNZajPJDJ+YnnADd9QsQCPu2ErknuVtbsUDV9YICfe7oUh6odMYSOIvBKkXtSeUeUqWGZpm1LMXfex+1uk5954wK+/6lLgInf49pzpWGKTNA0d9MiB+XCsnhVOIz5Wr6V8etU/3CfvUKNClLIPWFn1xpiOFia1tywDGccOBg9arP+Nh1Rlqnj8ORmdYOBYUeqx1dr5C6iafRvJYTc66gxWepm3fAZOhbYW5MN5ed++x7gvb/7v5kDhAW+hnw/542VFpkp23LP+N+IuEaPoeoxUZq7gFsPJodqVljuvYYBF7t+skJzcPTcAeia5G5nSpsYsLcbqiYmIcj4x1eAG4b5v6Mcqglyy1A5ce5JNPfwk4G1s4Fb9gfevlcqiCB3R9LcDTry1pXA5B96WfyMI5Y4LHvTHPEj4ESQu5BlKkjuGctCkTEUGfNHBFZGGrlobUeuN0B1qE69PhnB53cEt+kTfjYv5u9p9r/9baIsue7AqFPN5dOvOfhg4AdzgFp3pStRl3roZcaVfhJa7jkUotMPiHdna4ZHpgo4/xF1m/6NyB1Ma2P4JMQQo2afgd39H2GjqfkvAq//wbxPXFufoaovpxcVLSNQzAPXDkT9G78DEDKJCfDrf/taKRgjJfdohFnuYaiq5/9uDInEaG7w/07kUI0g97I07LZo7oxn4AOANTPN19aHhPrHPf9Ff2idCUmfEEVw950M3Hpg+H5FltFye7STQ7XgMDBHisIRsgxg+HA1K5JstXPcuCD+pkbLvWgmq5Uz1GMA4Ky7gNNvdbfFWI162xDXyGiLhmRCZJmQSLMq5KNlGfHNhRkAB1/g/x3Vfps2B0cZAoni3EPI/aHPAS9dE35f2ecirieMxBNdR2eUQ9W7fwEoNKHbWzxjLAUmMQlyd+t/8g8lzb2NqURKQNckdztXmiNQVPKmxfzfqoi85Yk09whyLyem1dSQksoyzAG2uSFtdYOkgkiygvcxh4Tf5br5HVyYVdSWKd+syHOkH/0jYLfD+bZaNzNjn70AVFaWEQ5VhzF/RKDIMoYMnDLCJthEwejPMRCxfj25TXnli9D5TdP/WYjRIdpwQJYx13UV8tGyjOe7CclweOYd/jOELHQNgLe1ml7mayjPIM+oNjiSS4UlOVS9f936F/JMUbPsTdDqU1TZt4/ZAzkq+pKuaa3mzma5E9EpRDSPiBYQ0ZURx32OiBgRBVcgqCSSyDLH/Nj/WzQYMcU/yQzAjIHck0TLlBMXHZaBLtG5jt/YFILxsqIHnUT6x52t9c8NJfeQkVIS4nMKPF3EcVf70siexwLnPwocy1eeMq09WS4s4pa7w5g/IrCkWYP688dNykmS88a44IOBiAG+UMrfPwu0bJdGgxmpfCVa7tooyEOJskwVtSYj96j0taIMYZOUAG65685f/Xz9HKVDTDIr29AuNy0EFk1R94u6EAtiOwk0dw3Ccr/ixD35BlE/8kiqM5I78Vk5dwA4FcAYAOcT0RjDcXUALgPwlr6v4oiTZbr1B479iXS822CEtR82AWrrKuBPx/K/y5RlyiJ3U0eVOF6e+efLUQuy5a4PRU1JyrzLhZF7SGNPMlIJs2D3PtWrZztq9aESYVtwLXcgQ9LkGFGn+jPqlmiAkBOQe1hIo+m5500GFr4MzH9eNRiSkLupwxj2KWDs14Az/6huFwZKICLM3P655v7/7X1pmCVFlfYbN2/dquqdhgaaXmhQBBpkbZtVBhRBEEFHfWhUhFGUURnFZfxUlMFtPtfRT2VEZ3Q+FcUFHWVwgxHccAZFZV8bRWzWhpEGeqm6S8yPiMg8EXkiMzLvWmW+z1PPvTdv3sxTmREnTr5ny1LuhnPPUO6LVvq/A9R4aW7x0zK+uVU0FJKb47//SfoY5lrEOQEBtIyDxXP1dTZzwVjsEwsUJVyfJMp9cE07QpaRtQDWSyl/L6WcBvA1AEwwOd4H4MMA+p8VUzRD1V0tfQsDdXSxGar5oZAsf54HjmIqwrmb31vKXQ+iu69Kat/TDlMAsMOe6nVyMTmeR8n6rjdHR6Rk7PifRPS96YdDVWrLvSPq6nr4fA6ug9AdLyET0hfSWIuAs38G7Pm89Pe0fpG1+GQ9WTIJQFEdOOmfFLWw+zFke1go5P8c+R48LicDOPdWIrcPr/iuOq9TNC22lE1Qw5zFYOGbW0UzVLk5RZ24saGjr/WYk70dSP2875R98KVXmnpXZvEjdMw+L1RKPoDS7TVClPsyAH8inzfobTGEEAcCWCGlzCyWLYR4jRDiOiHEdRs3MpErocjKUFUncj57Up1TxyUXnguFDIiWKRUnztV3CebcZbJYWfHG+ho0twA36EgGl2d89vnA2Fw7yqIo5x6i3N2QQAp9b0o5VKeeUJUHHZg2ex2pngg6TgXKdL9al5YpwbmzcdHayl66P7Bgl/T30bhDy3h8Au55siJaTrskCSDwhkLasrYhMIUxjCOPljHNYzLG5oJdgP3Xpf8HM8Yf0oX8dkw9/CtYc4vIYnVUy1G81OChGGOUe7s7y/30w1Zhl0Xmt8yTTaR11YgmMXF3Ox7tQogagI8DeEvegaSUn5NSrpFSrlmyhOm1GYqoXtByD1gtpbTb2XHWyV1XqtesyVWGluGojdC2XLKTXIu8axI/iproCh2PS0sQ+Aa172mHixJx0dqWjuQw0IN97UoT6lbgsfU/3qgqD5pm3xqRULXJ29qhKmMnn0N73PId4KN7pq20VBhfwMLDXR8aLcNNapoOX4tIeYicDNWs8Tw2qVL3AT/n7nyWEJhCI8ChapRXTrisqKWVqxnjG+9Qrzt5lHsQLZMzx9pNfi5QP5pr6KQ497b6P8/+Wfa5KFxaBkjCL2NaZnCWe4h5uAHACvJ5Oax2vJgPYF8AP9GOhZ0BXCaEOFlKSWK+eoioAbQfy9hBD9A1r1KOm5DV8peftIv6cMp1vVbuvbbcOeXuU4ap85E4dis1nas26AzoWqQGIp2IVH4zEelvXHDx3al9tlpp+Bb0vdlv6Vx7VIXgsXvV65MbgcW7x5ujWk1Z7h2JCG1I13I31+mbZxgh7OOmHJYBVlwe587WKW8lYymOwxc5yj0jZrpwJgAAIABJREFUu9PAfG+UGVV0UgK//JS1e1sKTMs6xkWecje0TI7hIWpp5drcqkrumjnk0jb1SZWEGKLc86zqTpO33Omi5GSbpix32VYyLc0I8XXBLX61upJ3CA7VEOX+awB7CCF2A3AfgHUAXmq+lFJuArCD+SyE+AmAt/ZNsQPJBfPB0DInqThU2vfQC+psAbKtk6zH0lIOVWcgRg1/vHn6hPyjJFuy1HGo1upqEaGWO5X/wrVERh/nrrN+feFxnY46vjuZDcy9MoqjiMOpoa396SeszVFNWe4Td12OM+tXoNXRE9drGTsLcspHU1K5036jbLciUlDKXB+aaMNBdvJ5W3OuiKFlNt4B3PBVW3RpaJkm6pm9A0woZI7aELU01djaBnz64LSMBo05WrmHcO4596PtUe7c8Xycex79xZ5Xn9O13NvN0YyWkVK2AJwD4EcAbgPwDSnlLUKI9wohTu63gCxErVisa4jCcC96lnLNnFxlLHfHBz02WSwUklobBmxVO8ehWqur/5M+OZTl3H2LofnfxrJpmVL1+ce1cp960tocaYfqwl+pbMV62yxATG0dVqYylrvnese0DDMG29OqLDCQ/C9ZlSuBMMvdgKNlmCqgHaGUeyOPlgkKKICSz/0fUvVnnPMYPtw37rlEOB98tIz1BONw7pHzlCM9CWhZaPksd8K5P74BuPWyYsctiaBRIqX8vpTyaVLKp0gpP6C3nS+lTEkppTy6r1Y7kG+5cxmHRUEfPVNJLj0OhXRpmbE5xaJlYp4wh5dM0TI5lrv1W4/CiRNbPPIay95rueshGFtaBSx34zic4i33jusU9zksF+j4ALP/tL1YlLbcKS3DjZn2dLIwmf+lVs8eQ0G0jL6GXCgkc+yOBKa15T65eQPwq3/hjxsSCuz73kRsuTIamDIAIVU7y9IynHLvtNRTp5nv1AAqqjfM04qr3GXbNl6+cXqx45bEzMxQzbNuXIRYOu6gj5xHK/f83uP0IBRybDLcOqOcO70mvugN+l0tUgOxGaLcPZZ1HAXgTNZN96nmBHdfpT7ncO7e42TB1Ep3lbtusyddv4Wv/EBjLrDT04HjP6A+u9cuRLmnwg07AGS2Q7U1lVBKjVDLPaMWuoFLy3BKjYqOGloyQiQ62PuK01S6PH0a+s8LgNu/F5bER89PYSl3kb7PZvHnivu5cufN/XbTY9xwyr2p/R2ucg+gv1LnZWgZE2Y7hG5MM1O5i6gYLVMmttSy3Ml7UctWQKUsd5eWmVus/ECcMk1pmYw64dRzHzlhpb7r6m2O4gnxMvVufqnrpdQ9yj2uVFkizt0ohGnbqRvVaog605i83+n/6rPcW9vUGHnGWSpr9ukvtr+/7zrg6n/MliWvmTrbRLqplGjUSKxsEakibl95Cd9PNqsWegzXcqdt5NLXuSMFWohQRxuNzQ/Z8gOqGNfXXmqHbWaenpGPFu3jrsWhr1Ovi3ZNfwcUc6i6yn273VS4r1G+Oz+dzIW23fOU0jJFLXcfLQP4F60+YmYq9xop3cnBTZAoZblT5U5ucu4N70G0zNhkgQWJxPSWcaiaGufuPi585R7aHkeRkd9Y1XmWexnO3XNfoxpwrGQae8ehkM7/2JrW10KorFn32v/oncBPP2Q/lX39dODGbyafvc3Ua/a5KdpT6vo0SMXDWl1nr14BXHl++jeFaBmGc2fGRUcq672ONoTU/wd3vwM6kanzM99vIZY7p/z3ewlwwSZgnidE2jJccsZKp2nfK1M0jtbGoZx7TReKo0ZjVm6GD75oGWAoTbJnqHJnHDYUtD0XEBjnHkjL5N1wRjk++ZyPAjvt6/8Np9zL1JYJdahSJ5KohSn3opa7uWbGqs6lZQi19J3X8/sGolYTENziVvPQMs2MEDwKSp/ddhnw7bOSz+71dpups5a7dqiOO8rdwOX+AWTWQo+hv3edhABLD7SRWO4xfNE/QDlahh6vTMQIHZd5+RztJiwjy9QVosrXUuLakKM9dn2lIzLP6+HcgUq5ByOPljngZc7+IZa7Y3G7WWYGeY+knOH+tOPtetdrz7a/T3HuRRyqIKGQgQ7VmBs0yl1m/w4I4Nw9haIMd5ur3Mnkv/7isOSo+GLbF71eE0kDagAdoe8f10MWAKY2hdFgZoJyT40py91JWmE59+l0ghe975xMRTj3mIYi945xNHakQBuRXd+HK30b6lAto/zZ/WiGahece1S3Y+8jYrkbzt3IHddhKkHLmOPXOVqmUu5h8EXLLFgG7PsiYC+3K0yJf9On0PNuuB40V885Pt5UjyJ7QLuJESnOfaKgQ5WpZJflUL3vN+qVtdxNtTxHgXktd7PdsSbNYDbdsXyhpT5aJsTSiVv22co9EgKS/E+bF+gEJ1/5ASpHFgynyrbU86T41zKUe3tKW4geCpCz0IvQMnHBPHLvQi13M6asMNkuLPci38fn8xgddCze8QOVx2JF07jKvWEr7lqdjHMS8liLknFfxnJvZVnuFeceBl9EQadt85cUeQMqVSnQQ8XkLRT6ON+asy7e1Kg7yt2N+W65SUyhCUyAxblzzZMpOm1V+/3X/6rP01CKgKNlAmuAey13t+aMLw4+TmJylXuR6AJbudecWO0bjv68/oKxZOMfFbDci7RrzKRlmulkGSoHt8AFxbkb5W6eVLJpGeVQrSHiaBmuNEXR+G+ffEXAKfetfwYuWQd89VR7/LQdzp2WewZ0wbZ2sm/EWO6laBmmcFis3AfQiN7BzFTurhPQgCaNcL/Jwh+vsT/XfNEyYQ7VKZlc2lrNUe5u5IibzRdaVwbQnLtTrx3gF78Hrgf+60L7PK7lbo7hKgEfz+nj3F0FmKncBR/Bkgun2bZGvSYwH8n5m3N2Um/Motrcll6sQqzJ5hbgvy/iG0K7xzM9WXMzVJ1kGfqeK8pWpPyAlEmGpAFTpK4NRcvUwfDadByE0jLmfjzjLN8O2b83oLRMh5HNKPBH7rTHT3uap2Xiz5SWITqjVkv+R0rLvOK7Ye0nfYXDgPR1v+nS/ON1iW6X4OGAOj4oMpV7wXUs8ky4XM5dDZpWhwzMesOeJCnL3VFkxqIOQZEMVUDV0InPM+53qAY2eAi33DMWLFEDHrrZ3hZiucct+2zlvtfDP8CxY1+OP8cNjE0WZHMLU0wrwLK6+VsqasbE7lO41+frL9cnz6JlptPJMtSQ4HjaIM6dtAh0ex+wljtQr49hfl0AZlfzGyvBzdAygU/BPoVYJhcki3PvOElCsmPvXxuzZaaJYtSh6rPcdz8aWPwUYONt2TL6yg8A6es+gLj3mWm5Z9EyPqui6COWxbOTy+TrIGOgB820JMo5GrcnsGu5p2iZBoIfXWkzDsuhGhA3HlvuTGq3q+yufDffnzJ2HDrbuQXLB07WEMs9trJsZbHqz/ZTWMzSmCzI5pY0NRMSh2wmrVuHCMj3SbChkNM25wvYRgXH0xahZSB17wMtw8Y7gK3pOkttKdCUNUT0PpjflLHcjWXulbNHyp3mbbQdf5Ol3OtI0TJcyCM1GlNPVFHyWx/YOHf9O1eZmwS8PmJmKndftExA3fBgUAuKdkj39X400IOqLSkNM25b4q7lztEyRSx3OiANsvIADAxdxFrujkyP3ct3lo+fGpztobSMD0GWO0/LCJeDjy13vahOM5Z7iHLveK4N4FfuJlrIR8u4BollufuUe87YiC13o9ynlewXrgWuen9q97ZOYqpZnLspI13CoRoXQ9PX280U7tZydwMIOi3g1n9Pvu+0YA1IUbNldmkZ81RJAzU6TrSMcMJLDZ4gZcLb07A6agHJ/XSNpYkF7L/ZS8xM5e6rnJdV7KcwLUMmGV1l85T79nsAAKbhDAyL83Mcpq4iK2K5Q9pWSLw5MOOz5kliCq2X7+s56f6+iB8B4BWoCw8t4165IFompOlI1j6+JK/pDOXentZj1hP+yNIyRZQ7oWUyZO9AOVRrXIZzGcvd7Dc+Hzj3JuAF/5z+H4qCynblu4F/PoyM+ybwPdJOotNyLHdX4dbtcc6FQrqLrvBY7g/dlLxvT6W/99EyleXugS/TsJecOz1OEeV+2iW4cJd/xJNwCmXR87uycMq9kOVOwrea24BH786PBaaycCFnoUkXHN8PpM/fF8udV+5uYcP4s8lUbG5JW1LbmP4AJ3zE/pxl3fuut0niCuXc82gZBHDuHC2TodxbUqAlI1u5tznO3YkA8sGMJxGpnqq0A5KRqyjcBeHhW/1+JenQMinjquGEPBrL3XWoMnPWHcfbHk/et5t2jDuQPKU79Y/iQnF9xMxU7lwBqI52ouT06gxW8mUt9zmLceu8w9DuZNQIN4q7oY+bqudeRzjn3kmug2wD33sz8KmDgK1ZzUwcuTjL3R2MPvjaknWt3EPi3M059bX+8guB2y5PW+5GuwuhioQ1t6Y5961/Th9/L6fvaZbl7qVl9HX00jKOQZLrUC0SLdPRT7lNPttVoy2N5U55a2O5T6e3hTpUYyrDeWrrlpaJt/nqILXt9UPUMqJlmsTpHalF/oKFwJ+u5f1u7jimRkiLsdzn6UitJ5xONJXl7gHXJd5N9/b9JlTJ0Ek2TnqM+krXEqgkGmejNSEE8M4HgHUXq49FnI8uJIi10QHu+YV6/8SDYb/3xbnffXXY77nSB9znwrWxi1juUlmnd18FfP1lqYcey5Ifm1TWtI9GsX7oGApZyt216lcerl6f+Wb1mhnn7vDBBr76QMHRMsZyb2Zm/HYk0JQ1m3PPSowr6lBNUXI9Uu4+v5JLy4gaQ8sQKjPm3CO7ZaP1G6M/nP+Fzt12M025GuX+uKPcG+7TTO8xw5V7gdRo38qbdw4AWHVk+jhZPxVAx9XuNYeDb8xJZOm0VATNLgfqfQs6VKkSMBYB7QebBVGzlZaZFD/7cNjvjZLMK5NbpJQvEGi5E1qGcPTumYTlzJ7D0zLL1qSP7y5IWQuOS+s05qj7adr/+QqHudEy9D2bZRyg3JfsrV7n74y4mXzGwtSWAm0Z2dEynC+lqEPVN+fKcO5GkVPlmRWh5OZu0DIQ1M+U4tzpokDnbIDl3p5KK//J7dQ92PJosm3p/pXl7gVLy+R0iRGeldcHi5aZBxzgxC1noCYE2q5y5/g7um3FM4AVh6j3so1CDtUO4QeNRbBpQ9jPXUVRdOJxNUi4z0VRVLmTSZa23GlYakNNaNdhe8Z/pI9fpN2eS4O1m0iVinaR51BlgwYClPtRbwXOuFwZJaaZPJd4ZcSACoVMyQ84c6ygQ9VruZeAkcPqg5pBy9Cng9Y2MuecCLFU+QFyTNah6vwvbYeWcctsCJGmclefwsvdY8xM5c7VCMmKJwZKWO50kSDKIcACrdVE+omRm5DCmdRxNluB8rcm7MtYNIY2ejKUlulSuXOlD2K5ukAIbdIhtIxZDGp1JhSSftCRVpRLrk/a4a7xvs5Y8v1PY3PSljt93Ac8nDvnUHVoGcvZnRc/TuTe7Zn6eA11j77yYu/uhnNPyU9fjTxAAcvdlEEoq9xphmpLfY5ynmzMvnQcT2+xFbggjlOrcJjjf3KftoEcy73JL2TutjKd4Upghir3LFrGx7kXtCLo45+oIeERQyx3WFUJk2OY92bQ06y5sWQSdJrhNIaZfFxLtSysPCwtF1Dc4jaOyDyHalGE/N6y3LVyF5FNw8C13HUsM30y8ClL9177KmPO3cFjuXsS4eJ9OFrGGZ9c79AikV+GlslAR2ZY7tYcC42WMXIW9HNlCtnSlEqOTwJIJzE1N9tOUxFBhRDLNOdO/98gWoZy7lN8XSjXSChT8rgEZrZy5wplhVruC1f4j3/uzbYlR8MFS9My1Ponj4gGEenj2NZWSgiMhRvX7vak0c/Z3v58xLm2LAY0tDIEhkv0Fc4qi5DmHRwtU4tQ8yUxAcRyJxaXb7K5j9hc2QEAmLMDY7k7VhzbQ5VzqDqKk+1/WsB/YWgZX0E9GMvdc17LoWqejruMlikDQz3SY/nGqfvEM73FVuC09k6bRCu5yZF5DtVanbHcmYXMvfdlOsOVwMxU7lxd7mDOPcCKmL/UWbUFUhEAWeIJATcS0tkjfaxaZGezFbXczf/FOf3O/jmw7GBHBGaBAdIO2jxsMens0p5QlrIvUQWwiOU+/aRKagGU5e7sZumiWLkHWO4mLt7FM98K7HFc8nl8njoe/f/bLXuscfczLhzmyVAFHAPGjMEC19MkMblZosa/A6DVQQYt4xhQQZSCGy3TS8s9srf59rUs9y2wGnLEyr1th6LWImfRz7Hcx+ba46g1lY5zB9I6qbLcM8DSMk5zBBcp507GBHHjYqnlHjC4o5qiZc5pvA8fX3URc3ynJCugBp/V0CJUuZMGBABvudP61fHhPXH/5lE1Czd+E/jFJ9S+NAqAo8m4c4QghB4yE/ieaxArlFrNX34ASJQ7vU5ZypILfT3gpcBh5ySfjeK0IjTyaBmhLXe3cFiA5V7keppQyKknkqAAAHj2+fHbpgTacMa1uf53/tDeFmJ1ur6BXlnu1AACMmgZR7nv+9eJDCJKVvtOWyWKUcudjguujwNV7tGYKp9tKjy2p/mFLKXcK8vdD3OxCkXL6H/V26iZHp9R7ihIy3Qkfo3VeHDePpww+oUca/NGYM0rVZnUI88tb7lzWZSczGaAu+d57I/5yv3bZwH/+Q+qkXNrW5KM5WvQXTQMEghzKpvz0UxOUcuOczdF54zF1ZgPnPJp/zncOkD6HBZlY95bZWcdWsa9B2NztOXoFg7rNede16GfU8DiVeQ8iRJSlrv7BNcG7vghcPvl9rYQxZSKluml5U6ulbcMNakKedaPgWdfkHaoAsA1/0/5jEwWcS2yLfGaqwNg30cz1370Ti1PqHIvMR9KYGYqd8FZ7jnOHnOz4y7zOeewnuUZvjzrp5qWabYlGnUuSoaxmmVH8fzP+5jq1F4kQxVIHDlcPDNdnLJkAID/+nS4M/Tx+9TrhE7yymvQXQS0IcP33pok