EEGnet_CNN_autoencoder_NEW(drop, batch, R_loss)-XP.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original data type: float32\n",
      "Normalized data type: float16\n",
      "test input shape (69, 60, 800, 1) Nomralized MEAN: 0.5083 min 0.0 max 1.0\n",
      "train input shape (272, 60, 800, 1) Nomralized MEAN: 0.51123 min 0.0 max 1.0\n",
      "test labels shape (69, 1, 800, 1) Nomralized MEAN: 0.31958 min 0.0 max 1.0\n",
      "train labels shape (272, 1, 800, 1) Nomralized MEAN: 0.28003 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') # or ('EEG_data.mat')\n",
    "# mat_contents = sio.loadmat('/home/amplifier/home/DATASETS/EEG_big.mat')\n",
    "# mat_contents = sio.loadmat('/home/amplifier/home/DATASETS/EEG_big4CNN.mat')\n",
    "mat_contents = sio.loadmat('/home/amplifier/home/DATASETS/KOS_100Hz_ICA->0.5-38Hz,Env=1.mat')\n",
    "\n",
    "\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=0.20, shuffle=False)\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 minmax normalize:\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",
    "\n",
    "leng = X.shape[2]\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": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 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",
    "# we first define the autoencoder's architecture:\n",
    "\n",
    "from keras.layers import Input, Dense, Conv2D, MaxPooling2D, UpSampling2D, BatchNormalization, Dropout\n",
    "from keras.models import Model\n",
    "from keras import optimizers\n",
    "from keras import backend as K\n",
    "from keras.models import load_model\n",
    "from keras.callbacks import TensorBoard, EarlyStopping\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "class LossHistory(Callback):\n",
    "    def on_train_begin(self, logs={}):\n",
    "        self.losses = []\n",
    "\n",
    "    def on_batch_end(self, batch, logs={}):\n",
    "        self.losses.append(logs.get('loss'))\n",
    "\n",
    "history = LossHistory()\n",
    "\n",
    "def Encoder():\n",
    "    input_img = Input(shape=(60, leng, 1))  # adapt this if using `channels_first` image data format  \n",
    "    \n",
    "    e1 = Conv2D(4, (2, 2), activation='relu', padding='same')(input_img)\n",
    "    e1b = BatchNormalization(axis=1)(e1)\n",
    "    e1a = Dropout(0.1)(e1b)\n",
    "    e2 = MaxPooling2D((2, 2), padding='same')(e1a)\n",
    "    \n",
    "    e3 = Conv2D(8, (2, 2), activation='relu', padding='same')(e2)\n",
    "    e3b = BatchNormalization(axis=1)(e3)\n",
    "    e3a = Dropout(0.1)(e3b)\n",
    "    e4 = MaxPooling2D((3, 2), padding='same')(e3a)\n",
    "    \n",
    "    e5 = Conv2D(16, (2, 2), activation='relu', padding='same')(e4)\n",
    "    e5b = BatchNormalization(axis=1)(e5)\n",
    "    e5a = Dropout(0.2)(e5b)\n",
    "    e6 = MaxPooling2D((2, 2), padding='same')(e5a)\n",
    "    \n",
    "    return Model(input_img, e6)\n",
    "\n",
    "def Decoder():\n",
    "    input_img = Input(shape=(5, 100, 16))  # adapt this if using `channels_first` image data format   \n",
    "    d1 = Conv2D(16, (2, 2), activation='relu', padding='same')(input_img)\n",
    "    d1b = BatchNormalization(axis=1)(d1)\n",
    "    d1a = Dropout(0.2)(d1b)\n",
    "    d2 = UpSampling2D((2, 2))(d1a)\n",
    "    d3 = Conv2D(8, (3, 3), activation='relu', padding='same')(d2)\n",
    "    d3b = BatchNormalization(axis=1)(d3)\n",
    "    d3a = Dropout(0.3)(d3b)\n",
    "    d4 = UpSampling2D((2, 2))(d3a)\n",
    "    d5 = Conv2D(4, (3, 3), activation='relu', padding='same')(d4)\n",
    "    d5b = BatchNormalization(axis=1)(d5)\n",
    "    d5a = Dropout(0.4)(d5b)\n",
    "    d6 = UpSampling2D((3, 2))(d5a)\n",
    "    d7 = Conv2D(1, (3, 3), activation='relu', padding='same')(d6)\n",
    "    return Model(input_img, d7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def correlation_coefficient_loss(y_true, y_pred):\n",
    "    x = y_true\n",
    "    y = y_pred\n",
    "    mx = K.mean(x)\n",
    "    my = K.mean(y)\n",
    "    xm, ym = x-mx, y-my\n",
    "    r_num = K.sum(tf.multiply(xm,ym))\n",
    "    r_den = K.sqrt(tf.multiply(K.sum(K.square(xm)), K.sum(K.square(ym))))\n",
    "    r = r_num / r_den\n",
    "\n",
    "    r = K.maximum(K.minimum(r, 1.0), -1.0)\n",
    "    return 1 - K.square(r)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define input to the model:\n",
    "x = Input(shape=(60, leng, 1))\n",
    "\n",
    "# make the encoder:\n",
    "enc = Model(x, Encoder()(x))\n",
    "# enc.compile(optimizer='adadelta', loss='binary_crossentropy')\n",
    "\n",
    "# make the model:\n",
    "autoencoder = Model(x, Decoder()(enc(x)))\n",
    "\n",
    "# compile the model:\n",
    "# sgd = optimizers.SGD(lr=0.01, decay=1e-6, momentum=0.9, nesterov=True)\n",
    "# autoencoder.compile(optimizer=sgd, loss=correlation_coefficient_loss)\n",
    "autoencoder.compile(optimizer='adadelta', loss=correlation_coefficient_loss)\n",
    "# autoencoder.compile(optimizer='adadelta', loss='binary_crossentropy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_17 (InputLayer)        (None, 60, 800, 1)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_43 (Conv2D)           (None, 60, 800, 4)        20        \n",
      "_________________________________________________________________\n",
      "batch_normalization_37 (Batc (None, 60, 800, 4)        240       \n",
      "_________________________________________________________________\n",
      "dropout_37 (Dropout)         (None, 60, 800, 4)        0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_19 (MaxPooling (None, 30, 400, 4)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_44 (Conv2D)           (None, 30, 400, 8)        136       \n",
      "_________________________________________________________________\n",
      "batch_normalization_38 (Batc (None, 30, 400, 8)        120       \n",
      "_________________________________________________________________\n",
      "dropout_38 (Dropout)         (None, 30, 400, 8)        0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_20 (MaxPooling (None, 10, 200, 8)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_45 (Conv2D)           (None, 10, 200, 16)       528       \n",
      "_________________________________________________________________\n",
      "batch_normalization_39 (Batc (None, 10, 200, 16)       40        \n",
      "_________________________________________________________________\n",
      "dropout_39 (Dropout)         (None, 10, 200, 16)       0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_21 (MaxPooling (None, 5, 100, 16)        0         \n",
      "=================================================================\n",
      "Total params: 1,084\n",
      "Trainable params: 884\n",
      "Non-trainable params: 200\n",
      "_________________________________________________________________\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_18 (InputLayer)        (None, 5, 100, 16)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_46 (Conv2D)           (None, 5, 100, 16)        1040      \n",
      "_________________________________________________________________\n",
      "batch_normalization_40 (Batc (None, 5, 100, 16)        20        \n",
      "_________________________________________________________________\n",
      "dropout_40 (Dropout)         (None, 5, 100, 16)        0         \n",
      "_________________________________________________________________\n",
      "up_sampling2d_19 (UpSampling (None, 10, 200, 16)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_47 (Conv2D)           (None, 10, 200, 8)        1160      \n",
      "_________________________________________________________________\n",
      "batch_normalization_41 (Batc (None, 10, 200, 8)        40        \n",
      "_________________________________________________________________\n",
      "dropout_41 (Dropout)         (None, 10, 200, 8)        0         \n",
      "_________________________________________________________________\n",
      "up_sampling2d_20 (UpSampling (None, 20, 400, 8)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_48 (Conv2D)           (None, 20, 400, 4)        292       \n",
      "_________________________________________________________________\n",
      "batch_normalization_42 (Batc (None, 20, 400, 4)        80        \n",
      "_________________________________________________________________\n",
      "dropout_42 (Dropout)         (None, 20, 400, 4)        0         \n",
      "_________________________________________________________________\n",
      "up_sampling2d_21 (UpSampling (None, 60, 800, 4)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_49 (Conv2D)           (None, 60, 800, 1)        37        \n",
      "=================================================================\n",
      "Total params: 2,669\n",
      "Trainable params: 2,599\n",
      "Non-trainable params: 70\n",
      "_________________________________________________________________\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_14 (InputLayer)        (None, 60, 800, 1)        0         \n",
      "_________________________________________________________________\n",
      "model_18 (Model)             (None, 5, 100, 16)        1084      \n",
      "_________________________________________________________________\n",
      "model_19 (Model)             (None, 60, 800, 1)        2669      \n",
      "=================================================================\n",
      "Total params: 3,753\n",
      "Trainable params: 3,483\n",
      "Non-trainable params: 270\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# SEE INSIDE THE ENCODER AND DECODER:\n",
    "Encoder().summary()\n",
    "Decoder().summary()\n",
    "\n",
    "# IN FACT, OUR AUTOENCODER WILL BE A CONCATENATION OF TWO MODELS\n",
    "autoencoder.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'Channels 11 in test samples 6 and 16')"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f274864c8d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# inspect channel 11\n",
    "%matplotlib inline\n",
    "plt.plot(range(0, leng), x_test[15,10,:,0], range(0, leng), x_test[15,10,:,0])\n",
    "ax = plt.gca()\n",
    "ax.set_title(\"Channels 11 in test samples 6 and 16\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Training time!\n",
    "\n",
    "early_stopping = EarlyStopping(monitor='val_loss', patience=50, mode='min')\n",
    "train_history = autoencoder.fit(x_train, x_train,\n",
    "                epochs=1000,\n",
    "                batch_size=40,\n",
    "                shuffle=True,\n",
    "                validation_data=(x_test, x_test),\n",
    "                callbacks=[TensorBoard(log_dir='/home/amplifier/home/NEW_DL/models/CNN_stim'), early_stopping])\n",
    "\n",
    "# save the entire model and the weights after the training:\n",
    "autoencoder.save('/home/amplifier/home/NEW_DL/models/CNN_autoencoder.h5')\n",
    "autoencoder.save_weights('/home/amplifier/home/NEW_DL/weights/CNN_autoencoder_weights.h5')\n",
    "enc.save('/home/amplifier/home/NEW_DL/models/CNN_encoder_model.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f26c1114748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# print(np.array(history.losses))\n",
    "loss = train_history.history['loss']\n",
    "val_loss = train_history.history['val_loss']\n",
    "plt.plot(loss)\n",
    "plt.plot(val_loss)\n",
    "plt.legend(['loss', 'val_loss'])\n",
    "plt.title('SGD+momentum+Nesterov')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(272, 60, 800, 1)\n",
      "(69, 60, 800, 1)\n"
     ]
    }
   ],
   "source": [
    "# let's predict a TRAINING sammple:\n",
    "decoded_train = autoencoder.predict(x_train)\n",
    "decoded_test = autoencoder.predict(x_test)\n",
    "print(decoded_train.shape)\n",
    "print(decoded_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training sample 0 : Correlation  [ 0.72512519]  P-value:  [ 0.]\n",
      "Training sample 1 : Correlation  [ 0.69653434]  P-value:  [ 0.]\n",
      "Training sample 2 : Correlation  [ 0.70283097]  P-value:  [ 0.]\n",
      "Training sample 3 : Correlation  [ 0.66329432]  P-value:  [ 0.]\n",
      "Training sample 4 : Correlation  [ 0.57558787]  P-value:  [ 0.]\n",