4Tx6N3CNLlts/nf69MDQMiKLc3/j9cDez/fLwbUHrEWOcp+wZQa05Z4RLTOxUF0vt9idz3J/6JakjHBRy91QZzTlnSwiLSnQca/aH34GXHKqI0sn0HJ3Haq9sNxbac7dFy5L78PEQjWXqeVurt8NuvXlEw+oV7fOEkfL0GvvUjBe5e7crwFx7jO0njt5rDLIC4U0yibmHgusnkKEh6FBK/eORLsjMRZxyp2hZZ7zHnufcb8DjIUZvFwmols/xmyjrxSZ8dwkTd+ktBtZvW3+uuDcrzwf+O2XgOVrgP3XAV98vlpUDj6Dd9rWGM6dVe5aMfja/xlwtIyo2VERnHJvt2znqHudJxcBWx4JsNz1MT9zuP9YWYgayT0zORC1espyb7vRMpvuSx8rq+oqRc9CIem52+p4eZ2qAJuWiS1uyrnr/8HcWxJpZYHrwMbFvsfyTPPjaUic+wxV7lm0jE+5G8u9SAs7DRr/GkTLAB0p0ex4LHfOoUpLHADAnicCJ34UuP93wPVfyZfRTNbpwJowWY2bsyz3L5EEDFM3xSgN2VYW5viC3oVCbrxTvc5dol6NopIyvWABSZgbAR/nri1319Hogvte+Cx3UwZCMpa7M24mFmnFnuNQ5a5j0VBIcz2iBvB3v1WRM1NJwatWB+i4D/EcdVDaodoryz2yFaWv6bRbOAxIFlrqTzOUG4m0ssD5TLIKALan+f+14twLgEbL/PxjSgHkce6GtjETUQhgr5MCz1eQlqmpUMhmu4NGxEwS9hGPyWxb+2qeFuCQNYGkRDAtA6hHch8euCF5b2qVmzC7TkdZmJ/YVxVeis/VheX++H389k6Lf8Ko5cS5t5vAgzcBV/+j/jLHvvFa7uTx29w7wwF//jjlQ8lKYppYqJRQc4vjUPU4Nq3zFwyFpDJs/xRg/k6WbE2JNC3je6ILsdz31QlTyw5Sr72gIYrSMu6TNrXcjTzG/2aUu6t0BWOlc/kqBr4499Rxq2gZP8zNmd4M/Pi9wD8fQuLc8zh3cvFPvRj4mx/y+1MUTmISaLYlpEQ4LeNV4oETObPTEUNfZCn3b7867JzG+htnHKrdIk5/J2niQKLc203P/yWsaJmDt33GbphtFos42zLn+nIO1SzOfdsmYMOv9H4Zyp129Hrsj8n7LIeq71hZqHmcuhYtIyDdccYpZDeyx4e9TgQu2JTU1SkLem9MCV6LlvHU+qFJTO7TA+Xc40XZZ7nncO4ufHHuWcftI2aoctc32HiraTeiYM4davC4yT0cCicxIS75m03LkGP56CKf8jnojKRAFJCv3Itw7qGIOXet3L1UTBnL3c14dZS92wTZoFa3zvYoFtq0TNHIjSDLnSh36vOwrGbnGtCkIuoQ7jUt40ukospdAm1XFXDXtj09MMWkZKB5E5xy99TXtzh3Y0jRaBknuSq23J1rwD15ZdEyVZx7D2AujtUlJo+W8ThUgwZrsSQmSgMEW+4hIZoUJ38SeNrx5PcZSmtyEVK0jOsg2v6pxc4PEM7d0DI9tNzzOju5scwGIsqOc6fv/+r/5MvBcu41e7tZmH/+MdVEIj5xRigkLSZlQvEAv0PVOn+RJCamD6hznmab4dzdvr6Amm+DVO4U3MLCce7ROKzCYazlro8zoZ+eDj5TvWZx7nH5Anrtae2bjs5KDrDcR0m5CyGeK4S4QwixXgjxdub7NwshbhVC3CiE+LEQYtfei0oQN50lK3deEpPPoRoyUawkpjDO3WAsNBTSTTtPds6Qy6l26GLVM5UDbf7OGZZ7FxENU060jI+WKcW5G0tdy80pe/Z8MqXcrdPTBSjEiufoMtdyN6UqfvtF4JefJMf30DLvvN8ub0Ebo8RKxESEDcJyZ2gZrs2h6RE6KNAbd9cVjKXMKPf6RE60DHGomjlnFvlUtEye5W7kE4ksbOjsiFruQogIwIUATgCwGsBpQojVzm6/A7BGSrkfgEsBBBYDL4mYliE3N69wmBsPbm5MyIW2OPcQyz15P85Z7tZxu4C1ODCK6uXfVg40ALkO1TIW2bTrUO2j5Z767OHc283skr9UxpDIKV+cO41dniD8OS2t68tQbcy1FclOJNHNKBSaIu+iaCgklZvZ3uxI26EaNXg+e5iWO6DmLL2XnIz1cZtzh0PBAMn1a2snLfck7X5maUwii6GIORovpZNKNCspgZBRshbAeinl76WU0wC+BsAqSCylvFpKacjG/wawvLdiOjATo1VAuZvt8YQuotwLlvyltEydi5YpQAllnY8Obi5aJouqcb3/ZRaa2KG6IC1Pt+g4Dk+Xlmk3+VDIdjP1rFOnqy1dJEIsd24BcK8VXQCs42eEQtJ7f8qFzG+0zF07VD20DHnfbDtJTPUJ/l4O2nLPa8fHce71CeXD2Kz9GLHxYsKnZfK/d5r2tXR1Ry7nThearcn5XbjznKO8+oCQUbIMwJ/I5w16mw+vAvAD7gshxGuEENcJIa7buHFjuJQuzMWiDSny4tzNjXInNL1ZZ/8MeNHn07+1koDylXtEaRmWcy+iUDPOR0sN5CmqtWdny1CLgJMz0vA5uLRMVgPpokhx7E3gSTJmfJw786g+MeYpNhWi3Ll7lFLUNCWeHD8rWiZeXCO75RqtLe7KmyWTD3mVKcE4VN0CWvGOA7bcVx5qf3Z1Pcu514E7fwBc/X712eXcabMTE15p4Cr33Dj3QMvdve4c5dUHlNUu7JIqhHg5gDUAPsJ9L6X8nJRyjZRyzZIlS8KldGEG2C/+KdmWp9zNjYodK4zlvnR/4OlMUwOrKmS+cqe7NFhaxvHglwVVbnmKau+TVHhaLEIt/ZpHU8zfxaYgDFdsBnRzc/o36uDZx6U470HV8s7cz5hzbwMfJU5fUy7XRWtb6mzWPZAFaRk2Ucq5p1alQrIoZcW5m9/4rEUziNgIpIJVITk5CZodQFJVICKez25PDdZyP+njwCF/SzY4Re1u/Jp6/dtrkm2u5exy7lS5tzMapQAehypDy8gOUe6c5V6gXWMPEaLcNwCgxc+XA7jf3UkIcSyA8wCcLKXsr/TcAMuLlonrOZdxbohClrtNy2REy4xNAsecFxZrz6GIcnfhDlY3+4/DnifYCnX6SRXl44amdoMxfTy3nrs7IXyce3NLOlrGomUKOlQ5OyYrqoI+cVjKwbXca+l96OcBWu5TbSeJqRbx1mVrwKGQ9XFg5/2Sz7JjUxrm/s/bMdnm+kjisMd68pv42jq0jDseQi331hRR7gEO1ZAWkj1AyCj5NYA9hBC7CSEaANYBuIzuIIQ4EMBnoRT7w70X04EvwQLI4NwNLeO2vAp1qJr3AbQM2Yd3qNJwvLcBux6Wce6M8xVWVPS4jOXuHmPCKYkwNmlP+qknEmUM2CF91rkKRsvU6mmr3F04/nxPkizkYM69V/uPXZSWYS1312Pr6XtaxnJ3HapcQ4pecO4E21qOQ9Vnnbe2Dd6haiUydfhFh8rr0iJuNBhV7nddYRsRmZx7ZL8CSY+EqE44d065M/HwA0AuLyClbAkhzgHwIwARgC9IKW8RQrwXwHVSysugaJh5AL6pU7/vlVKe3DepOQVu6jAHW+5MIpEPFucesDsZj5mhkN3CRwGEwKWlOOV+4keB7XZTdMuN31DXjta7nn4SWLA8mfC94tyjujoWVWyu8+w/3lju2FT5lqkzxMFruQdw7q6yjJzxSVPpfcfKQoDlvq0p0XA5dw7dOFTPvVmVZPiXY4r9jv6vUvLOyFqWcmc4d+t+keMFWe5kcp/8aeC+3wAQxSx3k/TXZwSRvlLK7wP4vrPtfPL+2B7LlQ1ugMU1mXOUu2sRBsW5exIXPAhOYgpCxr5UkRRVVG60TH2c4RwjYMUz1Pvdj1alHlyMTSbHcMsNv+SLwDfPQCGOGFD38E/XAhcdmWzLWzie+RbgDz/3WvMxrKedLgtaveF3KrTWlIxUn/riAAAVHklEQVQF7AU3qyaR13JnaBmXminaiSl+zy8K29od1GlVSN/i0U2G6qIVtuM4FJZy73ia0WSV8WAsd999z1TujI+uMUfNi9u/B1zxLn3+jFDIXY9QlU33fyl//h5jMNH0vQY3wEyDYt+qaG6uKTfwjLPUa+E492LRMpkO1RBk0jLUcu+SlqlPpI/hUgLcwjk2yVvu0Xh41usLPuOcR9+rh29JHKB5yv3A01WzkzzsR2qUh1ihux/t/27x7sDCZf5oGQpfUSpfxUBz2zutdORQrzn3lsu5U5louF+XoZClAgi6pWWchCXZgTdhz00kzHWoQuV4TG8GHl2vPnOlumkY9kGvyEhY7C1mj3L/6YfUq68RsLnojTkqauRQ7YUPVe4FmhNT/10jK869CJbsld7WC87dLFqc5e42seAmdmsbcagSy71W5x9lORzgWDL0/hrFlqfca3U+UsHF8R8AVmofR4hja4/nqIzSvHMb+MLcUrSMo3Tc7Zbl7nYcKpAEk9UNSmNrq5N2qBr8/XrgeToqrT3ltf6DUEa5u5x7Li3jOlTdaBnpV67uHPLFudOIsfH5atwvXKnmx0ImxSduWtIjGjAQM7OeOxsto+s6+BJ3TrkQ+OWn1KORdayC0aBBoZCBce5FcMDLgENfB7yPFDrz0TKv/S+76iCHuISDniz1CXswH3NeWllyyuGRO5P7QR2qtXr5R3iuIQPXYcr9TVALxSipDZ93TIM8OsHH4WZ18Ipr+/iit0gSk/s04Hs6YGXLt9yn2wLSopDIfnO2BxatVO9bU92F75ahwaz5JvMt94bHoUo5d58hlEnLEMv9Lbcnxp4xJpub+TBqIJkHvahrXwCzx3IH/FY7oOqrHP+B9G9DFG1jbukkpsyqkEXhWhw+Wmb+zsCCXbKPFRdfIx2J3Oa/Lrjr/rQTeFqG9qosw7kbGAWcZ2VHY6nF6P82T+P3Pe79qhnKHscVk8uHEIWXGnc+zt15/Jdtpq5OgWiLiKEWHHQg7MJh1B8gBPmd7I6WKfNb16G6L6NAQxyqtQBaJjSJaWwyWfDN65ZH/Y1f4hybHvSSLYAZqtw9k6loazogTLmPz0ORR2FKy3TtUN356ep1h6eRE+j/31cnJctidh2ptCORVYckoKMMADzrXckxLeVeD7u2Bm++Tf35kEvLRFalxTs6y/HZtqc36na7AqddUs7BxyGEQ02FT/o4d8dxR2kZsy/nVPTBV98GAOaq+PAxtOwMVVchBoRThslSQt24yv2EDwFvvNHZh044Dy1jWe6BtAzbWN7ZRn18nDMVSOZmt0mLBTF7aBkgv2Uae6zAASeLZKjmWO5FlN5+p6pEjp10rbZzbwLGtFKyaBnyv2dVeKxFQLudyGCswHrDttY5y5277hMLknrkTYeWKfJ/5j1phHDuO+6T2nzCvjuHy1AWZcoY+EIhXT/FpvuAjber92NzVD0f18GahawM1XVfwXUXvwsbty3C9ng8vV9cSZGhJwYGh3OvRcDcHfy7pxQsEy3jdahmNEpxF1gDS7l79E9seA2mYJjBDLXcPWJzzaFzjxU6WMtlqLLRMkUsdyESxQ4o/nOu5t19tEyWheCz3KNxe9BPMJw9d9yJhcSh6ljubA3skghR7lEdWPsaa/OnX3pQ9+fOQ4ijzFc4zMe5m9effhC4+EXqvXnSKG25OzKsWIuPbf9etBHZJX/dhScr07bfcB2qQLbxEmS5l1DuZkF195m/lJzbZ7lPJOceIGam5Z4Xy14EWdblm28ng6JIKGTynqVlegUrIYda7hm31VW4PlqGs464id2YT2gZGi0TFbPc8xDiUAWA1S8AfvW5eHNU68HCkgffuKOKxmu55zhUKbpV7owhY26pzbk7dZis0rcDtty5bE7uek8u1n4XX4Yq4dx998u9F9R4MnPNfUrbjrSu8LXKHJJyn5mWOx1gu5JEl3k7lThWxuRfsJRRcsUs974qFxo1QaOEsqwrc+3MQDNhgauOsAf9XKawG7do1GrJ5H/ioWR7NEbuU8FrwN2TPIdqHG7W0Gfs0yPwqmemt3EJZGtepZy2BqHRMjs8Td2TUz6VPmas3ItEy2Qr5mvWK0qNDYWME6rI7wbMG1v18d2iahR/v175bHzKPQqgZdwqk9R48mXA0xIdPlq4Uu4FwGWOAcALL+rjSY3lnr+n6AUNEYJj3pm8D42hPeot6tVMgj2OBd6xAdj18HzL3We1me2P3KFe9z4ZePG/lXe+caUecmkZ5/G7XzjzcubczDmPPNdeZH1x7q6yqI8Dr/whsPLw9DENVVZESQTEuQPwOFQZy33QtWWocs/Kwq5FSDUut76nce6eMWKazxjHvAlmAPy0DIWXltHzqqJlAkAH2JFvUpmQx72/XLRMKAokMUWDUu57n5S8Dy0/cMQb1R+FcQpR5T65OP1bn9Xmnvsl/1/do82PqM+9uByhdWv05Eu1jesnuCelrOYcQH5Dd+5aH3YOsOPewBFvKCBbRrQMgWQtd/1KHbiDpmUsy52Mz71PBm67LL2/l08PCIU01N9+64ATPmh/56NlKHy0DKWEBoiZabnTQTpnMfD8T/RXsQNJKj1tbOzBIGjeFHpRBMuKFw7oKGOwaEUcVmc7UksOL25x5GK733QL9+Ny5+w18kpLh3YOo5hYqEIBixSeCgxjZKtCGkVpJacNWGWYMiFUHkCl8XPwxrAHhELu80LVs2Dtq9PfxbRMAWduDFP3vYqWyQed/GUaO5fBcR8AXnYpsCw/+qI2DO3ei9TmvCcOqhwOPhN4BbGczFMEa9kUvB6hkyDjf+4b5x6KVIijcw3iEtU+ZcRMzTKUU2AYI+tQNa90zHdrue9xfNKQOgSLVgCvvkq9D4kI82afBmSoLlwOvOU20neYIKZlMsgOn3KPE9IqWqYYBpX1NTah6oyE7BoNQbln9UvtFejEfvY/qKcmg7gbU4lw1LLg/udBUWJ58GWkGuR1DgOUAuu2RHGg5X7+8/cFrtAf3PLYjbnAoa8H/vvC7h2qL/tG8d+44Yzuewq3/IBBSJx7FkJoGV/5C3PdK1qmIAZUYa0IokE/ugLlEriKgk5sV9H4GnVw++YhVEFzlvtC1TTssoUvx7tPWp3+flDIpWVymstw35UxZHIs9wNWKCftsatJvDaXmWoW0kE7VIEkOshKyPIo6CV7ASd8OL3d4txL/A/d0DKV5V4Sg6JlCmBsGLTMIK4DnRSuYt36P+r1yDcl2ya3U4/g+76oP/Jwym58HnDBJry1P2cMR6rEb0HOnfuuW1qGUWrfeu3h6EgJPPlA+jeWpazvdy9zF0JhFKtV+8hz3YQADjkb+MHb7O1WJ6YS8zOIlvE8NWiDAyvWFj9vF5j5yn3AldZCMJDEGReDeFqgisId5Ev2BvBdYJ+/TrYJYYdrloZAHIq6y4HA/b/T8ozwg6f7JOUq1t3/ChhfCBz+d/5jpLo0lbDc6X1iFHNUE4gg7O+4fsO+ZjeDgFHu9ZzaR1mgnHs3yLTcPU/PO60GzvmNqv8/QIzw7AjECFru9X5mpWbhqLcBZ13Vv+NnWW1H/T3wmp8CS/fz7xMK16FKH3fP/F73xx8EfA2xDebtCLzjXmCXA/zHcK3/rmmZDKNDZDhUgYRa4xp29xsrDlGlup9LwhOL0pDmf3nqser1tb8s9vtTLlSlLVZm9Dv2We4AsMNTB26MVJZ7H1AfSiwkgGed1/0xnvvB5DHSRRaFENWzFVU3GJvUjlqRPYFGGWXojF5Y7qEO0FzLXSt3OQTLvTEH+Jvv29vyGrO87lrVyctACOAN1ydZ7Duli8xlYtEK4MSPZO/jDYUcDirl3gfUhxEt0ysc+lr/d4NyprkWpolCaMxLf/emW4GpJwYjVxYOO0c1LrnrCv77MsrdfYLpp3K3slAZzt1QIsOgZTjkWe477qX+KBbv1j95gMEENRTAzFfuo0jLjDIX3A0GVVfER8twiWoLl/VfnhAc/wGVTfnBlfz3ZlE64tzwY7pWcimHauCCzOWO0Pu9fK2iRnzJQ4PGKCnSl3wR+O2XRicMV2MWKPceWZMrDunNcTDDLfcs9CILtgzMI3ivmmv0C7SIFIcLNmV/7yLVoLyPT04WLcNw7jutTlMjw8QoUSD7vED9jRhmvnLvxWp53oM9tUqHxrn3G40Cae/dYPXJwM2XJp8Nz57VRnFUMH8psOrI/P1CYCz3Ez6Sblbea1gOVcZyHzWMsmwjgpl7haJGsbrWWeixFTC0aJl+o9/1ewxWnwK8+1G7GTiQOMOOehuw+eHByFIUb7m9d8cykSnj81X56X4iL1pm1DBiFMgoYoTvXg7O+TXwcEbPzSFi9lruA6RFaHz2I3eq14NOV6+9iAqaCTDOy0GXlhjBrO++4NSL86m0GYyZexe3W6X+RhADVe5vulX11RwExgbMedcngdZWYOuf1eelfQq1HFWYhJteFIXLA0fLDLv4Wr/Rb6pryJi5yn2EMdBomYXLAAwoYmTQUUCNuUq5/9XbVeu8hcsHe36K57wXAy8nXKurZubdOrIPeS2wfE32Plyce4UZjUq59wGRL1rm9O8AD908WGFmMhpzgS2PAPufChzzjuHK4jY4GQRqdVXHvtvKp27jCQ4c515hRqO6i32Al5Z5yjHqb6bDNOboN154EfCfFwALhmixDxNRHWhiMDHdloNSvx/1bODTvg5MLBi2FCOLSrn3AbPWoQoAb10/uHj3XQ8HXuXJ+PxLgLGgB+JQFao+0F7PA+7W9YlG3dm453OHLcFIY5bG7A0XQ6kKOSjMW1JZS4NCXAZgQIvps96lqm5ufUx9HnXlXiETQcpdCPFcIcQdQoj1Qoi3M9+PCyG+rr+/VgixqteCziSIKga3Qi9golYGnRlsmlJXyn1GI1e5CyEiABcCOAHAagCnCSHcFjevAvBnKeVTAXwcwId6LWiFCn9x6LbReFnscqB+/QsLPZ1lCBk1awGsl1L+Xko5DeBrAE5x9jkFwBf1+0sBPFtU5muFCt1hUrXAG3R7Nhx8JnDuTcCygwd73go9RYhDdRmAP5HPGwC4VbbifaSULSHEJgDbA3iE7iSEeA2A1wDAypWe6nmzBO87ZR/sr/tTVqhQCqdeDFz/1YF38IEQwKLZPT//EhCi3DkL3E1dC9kHUsrPAfgcAKxZs2ZWp7+dftiqYYtQYaZj0Urg6JSLq0KFIITQMhsA0NY8ywHc79tHCFEHsBDA//RCwAoVKlSoUBwhyv3XAPYQQuwmhGgAWAfgMmefywCcod+/GMBVUrodFypUqFChwqCQS8toDv0cAD8CEAH4gpTyFiHEewFcJ6W8DMDnAXxZCLEeymJf10+hK1SoUKFCNoIyVKWU3wfwfWfb+eT9NgAv6a1oFSpUqFChLKoM1QoVKlSYhaiUe4UKFSrMQlTKvUKFChVmISrlXqFChQqzEGJYEYtCiI0A/ljy5zvAyX4dEVRyFceoylbJVQyVXMXQjVy7SimX5O00NOXeDYQQ10kpc/qGDR6VXMUxqrJVchVDJVcxDEKuipapUKFChVmISrlXqFChwizETFXunxu2AB5UchXHqMpWyVUMlVzF0He5ZiTnXqFChQoVsjFTLfcKFSpUqJCBSrlXqFChwizEjFPuec26+3zuLwghHhZC3Ey2LRZCXCmEuEu/bqe3CyHEJ7WcNwohDuqjXCuEEFcLIW4TQtwihHjjKMgmhJgQQvxKCHGDlus9evtuupH6XbqxekNvH2ijdSFEJIT4nRDi8lGRSwhxjxDiJiHE9UKI6/S2URhji4QQlwohbtfj7LBhyyWE2FNfJ/P3uBDi3GHLpc/1Jj3mbxZCXKLnwmDHl5RyxvxBlRy+G8DuABoAbgCweoDnPwrAQQBuJts+DODt+v3bAXxIvz8RwA+gulQdCuDaPsq1FMBB+v18AHdCNTMfqmz6+PP0+zEA1+rzfQPAOr39IgCv1e9fB+Ai/X4dgK/3+X6+GcBXAVyuPw9dLgD3ANjB2TYKY+yLAM7S7xsAFo2CXES+CMCDAHYdtlxQbUf/AGCSjKszBz2++nrB+3DRDgPwI/L5HQDeMWAZVsFW7ncAWKrfLwVwh37/WQCncfsNQMbvAnjOKMkGYA6A30L1330EQN29p1A9Aw7