      "Training sample 5 : Correlation  [ 0.58472562]  P-value:  [ 0.]\n",
      "Training sample 6 : Correlation  [ 0.59666264]  P-value:  [ 0.]\n",
      "Training sample 7 : Correlation  [ 0.61557615]  P-value:  [ 0.]\n",
      "Training sample 8 : Correlation  [ 0.59532315]  P-value:  [ 0.]\n",
      "Training sample 9 : Correlation  [ 0.60760808]  P-value:  [ 0.]\n",
      "Training sample 10 : Correlation  [ 0.58613318]  P-value:  [ 0.]\n",
      "Training sample 11 : Correlation  [ 0.68729746]  P-value:  [ 0.]\n",
      "Training sample 12 : Correlation  [ 0.73285389]  P-value:  [ 0.]\n",
      "Training sample 13 : Correlation  [ 0.6953128]  P-value:  [ 0.]\n",
      "Training sample 14 : Correlation  [ 0.67321795]  P-value:  [ 0.]\n",
      "Training sample 15 : Correlation  [ 0.67213076]  P-value:  [ 0.]\n",
      "Training sample 16 : Correlation  [ 0.63977569]  P-value:  [ 0.]\n",
      "Training sample 17 : Correlation  [ 0.59081036]  P-value:  [ 0.]\n",
      "Training sample 18 : Correlation  [ 0.56869239]  P-value:  [ 0.]\n",
      "Training sample 19 : Correlation  [ 0.5968284]  P-value:  [ 0.]\n",
      "Training sample 20 : Correlation  [ 0.65059972]  P-value:  [ 0.]\n",
      "Training sample 21 : Correlation  [ 0.6048426]  P-value:  [ 0.]\n",
      "Training sample 22 : Correlation  [ 0.68885905]  P-value:  [ 0.]\n",
      "Training sample 23 : Correlation  [ 0.66498047]  P-value:  [ 0.]\n",
      "Training sample 24 : Correlation  [ 0.64053309]  P-value:  [ 0.]\n",
      "Training sample 25 : Correlation  [ 0.56795651]  P-value:  [ 0.]\n",
      "Training sample 26 : Correlation  [ 0.60587609]  P-value:  [ 0.]\n",
      "Training sample 27 : Correlation  [ 0.62098885]  P-value:  [ 0.]\n",
      "Training sample 28 : Correlation  [ 0.58434832]  P-value:  [ 0.]\n",
      "Training sample 29 : Correlation  [ 0.58791381]  P-value:  [ 0.]\n",
      "Training sample 30 : Correlation  [ 0.59627151]  P-value:  [ 0.]\n",
      "Training sample 31 : Correlation  [ 0.51908231]  P-value:  [ 0.]\n",
      "Training sample 32 : Correlation  [ 0.50090486]  P-value:  [ 0.]\n",
      "Training sample 33 : Correlation  [ 0.69622874]  P-value:  [ 0.]\n",
      "Training sample 34 : Correlation  [ 0.64921844]  P-value:  [ 0.]\n",
      "Training sample 35 : Correlation  [ 0.61981648]  P-value:  [ 0.]\n",
      "Training sample 36 : Correlation  [ 0.57408589]  P-value:  [ 0.]\n",
      "Training sample 37 : Correlation  [ 0.59009141]  P-value:  [ 0.]\n",
      "Training sample 38 : Correlation  [ 0.54275024]  P-value:  [ 0.]\n",
      "Training sample 39 : Correlation  [ 0.5194034]  P-value:  [ 0.]\n",
      "Training sample 40 : Correlation  [ 0.51830995]  P-value:  [ 0.]\n",
      "Training sample 41 : Correlation  [ 0.50253659]  P-value:  [ 0.]\n",
      "Training sample 42 : Correlation  [ 0.52499706]  P-value:  [ 0.]\n",
      "Training sample 43 : Correlation  [ 0.53262347]  P-value:  [ 0.]\n",
      "Training sample 44 : Correlation  [ 0.62816375]  P-value:  [ 0.]\n",
      "Training sample 45 : Correlation  [ 0.57020521]  P-value:  [ 0.]\n",
      "Training sample 46 : Correlation  [ 0.45772502]  P-value:  [  1.11823617e-42]\n",
      "Training sample 47 : Correlation  [ 0.50384074]  P-value:  [ 0.]\n",
      "Training sample 48 : Correlation  [ 0.57739836]  P-value:  [ 0.]\n",
      "Training sample 49 : Correlation  [ 0.58500427]  P-value:  [ 0.]\n",
      "Training sample 50 : Correlation  [ 0.55523902]  P-value:  [ 0.]\n",
      "Training sample 51 : Correlation  [ 0.60016996]  P-value:  [ 0.]\n",
      "Training sample 52 : Correlation  [ 0.56619906]  P-value:  [ 0.]\n",
      "Training sample 53 : Correlation  [ 0.5837115]  P-value:  [ 0.]\n",
      "Training sample 54 : Correlation  [ 0.58965319]  P-value:  [ 0.]\n",
      "Training sample 55 : Correlation  [ 0.69768339]  P-value:  [ 0.]\n",
      "Training sample 56 : Correlation  [ 0.56241101]  P-value:  [ 0.]\n",
      "Training sample 57 : Correlation  [ 0.56337184]  P-value:  [ 0.]\n",
      "Training sample 58 : Correlation  [ 0.53263491]  P-value:  [ 0.]\n",
      "Training sample 59 : Correlation  [ 0.48121798]  P-value:  [ 0.]\n",
      "Training sample 60 : Correlation  [ 0.50621998]  P-value:  [ 0.]\n",
      "Training sample 61 : Correlation  [ 0.5073247]  P-value:  [ 0.]\n",
      "Training sample 62 : Correlation  [ 0.55404162]  P-value:  [ 0.]\n",
      "Training sample 63 : Correlation  [ 0.59487045]  P-value:  [ 0.]\n",
      "Training sample 64 : Correlation  [ 0.59939456]  P-value:  [ 0.]\n",
      "Training sample 65 : Correlation  [ 0.63612252]  P-value:  [ 0.]\n",
      "Training sample 66 : Correlation  [ 0.7091164]  P-value:  [ 0.]\n",
      "Training sample 67 : Correlation  [ 0.63655978]  P-value:  [ 0.]\n",
      "Training sample 68 : Correlation  [ 0.55317593]  P-value:  [ 0.]\n",
      "Training sample 69 : Correlation  [ 0.57483721]  P-value:  [ 0.]\n",
      "Training sample 70 : Correlation  [ 0.53423423]  P-value:  [ 0.]\n",
      "Training sample 71 : Correlation  [ 0.50506693]  P-value:  [ 0.]\n",
      "Training sample 72 : Correlation  [ 0.45087737]  P-value:  [  2.59030021e-41]\n",
      "Training sample 73 : Correlation  [ 0.50760466]  P-value:  [ 0.]\n",
      "Training sample 74 : Correlation  [ 0.50119436]  P-value:  [ 0.]\n",
      "Training sample 75 : Correlation  [ 0.55052817]  P-value:  [ 0.]\n",
      "Training sample 76 : Correlation  [ 0.48034614]  P-value:  [ 0.]\n",
      "Training sample 77 : Correlation  [ 0.63280076]  P-value:  [ 0.]\n",
      "Training sample 78 : Correlation  [ 0.59679335]  P-value:  [ 0.]\n",
      "Training sample 79 : Correlation  [ 0.56530094]  P-value:  [ 0.]\n",
      "Training sample 80 : Correlation  [ 0.5845257]  P-value:  [ 0.]\n",
      "Training sample 81 : Correlation  [ 0.62785161]  P-value:  [ 0.]\n",
      "Training sample 82 : Correlation  [ 0.61857677]  P-value:  [ 0.]\n",
      "Training sample 83 : Correlation  [ 0.59156716]  P-value:  [ 0.]\n",
      "Training sample 84 : Correlation  [ 0.59972721]  P-value:  [ 0.]\n",
      "Training sample 85 : Correlation  [ 0.61033767]  P-value:  [ 0.]\n",
      "Training sample 86 : Correlation  [ 0.64137357]  P-value:  [ 0.]\n",
      "Training sample 87 : Correlation  [ 0.62935948]  P-value:  [ 0.]\n",
      "Training sample 88 : Correlation  [ 0.66572851]  P-value:  [ 0.]\n",
      "Training sample 89 : Correlation  [ 0.6182729]  P-value:  [ 0.]\n",
      "Training sample 90 : Correlation  [ 0.5424692]  P-value:  [ 0.]\n",
      "Training sample 91 : Correlation  [ 0.5536719]  P-value:  [ 0.]\n",
      "Training sample 92 : Correlation  [ 0.5242883]  P-value:  [ 0.]\n",
      "Training sample 93 : Correlation  [ 0.46582344]  P-value:  [  2.52233724e-44]\n",
      "Training sample 94 : Correlation  [ 0.47921476]  P-value:  [ 0.]\n",
      "Training sample 95 : Correlation  [ 0.49751821]  P-value:  [ 0.]\n",
      "Training sample 96 : Correlation  [ 0.51756227]  P-value:  [ 0.]\n",
      "Training sample 97 : Correlation  [ 0.52042943]  P-value:  [ 0.]\n",
      "Training sample 98 : Correlation  [ 0.51317507]  P-value:  [ 0.]\n",
      "Training sample 99 : Correlation  [ 0.63788188]  P-value:  [ 0.]\n",
      "Training sample 100 : Correlation  [ 0.59675699]  P-value:  [ 0.]\n",
      "Training sample 101 : Correlation  [ 0.58199209]  P-value:  [ 0.]\n",
      "Training sample 102 : Correlation  [ 0.57996815]  P-value:  [ 0.]\n",
      "Training sample 103 : Correlation  [ 0.61365193]  P-value:  [ 0.]\n",
      "Training sample 104 : Correlation  [ 0.58961362]  P-value:  [ 0.]\n",
      "Training sample 105 : Correlation  [ 0.51670063]  P-value:  [ 0.]\n",
      "Training sample 106 : Correlation  [ 0.49099734]  P-value:  [ 0.]\n",
      "Training sample 107 : Correlation  [ 0.5050087]  P-value:  [ 0.]\n",
      "Training sample 108 : Correlation  [ 0.4923166]  P-value:  [ 0.]\n",
      "Training sample 109 : Correlation  [ 0.48752767]  P-value:  [ 0.]\n",
      "Training sample 110 : Correlation  [ 0.53144401]  P-value:  [ 0.]\n",
      "Training sample 111 : Correlation  [ 0.5395515]  P-value:  [ 0.]\n",
      "Training sample 112 : Correlation  [ 0.56106871]  P-value:  [ 0.]\n",
      "Training sample 113 : Correlation  [ 0.60903364]  P-value:  [ 0.]\n",
      "Training sample 114 : Correlation  [ 0.64635283]  P-value:  [ 0.]\n",
      "Training sample 115 : Correlation  [ 0.60837972]  P-value:  [ 0.]\n",
      "Training sample 116 : Correlation  [ 0.63307434]  P-value:  [ 0.]\n",
      "Training sample 117 : Correlation  [ 0.61865276]  P-value:  [ 0.]\n",
      "Training sample 118 : Correlation  [ 0.59451604]  P-value:  [ 0.]\n",
      "Training sample 119 : Correlation  [ 0.54284239]  P-value:  [ 0.]\n",
      "Training sample 120 : Correlation  [ 0.54827797]  P-value:  [ 0.]\n",
      "Training sample 121 : Correlation  [ 0.72562969]  P-value:  [ 0.]\n",
      "Training sample 122 : Correlation  [ 0.63151735]  P-value:  [ 0.]\n",
      "Training sample 123 : Correlation  [ 0.58248585]  P-value:  [ 0.]\n",
      "Training sample 124 : Correlation  [ 0.6090433]  P-value:  [ 0.]\n",
      "Training sample 125 : Correlation  [ 0.58180213]  P-value:  [ 0.]\n",
      "Training sample 126 : Correlation  [ 0.53818864]  P-value:  [ 0.]\n",
      "Training sample 127 : Correlation  [ 0.56627399]  P-value:  [ 0.]\n",
      "Training sample 128 : Correlation  [ 0.54348224]  P-value:  [ 0.]\n",
      "Training sample 129 : Correlation  [ 0.5294525]  P-value:  [ 0.]\n",
      "Training sample 130 : Correlation  [ 0.5255841]  P-value:  [ 0.]\n",
      "Training sample 131 : Correlation  [ 0.57575655]  P-value:  [ 0.]\n",
      "Training sample 132 : Correlation  [ 0.6493755]  P-value:  [ 0.]\n",
      "Training sample 133 : Correlation  [ 0.56349105]  P-value:  [ 0.]\n",
      "Training sample 134 : Correlation  [ 0.5252279]  P-value:  [ 0.]\n",
      "Training sample 135 : Correlation  [ 0.54541582]  P-value:  [ 0.]\n",
      "Training sample 136 : Correlation  [ 0.53632891]  P-value:  [ 0.]\n",
      "Training sample 137 : Correlation  [ 0.53316933]  P-value:  [ 0.]\n",
      "Training sample 138 : Correlation  [ 0.52274644]  P-value:  [ 0.]\n",
      "Training sample 139 : Correlation  [ 0.51830167]  P-value:  [ 0.]\n",
      "Training sample 140 : Correlation  [ 0.54493409]  P-value:  [ 0.]\n",
      "Training sample 141 : Correlation  [ 0.58640552]  P-value:  [ 0.]\n",
      "Training sample 142 : Correlation  [ 0.62140459]  P-value:  [ 0.]\n",
      "Training sample 143 : Correlation  [ 0.68396825]  P-value:  [ 0.]\n",
      "Training sample 144 : Correlation  [ 0.6366114]  P-value:  [ 0.]\n",
      "Training sample 145 : Correlation  [ 0.62140971]  P-value:  [ 0.]\n",
      "Training sample 146 : Correlation  [ 0.60613394]  P-value:  [ 0.]\n",
      "Training sample 147 : Correlation  [ 0.5946123]  P-value:  [ 0.]\n",
      "Training sample 148 : Correlation  [ 0.60112846]  P-value:  [ 0.]\n",
      "Training sample 149 : Correlation  [ 0.5492869]  P-value:  [ 0.]\n",
      "Training sample 150 : Correlation  [ 0.48686039]  P-value:  [ 0.]\n",
      "Training sample 151 : Correlation  [ 0.52602828]  P-value:  [ 0.]\n",
      "Training sample 152 : Correlation  [ 0.52388102]  P-value:  [ 0.]\n",
      "Training sample 153 : Correlation  [ 0.533777]  P-value:  [ 0.]\n",
      "Training sample 154 : Correlation  [ 0.61822748]  P-value:  [ 0.]\n",
      "Training sample 155 : Correlation  [ 0.64185131]  P-value:  [ 0.]\n",
      "Training sample 156 : Correlation  [ 0.63346148]  P-value:  [ 0.]\n",
      "Training sample 157 : Correlation  [ 0.62223411]  P-value:  [ 0.]\n",
      "Training sample 158 : Correlation  [ 0.66624641]  P-value:  [ 0.]\n",
      "Training sample 159 : Correlation  [ 0.65753311]  P-value:  [ 0.]\n",
      "Training sample 160 : Correlation  [ 0.59973383]  P-value:  [ 0.]\n",
      "Training sample 161 : Correlation  [ 0.57205391]  P-value:  [ 0.]\n",
      "Training sample 162 : Correlation  [ 0.56525034]  P-value:  [ 0.]\n",
      "Training sample 163 : Correlation  [ 0.53163892]  P-value:  [ 0.]\n",
      "Training sample 164 : Correlation  [ 0.51840734]  P-value:  [ 0.]\n",
      "Training sample 165 : Correlation  [ 0.53985906]  P-value:  [ 0.]\n",