T7+t6P9EneZYD+DGAZwG4XE/4UZDrHqSV+1DvI4AFWlmJUZLLkeU4ANeMglxIekov1uPlcgDHD3p8zTRahmvWvWxIshjsJKV8AAD06456+1Bk1Y90B0JZyUOXTVMf1wN4GMCVUE9ej0kpW8y5rUbrAEyj9X7gEwDeBqCjP28/InJJAFcIIX4jVEN5YPj3cXcAGwH8m6ax/lUIMXcE5KJYB+AS/X6ockkp7wPwUQD3AngAarz8BgMeXzNNuQc14h4RDFxWIcQ8AN8CcK6U8vGsXZltfZFNStmWUh4AZSmvBbB3xrkHIpcQ4iQAD0spf0M3D1sujSOklAcBOAHA64UQR2XsOyi56lB05GeklAcC2AxFdwxbLnUyxV2fDOCbebsy2/oxvrYDcAqA3QDsAmAu1P30nbsvcs005R7SrHvQeEgIsRQA9OvDevtAZRVCjEEp9q9IKb89SrIBgJTyMQA/geI6FwnVSN0996AarR8B4GQhxD0AvgZFzXxiBOSClPJ+/fowgH+HWhCHfR83ANggpbxWf74UStkPWy6DEwD8Vkr5kP48bLmOBfAHKeVGKWUTwLcBHI4Bj6+ZptxDmnUPGrQ5+BlQfLfZ/grtoT8UwCbzqNhrCCEEVB/b26SU/zQqsgkhlgghFun3k1CD/jYAV0M1Uufk6nujdSnlO6SUy6WUq6DG0FVSypcNWy4hxFwhxHzzHopHvhlDvo9SygcB/EkIsafe9GwAtw5bLoLTkFAy5vzDlOteAIcKIebouWmu12DHVz+dHP34g/J43wnF3Z434HNfAsWhNaFW21dBcWM/BnCXfl2s9xUALtRy3gRgTR/lOhLqMe5GANfrvxOHLRuA/QD8Tst1M4Dz9fbdAfwKwHqoR+lxvX1Cf16vv999APf0aCTRMkOVS5//Bv13ixnfw76P+lwHALhO38vvANhuROSaA+BRAAvJtlGQ6z0Abtfj/ssAxgc9vqryAxUqVKgwCzHTaJkKFSpUqBCASrlXqFChwixEpdwrVKhQYRaiUu4VKlSoMAtRKfcKFSpUmIWolHuFChUqzEJUyr1ChQoVZiH+F3wsEFH4xlQJAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fb7240caa90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# inspect channel ___ sample\n",
    "chan = 30\n",
    "samp = 12\n",
    "%matplotlib inline\n",
    "plt.plot(range(0, 800), x_test[samp, chan,:,0], range(0,800), y_test[samp,0,:,0])\n",
    "ax = plt.gca()\n",
    "ax.set_title(\"Channels 11 in test samples 6 and 16\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "12\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fb703be62b0>]"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fb79c032eb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(samp)\n",
    "plt.plot(y_test[samp,0,:,0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Training time!\n",
    "# Try 20 epochs using batch size = 3, then switch to 4\n",
    "model.fit(x_train, y_train,\n",
    "                epochs=1000,\n",
    "                batch_size=100,\n",
    "                shuffle=True,\n",
    "                validation_data=(x_test, y_test),\n",
    "                callbacks=[TensorBoard(log_dir='/home/amplifier/home/logs/CNN_autoenc_reconst')])\n",
    "\n",
    "# save the entire model and the weights after the training:\n",
    "model.save('/home/amplifier/home/models/EEGnet_stim_reconst.h5')\n",
    "model.save_weights('/home/amplifier/home/weights/EEGnet_stim_reconst_weights.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# let's compute some predictions:\n",
    "decoded_train = model.predict(x_train)\n",
    "decoded_test = model.predict(x_test)\n",
    "print(\"The shape of the decoded TRAINING samples' tensor\", decoded_train.shape)\n",
    "print(\"The shape of the decoded TEST samples' tensor\", decoded_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test sample 0 : Correlation  [-0.02817558]  P-value:  [ 0.4261232]\n",
      "Test sample 1 : Correlation  [-0.06385712]  P-value:  [ 0.07104796]\n",
      "Test sample 2 : Correlation  [-0.12066752]  P-value:  [ 0.00062584]\n",
      "Test sample 3 : Correlation  [-0.00561276]  P-value:  [ 0.87400317]\n",
      "Test sample 4 : Correlation  [ 0.34031379]  P-value:  [  3.85373541e-23]\n",
      "Test sample 5 : Correlation  [-0.01486668]  P-value:  [ 0.67459142]\n",
      "Test sample 6 : Correlation  [ 0.13818924]  P-value:  [  8.80735824e-05]\n",
      "Test sample 7 : Correlation  [-0.07329368]  P-value:  [ 0.03820862]\n",
      "Test sample 8 : Correlation  [ 0.03372513]  P-value:  [ 0.34074634]\n",
      "Test sample 9 : Correlation  [ 0.04543263]  P-value:  [ 0.19925474]\n",
      "Test sample 10 : Correlation  [-0.02016326]  P-value:  [ 0.56903476]\n",
      "Test sample 11 : Correlation  [ 0.01915395]  P-value:  [ 0.5885402]\n",
      "Test sample 12 : Correlation  [ 0.22961327]  P-value:  [  4.95460756e-11]\n",
      "Test sample 13 : Correlation  [ 0.0344271]  P-value:  [ 0.33079365]\n",
      "Test sample 14 : Correlation  [-0.16295047]  P-value:  [  3.60881063e-06]\n",
      "Test sample 15 : Correlation  [ 0.04000641]  P-value:  [ 0.25837842]\n",
      "Test sample 16 : Correlation  [-0.0067209]  P-value:  [ 0.84944916]\n",
      "Test sample 17 : Correlation  [ 0.21018748]  P-value:  [  1.93839966e-09]\n",
      "Test sample 18 : Correlation  [ 0.15094045]  P-value:  [  1.80977986e-05]\n",
      "Test sample 19 : Correlation  [-0.09324368]  P-value:  [ 0.00831628]\n",
      "Test sample 20 : Correlation  [-0.05968904]  P-value:  [ 0.09157798]\n",
      "Test sample 21 : Correlation  [-0.02558315]  P-value:  [ 0.46992615]\n",
      "Test sample 22 : Correlation  [-0.01332221]  P-value:  [ 0.70672423]\n",
      "Test sample 23 : Correlation  [ 0.01232105]  P-value:  [ 0.72786599]\n",
      "Test sample 24 : Correlation  [-0.12527914]  P-value:  [ 0.00038247]\n",
      "Test sample 25 : Correlation  [-0.08101338]  P-value:  [ 0.02192917]\n",
      "Test sample 26 : Correlation  [-0.19950593]  P-value:  [  1.26026469e-08]\n",
      "Test sample 27 : Correlation  [ 0.01844909]  P-value:  [ 0.60231596]\n",
      "Test sample 28 : Correlation  [-0.01527097]  P-value:  [ 0.66624576]\n",
      "Test sample 29 : Correlation  [-0.16327953]  P-value:  [  3.44712475e-06]\n",
      "Test sample 30 : Correlation  [-0.15169775]  P-value:  [  1.64063022e-05]\n",
      "Test sample 31 : Correlation  [-0.08282838]  P-value:  [ 0.01912443]\n",
      "Test sample 32 : Correlation  [ 0.08734056]  P-value:  [ 0.01346541]\n",
      "Test sample 33 : Correlation  [ 0.16352712]  P-value:  [  3.32995842e-06]\n",
      "Test sample 34 : Correlation  [-0.09837282]  P-value:  [ 0.005356]\n",
      "Test sample 35 : Correlation  [-0.02669345]  P-value:  [ 0.45086351]\n",
      "Test sample 36 : Correlation  [ 0.28147146]  P-value:  [  4.92381470e-16]\n",
      "Test sample 37 : Correlation  [ 0.04512885]  P-value:  [ 0.20227523]\n",
      "Test sample 38 : Correlation  [ 0.01096338]  P-value:  [ 0.75682694]\n",
      "Test sample 39 : Correlation  [-0.0804946]  P-value:  [ 0.02279417]\n",
      "Test sample 40 : Correlation  [ 0.01936706]  P-value:  [ 0.58438939]\n",
      "Test sample 41 : Correlation  [-0.02974323]  P-value:  [ 0.40081826]\n",
      "Test sample 42 : Correlation  [-0.07178703]  P-value:  [ 0.04236828]\n",
      "Test sample 43 : Correlation  [ 0.12433173]  P-value:  [ 0.00042377]\n",
      "Test sample 44 : Correlation  [ 0.09376502]  P-value:  [ 0.0079597]\n",
      "Test sample 45 : Correlation  [-0.10035559]  P-value:  [ 0.00449416]\n",
      "Test sample 46 : Correlation  [ 0.00248456]  P-value:  [ 0.94394618]\n",
      "Test sample 47 : Correlation  [-0.10609153]  P-value:  [ 0.0026601]\n",
      "Test sample 48 : Correlation  [ 0.15495697]  P-value:  [  1.06958332e-05]\n",
      "Test sample 49 : Correlation  [ 0.26191515]  P-value:  [  5.14037360e-14]\n",
      "Test sample 50 : Correlation  [-0.14287451]  P-value:  [  5.00036804e-05]\n",
      "Test sample 51 : Correlation  [ 0.07074714]  P-value:  [ 0.04545557]\n",
      "Test sample 52 : Correlation  [ 0.05974234]  P-value:  [ 0.09128767]\n",
      "Test sample 53 : Correlation  [ 0.16145897]  P-value:  [  4.43769113e-06]\n",
      "Test sample 54 : Correlation  [-0.07062158]  P-value:  [ 0.04584251]\n",
      "Test sample 55 : Correlation  [-0.05403332]  P-value:  [ 0.12675533]\n",
      "Test sample 56 : Correlation  [-0.32073265]  P-value:  [  1.34725483e-20]\n",
      "Test sample 57 : Correlation  [ 0.15572821]  P-value:  [  9.65342952e-06]\n",
      "Test sample 58 : Correlation  [-0.07890362]  P-value:  [ 0.02563354]\n",
      "Test sample 59 : Correlation  [ 0.02721792]  P-value:  [ 0.44202191]\n",
      "Test sample 60 : Correlation  [ 0.06305391]  P-value:  [ 0.07468043]\n",
      "Test sample 61 : Correlation  [ 0.11774267]  P-value:  [ 0.00084789]\n",
      "Test sample 62 : Correlation  [-0.1387012]  P-value:  [  8.28631601e-05]\n",
      "Test sample 63 : Correlation  [ 0.11874909]  P-value:  [ 0.00076435]\n",
      "Test sample 64 : Correlation  [ 0.00947988]  P-value:  [ 0.78889823]\n",
      "Test sample 65 : Correlation  [-0.10631439]  P-value:  [ 0.00260513]\n",
      "Test sample 66 : Correlation  [ 0.00083982]  P-value:  [ 0.98094559]\n",
      "Test sample 67 : Correlation  [ 0.22093353]  P-value:  [  2.65996808e-10]\n",
      "Test sample 68 : Correlation  [-0.12599179]  P-value:  [ 0.00035392]\n",
      "Test sample 69 : Correlation  [-0.05926001]  P-value:  [ 0.09393966]\n",
      "Test sample 70 : Correlation  [ 0.09771889]  P-value:  [ 0.00567127]\n",
      "Test sample 71 : Correlation  [-0.21329051]  P-value:  [  1.10405907e-09]\n",
      "Test sample 72 : Correlation  [ 0.01001101]  P-value:  [ 0.77741963]\n",
      "Test sample 73 : Correlation  [ 0.06644268]  P-value:  [ 0.06032229]\n",
      "Test sample 74 : Correlation  [ 0.30392414]  P-value:  [  1.47405353e-18]\n",
      "Test sample 75 : Correlation  [-0.0209498]  P-value:  [ 0.55404383]\n",
      "Test sample 76 : Correlation  [-0.02904866]  P-value:  [ 0.41192669]\n",
      "Test sample 77 : Correlation  [ 0.12628959]  P-value:  [ 0.00034258]\n",
      "Test sample 78 : Correlation  [-0.00149255]  P-value:  [ 0.9665482]\n",
      "Test sample 79 : Correlation  [ 0.15486057]  P-value:  [  1.08330996e-05]\n",
      "Test sample 80 : Correlation  [ 0.02686261]  P-value:  [ 0.44801733]\n",
      "Test sample 81 : Correlation  [-0.09088428]  P-value:  [ 0.0101142]\n",
      "Test sample 82 : Correlation  [ 0.43757901]  P-value:  [  9.48608575e-39]\n",
      "Test sample 83 : Correlation  [ 0.29618335]  P-value:  [  1.15887003e-17]\n",
      "Test sample 84 : Correlation  [ 0.09346405]  P-value:  [ 0.00816372]\n",
      "Test sample 85 : Correlation  [ 0.12600915]  P-value:  [ 0.00035325]\n",
      "Test sample 86 : Correlation  [ 0.23209566]  P-value:  [  3.02517074e-11]\n",
      "Test sample 87 : Correlation  [-0.01139497]  P-value:  [ 0.74756932]\n",
      "Test sample 88 : Correlation  [ 0.05761681]  P-value:  [ 0.10342634]\n",