      "Training sample 166 : Correlation  [ 0.5206185]  P-value:  [ 0.]\n",
      "Training sample 167 : Correlation  [ 0.57068968]  P-value:  [ 0.]\n",
      "Training sample 168 : Correlation  [ 0.5797441]  P-value:  [ 0.]\n",
      "Training sample 169 : Correlation  [ 0.57496035]  P-value:  [ 0.]\n",
      "Training sample 170 : Correlation  [ 0.58204609]  P-value:  [ 0.]\n",
      "Training sample 171 : Correlation  [ 0.61448634]  P-value:  [ 0.]\n",
      "Training sample 172 : Correlation  [ 0.61472112]  P-value:  [ 0.]\n",
      "Training sample 173 : Correlation  [ 0.65809888]  P-value:  [ 0.]\n",
      "Training sample 174 : Correlation  [ 0.66290534]  P-value:  [ 0.]\n",
      "Training sample 175 : Correlation  [ 0.69544011]  P-value:  [ 0.]\n",
      "Training sample 176 : Correlation  [ 0.58673865]  P-value:  [ 0.]\n",
      "Training sample 177 : Correlation  [ 0.58235025]  P-value:  [ 0.]\n",
      "Training sample 178 : Correlation  [ 0.53184861]  P-value:  [ 0.]\n",
      "Training sample 179 : Correlation  [ 0.54843426]  P-value:  [ 0.]\n",
      "Training sample 180 : Correlation  [ 0.56391793]  P-value:  [ 0.]\n",
      "Training sample 181 : Correlation  [ 0.47472656]  P-value:  [ 0.]\n",
      "Training sample 182 : Correlation  [ 0.51365399]  P-value:  [ 0.]\n",
      "Training sample 183 : Correlation  [ 0.55184138]  P-value:  [ 0.]\n",
      "Training sample 184 : Correlation  [ 0.54960787]  P-value:  [ 0.]\n",
      "Training sample 185 : Correlation  [ 0.5446164]  P-value:  [ 0.]\n",
      "Training sample 186 : Correlation  [ 0.62471741]  P-value:  [ 0.]\n",
      "Training sample 187 : Correlation  [ 0.60888386]  P-value:  [ 0.]\n",
      "Training sample 188 : Correlation  [ 0.54281926]  P-value:  [ 0.]\n",
      "Training sample 189 : Correlation  [ 0.59903491]  P-value:  [ 0.]\n",
      "Training sample 190 : Correlation  [ 0.5981192]  P-value:  [ 0.]\n",
      "Training sample 191 : Correlation  [ 0.62817872]  P-value:  [ 0.]\n",
      "Training sample 192 : Correlation  [ 0.61975276]  P-value:  [ 0.]\n",
      "Training sample 193 : Correlation  [ 0.63511634]  P-value:  [ 0.]\n",
      "Training sample 194 : Correlation  [ 0.62359536]  P-value:  [ 0.]\n",
      "Training sample 195 : Correlation  [ 0.6070025]  P-value:  [ 0.]\n",
      "Training sample 196 : Correlation  [ 0.61484879]  P-value:  [ 0.]\n",
      "Training sample 197 : Correlation  [ 0.59893882]  P-value:  [ 0.]\n",
      "Training sample 198 : Correlation  [ 0.57576293]  P-value:  [ 0.]\n",
      "Training sample 199 : Correlation  [ 0.53485334]  P-value:  [ 0.]\n",
      "Training sample 200 : Correlation  [ 0.49898109]  P-value:  [ 0.]\n",
      "Training sample 201 : Correlation  [ 0.51985031]  P-value:  [ 0.]\n",
      "Training sample 202 : Correlation  [ 0.51961273]  P-value:  [ 0.]\n",
      "Training sample 203 : Correlation  [ 0.50706023]  P-value:  [ 0.]\n",
      "Training sample 204 : Correlation  [ 0.51530701]  P-value:  [ 0.]\n",
      "Training sample 205 : Correlation  [ 0.61308146]  P-value:  [ 0.]\n",
      "Training sample 206 : Correlation  [ 0.03014516]  P-value:  [ 0.39449444]\n",
      "Training sample 207 : Correlation  [ 0.04253848]  P-value:  [ 0.2294341]\n",
      "Training sample 208 : Correlation  [ 0.02839145]  P-value:  [ 0.42260087]\n",
      "Training sample 209 : Correlation  [-0.04397366]  P-value:  [ 0.21408063]\n",
      "Training sample 210 : Correlation  [-0.30413526]  P-value:  [  1.39221363e-18]\n",
      "Training sample 211 : Correlation  [-0.37226036]  P-value:  [  1.06622351e-27]\n",
      "Training sample 212 : Correlation  [-0.23904441]  P-value:  [  7.37579095e-12]\n",
      "Training sample 213 : Correlation  [-0.07468358]  P-value:  [ 0.03468514]\n",
      "Training sample 214 : Correlation  [-0.04894844]  P-value:  [ 0.16662532]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training sample 215 : Correlation  [ 0.05555185]  P-value:  [ 0.11641396]\n",
      "Training sample 216 : Correlation  [ 0.34147102]  P-value:  [  2.68994059e-23]\n",
      "Training sample 217 : Correlation  [ 0.43255591]  P-value:  [  8.24623172e-38]\n",
      "Training sample 218 : Correlation  [ 0.46124902]  P-value:  [  2.15799964e-43]\n",
      "Training sample 219 : Correlation  [ 0.54922521]  P-value:  [ 0.]\n",
      "Training sample 220 : Correlation  [ 0.57126719]  P-value:  [ 0.]\n",
      "Training sample 221 : Correlation  [ 0.4697971]  P-value:  [  4.20389539e-45]\n",
      "Training sample 222 : Correlation  [ 0.45577818]  P-value:  [  2.75215018e-42]\n",
      "Training sample 223 : Correlation  [ 0.51641411]  P-value:  [ 0.]\n",
      "Training sample 224 : Correlation  [ 0.56202596]  P-value:  [ 0.]\n",
      "Training sample 225 : Correlation  [ 0.58367354]  P-value:  [ 0.]\n",
      "Training sample 226 : Correlation  [ 0.61344767]  P-value:  [ 0.]\n",
      "Training sample 227 : Correlation  [ 0.61795062]  P-value:  [ 0.]\n",
      "Training sample 228 : Correlation  [ 0.62971103]  P-value:  [ 0.]\n",
      "Training sample 229 : Correlation  [ 0.58921373]  P-value:  [ 0.]\n",
      "Training sample 230 : Correlation  [ 0.61705589]  P-value:  [ 0.]\n",
      "Training sample 231 : Correlation  [ 0.51662022]  P-value:  [ 0.]\n",
      "Training sample 232 : Correlation  [ 0.51210231]  P-value:  [ 0.]\n",
      "Training sample 233 : Correlation  [ 0.529778]  P-value:  [ 0.]\n",
      "Training sample 234 : Correlation  [ 0.55735368]  P-value:  [ 0.]\n",
      "Training sample 235 : Correlation  [ 0.57888454]  P-value:  [ 0.]\n",
      "Training sample 236 : Correlation  [ 0.56506413]  P-value:  [ 0.]\n",
      "Training sample 237 : Correlation  [ 0.5966211]  P-value:  [ 0.]\n",
      "Training sample 238 : Correlation  [ 0.57281822]  P-value:  [ 0.]\n",
      "Training sample 239 : Correlation  [ 0.54658842]  P-value:  [ 0.]\n",
      "Training sample 240 : Correlation  [ 0.56237382]  P-value:  [ 0.]\n",
      "Training sample 241 : Correlation  [ 0.56414115]  P-value:  [ 0.]\n",
      "Training sample 242 : Correlation  [ 0.67936641]  P-value:  [ 0.]\n",
      "Training sample 243 : Correlation  [ 0.62803185]  P-value:  [ 0.]\n",
      "Training sample 244 : Correlation  [ 0.55591792]  P-value:  [ 0.]\n",
      "Training sample 245 : Correlation  [ 0.50381839]  P-value:  [ 0.]\n",
      "Training sample 246 : Correlation  [ 0.51273537]  P-value:  [ 0.]\n",
      "Training sample 247 : Correlation  [ 0.5330767]  P-value:  [ 0.]\n",
      "Training sample 248 : Correlation  [ 0.51244658]  P-value:  [ 0.]\n",
      "Training sample 249 : Correlation  [ 0.52166748]  P-value:  [ 0.]\n",
      "Training sample 250 : Correlation  [ 0.57698989]  P-value:  [ 0.]\n",
      "Training sample 251 : Correlation  [ 0.63070387]  P-value:  [ 0.]\n",
      "Training sample 252 : Correlation  [ 0.59705973]  P-value:  [ 0.]\n",
      "Training sample 253 : Correlation  [ 0.58191091]  P-value:  [ 0.]\n",
      "Training sample 254 : Correlation  [ 0.53803819]  P-value:  [ 0.]\n",
      "Training sample 255 : Correlation  [ 0.58788937]  P-value:  [ 0.]\n",
      "Training sample 256 : Correlation  [ 0.63422292]  P-value:  [ 0.]\n",
      "Training sample 257 : Correlation  [ 0.59052527]  P-value:  [ 0.]\n",
      "Training sample 258 : Correlation  [ 0.60122585]  P-value:  [ 0.]\n",
      "Training sample 259 : Correlation  [ 0.62443751]  P-value:  [ 0.]\n",
      "Training sample 260 : Correlation  [ 0.56889868]  P-value:  [ 0.]\n",
      "Training sample 261 : Correlation  [ 0.55770761]  P-value:  [ 0.]\n",
      "Training sample 262 : Correlation  [ 0.54392529]  P-value:  [ 0.]\n",
      "Training sample 263 : Correlation  [ 0.53842884]  P-value:  [ 0.]\n",
      "Training sample 264 : Correlation  [ 0.64872473]  P-value:  [ 0.]\n",
      "Training sample 265 : Correlation  [ 0.61079443]  P-value:  [ 0.]\n",
      "Training sample 266 : Correlation  [ 0.58023387]  P-value:  [ 0.]\n",
      "Training sample 267 : Correlation  [ 0.5655061]  P-value:  [ 0.]\n",
      "Training sample 268 : Correlation  [ 0.55126512]  P-value:  [ 0.]\n",
      "Training sample 269 : Correlation  [ 0.56137615]  P-value:  [ 0.]\n",
      "Training sample 270 : Correlation  [ 0.56120503]  P-value:  [ 0.]\n",
      "Training sample 271 : Correlation  [ 0.55778074]  P-value:  [ 0.]\n",
      "average correlation =  0.551925\n",
      "Pearson's R: [ 0.55778074] p-value [ 0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAagAAAEYCAYAAAAJeGK1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvAOZPmwAAIABJREFUeJzsfXmYXUWZ/vvdc/fel6Sz72EnbGGRNaiIgo7MODqKI+IMOo7ijP7EdXBEBVwHccFBULYRFEVFloAsSSMJhLBlIfveSXc6nd6773ruOfX7o6rOrbPcc293upMOnPd5+unue+rWqVOnqr79+4gxhgABAgQIEGCiIXSkBxAgQIAAAQJ4ISBQAQIECBBgQiIgUAECBAgQYEIiIFABAgQIEGBCIiBQAQIECBBgQiIgUAECBAgQYEIiIFABJjyIaAkR7fO5PkxE8w7nmEqMYw4RMSIKH+FxMCJacCTH8FZCMN/jh4BAjROIaDcRZcTh2UlE9xBR9ZEelxeI6AYi+s049n8PEd04Xv0zxqoZYzvHq38JImolomvG+z4THUQ0S6xr+cOIKKX8f4F453nxfy8RPU1Ex3n0tUR8/8uOz23EXvTHiOgspc0CImLK/9b7Ufq9zdHvCiK6Wvl/KhHdSUQdYqw7xb1cYz1UTBQG5mhCQKDGF+9jjFUDOBXAaQC+doTHMyoQR7BWAgAAGGNtgimoFusbAE5RPntefPYDcX06gHYAv/bo7uMAesXvcugFMBJGJwXgKiKa43WRiJoAvAAgCeACADUATgfwHIBLRnCfAOOE4NA5DGCMdQL4KzihAgAQUYyIfkREbUR0gIhuJ6KEcv39RLSGiAaJaAcRvVt8Po2IHhFc6XYi+qTynRuI6PdEdB8RDRHRBiJarFz/ChG1i2tbiOgdot+vA/gnwUGuFW1biegmIloJIA1gnpAK3+m432+U/88noheIqJ+I9hLR1UT0KQAfBfBl0f+jynP8kYgOEtEuIvoPpZ+E4GL7iGgjgDP95ldVsYjv3UZEj4vnfImI5jva/ofglLuJ6IeS+Ho8j8XxEtFN4IfYz8Vz/NxnSP8iOPL9RPRFpb+ziOhFMT/7iejnRBQV14iIfkxEXUQ0QETriOikCtfKl0R/HUT0L2XmatTrZ7RgjGUA/B7K+hf3SwL4RwCfBbCwgnvdC2AREV1U4a37AdwD4Jslrn8BwCCAjzHGdjCOfsbY3Yyxn5Xq1G++iehyInpd7Nu9RHSDcvlvclxiDb2NiOYT0TIi6hHr8X4iqq/w+d78YIwFP+PwA2A3gHeKv2cAWA/gJ8r1WwE8AqARnHN7FMB3xbWzAAyAc3EhcA70OHHtOQC/ABAH3/AHAbxDXLsBQBbAZQA0AN8FsEpcOxbAXgDTxP9zAMxXvvcbx/hbAbQBOBFAGEBEfSbn9wDMAjAE4COibROAU8W1ewDcqHwvBOBVAP8NIApgHoCdAC4V178H4HkxNzMBvAFgn89cMwALlHv1ijkMA7gfwO8cbZeLvmcB2ArgGq95EHPEAISVObnGZxyy/W8BVAE4WbwfuQ7OAHCOGNccAJsAfF5cu1TMST0AAnA8gKkVrJV3AzgA4CRxzwfU+fAY46jWT5m17rqf+s7FuP4PwFpHm48B2C/u9SiAn/rM/T3g0tN/AFghPlsAgDnWrHyXSwDsAzAFnAgdKz5fAeBq8fcqADeMcF/7zre478nga3yRaHuF1zMpz3AJgBiASeBE7NYjfX5NlJ9AghpfPExEQ+CEoQuCkyMiAvBJAF9gjPUyxoYA3Azgw+J7/wrgLsbY04wxkzHWzhjbTEQzAZwP4CuMsSxjbA2AX4FvdIkVjLGljDED/FA4RXxugG+CE4gowhjbzRjbUWb89zDGNjDGCowxvUzbjwJ4hjH2W8aYzhjrEePzwpkAJjHGvs0YyzNuP7pTef4PAbhJzM1eAD8tc28n/sQYW80YK4ATqFMd178v+m4DP/w/MsL+y+FbjLEUY2w9gLtl/4yxVxljq8R87gbwSwBSGtDBic9xAIgxtokxtr+CtfIhAHczxt5gjKXAiYwnDnH9jAbXEVE/OONyvuM+AFfrPSju9QCAjxBRpEyfvwQwi4jeU8kAGNde3A7g2x6XmwF0yn+I6O+EdDtERE+V6NJ3vhljrYyx9WLfrgNnVkpKfIyx7WKf5xhjBwHc4tf+rYaAQI0vrmCM1YBzVceBbwiAc0pJAK+KDdEP4EnxOcClBi/iMQ2APKQk9oBLWBKdyt9pAHEiCjPGtgP4PPiG6iKi3xHRtDLj31vmuopSY/bCbADT5LOL5/86gBZxfZrj3ntGMA7APQdO5xRn3+XmYaTw7J+IjiGix4g7zQyCE5pmAGCMLQPwcwC3AThARHcQUS3Kr5WRzNWo108lD+2BHzHG6sElhwy4FA/AIpYXgzMQAPAXcKnucr8OGWM5AN8RP1ThOL4P4FIichLbHgBTlb4fEeP9Arhk7wXf+Sais4louVBdDwD4NIr73gUimiz2YrtYE7/xa/9WQ0CgDgMYY8+Bqyh+JD7qBt+wJzLG6sVPHSsanPcCmO/uCR0AGomoRvlsFrgBupJxPMAYOx+cQDDwjQvxt+dXHP+nwA9LiSnK36XG7NXPXgC7lGevZ4zVMMYuE9f3gxM8iVkl+h0tnH13iL/9ng8oPU+V9v+/ADYDWMgYqwUnytYhyxj7KWPsDHC16jEAvoTya2Ukc3VI62e0EJLqfwL4iWI7+xj4+fMoEXWCq3jjAK6qoMu7AdQB+PsK798DLil/x3HpWQBX0MgcgMrN9wPg6tiZjLE6cOlNvmOv9fNd8fkisSb+GZUT3jc9AgJ1+HArgEuI6FTGmAmu0voxEU0GACKaTkSXira/BvAJ4k4MIXHtOKHuegHAd4koTkSLwNWB93vczwYiOpaI3k5EMXA7QwZc7QdwPfmcCjbqGgAfJqKIMGj/o3LtfgDvJKIPCaeCJiKSqrUD4HYmidUABok7bSSISCOik4hIOkP8HsDXiKiBiGYA+Fy55xshviT6ngl+cD6oPN+FxN2o6+D2unQ+Ryl8g4iSRHQigE8o/deA20OGibsx/7v8AhGdKbjvCDihzAIwKlgrvwdwNRGdIJwOSjkE4FDWz6GCMfY0OIH8lPjoKgDfAle/yp8PALicuHedX18FcE3AV0YwhFsAnAtu21M/awDwf8JZgQTxdqqEVZSb7xpwKTVL3CX+SuXaQQAm7GuoBsAwuOPEdHCmJIBAQKAOE4R++T4A3xAffQXAdgCrhGj/DIQKhDG2Gvxg+zG4s8Rz4FIPwO0Zc8A3+58BfFNs/nKIgTsfdIOrcSaDc/AA8Afxu4eIXvPp4xvgUlIf+OHygPJ8beDG9S+COymsQdF+8Wtw21c/ET0sbA7vAz8Idokx/QqcK4boe4+49hS4LWQs8Rdwh4Q1AB4X45OH6IMA1onrjzm+9xMA/0jcu9DPLvYc+Lt9FlzNJe0Z14EfWEPgROdB5Tu14rM+8GfvQVHi9lsrT4AzP8tEm2Vlnn2062cs8ENwb86LxBhuY4x1Kj+PgD9DJTbB34JLMxWBMTYI4Afgjibys25wp5UsuPPEEPiaqIHCPDj6KTffnwHwbWF7/m9wgia/mwZwE4CVYi+cA77WTwff548D+FOlz/RWADEWFCwM8NYB8cDOhcImFyBAgAmMQIIKECBAgAATEgGBChAgQIAAExKBii9AgAABAkxIBBJUgAABAgSYkAgIVIAAAQIEmJAICFSAIwriCUp1kTyz6kiP52gEEX2LeLmLw17KgcapBAkVy3XsHqP+jtgcBRg9AgL1FgcRXUlErwgCsZ+IniCi8w/zMB5kvExDqsQYY0R0F/EM0Z1E9P9KdUQ8g7pB9npFSxxt/pN4BvUUEW0iomPE5xcT0XoRo9JDRH8WwZPye9OJ6C/EM4HvI6JPO/q9g3iWeJOUmkPi2u2OMeVErIy8fjzxrNYDxLOM/71y7QTxjvrEzzNEdIK8zhj7Jnj2CfV+aj2yA0R0N03QemQ++AFjbE6pi0TUKN5Rioj2ENGVpdp6zVGAiY+AQL2FIQ76W8FzwrWAp235BYD3j6IvF1c6hpzqDQAWggcrXwwe7Plun/YvMqVeEWOsVRnTNeDZEy4Hz9H3XvBAYQDYCJ5RvR4859o28PREEr8BDx5uEd+/mYguVq6vBQ/UdAU7M8Y+zew1lH4LESAt5ukv4IHBjeDZFn4jCSd4UO0/imvN4Kl0fufz/BKyHtnp4Al6r6/gO0cTbgOQB38fHwXwv8SzdwR4kyAgUG9REE/l820An2WM/Ulk39YZY48yxr4k2sSI6FbidW86xN8xcW2JkCK+QjyX2t1en43RcK8C8B3GWB9jbBN4xoWrR/HMIfDUNF9gjG1kHDsYY70AwBg7wBjrUL5igJdDgJA+loBnWdcZY2sBPATAqgfEGLuNMfYseGYCv3FUgaf1uVd8dBw4QfwxY8wQiWNXQmT/ZrxG0W7GXW5JHVclYIy1A3gCvESEcywxITGepHw2SUhfk4mnhHqMePLTPvH3jBLPVbKelvi/joh+LST1diK6kYi0Sp/DcS85h99gjA0zxlaAE25nxvQARzECAvXWxdvAk3P+2afNf4GngjkVPG3RWbBz4VPAufrZKOZYc30mDsBRqQ2JqAH88F6rfLwW/uqa04gXf9tKRN9QJLkZ4uck4sXkdhG3TVj7gHgevn7wXIXXgafHAYoJPNVEngSPQ78CfAA8L5ssYOeVHNTVtxhXFsDPwKXeikA85+BlAF53XhPZwf8Ee3qhDwF4jjHWBX5G3A3+PmeBz4tfsUY/3AugAE5cTwPwLgCjtV8dA56rcKvyWbl1EeAoQ0Cg3rpoAtAtEm+WwkcBfJsx1iVyCX4Ldg7VBM/llmO8aqrnZyID94pRjlPaTQaUzwbA86V54W/gB/tkcELwERQTcErO/13gReUuFtf/VX6Z8XLm9eCqtOvBs49DlKhYCZ4INk5Ep4v+1eznleLjAO5jxSDEzeD1wr5EPBHvu8BrAtn6FuOqA3AtPIiNBx4WRG0FeH7AUkTtAdgJ1JXiMzBe1+uPjLG0mIObMIp6RUTUAuA94AUaU4L4/RjFulYjRTXsawLwXxcBjkIEBOqtix4AzWXsRNNgr3fjrJ10kDHmVGd5fXYoGBa/a5XPasETe7rAGNvJGNvFeMG49eBqTJl1XRLRH0i1GXgBvMs8+ukF5/j/oszRRwHMBS8X8r/gWcD3jeRhhDRzEXjiYHkvHcAV4HatTvCEu7/36ls4ktwO4D4S2c19cIVgDmYzxj7DGMsQ0QWKo8YG0W4ZgATxbOqzwSXmP4vxJonol8IJYRCcAagfhWpuNnil5f1UrGv1S3BGYjQYhn1NAD7rIsDRiYBAvXXxIri66AqfNh0oZlEH7LWNAO/6NmOamoQx1geetVotNncKgA3e3/Acj1ShbQE3qlc6xjD4AVorxrKHMfZextgkxtjZ4FLo6gr7krgKwAuMVxEuDpKxdYyxixhjTYyxS8FLMpTqOwQuXU0vcb0kGGPPK84aJ4rPTHCC+BFw6ekxVixq+EXwzOlnM16v6ELxuZdasly9sByAZlasa1UrxzAKbAUQJqKFymcjWRcBjgIEBOotCsbYAHg5gNuI6ArBKUeI6D1EJO0uvwVwvTCaN4v2vynV5zjiPjGOBuJ1lD4JXgDSBTH+FvH3ceAlQv4CWOUOHgT3AqwRxv5PQpTVIKJ/IF43K0REk8DrBb0unSiIu4LXEFGUiP4ZXFV4i3LvKBHFwQ/viFAFOvfYVV5jJ6JFon2SiK4Dr/R6j7h2CRGdRrxuVq24Zx+ATSOaRX88AOCfwKXEB5TPa8Alz34iaoRPvSn41NNijO0HL53yP0RUK+Z4PvHSGyOGkCT/BF7aooqIzgP3Ph3r0iwBjiACAvUWBmPsFgD/D9zWchCcy70WwMOiyY0AXgGvj7Qe3H36xpHeR6iTLjiEoX4TvJz8HnBbyg8ZY0+KvmeJ/mVl03cAWEdEKQBLwQ8x1fZyLbh6qANcinwAwF3i2nTwcupD4M9rwl619VLwyq994KW83y1scxJPgR/m5wK4Q/wtJQ4Q0dvA7WB/gBsfA5cUu8QzXCIcGACgHpxZGBDzsEDce8xUqYyxl8AloGngHn8StwJIgLvirwKfn1J9lKundRV4KfWN4HP4EJSS66PAZ8TYusDn598ZYxsAQKoyD6HvABMAQbLYAEcURHQ9OKetA5heKlg3QGkQ0TfBGY0YgCrGC0Ie1SCiO8FVjgcYY/PHoL833Ry9FRAQqAABAgQIMCERqPgCBAgQIMCERECgAgQIECDAhMQRy+rb3NzM5syZc8j9pFIpVFUdHUmwj6axAsF4xxNH01iBYLzjiaNprMDYjPfVV1/tZoxNKtuQMXZEfs444ww2Fli+fPmY9HM4cDSNlbFgvOOJo2msjAXjHU8cTWNlbGzGC+AVVgGdCFR8AQIECBBgQiIgUAECBAgwTmCMoWCYR3oYRy0CAhUgQIBRYVd3Clk9CCfywxceXIMF//VE+YYCT76xH3O++jj60/lxHNXRg4BAjRFyBSPYrAHeMsgVDFz8o1Z8/ndrjvRQJjQeXsNTV+7pqSz+/P6X2gAAr+zuG7cxHU0ICNQY4fzvL8dx3yiZBSbAmwz5gomv/Wk9Hnp1RMnM3zRI5Tgz9uSGziM8komN6fUJAMCWzsqSrM9r5t5xO7uDLE1AQKDGDAeHcuUbBZjQ6E/ncfbNz+DLD60t23bVzh78dnUbvvfE5sMwsomHVM6vjFgAiXiEH7HpfGXalabqGACgo38sK9YcvQgIVIAAAju7UzgwmMPvXykvFR0Y5AfIYFYf72FNSAy/BQiUaR56GrhYmJfNqnS+pENF3zjboH741824eelYJsMfHwQEaozBRpjb8OBQDi/u6Bmn0YwtMnkDc776OO57cfcRuf8DL7Xh2gdeG7f+e4eLh0K5A6VLSMzhkFdZpImNx9ftx/tvW4lcYfQ203T+zU2gOvozmPf1pXhsXUf5xj6IhqUEVdl85QSB6h4eP41Me38Gty3fgTv+tvOQ1sDhQECgxhgZD0eJwayOgYw3p33VXavxkTtXYdP+wfEeWsUwTOa5cKW08NNntx3uIQEAvv7n9Xhs3X7s60uPS/+9qSKB2t3tb9SWElRGNw6rG/EtT23By7t7D6mPbz6yAWv39uOvGw6Muo/hXHF95Avu58/qxoiZtYmE1i28isoT6w/NxhbV+BGbylVGCPQCn7PxNBm09RT3z+b9E7sAcUCgxhj9aTchOv3bT+P87y3zbL+jixtDO/ozntePBL7w4Bqc8Z1nXAdMQag8Kt1s44WXdh7aAV0KPQqB2nHQ30gtOVzGgD6Pdz4e0A0TP122HR+8/cVR98EYAwmh74Xt3aPuR7VBOZmvrG7guG88if95auuo+z/S2NLJGcaGqsgh9WMwuWcqk6DyBt9b40mgVAZv3b7+cbvPWCAgUGMM52YdyuoomAxDJRZoXZJvgNEuyB8/vRUfvP2FUX3XC4bJ8MjaDgznCmgbsnPGklP2khLHGyqXvr0M8RgtelM5EAFEPMan0vEcrpiVPoWAjtY+0t6fsdba89u6R63i8SNQknjf++LuUfU9EXBgkD/DoRIKKV2nKnSSkBJUX1pH19D4OErs7U0jREBtPIw1ewdc1wcyOq6592V0DmTRM5yDMQa2uNEiIFBjDOdm3dZVPEzzhvtF1yUOjUD95NlteHkMYyaGs8WDZ/eAN4E6EtircH3bDowPgepJ5TGtLoFpdQns6fFXI6rvcjB7eOwxqoRXSmVcDmvFgfS5ty9Ae38Gz27qGlU/dgJlJ9DdwpZ3FGv4cEAQh87BQyNQulgnldqg8oq6eE3b+Eg3mzuHMLe5Cotm1Hsye0vX78czm7rwncc34owbn8G3H90wLuOoBAGBGmPs68vgyjtX4doHXsODL7ehc6DIBfVk3DtWmtgPjsIoWs72kcoVcM29L5dVV6nQzWKfXWn7eMeCQHUNZUdlYJeBjse21GDl9u5RH9B+6E3l0VQdRX0ygqEy3nkFw4QmHCTKtR0r9ChOHD2p0R2cUrr5pzNnAiiqmEcKVSJwvgvJbJlHMYXqEoSpa/DQpBhjhGrxvGGipZa7mu8sI8WPFps7h3Dc1FpMr09gv4dpoSrGi1ys3sVV6fe+uGdcxlEJyhIoIppJRMuJaBMRbSCi//RoQ0T0UyLaTkTriOj08RnuxMcvlm/HCzt68Ni6/fjKH9fbnB+GdfeGlXad0UhQKlHzMki/sqcPz2zqwnV/8I/rGcrqOPOmZ/DC9m4UFMngYIYTpNfa+pAvmJZ+fLTY2DGIs256Fj95ZuROFru7uUTzyQvnIaMb2D7Kg9UPPcN5NFZFkYxqZQ8U3TDRWBUFAAyNswT16NoOHBjM2ojSwaHRqRUHBTGZXBPH1Lp4WVVmKagS1P4B+yGu2ueORjDGrP3YNZTzVafqhulLxCTDV7ENqmCiIRlFTTw8Krs0Ywwv7+4t6aBimAz7+tKY11yFqfVxdA3lXGpeOdaJENtZiQRVAPBFxtjxAM4B8FkiOsHR5j0AFoqfTwH43zEd5QSHuhgk1yMN0U9vLHpKZQruRSOlktEsBlU6K3hsogPi+r4+/4W+oWMQB4dy+NFTW6ArUll3hmHdvn78wy9ewM+WbUPuECUoSayf3zZy43xfOo8QAXObkwDGR2rpTUkCFUa6jJ0tbzA0JjmBGs+YoB0Hh/G5376Of/rli2MiQQ1mdSQiGqLhEOZPqsaWA6Pz4krlCqiJh1EbD2Njh90DtUcQqCNpuzgU5A0TecPElNo4DJOh22euv/DgGpx187PIlLAxFUao4tMNE9FwCNPqEqMK1v3Dq/vwwdtfxBNveHsf9qRyMBkwuSaGaSLLhZNJ8dJOHCl39LIEijG2nzH2mvh7CMAmANMdzd4P4D5R6mMVgHoimjrmoz3CWLu33zNmyUkcPv/OhXjt+ksAcHFaIuvxjguCwxqNiu+Awrl5qd/29PKFl8oVfF1+JVcd1kLWs4QIGMgxrN03YN3L6x4bOgYsVUA5ZMUir46NvE5mJm8gEdFQE+c2u0qIQr5gor0/g/95aktF3GhPKocmIUGly/SvF1QJSse/3PMybnhk7HX1TwlX8N09aSv2Ciga8UeKwUwBtQk+/yfPqMOWzqFR5ZBM5Q3UxMI4fmqtK0RC2uTyhnnY4qUYY2MSWAsU1XHzJvG0Q39+rd2znWEyPLZuPwBgdQnXf0vFV6GTRL5gIqqFMK0+jr29Iw+neF3YrUoxvFJ1OakmjgsXTkIiouGelbttbbw8kVXm6HBiRDYoIpoD4DQALzkuTQewV/l/H9xE7LBiY8cgXtgxejdaL7z/tpX4yJ2rXIe98+A+fVYDGqqimFzDdckyH1fOQ4KSHFb3CCSou1fuwtV3r0absoC9iMduYehP5w3b4eaEJHQRjSy71tS6BAbzzIoHmtWY9LzH5T9dgQ/9sjK3Z4vLHEVsa0Y3kIhqqInzw7WcWu2N9gEcc/0TOO97y/CzZdvxgEjCWQpZ3UBWN1GfFBJUmQOlYJqoS0RABKzZ249lm7twzwu7R/RMleDFnUWG6PW2PjRXR1EV1UZ1eAFcgqoVRP6UGXUomGxUMXipXAFVsTDmNlehrddO/AcVDrzrEJ0MKgFjDJ+5/zW845bnbEzbaCFVXGfPbQJQjIlyYqdi293Q4faGA2BpJCpV8emGiYgWwhmzG7DlwJBNS1IJpKdnWPPeZJIRnlQTw5S6OJYcOwmPrO2weYh6SVDjGTjsh4pZWSKqBvBHAJ9njDlXtNdsuE5jIvoUuAoQLS0taG1trXykJTA8POzZz9VP8oP1F+9IIhkZ22j/B5cux5SqIm0fztsfdXjPG2jtIEyOFtAFoCGcRzuAgXTONdZ0ji+MVN7AX59Zjli4/Fh/sjyN/hzDim3FjdP6/Ao0xO38xobdGVRFgJQO/PChv+G986Ke/b20lY+hr68PL760GgCQQA66CazYwA2kbXt2IX2g2P+zy5ZbTgIAsGz5coTIf+ybdvD7DPT3j/jd796bAwwDa15eBQBYs2ELpqZ32tqoa2HpTjvH9+zanVgc21+yf/kO2/fsQn/axECq4DvGgaE0+kMZJMPAUhHMWRVBxc9Vat2q2Nlv4PmtWSyoD2F7v4mXdvViRjUhESO8vm0vWlu9D04/7NmfATP4OFPCxnjtfS/ixvMSIJ/35xzv3v1ZGAWGfH8O3cM6nnp2OaLiUNzeVjxUv//HlXjnrDCaEsW1Ixk8v/uNBL1ZE0+8wYnkN+5/DlceH6tofkthnwivyHXvweIWDXsO9Hr29WJHkei8vmkHWuFOkZXJ8oN9IJUtOR51rD19GVRFCDXD/Py66/EVOHcaP6Zf7yrgjnU53HpxErESBGhfJ5/71zdswfTMLtf1ZW2c+Oza+DqGdoVwbLSAJ/IGvvO7VvzdfH4+bG/LoiYCXDAjgrl1Idy2JodlL7yC3slh13jHGxURKCKKgBOn+xljf/Josg/ATOX/GQBcOUIYY3cAuAMAFi9ezJYsWTLS8brQ2toKZz99qTzw5NMAgEzjQlx2xgzbdd0wYZgM8Yg2sps9+TgAID7tWCw5rSggHhjMAsuexacvmo9/OW8OJtfGAQArhjfijRW7cMr86Xijuw2mFnWNFc8+iWSUSzmnn30uJgmpyw/T1j6P/v2DUAWaxWedg5mNSet/xhh6Wp/CP5wxE5s7B7ElZeJHS8737O+J7nXAzr0wI1U49fRFwAsrcdysFmzv34+t/fwms2bP5Trrtdzh4uzzLkAiogFPLgUALDzlbNv9vfBqfguwbTvq6uqxZMnbyj6nit+3v4qGwjDe/Y4LQc8uxeTps7BkybG2NupaeKJ7HYC9+N4/nIz17QN4+PV2XHDhRTaiqqJzgL/Dk084FjU9KTzfsdv9rhSEVy3D9KlNaJ5kWGqeGY01WLLkwoqex2vdOvHi0k0Ia7tw5zUX4uIftQIAZrU0oioWRltPuuJ7feexjThuSg0+uHgmfrT+ecysjmHJkrPAGMMXn1uK9mGG+YvOxqym0u+vtbUV511wIf6OCuDLAAAgAElEQVTt/17FZy+ej9imzaiLhHDh6TPxp21r0LTgVJwxuwEA8KvtL6FpeBA9qTye2KUjVNOM297D/aZyBQPHXv8krnvXMbj24oUVjb8c1u8bAFpXgAjYMBDBRRddhOeee67s/JbCa219wMoXcNZpi5BOHsSW1/Z59rV75S5g3UbUJyOI1jVjyRIP37DWvwJ6AXmTcNFFF3kSZXUtfH/t82ipT+Cjl5+Om1f/FcOJFixZcjIA4OYfP4dMIYcZx5+OE6fVeY791zteAnq60TRlJpYsOd51/Z67V2N2UwofePcSEBGWALhvy9OIN06x7vOr7S9hYbSAX/zbedjdncJta1oxe+FxWHLaDNd4xxuVePERgF8D2MQYu6VEs0cAXCW8+c4BMMAYK82ujjOeVyLkb3l6KzJ5A0NZHXf+bSd0w8TVd6/GCf89stIYqlpvzV57fIJUfc2bVGURJwD48FmzMLe5CkuOnYyIRvDSSumGabl1VmqIVMXtKeJ+TgeGrG5iKFvA1Po4zprbiI37B0smNu3PSCmuYMVtnDmnEUDREytfMG0qvpxuoFNRp5SLGwKKKr7cKFIDZfJcxUdEqI6Fy6r4NnUO4rwFTfjwWbOweE4DUnkDW30cAmTwcTwSQlU0jHzBtLnx64aJGx/baDkAcGM24dMXzbe1KQXGGD7329fx+LrKtoVpMjy18QAWz27EnKaklTKnqTqGmlgYqUrjagomfr1iF7700DoAXOUmmSAiwm1X8kPVb24k9vVlsGxzF6688yWkcgUko2GcObcREY1w/6qiK/JgVseiGXX447+/DZNqYtim9C3tOz8awywT0onh70+djvb+DN77sxXoGB6dQ8+9L+zGj5/mY6uKhTG5NoahbMHTlibtoPOaq3CwhCpT2qAKJrNinPycR3TDRCwcQlgL4ay5jXjw5b1oF/bTRJSfE4+u3Y/fv7zX8/tyj5ZKNtvWk8ZJ0+tshHJyTczmidifyVvxmZKJz+pHJgayEhvUeQA+BuDtRLRG/FxGRJ8mok+LNksB7ASwHcCdAD4zPsOtDK+39SER0fDV9xyH9v4M3ugYwI/+ugU3Ld2EJ97oxMrtPTBZ5UXEAG5clrjnhd245t5XLKIliUMsbJ/OBZOrsfy6JbjkhBZUxcLIOQJ1GWMomMxyGqjES84wGXpSefzdKdNw+aKp+Pw7ORfqtA/JA7cqGsalJ06BYTI8/Lq3sVem6knninnlFkyuxoeOiVgxGXmD2YIIswXTZojdXcFcSieJ3CiM8hndsDZLbTziS6AKhoktnUM4fkotAOCEqZzb3Objmi4dBRIRDckov4/qybd8cxd+tWIXvvXoRn4PkyEcCuGk6XV44atvx/tPnWabHyd6U3k8urYDn60w2e369gHs6k7hg4tngIgsojKtPo5oOFRxTJpKePIFE93DOYupAYALj2kGgIq8+WTGjFzBxHCugOpYGNPrEzhtVgP2KU4ogxkdtYkIzpjdiL8/bTp296StQ3k8inpKA/41F8zD4tkN2NAxiP/baCcYm/YP4orbVpbN+vHNRzZYXqbJqIaWGj5XXk4pQ7kCYuEQpjckbcyaioLBuKYBnDiv3N6N+V9fWtLuly+YiAj13ZcuPRaGybBKOGYlRT+3P7cDX/7jOu/vizXo5egA8PlPODRHLbVxm416IKNbBEq2LeWlON6oxItvBWOMGGOLGGOnip+ljLHbGWO3izaMMfZZxth8xtjJjLFXxn/opdGXyqO5JorLT+aOhNsODGNdOzdivqp42+ztrTzOoFcs7G++7wQsnFyNZzYdsDzcJOcsuVwvVEXDyDjOVCmtWASqAi6lN5WHYTIsntOA2648HZMtAuJNoBIRDYtm1KOlNsZVIR6QmzadNywvvnCIcNm8KFZ97R2ojoVdElRWN2yOHb97uc1XggCATJ5fH03Ab0Y3rc1SEw/7upnvH8giVzCxsKUaQFHK9ItXkQdnLKJZEq1q2JasheSk9QI3ZgPAtPoEEhHN9/lVV/9KDuk9wgnipOmcuMpDa1pdArFwqGKXfzVIe337AEwGtNQVCVRNPILp9YmKCuqpyXSlRAtwg7u6FiTxAoC5zVXIF0zLi/JQQxW8IKXaWU1J3P/Js3H5oqnY3GvaNA0/eHKz5cxSKaqiYbTUSgLlXjvDWf6cCyZVY29f2tMRQhfONABfTw9YFXO9vf6kmzkAnDStDrXxMJ4XtmbJOEl4aUTk3irl5JMtmFaNKgkuQSkEKl0kUPEob5vRDbzRPoCBw5R3UuJNmUmiL62jIRnFdHFwbOkctDgWNbXHSAJPe4UaYW5zFR769Lmoimr4+bLtvB+xKKLh0tPZXB3FYM6ZfJV/ryqmifGU37xSapkkCptFNc02BgnJ8cTFop7ZkLR5/amQElRGN6x+wuLwJSJEwyHohmk7XLK6YcXiNFfH8Eb7ID54+4u+7uyWBDWCQyqVK+CGRzageyhnEahyKj6pKk0KlUhtIoxYOOTryWip+MKatTlVbyZNqEQkU5E3TEQUh5aIFrKuLV2/H/c6PPpUAqWmLCoFmdBTeoBKCaouEUEsolWsDm5XJBuZGFSVoADg2Ck1Fan41HHnCibiotbRpOqYTZrOFYqHbLFCLJewVeI8VnFSXWJtVEU1xMIa/v2i+WAAXvAICfGTop1IxjRLg+BFoFK5AqrjYZw4rRaMwSUVmSYDY8V0Zum8YaneSq3FvML4hEKE954yDUvf6ETBMC2GQGLnQbfWQu7fUirgTN6w3pvE5NoYDoqce4bIG1onYvyiWggh4t97789W4MpfrfLsd7zwpiRQ/ek86pNRhEKEeZOq8NcNBywd6jolOWIlEouEVCM0VcVQl4zgfadMw+ttPAeeJCx+BGpKXRy9Ofv93BJU+UNHuolKyUne00mgstaBy6/Pakx6Buzqhon+dN5yHpBcWURTD1/ykKBMK+faY587H/98ziys2dvvG++RlTaoEQT93fvibtzzwm6092esDVoTD/vGQUkCGFGI7JS6uK8LslwLiaiGenGgqGoSScDkoaobJiKh4vsOawRd3Pcz97+GbzpiotRcgpWEFLT3ZVCfjFjS3E1/fzJOml6Lc+c3IapxFV8l5SzU+K+1wnba4iBQs5uSFbmtq67IqXzB4sQn1cQwlCtYa043TEubMFfEEu0SjKHKnAyOUbqq9r4MptXHLbuKdNZRJea9Yu2XK6OiojEZtWzKXu7yUlKU8VJ7HWVgZBYJSaCGcwXrHClls80rEhQAzG3iEmhGN1wE3Wv+5FmU9siEwhhDtmC4nMMm1/CA5N5UHkNZ3UZUiQiJiGYxIBs6Dm9ZoDclgeISFJ/gBZOrLf3wKTPqbFnFK5FYJKR6Q6bfb6mNoyeVh24UD24/Fd/UugT6svYFJlVCVSOwQclNN6mabxyLQDmkQUvFJw71+ZOr0d6fccUzrNvXD91gOHsud4qQQZZh5fCVEpQ6XzndQM9wHlVRDVPq4lg0ox6AveifE0UbVOXzrqoU4paKzz9Xnnwfqk1wWl3CN61PVnGSkBnmVQIl1TcFk3uAmqxIAAH+7v3Wk1rioFwWiPb+DF5v68eCSdXWZ8e01OCxz12AybVxxMIhmMw7e4gTHf1ZS2KSKukpdXYCVR3jmTPKEbxuW2qtYrVYGe8nY3Z0g1lzM6k6hngkZElyqgQ1VvkU2/szmN5Q9ECsjYehUVHiY4xZBNgvYNv5/GEthNp4GPFIyJO5GRIqvlJETMY41loSVMFiMEup4GSgroTUgGR0w/Wdguleb3LtezFwecMEY3BJYlJK7BrKWoy8aqeKR7SSNrbxxpuUQOXRIETUhZOLm1x6pkmMSIJKFSUooCjBdA/nKlLxtdTGkSnY7RoFpwRVAYFSA+2AIlEsJUHJhXbeAm4Md2bCkNmtLzpmEoAiV2aXoELIGQ4JqmAgVzAsz6Lmaj7fXofvizt6cONjG63FPyIVn6KqqBaq0Oq4v4pPSqbq+zhzTgPeaB8o6cnopeJTOVQpGRomsxgLt4rP+7kKhonH1u23gozLReWf971l2Lh/EIsd61UiJiSXihiaoSyOnVKDmngYu7pTiGhkpWiSSEbDYKy8p5Yzs7eUoKTEsrcvbamJVOm1IRm11MjqmMeUQAlVqLxnTZTQM5zD+n0DeK2t37pvuw+B8ppPIkJNPOJ54EsJqjoWRjKqWY4U+/rS2Hlw2GIg6hSJXDK6pZgZ3SFByf2b002Xo4LuUR2haINyj1e+X6cz1yThCNI1lLPWsBroG49oIw4YHiu86QhUKlfAULZgHeAnTKu1rp08wx47MBJ3595UHomIZnEfk8VLPTiUsxa2H4GSh5NKoORiqB6Bm3l/Wkc8ErLGIe/p3FyWDUos8LlNXA3h1H3LHHfykBlQ0h5JRLUQ9IJpG99gpiBUOXwhNwrC7XX4fuTOVfjVil22zWmaDL9b3YY5X33cN8peVVXUi4O1Jh7GUK5Q0oaRd6j4AO5sYDJgl4feHihu3kRUs+7Tr5SRkGPUDebpFBPRuFTjNabnth5Ef1q3xpXVDTy4JY8n33C7nKvpet67yDtbWCmmxAu9w3k0V8fwwTOKYYohRyyYtIH6pSUayjM8utYe2hhT1McA0Nab9iTe9cmoJY0eigS1ameP65BO5wvoTeUxoyFh+7w2SjgwmMP7fr4CH/hfXi/trLmN6B7OlzxsVSKkErxkVPOsgTYsbFCA9ITj/Z7//eV4+/88Z3nESgJlz/zi7q9gmC7JXDIBGd1ARjdw4TGT8PBnzxPtSxMoL1V7TrefCRJNImVXn9AIAfa1nYhq6BgoEvbDWSn5TUegpNfSfKEXPkkEtC2cXI1jWmpsbUfiTTaULXq2AMBUoSbZ25uxuCEnZ6LCi5C4VHwVSHRDWd3KR6fes5SbuVyMyZh0dbUfQn1pHvMgiaSVl085xKJhrr7KF4qedFJyjIj7y0Xe6+MAoLqi5w0TX/3TegDubNgq1EOsXqjeauMR5Asm5n99Kf6yxu06L9WdKsPQIMZX6lBUJaiqqIZwiDxVfPmCaXGu6hzJA9lLipKqsesv54GT6byBJ3bp+PRvXnMZ1uX6OH9Bs+XB50RMctVlGBrGeEhCc3UUHz93NpJRDVefO8fVTr5Tv/ROj+5wv1e5tlpq44hohH19Gc8Drj4RsTxFVQJVScLbVK6AXz63A3t6UvjwHavwiXtW265LlZ1KUABOoF5rs9dJu+5dPLD7sXWuHAIAirXQ/uPtC/CXa8+zPk9ENM+5SSneijMaEtjcOWQ7vJ0SlKpi9jp7vCR/1c07nS+gJha2mF0vFZ9cP3ydumMjATeBksxuVjeLXryKBJWIaDaNxeEMiXrTEqgFQrU3uTaOu65ejIc+fa5lzHzbPJ5jayTG+lTesA552X84RNi4f0CxQZXOTGERElvw58hVfIPZgrVAAdUGVULFJxZfRAshGg65CFR/Wkd9Mmq1KzpJOCQoQaCaqqMIhwi9qbzN1iATpzo91NSNqDJeKjH2i01RbYb1CX6PWuX57/fIsZcXOQ+dhyQA9JcgUDI5rAwGrktEbMRMqhqzejFWzCllyusS8rCSm/u9i6YBsBPxax1xUZJQXnJCi+c4geJaKsfQpPIGciKp7eymKrz2jUvwX5c7CxEUGSS/4F9Ji8+aW1Q7yoNOCwk1nlgTgJM5iFjea+oab/dw2tk/kMEb7UVHppuXbsJ3n9hsJeJdtdPunv3b1TxgdbpDgmpOkEsNfOacBiSjWsks4VKCOnF6HZqrixldElHNMzRgKFuUoN51Qgu2dw3bajjplgTF20gCNakm5rnXvSR/i0CJXJGxSMhyznGq+BjjsYryPHE6SkgbsNPNXHr1ZXSjqOILuaU4q5/DGBL1piNQUsUk9aoA8PbjWlCXjCAW1vD8ly/G3Z84E8DIJKh0roCqaPFgjEc0zJ9Ujc37hyqyQXmpZUaj4hvKFmwSVCl1j1SFqMbO6pjb+40TqIi1KeT3wg4bFK8HxaPcG6ui6BnOc1dr8b1kVEM8ErLc8Z3jkKiKurn/Pp/YCrXCr3R8aVIOD6/MRV5eldLxYaAEMexN51ETC1vfqUtGbMRMctDpvOHJicp5UImP5EYHMzqIOCcd1UK25Lx7+zLoT+fx7KYDon9BKH3ScJViSpyQ8UFyvkql9rICk30kqHSBMyHf+rsTrc9UjUFDMiqYFvchW5+MWsRe9VT1sge97bvL8N6frQDAHYJkcPlykbA14shB9+sVPN/cHKHClphcVWx351WLcduVp1sBz6UqB8i94cy2n4y6Jah8gYdd1Ii2UtpVVchS3Vsd50mFpQfhzIaE59njtW5VJ4mc8MCTe9MpIcnvS02Dk+HIKloCFdKmyZkvPmZ1nhNR+3x4Jb0eL7zpCJTkTJxUX2JmYxLxiIaIRiOMxzEsXb1EbYJnvZb6ZF8C5aGKswhUvHIJajirW5uiVL9AMQuCetBVxTRPFV9DMmqpq+RGVF2oI+EQzyRRMBELa2iqjlkejNIGRURoqoq5bFDOTSJVbapOv89HLagS1KR4bqlOBDhH7TzovLwq6zxcx1X0pvLW2GT7g4M5PPTqPuQKhiWtZHTDkxOVB7JKbOX7HBTeXqEQIR4JWR5Ri2bUI18w8fG7VuNf731FeFHZ49e8IL3nyklQz4hy7idMrfVtV1WC41bRn2OYVh+3Hd4qwZNSkpcUMFV4vA5kdGtO5k2qKlun7Jd/24lU3sAcJUegmpDYNBmIgCtOnebKYTlZSU57yQktuFzY8yZVx0q6+adKECgvFZ9sK+dO2nB3eUhQEY1QFQ2jJ5UHETC1PuHJXFgEyqFeA3iIRlbnsWeSQDkrasu5l+MvybQ61lYsHAIRZx6k2lB9f43JiK19IEEdAiSH5ufyDfBNXokE1TmQxbcf3Yi+dN4mQQFcDC6YZkVxUF5cr1PF55dhYN2+fsz56uN4ra3fW8XneJaBjI6oFrIdolXRMIYdh9DBIV4DSfYjCYcqQcm4GxmAWZcIYzCjW6UBJBqroi4Vn9PwLlWBqstyqbxhACdQ8yZVYXp9wrIrqhIUAJz//WW2/70k2liYpzDyUvHd8vRW/GVNhzU2gKsEV+/uxXV/WIsXtvdY7yaTNzylU8lxqs8r16Ja4iIR1Sx3ZOlhKt2/dx5MWdk2/CQoS8VXRuJev68f0+sTNkchL1ipeHxUfAM5hknVMdvac0pQfWnddihLnDarAYzxHJZyHltq4lZIgxdyBQPdwznMbkriyrNnKZ+blmTSm86DMeDUmfWu7x/T4D1/zdUVSFBxB4GKhl170yltNVVFkYhoWLuvmKdzQKSOiWghy8t1RkMCVVHvs0f3UfFJr1k/FZ/sUxJNZxhCtgTzTkSIhzURqO+2QUkHqGI/gQQ1auQKpuAI/FP5R8OhilRqn7jnZdy1che2dQ1bHLyEKlkA/kTRT8UXj2iIhUO+Kha1KKB6SEjJx8mR9ae46k6dh6pY2HaApvMFdA3xQ0BuirSHii8aJssGFQ2HEAtryBkm9AJzESink4TzmaT7vxozUk7Fd+mJU7Dyq2+3VJtys0s4nYr0EgxDQzLqWdfmp8/yEvSqZFalvOusIjXlDdM6yFUJQt5LLRVvSVCZghULk4holopvgRICAQDbu4ZtKapKoRRT4sRwrmBz7CkFS4IqR6BqYrZ5idkkKIcNSlkXx4iUU229aeQKJoj4GnYGpqsSwUBGt3LCfeD0GXj3iVNwlnC7l3MkvfGm1NntTwBQGyP820Xz8IMPLLJ93lIbQ+dA1tMTraSKL6K55ka2lXuRiHDitFpbBW3V4eh/PnQKrjh1Gn7y4dPE2VN81p6Mid+/steT0ZXSznDOgG4wxMIhyzHJ6SQhvy9Vts7rVjqvsHttxSMh4SThJpJNjv3mVRl8vPCmI1BZ3R0p7QUpFZSD6mVV5RCNIyFe4K+o1ihNFL0OlbzCbdaUSYCqLhLVBiVTETmfpS+dt3TR1vhjYZuKT7q9zm6qsghdxkPFF3XYoGLhEHK6IdL92Beyk0Clck4CxcekuruXCrrNins4DwwpjXzqwnloro66VFil3sfMxoSvWkmVoFTPQmmglpBqQi8Vn02CsgiUbjl2xCOa5QK80OFV2j2ccwVYeyGuGM79oBrx/WCldirBKJgmw2CeEyj14FIlqLlNVehJ5a3sKmo7ydhl8jzbRDysIR5xOx6o639QIVBN1THc/rEz8L5TuZOJdGiRBGqqI/BY4mvvOR4fOnOm7bNZTVUYzhU8vU2lvdOl4lNsULphYnvXsELMinvswmMmecZ5hTXCGbMbceuHT8PpsxoQ1ewS1K2v5fDlh9ZZqkebm7kgJtJ2Go9o1l51SlBS5SvH73RDLwaju9dWQryPgoeHqrovAMBHEzzmeNMRKClBlUMsUnlGaIkqpwQlvNtyIrWLn9Tm6WauiPTlEqCqi22hg/OOae7kodI7T0VNLGzzipNeVDMbk0UJIF9ARCNbrIx8Thnlrrqdq5xyU1UUPamcjTt1cp7SziNjRgCUVPU4uVSJUIiw8+bL8LX3HIdz5ze7VFOlVK5++QgBYE5z0dD+X5cfjw8t5vVvOIEq7kp5uHk5SajxJ1JC70vnrU2uEh7ne8wXTGQqcJKoqsCpARAONbHKCBQR0FuCQPWl8zBZMf+jRK3CKL3/NE48VopAcJVxUd3YcwXuiSY5dhVZRaMxkNEFYS/eIxmxP/f+QX8C5QVpz9rjsQ5SuQKI3ElZVS++rzy0Du+85TnL1qTapZ3vU3qnqh5xgPvsGRLFMiXTZneSCIm++LuJhUPWWnMSIDkvcr84nShylnOP+3yMRzRkFdd0lUjK9z5NzHOg4jsEyA1QDlGPQ90JJ4fnlKDCGqEgVHx+9ifA3808Gg6VzS+njsVpU5AEQ0V/Jm9JKxK1iYgtO4Jc9A3JSNFJIme4VJVSQrOp+HTTqokk0VgVQ1Y38eqePktd4wwYlGotKcnUxEoTZinNJaPuQzYUIl4bKh52OX6UUrnOakzi4FCupK3vgoXN1t+nz2rA9e89wRqHl1NH3MsGlVNtUHwcfemiA4ZKeJw58XTDrEiCKkok/gRqOFdwEXcvaCHuVl/K3b+YvcQ+XtW1W6pupRenKr1qIS7lS0JvSVAOFbtKsLiKr6gaBdSAYqniy0ALkcsm6QcZVOyVe3AoV0B1NOxiNGviYegGQzpfwJ+EV+FzWw9a1yScBTuttGEOSV6mxZKMnDyupPOMum5lslZpO41HNGghApFbhZfR+f0kUXcGjRede9xrKxbRkFGqGagE6oJjmvGd95+IOz++GG8/bjLqYmNbodwPb0IC5c7W64VyElQmb+C4b9iLGjo3glxozvQkXvDKOq5yK+UydMtN+Zkl863gY6tvQUCyuoE2kYRSZnRX0ZCMoD+tWxtDqiBq4xGL43Umq5Tjy4tcfNFwyNKhO50kpA3nH29/Ed97YjOAoreTPJilQVu63C5sqS753EWduU+GDo954+7v5Dpo5AHyVaWWjtzE/7R4ppVPUMIyUOvci09y1r1pN4GKOmx48numydCX1q30QrKPRBiu6r5cgirvJCEZpXJFC9VMB+XQmHSrZyU2tHM1t0zvJaE5pOxERLO8OJ3MQTLKD8Cszss9JDxUfKpNeDhnYDBjD46X7s7yIO4cyKGlJlaySrIXZB7CUuUznFoSoChBdA/lrbmXKcNUFd/MBjuBsuy5ITfDBxS1KZIXkSpLdf/JZK2qBAVwFbxTxSfXjiTqzut+Kr54hNvkvVIdxcIaPva2OThxWh3uuvpMLJpU2ZoaC7zpCJQMZiuHchKUV/kBp1E7ooVQMBgKBrPpbD3vV9YG5a/ik5z1F991rCtVjSRQd6/cjQt/uByPrO2wMrqrqE9GUDCZJanJoNyaeNiVtkdFTJWgtJD437AF6gJ2XbVMjSI31lNfuBB/+9LF1ubY1Z1CXSKCltq4jw3KO/JdRVUsbBFLCafqUUISqIfXdFjEWaog50+ucrWPaCGEQ2Rx/pMcSVFtXnxhtw1KNxgGszoMk1kSlMzlmBSSp6qCy6sSlM8zS4nST8XHGHNlHfFDfTLi6U25ry+NL/5hLQBeagbgGTFueJ874LcmHrb6cK4h6aqdKxiIhTXEIhqyuj0juypBHRzKIW+YaKxSVHwO1WZPKjci6QmAK2+eikFHthgJ+d4PDmehSw9CQcxVBqDOobGQ6lqnBOXUpkTFfpZE03mWJKIaBkTaLengwLU3jtCSvJSgvDNNSAIW92D4JMMgiVokNDFIw8QYxRhCboByKOdmvrO7WDdmkcjh5yRQYY17txnmSAhU8VBR08LUxCO2oFQnMrqBaDjkyS1KJwbpofar53dCN5hLxWflmBNEYyCjozoWRlgcxM6xSthsUNJJQjhN2HTVSiyKlGr603lEwyHMaEhgVlPS2pyD2QKm1MZREw/bqhWrKBX5rkIahFU1H3fHda+BmY1FlZRMgyMPOy/OGeAbN5Pn0ql0xpAlB7ycJFSVpm6Y1kEmpUuZOVokF8CfP3seLl80FY1VUeSEFAz4q/ii4RAiGvnmMMyJlExOg38pqPnyVKg2O/kM11wwD1efN9fVtiYetp7XRaBEPjspQcU9Et6qXn1SBWfL6OCwQeWUIpaVgogwuSaG7R51oQYy/gSqoz/rOjOSjvs/cu15VjqplIfDEeBOTya7OOjhJAFwBs3pmBMOEQomQ8EwceszWzGQ1i3mRkpQLieJgoGIRrYMKNZzRDUM59QsKYdPjeeHNx+B0itzkijnZi6Lgd119WL86qrFuOVDp7hsBlL1ZZjMJdV43Q9w2KAKlav4snnDZbxV+84bprW41om4GqcXX4MHgZIbUuq11bGq/ZuME8mIlKAsJ4nic58ysx5//PdzceExkyxbFw8ELrq7q8xDXTLiWzrDTyUhITlYde7SeXdZa4Craml1HasAACAASURBVC49kacQkirGUq7FEnF5sBZMzJtUhWRUsw5tVZXsZYPKG6Zlh5C2ikliDUk784LJ1bjtytNRn4gIFZ8BLUS+HqFA6fxwElKScap5S0FKyU7IcIApSbfK1InqeKTIgTvGL1V8koGMewQbZ5X7t3kQqKIExedUxgWNFHXJKJ7behA7leKljDH0pXTUJjxUfIJA7emxJxquimqufb9oRj3++Rwet+WVlQVQAq3F80pTkVyLzvaJiGaFYsjvSqZx1c5e3PrMNnz9z+sVJwmp4nO7mZcyf9Qlojy20cMGdSQxMUYxCnQOZHHNvS9jQ7c731RFbuZlyma392UwrS6Otx/Xgsm1cfzD6TNcbSLCScJgFUhQnnFQYjGEQzydik9NnlKHrnwW7gZu/65TxWdlUxDqgsFMwRbHIRelUz2mxkhJGxRjfAM6F/IZsxtQn4hYB3NfWrdy6AGwHSiJiMZVCyXeg6Xi85GILQlKUa1ldcNTAiEi3P7PZyCqhSw1ncye4OWIIcc4JNR0iYhmZc2OhkO2wynqIUEVhGEdKEpo0p4RdiwXWXNLvudyxMAZ0+ZE9xB/x86YsVLwcrQBiu72N5zrjjVyQs2R6Jz/hIglkip4uUdVR4lsGQmqGK8lJKgKPXad+Ng5swHYq+v+fNl2bDkw5Hl2SKcD6cQg47r0Etn0pb1Zvh/n2WClL8rL2Dr+uRUk76fikxKUOHvk/6+39VnzIt+Dy0lCNz01CwCs3JMFxewwEXDUEqi6RATLtxzE5l6HobVCCSpWYkNK8CJo/ptScjGFCiQo+cJL2aCSUQ2GyUqOKaP7ECih4tMN0+Zy6+Sek46N0dGfsYqtATyuC3A7JagSVThEFhfnjIOSqBWZJgBZ3didfR3gh1Y0HIIhVBVOqEUES0ESqGGHBFVK2pTVdeXBO2ylrPFun4holr0qGg5ZRN8VE+fIZQhwDlYSQJmFZPGcBlx28hR87jS7NC6ZjEyFcXzJqOZbvbjbkYevHCKipIoTBwazqImHEXdSVA+oHm3OrCuJaBgZ3RRERbPeqUqUcl4SVE1U6cO+frN6Zep8J5Ycy2uf7VdSZD2wug2Adzb+mFCtSxWcVPWWMhEUQzakBOVQ2YXtz54TjKUkMFrI2V6zmFnJrIVDIehmUSXcMZC1bF6Wk4SDQOV0o+Reqk/ymlfFTDITgzRMjFGMAomohmNbarBz4BAkKJ9cZh0DGUyr9ydQYS2EgslgGAxaGY6XiBAN2dUYlhdfKFT0UCpx6PgFIEfDvKCg9Ca8/GSed8wZH1LkWk0r4PD4qcVgUUlsXCo+hZsKayGbFOSlCqiNRzCY5d6CMhltsS97lLxT9dmfzuODt7+AzZ2DvpHvEpaKT1GtZfL+a4ATKH44WTnVSkhQ8ahmEduoFrKyorti4pQ4Mgk164Q8XJurY/jFR89AS5W3pySX/spvy6pY2KZOdMJyDR8BgXJK4ADPLl5pnFFNTHUJ98jGkCtYh6S1FlUVnwex8ouDGq0E1ZiMIqqFrDgqABaj5lX6hYiQjGjoFh6KMjHsmXMaPPuXzKil4vOQiIDi88rcdvJ9lpK4gKIEJbU36vyl8wbCIbIYWSfT53c2yj0qJe9yGqHDhaOWQAHA8VNr0TFs31SyDHM5xMJaSWnFMBk6B7JlCZQ8uHMFoyJX15jm9PIyEQ7xoNhyGaX9pIKoFoJhFt2+f/aR0/DsFy9yxWVYXGvewJ6eFPKGieOmFAmUdId1Eh2nBGUrSe3BkdUmuC0iq3OVlXpYEZF1qCSimstg/PTGA3h5dx9uenxTydxhKrycJDJ66bkC+KEtcwamHCo4JxKRkKWujIRDllTqXGNFG5RDgrKcMPyZJikFZ/IGkpHy67dciXnp7u1MU1MKsbB3ReDOgaxnKiEvSAlKBnOr4CpJw5KgEg57ElAkSvLdEdkl7rAWQlQL2QnUCJ0kAB5DN6Uujv1K2Q1Ztv6mK07y/E4yplkS1InT6nDHx87APZ84y7OtfPZ0CS8+tYQGUFT1py3pxdnenVmCM8emjain81y1XUwm61bxldLCSPW/lLwDG9QYYHp9HP25onqIMYbhbGXBiTJdjxe6h3PQDVaRBAXwF18JgYpqZCNAqpt2OQKVKWFXAbiDAw8Y5v2FQoT5k6pd7VS9v2X8ri0+oyS4Xl58EmGNbBKU04sJKHK9g1kd6XzBRSwsAhVRJChxOEkPucGMbr0fv0OolIrPz7urPhmxgm1lKqZSTE0iUpSgIlrI4jSdBK1og1IYkEKRQJWycVnfF3kdM7rhm8lcgqvkSkf0D2V1aArjU74/8lRZ7R/IYkptZVKYNM57STUyjCIrJCjpXj9sC2zmcyWZAC9bHE87VLDaj0aCArgUrVbWzeoGTptVj3MXNHu2T0bDFoGqiml414lTSjI1RQLlHQfllB7ltEvzs/MsUddyTPHi08V6kRjI6EhGNet+zmSxXLPgPV9SLdiTyiFE7jEcKRzVBGpafQIMwF0rd1lBgAWTVRT74WeDareqdPqrNuTBnS0YFYnEMc2uxsgXTIvzVoNCveCntgprBMNktvIXXlDv0Z1yG9EtFZ+PBBUJhWyFGb0OXukJNZjRheTnOMwFF5iIaNa9JPcs537/QLZs6RSgSCiGHSo+Pzftxipen8g0i04MyVI2qKhmxYtFtZAV6+LcwM5kuwBnQKTaphyhsGxQecPGMZdCpIwNNZ03kKzA2cLqT3NLULph4uBwrmIJSqpbvchmVUzDcK5gqaq9vC/l2m8QsU9ec6bWZqo0a4wXptbFbWXMnWmVvO4riUG5BLzSrVxK9U6HA2cuRaelwUnQ1LVsBepqIRQM08Zk7+tLIxkNFyUoZ7JYHxWfPBsGM4UJIz0BRzmBmioknJuXbsaNj2+03JUriZ6XB4KX19yn7nsVAMpKUHLhZXWzrJMEAMSEBGWarEhQwlKCcgdf9qfzFqef9VFbaaGQ1Z/f4rI2Rt50FbMDinrnchKUeph7Hexyo/eL+j9OaUZ2p9qgJDGS+ci6hnIYzPJCf35Z4qu9CFQZFV99MgqTcQlvuJwNKqIprtMhy6vM6UHnlSxWN02k8jx1VLlNz1V8hq8zjL09earkJDKOCtBl+wtze6qpcN0Hh3JgrPJcd1JzYXrsqepYBCbjjgOxcMhiIlXJN5XncTpqaRInpLs6Y8xSF44GU+sSODCYtZ53MOuf+V1dT6pXqhdCIkxAEiAnM6M6iBQM00XQnSo+lajIv8Maj4NSbVC7e9JIRLSS5Tiyeun5kv32pfMVS92HA0c1gZqnJPfceTBlGcprKyFQGo/tcYrBg1nd0sN6qclUyEMnpxtlnST4PTkBOuvmZ/D+21bYCIqqk7/+4fV4ZuMBfO63r+O07zyNvlTeV20lg/bKESgZX5MVtXa0EFlGf/V5vOKgrHtpIVv2A6/FLA8qGRnvbCOnnJcZsaeA6lIM1zu6hstKAVKN9Ub7gBVykCmj4pMBzH1p3ZrXUioNZ82n2cKu5yzMqDIrEnqBS2iVEArp5l1pNn4viUdFWndLruX6A+xxetJhYEptZQSq1o9AKXsyHtEsxmJQiYGTmS+sdFAe85CMcvd6Oc5Rq/hqY9ANZqWtGsx4x0BJSCemELmTF3shqoVsjI2tLzWFloda1eVU4ZFSi6c6stugDg7luIqvREFDPy8++XlfKj/i4OfxxOFLqjQOmKG4ge/tS1vqgopsUJGi7UNdQKtEjq3ffeqcslyvXEhZvVInCUImb6B7OI/u4TwWTKp22aBeb+vHb1a14Ter2mzP5ud+rIVIuKizsh5gssxBXyqP5uqoTfKTm93lZq6mQQqRTYWa8DDoS3221PE7CZQ8vxIRxUnCMGGaDAeHcjh5eh3Wtw9g9a5ezHDkN/NCVSyMZzZ14RkA553dX1YKkVLjQ6/uxXCu4OvA4ExnJDOeOwkUjyPjdgHpOq8bJoayhZLSmYqosClp5M9kWGPR3LnYVKRzbttfufsDXK0n19m+Pu7qPaUujgOd5fuQ64Lg3gvVyhzHwiFPyVfG5Uli4Cw1zj/TLGcL2ddo0CjWQF8qj6aqqK2opBdkWEFdIlKRtiQSDlkBTi6vPIVAedn9StmgQgTr3mFhM5RSGhHfVzYnCVccVOkzRDpfpPIGWkaQHX68cVRLUESEBfX8ETr6M1Y25kpsUE7bh8TK7d1IRDScNstdpdPVh6KeqsxJwu21JDlvuWF/IornqURBpsApZVfhEpQJveBvgwL45sjkDazY3oOz5jbZrlUrXlhezwnwzaMyAF6HoNzoUoJyHjSSw05EQzYniV09KRRMZs39YLaAOc3lCZQq0e0QwZd+yXvPmdeIKbVxvLSzF6mcd4JQCXXOo1rI8vb6zMXzXW3VQGduxDaxszuF2U3ln0FKULrBKkozE9H8kx37eX169+euMXTDIxsAANMqtEEtbKnGKTPq8OuPL3ZdU5OqyozcVVHNZoMaEkRC2uC8bHHSBmWFIIyS25dSdG8qj6zO573WR8Un10ilmTnUPeRW8RVV7V52xFJOFbY2Wgi6UPFVx8KYLswRNicJV6oj00eCUu3KE0eCKkugiOguIuoiojdKXF9CRANEtEb8/PfYD7M0rj8nga9fdhxMVkzhU4mIKhe2c5NvOTCEE6bVVqTbtpwkRiBBdSucd3t/xupjRkMC1168wLqmLtzhbAG6wTw95oCiBFVOxQdwUX7/QBbdwzmc5iiVLTl9J4FX+4xoIZu6xk/F1yk8BZ1tsoqxuUjkDby0k1cNfvdJU6y2ao2mUlDHI1VGfvMQC2tYcuwk7OpOIZXzV4U5az4REXZ/73J8/p3HuNrKe0Y0spIRbz8whGMchQm9IAlOwTQrStQpqxyXQlo3PCWQ0v3x59QVj9hUnnu2OZOglsLkmjj+cu35np5wqpeklHqq42HLBlUQ0mZNPGy9D6/3khRefDKGcbQSlCQ0fWndWjN+EpSsheTHzKiQ6zoccqeI0kSoRsZDgiIPDzrp1akKRLJYKnd8CFn2cqmu9irHkfXRLDgD6CcKKnm79wB4d5k2zzPGThU/3z70YY0MkphIG1S50heAKkHZvea6hnJWQs9ysFR8FUpQMa1Y4gIAuody1liJCNddeizW/PcluNERiyFTE/lLUJURqEREs4JUnTEych85jcWqF1JYI1QrB4fXmGQJ+wMy67ejjXQEaUhGrY2xelcvvv7n9QB4uiSJuU3lCZRKRGTi2XKR8HOaeQXYzsGMTf3khNMG5YcigeJlubuGskjlDatInh+k045uMEQqyNpQ1gaVK7iyXfj3Z890MpQrIF8wraDvQ4VKoOT7qolHMJTT8d2lm3DsN57EK3v6hIpP2KA8nSTCyOQNxX1/lBJUlcxLmbfCCPxsUDOE7dHPO1SFPF9KScO8YKPbBuXlDexFMGSqI5lNQ0pQkilxluNgjPmaCdTPR8LYjDfKnuSMsb8B6D0MYxk15CEnubFKdPiqDaq9P4Ov/nEdMnkDBwdzmFxTmQ5WumVXks0cgEv91p/RXWOtT0ZdzhnFTMY+XnwGc5W/8EIiGsbeXk6gnKWc5XnnJFCquiIcsuegKyV91CYi2D/I71NK8musiloEetVObvu7+tw5Num1EglKzb4xZLmE+7+POYLwbegYrFjFV94Tj6x2EY0sJwNnoT/P70oVX8F0qXi8UNYGVcbV3uv+QFFylzE/zSMsZ1EKqpRrSVCxMPpSOn75t51W3riaeMRaL17SkcxZKaWeci7fpSBVfD2pfEUSlLzP7MbyzAagMCsl3qW0BUsGWc6/F6PrRaAiGk91lNNNJKKa5Wkp16CzHAcvkFj6DLFLUBPH8jNWpPJtRLQWQAeA6xhjG7waEdGnAHwKAFpaWtDa2nrINx4eHsbOjq0AgA1bdwAAXn35JexJ+E/y1gOcmL3w0mrc9UYeewZNNOldGMoVMNzdjtbWg2XvvVFJVNvb01P2ecjIA4oBOZ03kBoacH3vQMrOVa3ZyO1Su3dsRWt2l6vfjvY8coUChjMGurs60draV3IMWi5rGVZ3bloHo724YA928wN1784taE3twPDwMFpbW23j2bzxDcS7N1v/r1290tPLLmzmsbeXHzr7tqxFps39Pja+vhoDOdHmYD8A4PzqLrS2HkRc4ylgOretRavHd1V09RVLQmzZtRcAsGP7NrTmdpf8TveQVGUBqYHeku9uV0fRRrLmtVfRva30WAo6P9QL+SwKJrCni0u+bds2orV3i62tnFuJ9jbeNp0voLNjX9n1t789j6xeKDnu/lQG/d0HKt5jWzvFfli1GntrQlaOy46dm9E6sM013pFCvmcA2Lp5AxI9W1BIZ/Fij12DMdTTiX3DfI4HPcbf3ZnHcFbHytWvAQC2bbSvYYly42WMoTFOuO/5rSh0c0Zt+8Z1MDtKMIEmw2VzI7iorvRaUZHNcObMNEu8IyOPPfs6sGo1