      "Test sample 89 : Correlation  [-0.20728451]  P-value:  [  3.25651950e-09]\n",
      "Test sample 90 : Correlation  [-0.06233543]  P-value:  [ 0.07805606]\n",
      "Test sample 91 : Correlation  [ 0.03277764]  P-value:  [ 0.35450271]\n",
      "Test sample 92 : Correlation  [ 0.06299237]  P-value:  [ 0.07496461]\n",
      "Test sample 93 : Correlation  [-0.03711733]  P-value:  [ 0.29439181]\n",
      "Test sample 94 : Correlation  [ 0.11100686]  P-value:  [ 0.00166334]\n",
      "Test sample 95 : Correlation  [-0.15157147]  P-value:  [  1.66769369e-05]\n",
      "Test sample 96 : Correlation  [ 0.05800793]  P-value:  [ 0.10110345]\n",
      "Test sample 97 : Correlation  [-0.12033322]  P-value:  [ 0.00064817]\n",
      "Test sample 98 : Correlation  [ 0.01669632]  P-value:  [ 0.63725448]\n",
      "Test sample 99 : Correlation  [-0.12300827]  P-value:  [ 0.00048845]\n",
      "Test sample 100 : Correlation  [-0.00144946]  P-value:  [ 0.96746439]\n",
      "Test sample 101 : Correlation  [-0.06337675]  P-value:  [ 0.07320199]\n",
      "Test sample 102 : Correlation  [-0.05010137]  P-value:  [ 0.15684889]\n",
      "Test sample 103 : Correlation  [-0.04842677]  P-value:  [ 0.17118929]\n",
      "Test sample 104 : Correlation  [-0.17897169]  P-value:  [  3.48080306e-07]\n",
      "Test sample 105 : Correlation  [ 0.23198223]  P-value:  [  3.09451284e-11]\n",
      "Test sample 106 : Correlation  [ 0.06574046]  P-value:  [ 0.06309624]\n",
      "Test sample 107 : Correlation  [ 0.18269157]  P-value:  [  1.96029688e-07]\n",
      "Test sample 108 : Correlation  [-0.15566421]  P-value:  [  9.73624537e-06]\n",
      "Test sample 109 : Correlation  [ 0.02006615]  P-value:  [ 0.57088184]\n",
      "Test sample 110 : Correlation  [-0.08176384]  P-value:  [ 0.02072888]\n",
      "Test sample 111 : Correlation  [-0.06794148]  P-value:  [ 0.05474612]\n",
      "Test sample 112 : Correlation  [-0.00983235]  P-value:  [ 0.7812596]\n",
      "Test sample 113 : Correlation  [-0.13724558]  P-value:  [  9.84962608e-05]\n",
      "Test sample 114 : Correlation  [-0.08760212]  P-value:  [ 0.01318774]\n",
      "Test sample 115 : Correlation  [-0.11423878]  P-value:  [ 0.00120916]\n",
      "Test sample 116 : Correlation  [ 0.03688187]  P-value:  [ 0.29745355]\n",
      "Test sample 117 : Correlation  [ 0.25051287]  P-value:  [  6.50586518e-13]\n",
      "Test sample 118 : Correlation  [-0.0545631]  P-value:  [ 0.12306657]\n",
      "Test sample 119 : Correlation  [ 0.09739538]  P-value:  [ 0.00583336]\n",
      "Test sample 120 : Correlation  [ 0.25634897]  P-value:  [  1.80252311e-13]\n",
      "Test sample 121 : Correlation  [-0.11412847]  P-value:  [ 0.00122257]\n",
      "Test sample 122 : Correlation  [ 0.34707955]  P-value:  [  4.60946179e-24]\n",
      "Test sample 123 : Correlation  [ 0.10690811]  P-value:  [ 0.00246362]\n",
      "Test sample 124 : Correlation  [-0.21191621]  P-value:  [  1.41816603e-09]\n",
      "Test sample 125 : Correlation  [ 0.12805133]  P-value:  [ 0.00028218]\n",
      "Test sample 126 : Correlation  [-0.06242144]  P-value:  [ 0.07764775]\n",
      "Test sample 127 : Correlation  [-0.37524667]  P-value:  [  3.75677889e-28]\n",
      "Test sample 128 : Correlation  [ 0.00678539]  P-value:  [ 0.84808195]\n",
      "Test sample 129 : Correlation  [-0.00956028]  P-value:  [ 0.78719723]\n",
      "Test sample 130 : Correlation  [ 0.00443297]  P-value:  [ 0.90032864]\n",
      "Test sample 131 : Correlation  [ 0.13236997]  P-value:  [ 0.00017357]\n",
      "Test sample 132 : Correlation  [ 0.08969826]  P-value:  [ 0.01114263]\n",
      "Test sample 133 : Correlation  [-0.0130454]  P-value:  [ 0.71257037]\n",
      "Test sample 134 : Correlation  [ 0.07486453]  P-value:  [ 0.03424739]\n",
      "Test sample 135 : Correlation  [ 0.05017691]  P-value:  [ 0.15621938]\n",
      "Test sample 136 : Correlation  [-0.15520227]  P-value:  [  1.03531111e-05]\n",
      "Test sample 137 : Correlation  [ 0.12615326]  P-value:  [ 0.00034772]\n",
      "Test sample 138 : Correlation  [ 0.23430751]  P-value:  [  1.93979277e-11]\n",
      "Test sample 139 : Correlation  [ 0.04358245]  P-value:  [ 0.218191]\n",
      "Test sample 140 : Correlation  [ 0.34505028]  P-value:  [  8.76274819e-24]\n",
      "Test sample 141 : Correlation  [-0.37862608]  P-value:  [  1.13871306e-28]\n",
      "Test sample 142 : Correlation  [ 0.04779853]  P-value:  [ 0.17682151]\n",
      "Test sample 143 : Correlation  [ 0.07259861]  P-value:  [ 0.04008172]\n",
      "Test sample 144 : Correlation  [ 0.03590218]  P-value:  [ 0.31048742]\n",
      "Test sample 145 : Correlation  [-0.0084408]  P-value:  [ 0.81154001]\n",
      "Test sample 146 : Correlation  [ 0.01049535]  P-value:  [ 0.76692975]\n",
      "Test sample 147 : Correlation  [-0.07681801]  P-value:  [ 0.02981421]\n",
      "Test sample 148 : Correlation  [ 0.09281662]  P-value:  [ 0.00861885]\n",
      "Test sample 149 : Correlation  [-0.11449277]  P-value:  [ 0.00117882]\n",
      "Test sample 150 : Correlation  [-0.15229937]  P-value:  [  1.51702789e-05]\n",
      "Test sample 151 : Correlation  [-0.1266796]  P-value:  [ 0.00032825]\n",
      "Test sample 152 : Correlation  [-0.12930021]  P-value:  [ 0.00024556]\n",
      "Test sample 153 : Correlation  [ 0.14596784]  P-value:  [  3.40776678e-05]\n",
      "Test sample 154 : Correlation  [-0.01373227]  P-value:  [ 0.69814003]\n",
      "Test sample 155 : Correlation  [ 0.23324311]  P-value:  [  2.40363476e-11]\n",
      "Test sample 156 : Correlation  [ 0.07469324]  P-value:  [ 0.03466088]\n",
      "Test sample 157 : Correlation  [-0.13708568]  P-value:  [ 0.00010037]\n",
      "Test sample 158 : Correlation  [-0.08694852]  P-value:  [ 0.01389039]\n",
      "Test sample 159 : Correlation  [ 0.36039308]  P-value:  [  6.05218585e-26]\n",
      "Test sample 160 : Correlation  [ 0.08474332]  P-value:  [ 0.01650896]\n",
      "Test sample 161 : Correlation  [-0.29050565]  P-value:  [  5.05572062e-17]\n",
      "Test sample 162 : Correlation  [ 0.19378476]  P-value:  [  3.29589227e-08]\n",
      "Pearson's R: [ 0.19378476] p-value [  3.29589227e-08]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'Original')"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdb65f87860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# set FALSE if you want to reconstruct the TEST samples:\n",
    "reconstruct_training_sample = False\n",
    "\n",
    "from scipy.stats import pearsonr\n",
    "\n",
    "\n",
    "# now let's see how the model performs:\n",
    "if (reconstruct_training_sample==True):\n",
    "    reconst = decoded_train[samp,:,:,:].reshape(800,1)\n",
    "    orig = y_train[samp,:,:,:].reshape(800,1)\n",
    "    R,p = pearsonr(reconst, orig)\n",
    "    title = \"Reconstructed input based on TRAINING data, \" + \"Corr\"+ str(R) + \", P-value\" + str(p)\n",
    "    for samp in range(y_train.shape[0]):\n",
    "        reconst = decoded_train[samp,:,:,:].reshape(800,1)\n",
    "        orig = y_train[samp,:,:,:].reshape(800,1)\n",
    "        R,p = pearsonr(reconst, orig)\n",
    "        print(\"Training sample\", samp, \": Correlation \", R, \" P-value: \", p)\n",
    "    \n",
    "else:\n",
    "    reconst = decoded_test[samp,:,:,:].reshape(800,1)\n",
    "    orig = y_test[samp,:,:,:].reshape(800,1)\n",
    "    R,p = pearsonr(reconst, orig)\n",
    "    title = \"Reconstructed input based on TEST data, \" + \"Corr\"+ str(R) + \", P-value\" + str(p)\n",
    "    for samp in range(y_test.shape[0]):\n",
    "        reconst = decoded_test[samp,:,:,:].reshape(800,1)\n",
    "        orig = y_test[samp,:,:,:].reshape(800,1)\n",
    "        R,p = pearsonr(reconst, orig)\n",
    "        print(\"Test sample\", samp, \": Correlation \", R, \" P-value: \", p)\n",
    "\n",
    "print(\"Pearson's R:\", R, \"p-value\", p)\n",
    "%matplotlib inline\n",
    "plt.subplot(211)\n",
    "plt.plot(reconst)\n",
    "ax = plt.gca()\n",
    "ax.grid()\n",
    "ax.set_title(title)\n",
    "plt.subplot(212)\n",
    "plt.plot(orig)\n",
    "plt.tight_layout()\n",
    "ax = plt.gca()\n",
    "ax.grid()\n",
    "ax.set_title(\"Original\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set FALSE if you want to reconstruct the TEST samples:\n",
    "reconstruct_training_sample = True\n",
    "\n",
    "from scipy.stats import pearsonr\n",
    "\n",
    "\n",
    "# now let's see how the model performs:\n",
    "if (reconstruct_training_sample==True):\n",
    "    reconst = decoded_train[samp,:,:,:].reshape(800,1)\n",
    "    orig = y_train[samp,:,:,:].reshape(800,1)\n",
    "    R,p = pearsonr(reconst, orig)\n",
    "    title = \"Reconstructed input based on TRAINING data, \" + \"Corr\"+ str(R) + \", P-value\" + str(p)\n",
    "    for samp in range(y_train.shape[0]):\n",
    "        reconst = decoded_train[samp,:,:,:].reshape(800,1)\n",
    "        orig = y_train[samp,:,:,:].reshape(800,1)\n",
    "        R,p = pearsonr(reconst, orig)\n",
    "        print(\"Training sample\", samp, \": Correlation \", R, \" P-value: \", p)\n",
    "    \n",
    "else:\n",
    "    reconst = decoded_test[samp,:,:,:].reshape(800,1)\n",
    "    orig = y_test[samp,:,:,:].reshape(800,1)\n",
    "    R,p = pearsonr(reconst, orig)\n",
    "    title = \"Reconstructed input based on TEST data, \" + \"Corr\"+ str(R) + \", P-value\" + str(p)\n",
    "    for samp in range(y_test.shape[0]):\n",
    "        reconst = decoded_test[samp,:,:,:].reshape(800,1)\n",
    "        orig = y_test[samp,:,:,:].reshape(800,1)\n",
    "        R,p = pearsonr(reconst, orig)\n",
    "        print(\"Test sample\", samp, \": Correlation \", R, \" P-value: \", p)\n",
    "\n",
    "print(\"Pearson's R:\", R, \"p-value\", p)\n",
    "%matplotlib inline\n",
    "plt.subplot(211)\n",
    "plt.plot(reconst)\n",
    "ax = plt.gca()\n",
    "ax.grid()\n",
    "ax.set_title(title)\n",
    "plt.subplot(212)\n",
    "plt.plot(orig)\n",
    "plt.tight_layout()\n",
    "ax = plt.gca()\n",
    "ax.grid()\n",
    "ax.set_title(\"Original\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}