f8+xkIk8/4Kr/bau4hqU13q6sxgaNrGvMws9x6AN7AMAdHS0o7W1G8QM7Gzbi9bWLgBAShf7bM8utLbu8xxzOMSDhgf7un2f8VDXwkgwFgTqNQCzGWPDRHQZgIcBLPRqyBi7A8AdALB48WK2ZMmSQ755a2vr/2fvu+PlqMr+v2d7uyW3p/dCOklIQoDkJvToC4KIoIL4iqCiIioqoCD6/gSxvIq8NAsKIsFCUTqEXEogPaRAem83t5ftO7vn98fMOXNmdnZ39t699+6N8/188sne3dmZZ2bPOc/zfJ9ycProycDWTaioHQbsP4TFZy/K7QXtagI2r8f0WXPQvmkdgBSa7NUAjuLMWVNRPze9e7ke3v2twIY1AIDammrU18/NevxLB14DoN1aorqyAvX1CzTvxaQkvveOuptv9dARwL4DmDFtKupnD08779roTuDIARC7HaNHDkN9vXG7FgBo6PoQm5oOAgDOX7xIU+v1pwPrgOZmnD5rBuqnyAZEfX29XDz7zpsAgNNnz8KSSdV4Z2YYZT5nRquz7qPVOBHqgN9lx+UXLdUqsVdeBABcdC479yp0S3aUuIHzli0FALw9N4qj7RHMGWXc70yE9M7rgDy94S+vAo41YtppU1A/b2TG70QTSfzwvVdAKTB2xFDU188yPm57I7BVrotbtHA+xmUpPSjd9BaawkGUBvyQUhQHle0Zzjt7Pibq4lDs2TLstu0D9uxEkgLjxoxGff2UrPe8Kb4LLx/ciyVLlqQZCN3RBEKvvIYzpk5AfX16MocRpI9OAh9swKzT52DmiHJ0bzkOrNuM886ej0m1JWny5otIPAmsksf0GXNmY9H4Kvzt2EZsb21Euc+pUM9RnDFtotxaatceTJ0wBvX1kzXn2SLtwUsHdmP4uMnApq1YetZCvpGiCDPyXk/24r5XdqF8+ARg44eoP2tB1t/3vGXm73fIh6txqKsDXrfbUI4hm99GaYUP02eOBdasQUWJD92tYXjczrTjXftagE1rAYB/9mLzFhwKtcBf6odTSuHbn16EadMaMWdUOWpKPSh7fyWGVFXxcd3UFQVWrsS0KZNQr3Rz18PX8KqcmDRiOOrrZ2S8t96OhXzQa1+OUtpFKQ0qr18C4CSEGPcL6SMwuo5RfNkKO/l3FBqpPRTnvdb+vlG2LKYOKzV1XUeWTB0juAyOMZLV7bBrqAvWPicTxSRm8eWioYYLCkmfkcRoFptuwRNjL6zj+cgKX1ZKhH02obYkbQG9cp6s/AkhnCIMxiSUC7un1pR4TCknQKWl7ERNksgVh/Q47Tw7LVuSRD4UnyYGZSc8nV5PpRpBQ6OaTDOnNH1LBQA41Cp7lGayB/n5HGqaOaD2ZDPbbDYXxOwxNvdYc9nx1QGeWDOs3CvEhNLHF4s5sQxRMzWPmTBBUUbbjnVmvF5P4eR0b4YYlEveUZiNXRajMwoVMFquRtgQVM3iU+NKF02v401v5Q0iVc+LNQ/OVlLBzjPYkiSyghBSR5QViBAyXzlna2/Pmw/0SRJmYlBsAdul29rd67RjYk32Al1+jjwVlNHYyCTrW7fW47VbFgMAgjHjegrx2inK0taz37vYQFYfo/j+xVMwfXgp5o2p0Lyf730C6mSfZPAs77tiFg7e+zH5OI+DJ2dUmEzh1ePx/56P688ei3I34QrKTByHpbBnay6sKZLMofR4DErYlsNlt6Xty2X4XXHzQ7N1NpDTwl/adgLPblZpG7aX0iiT8RJAjUGw7Ljm7hicdtLjGI8eopHClBWLS42p9PPAfLnPiclKA+Npw8rSzsOKntneTGa3tDcCi29uPdqhnKtwyQFs/GVMknDIWXzsebNrG80vFsdaLiSsOJUYE9uhWA+5oFmlT1/cegIOG8HZE6ozyszWDjOdePoLOSUhhDwFoB5AFSHkKIC7ADgBgFL6MIArAHyFECIBiAC4imbada+P0KMkCeU7LBg8oSaAvU1BTB9eanovFHHwmVJQBtlZRvspAXKyhL5TdzYPCpAL87L1rQPUvWyMMG1YGV74+jnpMuZp3QNqoHVCDmXvsNtQ5nWiI5zgmVX54vRRQ3D6qCF4afNBnsVnZsM11ig3W92RLw8Pij0bl51gwdgKrDvQhhkjysx51478njH7zePJFL76pByPuex02TNl3RGqS8x7P35dq63m7hgq/W5TRan5ghmUbFxPqAngcwtH4bZntmHmiHKcOa4Sc0cPMUzPZ79HY2dM02y4JxhT6YfLbsPuk0F4nLYet00yAhtTmQwlr8uOtlDcwINKP37GiDI8dt0ZOHuiSkw5bDaexWeU+MAKmhlag3FUBdxZxwTzxitMlhX0B3IqKErp1Tk+fwDAAwWTqAdgAyuYoTmj8XfkgcA2KDtzXCX2NgUxY3juAl0GccEy1erIYKxmk9WhFHxm2gqaX1sY1Lkm2cgKeVGePtwcjSnLKNZBmVuwbjl/Euw2gv+aNSznsUN8LllB9dCDYnDbCTpM1EExLJ9Rh39uOpr1up680szVRenq+aOw4WA7rl4wyozoWhrVxDN26Sg5Eaz41cy2MwzMM2F0cnMwlpeCywfMs2Pj+ewJVZgxogyvfHMxPyZT7ViFX5Zpx4murGnhZuBy2DBteCk2H+7ISlf3BGz8ZWI9PA45i4/RmYwGzuT4L51Sozs/4RsWGikon8uu6XYSzlLoz8AMqZ4ain2B4vHlegEWg+qOJuC0pxfGGX5HWciZgrrurDE4a0JlWneFbNDXB+W8pnAMq4jPFS9zO2zcM8xkjYmTIFfhIiEEb9+61HTxJaC7TxPUGSA347zn8pmmjh3ic+IAzFfpZ4LLDnQFzXvR555Wi9dvWYxRWWI1PYpBOeTCyaduWGhGbADa381sqyNAq6CCMQk2Is8DG8mvRkjvQbUEzZdbmMWP/muqpkHwdy+agnNPq8WMEelUXiacMWYIXHYbjnVE+NbrvcH0YWXYfLijoPQeoM7JTN6w12XH4bYwjnVEQACMVFp6pTKXtmnPz+qgMnSH8LnsOCr06YvEpZyxJTbPzVDS/YVTQ0E5mIIy3yqeTRK2cV2Z14mLpudXlCheS59YYCyn+ppxxLnkdTvtQpJE5hiUenzu+8+2IBtBVPhmPah8EFCs1wp/76xYl50gqcxwMwYDgLTsOj00e/HkjEGpFF++0FJ85gp1AW0nlK//dRNW7WrG1fNHIuB2mN5qAxA8KMVbb+6OYdpQ84rDDK47a6zm71KPE0sn12Q42hg+lwPjawKyB1UAr2eSEu8KZtmduCdg1H0mD6ozkkA0kcJv39yLcjfhhpBRo10jOGw2pZt5kvfRE+F1OjT1gWZaXzFlWl7AZJHeongqsnoBMUkibwWlZLfkQ4cw5B2DUo63ETWbJqeCctj4opHJGhPlKCSPbnytwg8ZZr32NiAvJqEUSpGKCirXgi9m8eULcY8tM99nVjMzsABg1S65pua1D0/mnTwgbqeeSlG0BOOoKikeS1oEy0QtRNbdDGX7djPtqPKBk3tQxmNmy5EO/rrcre4ybZSVaXh+toNzhn3iWEsohmwbnjKwYl8z3fT7C6eIglKtSbOLgxiDctpJTuvYCPlmtzHnZsmkap7anKutjdthU2NQWbL4xOP7EmY6ZuQLtuCwjhk9hUsTxynMc8jneTKruSfXzjcGxejQgy2htM9aQ/G8KSuH3SYbQ3EJ7eE4kilasBTzQoPtYtCbFHOGWSPK8K+vnYVfXTm71+cSwQy5TJ0krpqv1uh5HKrBbNqDEsZYJopPTJLItQUNAPzqytn4wcdOKwh1WiicEgrKkwcNw8AGhJSiedMhDPkmSdT4bLj/6tPx28/M4e52rhiU3BIllXY9EfnEoHoLMzU9+eJzC0ajttRtKqEiG0QPymysLBfyyWJTa196YOyIFJ8J2dnvcMBAQQE9S5n2ux0Ix5K483m5EUyNyX2g+hssySIUT+Y4MjcIIZg5orzgCSFsLGQyXG+9cArW3HYu5o4egs9PdQsKytz5xTlv5PF4nHbEpBTfkNEMxVdd4sb154zr0VrYVzhFFJQN7PcyS+04bIR/p6e1FBqKz+R1L5k1DAG3g1s9Zig+o+uJ0GTx9bF7brabcz4YVenD2tvPM9V3LxvE+ecy0XC10OAxqB5c26VJksj9faag9jcbK6ie0KU+lx27GruxcudJVJe4sWRS5pqZgcRZSrd0Rs8XI3LVQQHyPlv//MoiDA3YOMVrpskAoF03jFLtea9RJYkm396MxYJTIkmCEAK/y5FXDIoQApfDhmgiZbp7uR6i+27Gg9J812S8wq0p4Bx4D6qY4epBtqEZ/PVLC3iWVTaoqcX5XzvfLL5yJQtzv+JBDSvzoMznQnWJG2/vbtYUZJuF3+XAuoNyX+jXvrmoT4yRQmBSbQC3XjgZF0ytHWhRMsKRpzfNlEyuOkb9+TNdg60bsUSKN6Y12ly02DH4JM4Avzs/BQUA5V4XqNbdgQAAIABJREFUGhPRHlMZNhvhezGZ7bDAwNKDc1ExYlZeZg/KUlCAjuIrYLbhovHmOnf1JklCNETMJKK4HfK26fub5Q0an73pLNSWeviWJT3pQs4y+apL3HlnevYnCCG4Sdg7rRiRqw5KD1VBmfNyRGPVWEHJ78WSSVDqQDie3w7LxYJTZjVjkytTZwYjjKuWKaXaXtR7MOWQr4Jq6Zazr3L1SzND8eXiowuFnmQ69ifcwnMwS5UUEizhxcx+TnpoCoJNjqUyr5PvJ8QovWVK2vbkHmSlsVqonracsqCCF22bNJTYeDVL0YtjzIgSFltXRRJJpGjf0PN9jcEncQawxdOdx8LEsoGGlfdCQSnUXr4UH2vGOboie9xFtKwzLbr94UFtufMC03G2gYLY87Uv0uFzgVm1PVGOmpZKJn9DVqrgcdq4gjtvai3eurUeo01s9JhJhvIianUzWMFoXrtJupdCTmYwO39FGtlovKk7Vad4duxg/F1PGQ+KWX/5NDq8ZuEYXHvmaHwqy7YMucD0Ur6L97VnjgEADM2hHLUelIk6qD7yoMp8zsHlQQ0A1cl2Pq7sQaajmALsN7mjKfs9yr3a6/VEOQGqhd0XmZr/aci3Do95wqYpPrMxKCnJd/EupgJcsyjuFScPMGsyn/TaGSPK8mqzYgRWtpBvgezN503EN86dkDOlUxODylgHpR6Taffa/wSI+jOf7c4LhUZlB926Mm+OI9MhprObHcNMoRTKMmYeVDH1YhusyNeDZxTtNJNb/Wj6Y2bJ4otJKd4xvVCd6fsTp5CCkm+l0D21zKIne6jk0zMQyBx8Fxfjgbr/YkDAqT7PgajlYJ3xWUV+T2HWU2W/daEUFBtfhdrm/T8Z+Ra0zxk1BH/90gLM1211k/H8ORo4i80LOhXPPp/+m8WCU4biY325erM/TE/A1kGz6aH5gg00QjInYojNHQci9lIs8DsHNkb208tn4BOzh5ne8DITzAazM1F8PQVrqVVXpAW6gwk9yeRcNL7K9PwVlVKmTU8B2YNiFJ/lQQ0gKgPyJM03WaG3YFfrq+w5FlPKVAMFyN3ALQAB18AqqGnDyvDrq07v9XnMxs8KTfHxWIU1nnoNZlj2RWswQL8FTpY080TSUlDFAOY5dUd718+tp+irbZLZQGNZPkYYjAOvL+A/ZUazOTAPqlDGEd8PyEoz7zUunj4Ua/a34lNzR/TJ+U0rKCmFkLIjt89k8k0xYfBJnAHMi0iYbWZVYJjZ5qJH5xW29s6E/2RaT0S+tWiDHWco8QpWz9db3H3JNEyuK8GCsebiIBYyo8znLIg3nQkOTcaqUQxKpfgiiSQ8TtugnB+njIL62Myh2HKkA19bNrFfr8uC8X3lQQ3G4rqBxJhKH+rz3GNosGLxpGq89/1lqClQo9OaUg++ed6kgpzLQt8ipwelGMzbj3UikUwNSu8JOIUUlNthx92XTu/36/Z1DMqsgnry+gVW7ABAw61LB1qEHmPd7efmzQAMK88/pd3C4EeuOii2bvzpvYO4/PThg7LNEXAKZfENNPpKQQXc5s571oQqTBtW2B1QLfQvako9fG8sCxayIZcHFXA7+PY1zcGY6eLvYoOloAqEPqP4lIFVRFu0WLBgYYCh2W4jQwx6+fQ6AMCHx7sG5VYbgKWgeo8+roOyYlAWLFjQQ7PdRobmxOOq5Z1x20Jx3mlnsMFSUL3EsilyQL7vKD5LQVmwYEGLXBQfIO+UwJiXwbgXFGApqF7jvitm4u1bl/Z5koTF8FmwYIHBqdmc03h18DjtPKZpeVD/oXA77H26uRsbWAPRW86CBQvFCXG7jWxrA9tvrnaQtq8anH7ffxC8TjtmjSjDjUvGD7QoFixYKBKY3c6DeVBTh/auP+RAwVJQRQ5CCJ7/2tkDLYYFCxaKCGYZlVsvnAKfy4GLlIy+wYacFB8h5I+EkCZCyPYMnxNCyP2EkL2EkK2EkDmFF9OCBQsWLOSL6hI3fnTJtD6Lkfc1zMSg/gTgoiyfXwxgovLvBgAP9V4sCxYsWLCQC4O1Q4RZ5KT4KKVvE0LGZDnkUgCPU0opgDWEkHJCyFBK6YkCyWjBggULFnT4+5fPxMghfZegVQwoRAxqOIAjwt9HlfcsBWXBggULfYQzTO6+O5hBZMcnx0GyB/UCpTStGysh5EUA91BK31X+Xgngu5TSjQbH3gCZBkRtbe3cFStW9Ep4AAgGgwgEAr0+T39gMMkKWPL2JQaTrIAlb19iMMkKFEbepUuXbqSUzst5IKU05z8AYwBsz/DZIwCuFv7eBWBornPOnTuXFgKrVq0qyHn6A4NJVkotefsSg0lWSi15+xKDSVZKCyMvgA3UhO4pBMX3LwBfI4SsALAAQCc1EX/auHFjCyHkUAGuXwWgpQDn6Q8MJlkBS96+xGCSFbDk7UsMJlmBwsg72sxBORUUIeQpAPUAqgghRwHcBcAJAJTShwG8BGA5gL0AwgC+YObClNJqM8eZkG8DNeMqFgEGk6yAJW9fYjDJCljy9iUGk6xA/8prJovv6hyfUwA3FUwiCxYsWLBgAVYvPgsWLFiwUKQ4FRTUowMtQB4YTLIClrx9icEkK2DJ25cYTLIC/SivqTRzCxYsWLBgob9xKnhQFixYsGDhFISloCxYsGDBQlFi0CooQshFhJBdShf17w+0PIBx53dCSAUh5HVCyB7l/yHK+wPaBZ4QMpIQsooQsoMQ8iEh5OYil9dDCFlHCNmiyHu38v5YQshaRd6nCSEu5X238vde5fMx/SmvIoOdELKZEPLCIJD1ICFkGyHkA0LIBuW9ohwLigzlhJB/EEJ2KmP4zGKVlxAyWXmu7F8XIeSbRSzvLcoc204IeUqZewMzds1U8xbbPwB2APsAjAPgArAFwNQikGsxgDkQum4AuA/A95XX3wfwM+X1cgAvQ97NfSGAtf0s61AAc5TXJQB2A5haxPISAAHltRPAWkWOvwG4Snn/YQBfUV5/FcDDyuurADw9AOPhWwD+CrlNGIpc1oMAqnTvFeVYUGT4M4DrldcuAOXFLK8gtx1AI+RC1aKTF3If1QMAvMKYvW6gxu6A/EgFeIhnAnhV+Ps2ALcNtFyKLGOgVVC89RNkpbBLed2jFlF9KPfzAM4fDPIC8AHYBLlzSQsAh35cAHgVwJnKa4dyHOlHGUcAWAlgGYAXlMWmKGVVrnsQ6QqqKMcCgFJlESW694tSXp2MFwBYXazyQm3+XaGMxRcAXDhQY3ewUnyZOqgXI2qp0vpJ+b9Geb9o7kFxy0+H7JUUrbwKZfYBgCYAr0P2ojsopZKBTFxe5fNOAJX9KO6vAXwXQEr5uxLFKysAUACvEUI2ErmpM1C8Y2EcgGYAjykU6u8JIf4illfEVQCeUl4XnbyU0mMAfgHgMOQdKToBbMQAjd3BqqCM9jsebPnyRXEPhJAAgH8C+CaltCvboQbv9au8lNIkpXQ2ZO9kPoDTssg0YPISQj4OoIlqO/pnk2fAny2AsyilcyBvQHoTIWRxlmMHWl4HZCr9IUrp6QBCkCmyTBhoeWUh5LjNJQD+nutQg/f6a+wOgbzH31gAwwD4IY+JTPL0qayDVUEdBTBS+HsEgOMDJEsunCSEDAUA5f8m5f0BvwdCiBOycnqSUvqM8nbRystAKe0A0ACZny8nhLCWXaJMXF7l8zIAbf0k4lkALiGEHASwAjLN9+silRUAQCk9rvzfBOBZyAZAsY6FowCOUkrXKn//A7LCKlZ5GS4GsIlSelL5uxjlPQ/AAUppM6U0AeAZAIswQGN3sCqo9QAmKpklLshu878GWKZM+BeAzyuvPw851sPev1bJ2FkIk13gCwVCCAHwBwA7KKW/GgTyVhNCypXXXsgTaQeAVQCuyCAvu48rALxJFaK8r0EpvY1SOoJSOgby2HyTUvrZYpQVAAghfkJICXsNOU6yHUU6FiiljQCOEEImK2+dC+CjYpVXwNVQ6T0mV7HJexjAQkKIT1kj2LMdmLHb30HCAgbzlkPOPNsH4I6BlkeR6SnIvG0CsmXxRch87EoAe5T/K5RjCYD/U+TfBmBeP8t6NmRXfCuAD5R/y4tY3pkANivybgdwp/L+OADrIHfT/zsAt/K+R/l7r/L5uAEaE/VQs/iKUlZFri3Kvw/ZfCrWsaDIMBvABmU8PAdgSJHL6wPQCqBMeK8o5QVwN4Cdyjx7AoB7oMau1erIgoUCgxByO+SJen0hjzVxLgpgIqV0b2/PZcFCMcBSUBYs5AAh5DoA3wYwHkAX5BjNbVSOhRUNLAVl4VTDYI1BWbDQLyCEfBvAzwDcCjkAvBBykeXrrJped3whdqm2YMECLAVlwUJGEEJKIfPxX6eUvkIpTVBKDwK4ErKS+hwh5EdEbrnzF0JIF4DrlPf+IpznWkLIIUJIKyHkh0RuK3Se8hk/lhAyhhBCCSGfJ4QcJoS0EELuEM4znxDyPiGkgxByghDygJGStGDhVIGloCxYyIxFkIPAz4hvUkqDkFvRnK+8dSnkVOdyAE+KxxJCpgJ4EMBnIXcLKEPuosuzAUyGnEF1JyGE1XslAdwCoApyNf+5kFvNWLBwSsJSUBYsZEYVgBaqVtCLOKF8DgDvU0qfo5SmKKUR3XFXAPg3pfRdSmkcwJ3IXch4N6U0QillmXWzAIBSupFSuoZSKime3CMAlvTs1ixYKH5YfLkFC5nRAqCKEOIwUFJDlc8BbVsaPYaJn1NKw4SQ1hzXbRRehwEEAIAQMgnArwDMg5y27IDchsaChVMSlgdlwUJmvA8gBuBy8U2lmPViyLUrQHaP6ATkynv2XS963qvsIcj1KRMppaUAbodxqxkLFk4JWArKgoUMoJR2Qk6S+C2R9x9zKo11/w65EPsJE6f5B4D/IoQsUhIa7kbPlUoJ5DT3ICFkCoCv9PA8FiwMClgKyoKFLKCU3gfZU/kFZOWwFjJldy6lNGbi+x8C+DrknnwnAHRD7rmW87sG+A6Azyjn+B2Ap3twDgsWBg2sQl0LFvoRSvf4Dsg03YGBlseChWKG5UFZsNDHIIT8l9J80w/ZE9sGeYNACxYsZIGloCxY6HtcCnl7guMAJkLeOtuiLixYyAGL4rNgwYIFC0UJy4OyYMGCBQtFiQEr1K2qqqJjxozp9XlCoRD8fn/vBeoHDCZZAUvevsRgkhWw5O1LDCZZgcLIu3HjxhZKaXXOA01sXvVHyGmx2zN8TgDcD3nDqq0A5pjZiGru3Lm0EFi1alVBztMfGEyyUmrJ25cYTLJSasnblxhMslJaGHkBbKAm9IQZiu9PAC7K8vnFkAO/EwHcALna3YIFCxYsWOgVciooSunbANqyHHIpgMcVxbgGQDkhZGihBLRgwYIFC4XH4dYwY8GKFoVIkhgObbPMo8i9nYCFIsWek914sGFv0Q9cCxYs9Bzv7mnB4p+vwr+3nhhoUbLCVJq50n/sBUrpdIPPXgRwD6X0XeXvlQC+SylN67JMCLkBMg2I2trauStWrOiV8AAQDAYRCAR6fZ7+wGCQ9btvh9EUpnhgmQ+Ih4peXhGD4fkyDCZZAUvevsRAyPqP3XG8sD+BT0xw4hMT8tvzshDyLl26dCOldF7OA80EqgCMQeYkiUcAXC38vQvA0FznPFWTJEZ/7wX6s5d3GH5WbLIaYenPV9HR33uBbj3SMSjkFTGY5B1MslJqyduX6I2se05209Hfe4G+8VFjXt+75enNdPT3XqCPv38w72sWW5JELvwLwLVExkIAnZTS4vYb+wjJlOyNPtiwb4Al6TmqS9wAgMNt4QGWxIIFC7mw/qCcHvDahyfz+t6+5hAAICGlCi5TIZGzDooQ8hSAesgbtx0FcBcAJwBQSh8G8BKA5ZDTzMMAvtBXwhY7YlJyoEXoNZiCOt4RwcQBlsWCBQvZEU3Ia47Had7XCMclbDnSwV8XM3IqKErp1Tk+pwBuKphEgxiR+OBXUC67PNC7Y5JihliwYKHY8NHxLnicNsQUD8jjtJv+bkt3nL8OF/madUq3OlqzvxXT7nwFHeF47oMLgEiiuH/sbAjHJVBKEU/KAz4YLW7LyoKF/2Qsv/8dLPvlW9yDcjvML+UdEUtBFQUeeHMvQvEkthzt7JfrRQepgmrujmHqna/iD+8e4BZZMJYYYKksWPjPA6UUwVh247Arqs7NaEKer+48PKiOsPr9UI5rDTROaQVFlI21aT/V9ETixR1wzIS2kGxRPbHmkKCginvg9hWCMQn3vLzjlKBrLQw+/GXtYUy/61Uc64hkPGbPySB/zeLe+VB87QKjFC5yo/oUV1CyhuqvmtPoIE2SSCi0XlswjpgyYIOxwXkvvcXT64/gkbf245G3C5+JufVoB55Yc6jg57Vw6mD9ATkrr2FXU8ZjmPHocdq4B+WyE9PX6IzIHtTwcm/Re1AD1s28P2BTfjOW/t3XGKxWN+Ohu2OS6kFF/zMpvlKPPCW29gEtfMkDqwEA1ywcXfBzWygMVu44iUSS4qLpdQNy/fHVcgGs6CXpwdYZp93GDcp8ljhG8dWWuot+zTqlPSib4kH1lxvb30kSW4504Ft/+wDxXtYyiKmm8SKh+LoGSEEmkvJMf3NnU5/FFPvLYMqGdQfa8Lf1R3If+B+GL/55A778l7QmOP0GKSXPv2zjX0yMYKyNlMeY6ook4HPZEfA4uUFarDilFRRzeiNxCdFEss+TGPo7SeJ37+zHM5uO4d9bjvfqPKIVxTjtvsrim/LDl3HbM1uzHvPC1uOY+aPX8OHxdC/maHsY1z22jtMUhYaorFlsrtAYaOUPAFc+8j6++8/sv4OF/gdTGM9tPoZ/bDxqeAwzhF12GyTFoEoJCur0H7+GKx9+P+M1gjEJAbcDHoet6BO7Tm0FxTyoeBJn3rMSi+59s0+v19/u8rgqedOwQ73s+hDWKCh5gsSTfWPlRxMpPLUuu+X+T2ViHmtPDxTfv3IPGnY1m1bKe5u681JmovLorWdq5hr5Ys/JbhxsCRVQGtlaX/jTldhwMNumBcWPrmiioMr/YEsIE25/CbtPdmc8prEzWtB5HxUou+/8fYvhMZzic9g4tZcUAu3t4QTWZfktu6MSSjwOeJx2y4MaSLAYVDieRHs40WcWMUN/WyMJZXSKC2kyRXnSQzJFDRfZE50RPLf5GP9bpEDZgGXn2N8cxMvb+qZzFaUUneF05cEymAIebYj09me34W8bZOUlJc1NrPN+9TY++dB7pmUyUtaFRk8D08kUxfn/+zbqf9FQMFkopdh8uAONXVH87xu7C3ZeM7j2j+vwmzf2FOx8M3/0Gi594N2CnItSipe3N0JKUTydgQpNpSgW3rOyoJRgLJF7zIkeFMtQzoc27o5JCHicSpKF5UENGNiP1lVgOiiaSGLOT17HS7qFOyIMrhOdEfzwue149cNGPPBm4SYhA6WUD2ZxkH3yofcw8Y6XAQA3PL4Bk37wctp3b3xiI7759Ae8gDmi0FqEqMqOKahlv3wLX3lyU0FS9RM6pXL/yr2Y9ePX0gyHE51RAOD0BcNf1x5Wj+mK5rwek3lvU5B/f92B7F5CX3lQ4vPr7iF9elK4Z/2z7CliUoore4et8MtBWyiOp9YdNvzs7d3NBVOKBxSvkvWYM8Ly37xjWiFGEkn4XHLqdqZi1oOt8rXe2t2cj6hZoc8ENjLE2Hy32whSOgVlZp52RxMo9TjgdtgNFRSlFG/tbkYimcJdz2/Htn6qIzXCKa2g2I/NFiigMF7Oya4o2kJxfPPpDzTvi0kS//PCDjyx5hBufGIjfvGaPAlX7IwXhEb56HgXxt72Et7cKTeIFC39D5QeWwCwcqdxqipT2I3KgscmoJ0QPtD1C2AhKs71SSR/eu8AAKAtFNO8z36jbIvwyc7cCiqqs0Z//cZu/CVHmndYVFDJQlI3qixGNNSOE105PSuxgW+4F2UA4qIXiScR4llh5lOVGY53RNIWxdN//Br+34sfAZCNpNue2ZZW11Po2sRczY1TKYqPTnSZVoidkQQ3UCIZ+tXtOJGZ+uspookkhpV58NkFo7gcejCKL5miKsWnvHhm07G04/UIRpUYlJCmLmLjoXZ8/o/rcPOKzfjz+4dwxcPmGYhC49RWUMrDbwmqC2BzdyzT4abBFmu9hS0qPwrtBIxLKbxyMIErsgQvzWL13hYAwMFWeVLG8lS6VQG5IeyJDq2CklJUUFAUEUm9h0LQo3quvl2h97p0HgXLpMumoMwozJBuYZFSVFOkaASx/ssM3SIlU/jCY+t4V+nM51Vl0SuiRDKFi3/zDr7y5Kas5zgiLMLBHjb5fGV7IybcoXrVobjEPel8Paj/W7UXi+59U+NBUErRHk7gd+/IxsfWY7L1HU0kISVT2HS4HUDh6dOWHPO6I08W5a7nP0TDbtnAy1QT2CoYVvl62y3BmOGYiUkpVJW4MX9sBQBjuZmhl6SUZ/0lUxQ7TnTh2xniVuyaJzojCMbUGFRUSqYZC4zBeGlbI5dpoHBKKyi2qIqLQ64FygwyBWLFBdjj0FZ2t4Z6rxgzXT/XANLz0xV+eYOy452yVStmrokUQ8ORwma0ZVIqmShYMVEjlaJw2gnOmlCJ2SPLTaX0i79HXEohIaVy3kckIfHYZcxAQR5uDeNnr+zEic4Ilv6iAe/ubcGqXc24RedN6yE+Y32GJPOG3t/XkvUcrYLsZuNYTV1RNAre5toDrZrPI/EkfyZ2W34e1CrFQz/eoZ5fb2ywhTsUk/CblXtw+YPvYfuxzoLT7qIRamTYNHXn9rhFvPbRSazeKz+rTLsUiPPwSHt+iUpXPvI+PvXw+2nKIZpIwuOwo9wnz1GjPqJs7KdSlBtzSZoeb9Y/h3N/+RbOvOdNdEclBNxOeJx2UAref5PhpAn6vL9wyhbqRuJJzhGHBAuoEBk3mRSU6EHpF1BxEuvx25V7ML4mgOUzhpq6vn5xysQjM8SkJHwu9acOuOXXrGBPVByUysHXeDKFg53q+7kW9l2N3agr9aDMl7kFeqZnnynLTtyrJhiXkEhS1E+qQcPuJlMelHjMya4oEqncCiqaSKHM60R7OGHoQV3/+HrsPhlEKCbhQEsI1z22HgBQ6sne+l0cM90xCa5oCqt2NWHp5Bru6eWKc4vPyayCmv/TlQCAg/d+DAAwrMyr+TwUT6JdeSb5br3AxrgoS2vQ2BALxiRsV7yp4x2RvJqbXv7gaoytkgtYNxxqw1u3Lk07RlTe+vG0Zn8r/r5BTdlOpShsBso4U+JNJu9INDQOtoR4ka0Z7FdiZfp1IialEHA7UKIkCHVFZQ+3xOPkBgSb71KKcpmTKdWbEuUb4ld3y2XPhXlQ7DeISSm4BYO6mBTUKeFBGWWw7Gjs4i2ORKonkkji9+/sx/V/Xt9jZaW3gBPJFCLxpGawtQa1C+HxLL21fvn6bnz1yU1Z04cTyRQfmHrqink9Yi2EaBXpeWZ2lMqxa5+DW9lb5kh3CrWlMh2Ya2G/8Ndv49OPZqcvIwlVbtG601vdRscwWd1OG7xOu6nfTlxwT3RGISUp2kLxrPGPaCKJUq+sbPSWJaDSoqwInOGjE104msWKFo2kUEzCLzZE8YXH1iORTHFFmhLk6gwn0uI22iafue/fiM7WL4jhuIQ25bz5JG/saw7iw+Nd8vdEBZVhnIRiSTiUrVwiiWRehdibDnfgn5uO4p+bjuJQq/EzFim+Dl1m6FWPrsE/N6kKKhM9avR7G72/+2Q3nv/gGEIxCYTI2cLv7Ws1/G4u6J95NCErC8bAtIfimP3j13HPSztUeZS5oPGgUjRtTGRLuWcUn3xN7feaChAGKRQGrYKKSUlsPNSGl/bHMf72lzT8PCDTDzYCjBii7Tf13OZj+J8Xd+CNHU1Yc6Bng0o8XzJF8aXHN+C0O1/R/NAtOkovU/NHcbH8+Wu7Ml7zjme3YcoPX8Hh1nAaJ84sfXGhiMZFBaU9nk049r/eG2ED92SYYsbwMgByjcnzHxzDWfe+mba4sPPvbMweNBab6TZ2RjmVJtI9olJKJFNo6orisgdX84XJZbfB63LkTfEd74hASlHEJFkhJFOUGw2UUjy59hC2H+uUFZTiDekzGgH1Gf/pvYNp1/vf1+UMsev/vB6XPbha85loVARjEprC8u/eHo5zRSrqzdue3Yqz7n1To/TE56Q3UoxwuE01eNi96n/riOBB5aohuuYPa/GMstDf9fyHqiwaD0pVUKLBFI5LcCg/eHN3DO0h7RhKpSjW7m/tcQF2i8aDiuN3W2M8VqtHpgQTvSG3eFI15o4ekuZBXfC/b+PmFR+gOyphWJkXy6bU4vWP5IQlvfHTForjSFsYf117GDN+9Gqal9atm0sxKSkbYUoGIYsHPfeBmvwQF2g9saRE7wFnMwI0HpTuvgtNv/YGg1ZBtQbj+ORD7+Nvu+WHeUDnfXx4vAuT60oxtMyjoU6e+0At8GxTJtM7e5rxXg7+X4Q4kcNxCQ275CCxaA2Z9aDEhdaZJQbwrFK3dKgtpMk0A1QPKlMsSa+g2IDM5EF5FQWVpMCwcpkS6opIeGrdYRzriPBCWgazA1qUT8y6Enl2cfGIJylWrD+CzYc78Lu39wOQ+4/5nHZTdJS4GIsGQlsojt+s3INF976JYx0RvL2nBXc8ux33vrwTMUmm+AB5sXhm01FMvOPlNAPICAG3/Nze2NGEzYc7NJ+Ji3gwJkFhWdEajBt6Q298JMd3thxRU3w7InEePzRD8YlKYNG9b6I7mkjLSAsJMahstOk/dsfxzp4WfOtvchCeLaBMlmBMgpRM8bFgI1pPJRiTuLXf1B3DvmY5s5alcr+0/QQ+/ega/O/rPUs7b+mOwaUsuFuPdmL1cQmf/f1aw2NDccnw99THmr68ZBxqStxpdYYMR9rDKPE4UFPqRjieRFxKYextL+Hht9RGw2fesxLn3LcKtz+7Dd1RKc1YR2XcAAAgAElEQVQIMIrZue02vkOuEd3GqO9kSm1xZOhBCefWrwEBt5P/hnpjx8hQSeXin/sIg1ZBDS3zwC9MEqddeyudkQQq/S4+aI3A6J5r/rAOn/md8WAWseFgG5q6ohpFJFp84qDXW4KZYlAiHaGnR4IxiS9EqjueSuu7xRSOaAmJAzKakGtd/ueFj3C4Nax6UFIK+5uDabKK20eX+1zwu+zoiiZ4HGuXzlMya/WKynh/c5AbDiIVFdT1BWS0F7tnl0O2Lk3FoITriRRlezjOM89OdkV5ynprKI5oIoVSr4Nff4VSpHmkPWwYo1j1nXr+2ud2ZIzBsN/RZbchGJXgUVK620JxQ2XLxq2Y1NMZSWC4YjCY6ZigTwg62RVN8zx/88ZuTulkU/ov7Nf+xqGYhKlDSzG2yo+OcALT73oVd//7Iy4XIUTr8cXUbMGOcJx728yKP9IWMZTZCEYUbUswhhHKszHqQCLisdUHcM59q3hGIYPegyr1OOFy2DQUnziO9jYF4XfLnkhMSvJ58afVB/kx+gQmpuyYLcrWkvtX7sFlD65GIpmCy2HjFB/zoMRbVr0mtYYtSVUP6p7LZ2jOrZcbkIvgRw7xAUAabWpE9WaiP/sag1ZBEUIwobZE+Fv+f83+VuxvDqIjHEeZ16kJ/unRGopjT1PmrsEiDreGccXD7+OSB1ZjZ2MXf/9kl7ogHe+MZgz+tugWrlRKppV2CW1Umrq0x0y/61VcolTGM3okkkhqYhUAsKcpiK1HOzSTQZxsUSmJD4504PfvHsAdz21DXLEUV+9rwbJfvoVdJ7v5+QHt3jIlbgdKPE50RxN8suhbv5hN4RU9NXFSiM9Q9A5ZbA8QFnhFQZmpZ8sUvG8NxfkCkUpRbkFKyRRiOoqPWaFuh43fv4ixSrspQK4jMzpGlkWWt6bUjWBMgsdBuCxGypYtCG2hOF/024Jxfj1xwWnqjuKrT25Mo4vYYv/INXMBAC1B+Vpuhw0/vnQaALmwlY1Ns7VuiWQKJzqjGFvtR8DtwGqFfXhizSEuazJFNfWHwViSf9YdVT0YNmaZkWInuTMJeTsuKYXH3z+IRFJOfhk+RFZQLDsVMFZmK5RWW/rU9HQvwyEnDAnzSszGbQ8nFAUltwzarvSOHDFEm4hiJDtL6We/2a9e343NhzsQl1Jw2lWKr9FgPIm0XkLoxcfq2WaOkGl50YjRK6gSjwPja+Skjr26NTAYk3D2hCpcOW8ErjpjpEbu/sagVVAAMLFGzZqJSSkcaAnhqkfX4OYVH6AzIqHU64TLnn6LtaVuVPhdaA3GTFE3gFrU2tgVxeq9rZxqadalrzJ6SA9xgCSSKTyz+RjueHY7fqnEncZW+dEsLKL6yni7MqCj8aRhGu1L2xo1FIXotj+55rCm/ooNtsOCkigXsu9EBRXwOFDqdaArInG6YceJbiSSKfx7y3Hc/uy2tKA0pdRQgYgLoH6BBeQYxx/ePcDfZwshAD75XQrFl0hSfCVHixmmlOpKPRrvtD0U5wuhlKJcrkQyhaiUFCi+FJ/k0UQqLZW4KiCPgevPHgtAntj6xIRoIomOcJwry9pSj0zxKY+4LRhL81wiCl0EyAvX9LteRSSeRHMwhmHlXpR6HHhu8zGM+f6LaOqO4v6Ve/DStsa0Is32cAIOG8GYSlmptQRjiCaSmFATwKeVhYfBYSMIx5OmqJyJd7yM4x0RDCvzwO+2c4t7iM+pMQpYhiMgGwvMe+uOSjw+Ek3IdThsDMSkFB55ax+uevR9dEYShl4rYwr+uPoA7nz+Qzzy1j5IKcq9S9GDEik09nsxbzxFZSU65vsvYvuxzrTfocTjkD0oYXFu6dYt9IoHFZdSfLwxNseIhmVzlJWc6b2VYEyC027jhi5bd8Rfhc3fFFWVFRvHhADVSp2jaLDo2ZkStwMBtwMjK7xpnSKCMQnjqv2474pZmKbEoDOl2vc1Th0FlUhi5wnZszncFkZnJI5yn9OQ4rtoWh1qStxoDyf4AMiGfc1B3PUvNSgcjEmYN3oIgPSMl9pSj+E5xAU5HE/ylFsWB5OpkjhfIA61amNqrMo/KiUNq79rS90aK0ekV8QMphKPg084kSoUFatGQbkdKPU40RVNIByXMKzMg0giiS1HOvD1pzbjr2sP401dx4rHVh/ElB++kma1sQWq1OPgE8bnsqOpW14439nTwik1QPYi2CRjE9nlsMGnBHBe3t6INz46qemewRBNJHGyK4ZSjwPVJW60CDHBtlCcpxlH4km+kARjSSSSFD6XA3YbQVxSrx+JJ/k5SjwO/PyKmdjwg/MBAD/4+FTUlXoQ0ikoSimu/cM6zP7x6wjGJbjsNgzxudAVSaid9hOptNiBUc3O8vvfQSJJUVvqRmXAzYu0d5zoBlvD9SHMjnAc5T4XX5hbumMIx5PwOu1phlt1ibyo6VvtvLOnGb94dRcIgEm1WoOwrsyLmhJ1vIdiyYx0bzgucW+4O5rgx6WoPA7Zc4tJSdz7yk6s2d+GD493ImpguUelJNYdaMO9L+8EIGdQAupeSmJ8U/ScH/7cXM15uqMSVu6Qkxue23wsLfmohFF8ggyidwYAfredZ72KadzfeGozfvDc9nTZEywDT/5bX2SfSFI4HQSEELgdNsNMTFUpqXR/KkURiknwOu08C7Vb40Fpz1OisAQLxlZi7YFWvu5QSnmnCUClYJu7YwOyd9SgVlCTBIovKqW4ixuJywuNTPGl32KZzwWfQhMxF9rI02K46clN2HGiS/PedMWy0NNymRSUPrGCeSPHleuPqvAhRdVBxRQfiwfZhQVV9E4uO324cs6kJgYlLhTivYViScO6Do2CEp5ZiUeuyeiKyq1fTlcUs7gVgL7XGks4YYpLSqZw84rN2HSoHTYiFwozqmRUhQ/dUSnt+QJAQlJpC34vDhvGVvn439c/vgGf+L/VGmu1LRTH3J+8jifWHILf7UCZ16lZqNoEii8cT3IPilFdHqcNLrscV2ALSiSRRKdCmb357Xp8ap7WAwl4HAjFJY0XHJNSvKv0zhPd8LvtqPS75Ninckw0kcQJZdETs9z0YB51ZcDNvXdAXkAYjcXSvhlagnFUBVwo97lgI7IVHUkk4XXZead/Bqag9Mry839chwdW7QUFcMaYCs1nw8o8GiMxnkxhZ2M3Rlf6MH14KX/fbiMIxpJaDyqi/l4xKcXpyJiU4vGWpq6YoSceTSRx5SMqI8A8pgmKAm0XPHqmrG69cDJGDPFBRDAm8XKBFE33eFwOG1wOm6Zg+8k1h+C0E76TQMCthhFY94z2cBz/2nKcJzaJYJ5IIqUaiTFdz0m3Ml9FpkR8DiqtJ9B9lKIjnMAQnwtuhw0OG9HFoLSGA2NMpg4tRXtYNRhikqz0WKNmtn5+7P538dnfr0m7n77GoFZQs0aW89fRhGoJM/5+SAYPyuu0w+dyaBRFPJnK2BHYiFI7fZR8bX3tS12ZO6fcoVgyzXMbXSlPHtbdu0n5nCkOuxCDEj2lMq8TNiIrPXEQi4scex6Ta0vQ3B0z5JOnDSvjr8UMrRKPAz6XA8GohBQFxihyvqak1eoRk5KYUicvTmxxONQWxvMfHMfKnU3wuRzwux08y5EtGpc9mN7vS64RklDiVouMXXYbJteVph0rZjvtPNHFFduJzihKvQ6Np9sWivPnGYpLaQuTx2nnljNrWRVJJPnCV25QjOx3O9Ad1XpQ4tjYeKgdfrcDlQFZQTFbIppIchpXSlGF7pLPwZSGiKqAC0N8qoJy2Ahf0FesP6K5fkswhqqAG3YbQYXfjZagbAV7nelxWUYLiVby3zcc0WTATqgJ4LaLp/C/h5Z7MblONRIBYNuxTvhdDpx3Wq16XJnsXbJzM0XJvbZEkivGmJTi8jV2RTMoKO34ZXHcCUKh7Pmj5THDaM+A2wGfW3vfwajEY9cU1DDxxK3EoJgRcLA1jAum1fH7DrjtaUaw3mgVEUukFFpT/juRpAgntOsOowjZs18wtkI2QCWVigaUVkdJNYmoLRRDhd8FQgjKfU5NdmxbKGYYZ2YFwUyZsf+ZByUat5sOd+Bnr+zMeG99gUGtoERLMial0tIlx1cHNEkSbFFyO2zwOO1p9EqmLKYSgy4Bpw2VF0l9enudgQdVots2IhyXNMFPGwHnzzsi8qBiiQNEIYOY8owmUppJ63XZ4Xc5lAGsrTMSsWxKDaYNK+WekIgxlT7NFtdeDcXnhNtp41x+iceJKmWRBbQZf4C8VTWzDtl1RG/O47RrFNSoCq1VKyKRTCEcS6K2TH2mTrsNw8o8WDS+UnOseO8s8WXasFLceuFkLJlUrTlW9qCUvcJikkEdmBwDiCdVaz6aSKI9HEfA7UjLGAUAv5JZKFKrYvFmMCbB73KgMuCGlKLoiMknbg3FsUWgKONK7RcAXoPGcPvyKVg0vgqVwriPSyn+vAFt3KElGEOlQu9VBVxo7paTJHwuAwXFPChhDtz6D+2Ghj6XHV86Zxx++alZuOW8SZgxvAz1k2vwk0un4bEvnAFAHqcBoQgUkOdERzjO6Sg2HphSFGN9MSnF9zZq7Iwa0tl6pRVNyJlvrMckAIwrk6//L2XfML/bgYDLoTFYxYxRauBBAWo2ZSJJkUpRdEUTGFfl50Ycy+ITkS3jLSaleIcb+bwp6HeccerOt1gZv8x41SZJqEW7baE47xxRU+LRKMquiJS2DgHq2vbWnmZNMwC2Brh18/uhhn3oTwxqBQUAv10mL3A/fG47Xtyq3f5iYm2JZkAya8DjtMPrsiMSlzRKyage5c/vHeQxjm8sm8Dfr/S74LLbeMsSBpHi+95FU7Du9nPTLO7uqKSx6P1uBx9YnRG5IPYJpes2G+xsIOp3Bvbxe0lqApl6D21CTQA+t11jicnyutFw69KsSRIep50vvC67DUOVdjk2Arytazvzx9UHEI0za1ixmIX4j89lR4nbwe8rW8ZTPCkbHaLSdzlsIITgz/89X3MseyaUUvz4hY/gtBP8+2tn46alE7BgrFaZtYcFBZVIphk2fre8kMUSKU7FReJJdIQTht4TAGXzN20MZtMhbRqz323n8aBORUG9ur0RraE4V7jH2iN4fM0huBy2NO9k4Tj5mIqAYJglU5o9tcQx3BqM80W7KuBGayjGKT49mIJi8+G1DxvTjvG5HLDZCD45dwRuPm8i7DYCl8OGa84cgzPHqc+Y7dbKUOZ18vjdUMHYqCkVPSjFghe6iG863G5YbGrkVXmdWk9miEdLYQbcdthshBuCgOxBseclpVJZFVQ8mUIwLoFSaEIHXpddM19yISYlsa9JXTOkZAohSetB6cMNjF050h5Bw64mjYHJlL6UomgNxbnxUlPqxkkhlhmTkoZylipK64fPbcdf1hziNCwbIy57+nfy2Xuqtxj0Cko0CkQO/qJpdWkxKDbY3A45EyyS0LYn0i9UxzoiPDni0tnD8PVzJ/LPCCGoLnFrApGA6lkBUIr4PPC7tJbL4bYwkinKacLuqIRyhcrrCCdw8wq18SgLojLeWY5BqQPU67LD57IjlMODOmNMBfxuB4IxScNN+xVXXvQKAgKlVuJxwOOwa+qQGO3ocdo1NNQ1C0fjmU3H+DYfTM42TWpunF8T0MYRRXicsoIIx5KoExY19nvqvRh27zFJpmqXTanhiRClQnzN67SjVYgBhWNJ2UsrVe9j6eQaNfbAPSg5TiLSayLcikITFZQ+c8rvdmisfECNObKkm39vOYH9zSFct2iMZjEH1KLWCp/WgxKNEeZBheOyZ8iuF5dS2Hy4A83dMXid6ZZ0jWIEsESBG55Iz5A08rzE+2eJPPJWDuqxPreDx/euWzSGv88Mne6oxMeXGMPberQT31yR3oTXaBsZj9Om6a9X5iLY9MPz+d9szIm/8xs7TvLtN0KxJIKxZJpyYOMsLqmGQKlQvpJI0rz6CsaU2kOGRMqA4lPOd8PicbhxyTg+5m58YiOue2y9YWuwVIqiPaQWcteUuDUelNxvL11OkR3qjCRyelAAuPffHxj0Csqhy1yqKXFj4w/Ow8NK7Yfo1rLx6xG8DpFzj8ST2HOyGwt++gaOdUQ06apjq/xpiyJbOMV9dCYIQWP2I/OMGGXOsgF6sUCrsSarLIYwsSaAq84YmeZBRRJJTaaV12WXW//EJa7MRlZ4sUFnvZ8zsQp+l5zBJ2bvMdlEfrpKsND9LoeGxnM51Cp3t+LNMHzpnHEA1OJC7kEJC3V3VNIoqLoyD6eHRFT4XAjHJYTiEmoEJZip8PqqR9fg9YMJ/rzEgL44BmpK3egIqxOxK5pAKC7xDLBzJlbx2pbdjd38fJFEEs3dMcO4EHsWMUlWUOz5tAbjmmJyv8uRsaEoUxAseeSW8yZh2jBtrI0t+hU6iu9IW5h7op/5/Vo8/8Exng7NKL5ZI1W6kCma129ZzN9jXmq2riBGnhcDIYT/rn5BQbnsNgTcamG1qKCZZyAaMPq0a6N9nl7YcjztPb13UOYmmufEZBPfO6mhwBLoCMfh18WpuAclqcYHK+AF5DGup+SyIZZIYV9zEMPLvfA67bIHldB7UPKcun35abjt4tO41y7WUuo7z8tsg1oiUVvqQUswxr2dWELbEJZBnBuJZIpnHoq/nx5tUUtBmYY+GyngkXl+hkphsWW0jlsp9mQeFFukY1ISD721Dye7Ynjjo5Ma+o9Zs5+aOwLfu0gOFLMYQanHiTe+tQQv33yOZqKw12xylLjk67PCuIXjKnHv5TPwm6tm84HFJuTnFo5GXZmHc99MQW0+0q6pKpcTPuyaGNQn54zgn//iU7Ow5c4LeOxHD2adiXsBiR2Q7TaiuScWvxPv79PzRuLWCyfzDgx8UijyvLunBcMEbyAgLAIlHgePRYgo97nQFk4gRbVWnmgk6Gs6Xzig0kOiIhO/U+pxynSooqDawwmE40lUBtx49ZuL8btr5/Hv79EUmiZwsiuaMUuTUXxd0QTvGN7UHeXbJgDyOBAteP4M3A6+MLYqCRwepw1zR1dolDfr5OEQDKLWYAxdUUkTZ7t5xQf4xorNANQ4z/cumsINJqZoJtaWcCVR4VdSk6MSNus6LKhyZu/YzuaR3IhUNWJEj1xUciz+yCjgbJm0DMum1KDM5+Lzhcmv397G59RTfLIMIzPEPI93RvHvLccxe2Q5Pj5zKJ/j7LyXPbiax13LvE7MVwygacPKTO0bxoyN1fta8NwHxzGu2g+nnWDr0U7s69B+X28Ij6tKN2o8OqXIFDszPmpK3EhRNc0+JiXhctjw9A0L8fLN5/DviQrqQEsIW5WaKI+BB3WzwiC19qOCOuW22wjoFmGRkiGiB6UUe3ZHJZR5nQjGJMQSKd6tOplSi01HVfj4Vhg//9Qsfj5mlbaG4hrPiYFNUpayOcRN0BKhvM1LXakHM0eomYhep50XDpd5nZxyjCdTSCQpqkvcvCUMg0+h+LqiEvdcZgnnHOJzcu9MVAzM4mcKXFz09AuFxoOypyuon10xE0D6dgWxRAqJZApr9rfixiXjcfaEKrgdNr7PjiyTw5DTrvC7eAKKaNWKiodAW8DImsGy+zNCwO3QpI+3h+QiWr/Lron5uHXPYNXOZrQE44YKhl0vJslB5nMmVmN/SwgtwTgmK3HQuJSC3y2nd08dWoqPTnShQkk5L/c7OdffHpK9LmZ41QuKhykYtpkdIGdIAulUKYubMo/FYbfhtKEl2HS4Q5ME84fPn4Gn1x/mmZcbDrXh9me3AQDOHFeJKUNL8JjSuidgEGQXwc7rdzngFrLEynQUKwN7lszDrvC7OF35uYWj8Jc16VvFl3oc2NMkF/B+8eyxON4RwaHWMB+jr92yGFUBN7au12aFMuNs5nB5bgwv9/L+jGVeJy9zuHHJeB7rA9R434nOKF7e3siPP3N8Jdbcdi7qyjx4drO2N6UeM4aX4Z7LZ+Djv30Xzyu9QG9aOgE3PbkJaw+kb1qoZwmMPFeP064pwdC3RKtW6tNOdsVQU+pBPClTfAvGaeOxIv39vpDU4+W/n/r5F88Zi5e3n4Cd9F8z2UHvQemh58lFlx5KRhzbsgGQJ8cQv9o5gAUW73t1J6clHv/v+YaZfGMNLBsRzDUPKJavz0ngc9lxojMKh41oPD12/CFBQTGXPJqQu28vnazNRgMAr8uBUq8T3ZEE77DNPBlAa42Je0Kxljnl3nQFpd8rR/SgXA712emVgMOutZZjUhInu6JIUWB0hQ9nTajCvDEVGoXjc9kN4zplXidfrMQYnqg89XotkYShByXC73YgRdW6tLZQHJG4dr8sQGs5iouZUZamfLxdoQ5TGFvl55Sb12Xn1i5bJJ+6YSG+PNONWUpLmiE+tWdkWyiuGWsiQ8AW4aFlXmz8wXn8eEBt6quHyCCwuSAueBNqArjjY1Phc9nhsBG8sEVNNJo/tgJ3/dc0/rdRFpgIJqrfrW4XUeJxarzIUq8TV84bgYBCo4r3IM7Va88cg7lKXE5EwCOXPITiSfhddn4vTCFOqi3RzXnle8rvu3xGHe67YiZ+qvSr87nsuGCqnBLvtJO0a4r08tr9rco9yOdiFP9F04bicqUe0QjPfHURpgjGz0XT6rBwXKVmzonrllGW6KPXaIuM9ZQmU1BsbjLlzzt0JFKG8SSn3Yb1d5yHuaOHaOrHvIInxlDqceK1W5bgjLr+82tOCQUlDg59QoKhB+VQB3ZcSvFjuqIJ3iMumkjx5o+ZuPdsKdKAWuPDExFsqjw1Je40HrnM6+Tth0q9Ks/NFtPRlX58cs4IPPTZOXzAe5SkhQNK6uoFU2s1VJ44CUTlwWIhLE4kUnz6fmgifaKJQWXJCgLkjLRP/J+87cRQYQEV5SCEGD5fn8vOPashfide+PrZuGnpeI3i+c1VszVB90RKUFAG2UeAusiyGpG2UByhuJQWexAXCTFzT1/sySAq6+qAm1M6Pped/87svsu8Tiwc5uDtq8ZV+fn3W0PpcRAGUVmJCg1QKTo9RAXFFJ/RzrmEEJR6nZoUab1CyqmgFAOwxOPgc63U69A8v+qAG/ddMQvb776QL7IstiLKWupxprEhgDyX2ELqVpgQIH3BTv+end/nlfNG8vkTFuI2w8u9acpBjDmyLhL6dmZelx2/+vRseJw2XD5nOB767Bzce/kMzB5ZjvHVcuzaYbfx34z1DBTnnKigjKjOC6bVaf7Wl3ewZBuuWEpVDwpI35RQf49nT6jS3pPyPI2UZX/ilFBQ910xk1u2Pt2gri5xoyrgwsXT63iShOhBAeAW3t6mIJIpyr2Lo0qSRCYFNUSZeJnoJMb/M2rNaVetxNqydEu83KcuEGIGIkuFddoJfnnlLFw8Y6imbqXc6+RxqcvnjNAoaXGAiYprTJV2oRUVmd1G8Mg1c/G12fI13BkoPqP7FikDRnMB0MSgjGJheojHlHldmD68DLdeOEVzzKWzh+OW8yfxvymAm5XYi96D+sPn5+G0oaUYXy3/toweaVQ8PP1vLMYfxdRkIyoX0D6LyoBLk83JFK3eu2eL5PXnjFO7l4fiac9Hn80n3h+L3wQM4kNOO9EsSqW6okw9SnUKSK8gsjVeBgCPS13UGNVaXeLWGIlVJdouGICqZMX4HutgokdAM7aJqqAyzEFmwDh0Cy3zDL68ZDwfs0bxxUq/qqCiiRRsJP25MOz8ycX41ZWzcfGMobhq/ig8d9NZWPntev45M57YeNLUZAnZwJmUwj+/ciZ/rVfIQZ0HxZRodzSBOT95HduOdWbNNjx/aq3mb70CHCgUhxS9hMNu45SKX7cIeJx2bPjB+Xjoc3M1SRLiYsFSvFnM47MLRgEAjiq0ji+DdUYIwRNfnI9Xv7lY8z4bwIwqY0rT6yDcmjSiikTLTFRQbPCJA/fuS6Zh8aRqLBhbqbFQh5d7NQuc+B0WUAaAG84ZjxsXj8O3lAXeKXpQNoILp9VhXh1rd6L3oOS/jbavymRljxG6fueKZQDaYsdMtUeybNohrN/GgeHc02rx8s3ncMsSAGYLnUj0njdrgDuhJoAffnwqfz9zDEp9RpUBN6YqCupIe5grKL3i+c6Fk7H/p8sxfXgZX6ykFE1bAF+5eTHe/PYSzXvMymaLu5HXpV/EWIKAmHUqQv+7mPmdRPz3WWMAyPd/1vhK3LhkHP7nEzM041qkUpnhw2lKQRH7XHZDWl2UyWFT57GRNw8AP7pkGt/uXgQhBAfuWY7vXzyFj5VSb/r17DaiKQov9TrTErPyBaMGxcxZsXQkEz09d7Qae9R3A2EGKjO0mMKOJlL8+Wbbekg/x8Sx841zJ+LGJeMyfrcvccokSYiV3ZnAhoONEG7tAWrWGqvwZm1/jrWH4VLc80w4Z2J6XKjh1npN+jpLHvDaCYYowUsja43FgwBtDIoVK4pyjKsO4HGlWFVcAEZX+TQUgTgJRD65zOfEbctPU4/TeVAi0tPMM1vSXlf68193+7kZ66wYXr9lMVJU3jp+wdgKnBA2GCw3WDgYMlmFmSajKPviiVU8mUDv3bB0659eNgMjhniVXm7ejIuT+PyG+Jy8YPJIW0RNljG4b5vQ3YShVLcwlwmJLgyEELjsNrQpVKWYYXf5nOF4ZtOxtEXs84vGwOWw4er5owzvQR+HE5NtzODS2cMxe2Q5RlX4QAjBbRfL40tKGWe5sd+CeYF1ZaqnSgjhDIUIjfHlsPF53BOVwX5LFgv+5BzjONJfv7QQ5/3qLextCmbcrSAfMGWQKRfOTF2VniliYL+5wy734xPrPLOdVx+3E4/9lsBS9DdMKShCyEUAfgPADuD3lNJ7dZ9fB+DnAFh3xAcopb8voJw5wR6o3oMSsXB8JZ7ZdAw+l10zeasV7vtgSxiEgFu/LcF4Gu1hBpmKMb0ONSvIKDGADVyv0guOcfJfULYtcGWwfBkF+Km5I1DqcWr2wNFkvRGCry+bYJammgkAABNnSURBVOjliIrMpo9BpaWZy+c02Gon7flPrAlovBbxWmJD0YlKFtob31qC0ZU+TtUBmbcwYfdkhEwKSpx4HpcdfqXIWW/YsD2uakrcIITgpqUTkA3iQjCs3MvvcfmMOr7jcq5CV4ZMxct6uBw27l373Xbcf/XpeGnribR0cgan3YZrzxyT8XxMgV52+nD8/IqZ3CCqCrg03eCzYXSlP+29mhIPHvzsnLQFUk/x6alMpjhcdnXTQFHJO22Esxu9SXz+xOxhmDmiLOtzF+OHvQWb+0Y9PgEt25AJmVgd8Tf3OO1oF2oQs1G04nr4xbPH9tpLLBRyqmpCiB3A/wG4GMBUAFcTQqYaHPo0pXS28q9flZOITJYFIFvDr35zMSoDbs1iwRRKMCahwudCqdfBF1q9VdkTsMD68BJ1nxcjq5JRDIz60NcHZVp0Pz5zKJZOrsatF00GoF20HTpv6NsXTMYNi8ennUPzHXtmBeWy27kCM4rN6d/Tt+sBgKnDSnHlvBFp2x8AMqXmtNvwk0un8yax2TxYBj31lsla1NSpOez4xrkTQQgwrlq7KLAMzpoMlJ4ebGPD71wwCU67XMC85c4L8OtPn87PLQbF9RC9hxkj0pvhGt+Ljf/vsNtwyaxhePiauXwhMmoKmw1sTlT6XZpn/s53l+Hh87InBOXC8hlDce5p2jiHy26DXbHyXXabJkkCUAvGqaB+RE/cabdhrJLsM8VgnJmFw27LaRQwBVWeoZNIPmCGKItJeR3Aw5+bwz83owQzxcXF39zjtGk6cySNLEoF4vwXKe2BhpnVdz6AvZTS/QBACFkB4FIAH/WlYPmCPftsVqrHqda66JMk7DaCZIqi3CdzzMtnDMXfNx7NmFGVDz47fxQm15YgfGgrKifU4bdv7sWyKTVpx3HXX7mXKl3XgkyFkjWlHjz2hfmGn/UkCyfdg7JpXrOYyoW6zCJAjeXMH1OBtnBck2XH4HbYcd8Vs9LeF1EZcGPb3Reaknfv/7sYxzuiWPzzVcI1Migo0YNy2vHfZ4/FjUvSFfb/u2w6Hn17v2kDhSk0TXKH8nt+ZckE3PTXTVn7DgbcDiwcV4GNh9qxaHxVxuNElPtkzyYtmYErrvzGLjNmhujoHjlVvvAWNSEE1QE3Grui8LvVmBOjqFmCQsDtwO3LT8OMEWWaHYsddoIlk6qx/e4LszInhQBbB+pMGizZwKh8pqC+MM2Ni6YPxR3LT0vbeTsTMikoLdthx0GhmfVHx9O3tBFx6exhmo7wxQAzs284gCPC30cBLDA47pOEkMUAdgO4hVJ6RH8AIeQGADcAQG1tLRoaGvIWWI9gMIiGhgY4o7Iru27bboyKHcz5vZaI6sHs/vADOAhFEgBJRNDQ0IC6lGwRh0LhgsjJZMXezfjTRX507t+Chv3az481ytdMJBJoaGhI2656387tcDTtyOuaG9atwT5Pfkpqy+ZN6D5g58/2ZEh9VhvWvo9aO/C12W7UhfahQXcTLSeVanvaha/OcSN4cCsaDuZ1+R5BX+y7acN6HPWl3/f+DpWKO7B3Fxoi+9OOAeRBf/cZxPRvP9udwsKhdgyNHERDwyHNZ34Av7/Ah4Pb1+Og8h57tiK+MI7i2rFerFn9jqlr2hNynM6FhOZcx4/Iv0Ek2JXX2D16Ql78m48eQEODtvjUSN5CwEdkWe1UwpaNawEAVR6KhoYG7FN+q1JHEtXBfWjcCexoVX+/3Tt3oKFjj+F5Cy1vd7usOKLtJ3t93jWr3wYhBCFlzaKJKBoaGjARwEQf0NBgvJWNiNZGtd2Tjag1gevXvMe7aKQSUezvUOdFBcn+TC6rA4BONDSk72Mloq/GghHMKCgj00nvK/4bwFOU0hgh5MsA/gxgWdqXKH0UwKMAMG/ePFpfX5+ftAZoaGhAfX09JswKY81v38VXPr5As7dRJrQGY8BbbwAAlp51Ju5Z/xaQTGF0XRXq68/AzFAcv938Om66YBrqF47utZyirJng2d+KBz9Yg4l1ZaivP1t+89UX+eeLzzxDk76cFa/I3zvn7LPSYmK5vjP/jHmYPryMy9vYGQXeWQkAOH/ZEjjtNpyf4RTrojuBA/swfeI41NdPzHBUH+E19VktOXuRYSJK9fFOYM27AIDTZ05HvdIhpBD4hDmHD0DusWAGTx/diF3tjRhRPQT19WoK8kfYC+zbhcqKIaivX2j6fH85tB442YQFs9OfSyHkNcLfjm3E/m2NqCoL4NILzkGTdz+WzxiKEUN8OFNKYnN4C245fxKv2/PubwXWyxvnzZ45A/W69Oi+kndddCfeO74PQ0eMRH39abm/YIDvpPbgX1uOY+lSOSMzqsy3oeVe87Iq35k0fgxePbQXgOwpsRjosvrFas/Gre/gREj2mn5+xUxcMntYzlIBM+irsWAEM6b1UQDi9qEjAGi6NVJKWymlzDf9HYD04EIfY8QQHz648wJTygnQxpZKPA5OqzHNW+F3Yc//uxifW2Cc8dQXmD+mAg985nQ8cs08/t52geYySoPNhUJTfLnOx9Jls1GtfYUpFdp6LSOYVtaDACzzqjItA0t+9kYFudnAapc8/fjb1ZXKtCfbaO+GxeN5zNbtsOOBz8zRNNgVY2P6WGlf4nKlv+X5pxkrRDP42rKJeO2WJWnvVxt4+rkgUnli2YA4P8VjLphWVxDl1N8w82TWA5hICBlLCHEBuArAv8QDCCGiuXUJgPx4qAGAGKMIeBy49UI5weD/t3f3MVJVZxzHv88ub7ssCiiuIIgSBF/QwmIRtCGiRcQaIE3bQJVqorHR2lb7h4U0JfElqTZNa0ybatPSGNOCrcWWUA01Ck3TJlAQVCiiqFTxDapVQxsr6NM/7hm4Mzu7O7vMzD139vdJNnvn7s3ub2fPzjP33HPPSU+FX7jYXS9NTcaV540pWl4i3bfemxGFhZsBSwdJVKLzMPPKG3ZhpvUsbvRbNqOFR2+6kIVTx3R5obn9uCEsnDoGODp6LK8Khal0VGahbfe27X7UwzyGtVAYuVfpz0y/8ahkctlqmXhSG3vv/hznlyx7Xw0tfbi+VzTzRHjumpus6H+30O29bP6ZVRl9mIUeX/Hc/bCZ3QysJxlmvtLdd5rZHcAWd18LfMPMFgCHgXeBa2uYuSrS880NbG7iqgvGc9cfd3HJmX1/h1QL6ReZ0ptJu/PwV2eycfeBimZtKFVaoHrzglVYrLC3F+irZdqpI5h2auc53NJunjORP79woOxAlTw5N9ynVDqBcOHv1dtCc+eiKdz9+PN09PD8VVPhzdjhChfBS5819eXNV0zuWzItrGzQefmQnrQUnUE1hc/Fz8cHqVsl8qqiVy93fwx4rGTfitT2cmB5daPV3rcvP5OOsGhgy6Bmtn13bo/zjWVhUnsbL7x9sNMkrt0ZO6KVq/t47ay0QPXmnfiXPj2ONdte58KJlY1Ey8IZ7cPYvuKyrGMcs4snj+LK80Z3Gik5e9Iols4cz9JZvfv7T2ofxsprO6/NVUuFe4LSq8R2p6g7q45nerWw4FPJmfzGjX0oUGUmly3tfn85jOArd6tHXsT3alxHN15cPLy4dHhtLNbcdFHRDXe11pebkwtmTjih7NQyUn0Dm5v48Zc7Ou1vP24Idy6akkGi3itMvzV7UucZWcopug+qm/vKGl1ryZyE0LnLc8mMU1m1+dUjEw/kUb8uUHnRNnhAlxNU1kLpMiAitTJuZCt/W3ZJl8uYlMpqkERs0t23XZ1Bfe/z53LXoinRzArRFypQUiTn3fqSQ12tZVXOwKbyI9b6m/T1t8IgiYFlBlv0diRnbFSg5IjtK+b26jqXSL0VT3XU/9rqA0uns37nW8Vrg3VxBtUIVKDkiO7mGbv/6o5Os2yL1NuALu756S/mnXMy8845mc2ppeIHdHENqhGoQElFLp9SvRkXRPpqoK5BAcVd8V1dg2oEjfcbiUjD0ii+RLorvjBgohG75/vvX1hEcqe5zOCA/ig9HVmhaDfiCWX//QuLSK5lMd9jLJrLFSidQYmIxCHP9/ccKytzDaoRnw8VKBGRnEl38Q0K9z81q0CJiEjWmsvcsNyIY0Ya8FcSkUZXugZWf1NumHlzA1Yo3QclIrnyl9vm9PubxkuXCwJoyWANtlpTgRKRXBk3sjXrCJkrugYVxpf3Zr24vGi8kisi0uDSAyI++jhZ7LF1cOMNu1eBEhHJmfSAvQ8PJatY6wxKREQylx7F97/DSYFqVYESEZGspa9BfXjoEwCGqotPRESylh5mXthsxJGNjXdOKCLS4NLDzG+dO4lRxw1m0bRTMkxUGypQIiI5k+7iGzF0EMvnn5VhmtpRF5+ISM6kh5k34hx8BSpQIiI5Y6lX7gauTypQIiJ5U3QG1YDrQBWoQImI5EyTuvhERCRG6YnLm3QGJSIisWhq4LOmNBUoEZGcaeRuvTQVKBGRnOkn9UkFSkQkb6yfVCgVKBERiZIKlIiIRKmiAmVml5vZbjPbY2bLynx9sJk9HL6+ycxOq3ZQERHpX3osUGbWDPwEmA+cDSwxs7NLDrsO+Le7TwR+BNxT7aAiItK/VHIGNQPY4+4vu/tHwGpgYckxC4EHw/YjwKXWX67iiYhITVRSoE4BXks93hf2lT3G3Q8D7wMnVCOgiIiUN7l9WNYRasrcvfsDzL4IzHP368PjpcAMd/966pid4Zh94fFL4Zh3Sr7XDcANAO3t7dNXr159zL/AwYMHaWtrO+bvUw95ygrKW0t5ygrKW0t9zXrgv5/QNshoGVDfzqpqPLdz5szZ6u7n93igu3f7AcwC1qceLweWlxyzHpgVtgcA/yIUv64+pk+f7tWwYcOGqnyfeshTVnflraU8ZXVX3lrKU1b36uQFtngPtcfdK+ri+ztwhpmdbmaDgMXA2pJj1gLXhO0vAE+FECIiIn3S45Lv7n7YzG4mOUtqBla6+04zu4OkCq4FfgE8ZGZ7gHdJipiIiEif9XgNqmY/2OwA8M8qfKsTSboU8yBPWUF5aylPWUF5aylPWaE6ece7+6ieDsqsQFWLmW3xSi62RSBPWUF5aylPWUF5aylPWaG+eTXVkYiIREkFSkREotQIBepnWQfohTxlBeWtpTxlBeWtpTxlhTrmzf01KBERaUyNcAYlIiINSAVKRESilNsC1dMaVVkws5Vmtt/MdqT2jTSzJ8zsxfB5RNhvZnZfyP+smXXUOes4M9tgZrvMbKeZfTPyvEPMbLOZPRPy3h72nx7WIHsxrEk2KOzPfI0yM2s2s21mti4HWfea2XNmtt3MtoR9UbaFkGG4mT1iZs+HNjwr1rxmNjk8r4WPD8zslojz3hr+x3aY2arwv5dN261kPqTYPkhmtHgJmAAMAp4Bzo4g12ygA9iR2vd9YFnYXgbcE7avAB4HDJgJbKpz1tFAR9geBrxAst5XrHkNaAvbA4FNIcdvgMVh//3AjWH7JuD+sL0YeDiD9vAt4NfAuvA45qx7gRNL9kXZFkKGB4Hrw/YgYHjMeVO5m4G3gPEx5iVZmeIVoCXVZq/Nqu1m8keqwpPY4wS2GWY7jeICtRsYHbZHA7vD9gPAknLHZZT7D8DcPOQFWoGngQtI7mgfUNou6MMExlXOOBZ4ErgEWBdebKLMGn7uXjoXqCjbAnBceBG1kv1R5i3JeBnw11jzcnTppJGhLa4D5mXVdvPaxVfJGlWxaHf3NwHC55PC/mh+h3BaPo3krCTavKHLbDuwH3iC5Cz6PU/WICvNlPUaZfcCtwGfhMcnEG9WAAf+ZGZbLVkWB+JtCxOAA8AvQxfqz81saMR50xYDq8J2dHnd/XXgB8CrwJskbXErGbXdvBaocgug5G28fBS/g5m1Ab8DbnH3D7o7tMy+uuZ194/dfSrJ2ckM4KxuMmWW18yuBPa7+9b07m7yZP7cAhe5ewcwH/iamc3u5tis8w4g6Ur/qbtPA/5D0kXWlazzJiGS6zYLgN/2dGiZffVquyNIVkg/HRgDDCVpE13lqWnWvBaofcC41OOxwBsZZenJ22Y2GiB83h/2Z/47mNlAkuL0K3dfE3ZHm7fA3d8DNpL0zw83s8Ks/OlMR/KGrx9PMtN+PVwELDCzvcBqkm6+eyPNCoC7vxE+7wceJXkDEGtb2Afsc/dN4fEjJAUr1rwF84Gn3f3t8DjGvJ8FXnH3A+5+CFgDXEhGbTevBaqSNapikV4r6xqSaz2F/V8JI3ZmAu8XTvfrwcyMZJmUXe7+wxzkHWVmw8N2C8k/0i5gA8kaZOXyZrJGmbsvd/ex7n4aSdt8yt2vijErgJkNNbNhhW2S6yQ7iLQtuPtbwGtmNjnsuhT4R6x5U5ZwtHuvkCu2vK8CM82sNbxGFJ7bbNpuvS8SVvFi3hUkI89eAr6TdZ6QaRVJv+0hkncW15H0xz4JvBg+jwzHGvCTkP854Pw6Z/0Myan4s8D28HFFxHnPA7aFvDuAFWH/BGAzsIek62Rw2D8kPN4Tvj4hozZxMUdH8UWZNeR6JnzsLPw/xdoWQoapwJbQHn4PjIg8byvwDnB8al+UeYHbgefD/9lDwOCs2q6mOhIRkSjltYtPREQanAqUiIhESQVKRESipAIlIiJRUoESEZEoqUCJiEiUVKBERCRK/wdRiJLo6x9guwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f26c06752e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "reconstruct_training_sample = True\n",
    "chan = 13\n",
    "samp = 180\n",
    "\n",
    "from scipy.stats import pearsonr\n",
    "# now let's see how the model performs:\n",
    "r_accumulator = []\n",
    "\n",
    "if (reconstruct_training_sample==True):\n",
    "    reconst = decoded_train[samp,chan,:,:].reshape(leng,1)\n",
    "    orig = x_train[samp,chan,:,:].reshape(leng,1)\n",
    "    R,p = pearsonr(reconst, orig)\n",
    "    title = \"Reconstructed input based on TRAINING data\" + \"\\n\" + \"Corr: \" + str(R) + \"P-value \" + str(p)\n",
    "    for samp in range(y_train.shape[0]):\n",
    "        reconst = decoded_train[samp,chan,:,:].reshape(leng,1)\n",
    "        orig = x_train[samp,chan,:,:].reshape(leng,1)\n",
    "        R,p = pearsonr(reconst, orig)\n",
    "        r_accumulator.append(R)\n",
    "        print(\"Training sample\", samp, \": Correlation \", R, \" P-value: \", p)\n",
    "    print('average correlation = ', np.mean(r_accumulator))\n",
    "else:\n",
    "    reconst = decoded_test[samp,chan,:,:].reshape(leng,1)\n",
    "    orig = x_test[samp,chan,:,:].reshape(leng,1)\n",
    "    R,p = pearsonr(reconst, orig)\n",
    "    title = \"Reconstructed input based on TESTING data\" + \"\\n\" + \"Corr: \" + str(R) + \"P-value \" + str(p)\n",
    "    for samp in range(y_test.shape[0]):\n",
    "        reconst = decoded_test[samp,chan,:,:].reshape(leng,1)\n",
    "        orig = x_test[samp,chan,:,:].reshape(leng,1)\n",
    "        R,p = pearsonr(reconst, orig)\n",
    "        r_accumulator.append(R)\n",
    "        print(\"Test sample\", samp, \": Correlation \", R, \" P-value: \", p)\n",
    "    print('average correlation = ', np.mean(r_accumulator))\n",
    "\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.set_title(title)\n",
    "ax.grid()\n",
    "plt.subplot(212)\n",
    "plt.plot(orig)\n",
    "ax = plt.gca()\n",
    "plt.tight_layout()\n",
    "ax.set_title(\"Original\")\n",
    "ax.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test sample 0 : Correlation  [ 0.47509801]  P-value:  [ 0.]\n",
      "Test sample 1 : Correlation  [ 0.51149452]  P-value:  [ 0.]\n",
      "Test sample 2 : Correlation  [ 0.49352029]  P-value:  [ 0.]\n",
      "Test sample 3 : Correlation  [ 0.6073795]  P-value:  [ 0.]\n",
      "Test sample 4 : Correlation  [ 0.59454221]  P-value:  [ 0.]\n",
      "Test sample 5 : Correlation  [ 0.59018493]  P-value:  [ 0.]\n",
      "Test sample 6 : Correlation  [ 0.55333585]  P-value:  [ 0.]\n",
      "Test sample 7 : Correlation  [ 0.5507409]  P-value:  [ 0.]\n",
      "Test sample 8 : Correlation  [ 0.55475879]  P-value:  [ 0.]\n",
      "Test sample 9 : Correlation  [ 0.58399832]  P-value:  [ 0.]\n",
      "Test sample 10 : Correlation  [ 0.57415867]  P-value:  [ 0.]\n",
      "Test sample 11 : Correlation  [ 0.58130622]  P-value:  [ 0.]\n",
      "Test sample 12 : Correlation  [ 0.56964087]  P-value:  [ 0.]\n",
      "Test sample 13 : Correlation  [ 0.54215288]  P-value:  [ 0.]\n",
      "Test sample 14 : Correlation  [ 0.53503907]  P-value:  [ 0.]\n",
      "Test sample 15 : Correlation  [ 0.49753621]  P-value:  [ 0.]\n",
      "Test sample 16 : Correlation  [ 0.4973498]  P-value:  [ 0.]\n",
      "Test sample 17 : Correlation  [ 0.52664822]  P-value:  [ 0.]\n",
      "Test sample 18 : Correlation  [ 0.58851081]  P-value:  [ 0.]\n",
      "Test sample 19 : Correlation  [ 0.51916337]  P-value:  [ 0.]\n",
      "Test sample 20 : Correlation  [ 0.53890759]  P-value:  [ 0.]\n",
      "Test sample 21 : Correlation  [ 0.57323885]  P-value:  [ 0.]\n",
      "Test sample 22 : Correlation  [ 0.58248711]  P-value:  [ 0.]\n",
      "Test sample 23 : Correlation  [ 0.54993343]  P-value:  [ 0.]\n",
      "Test sample 24 : Correlation  [ 0.57451522]  P-value:  [ 0.]\n",
      "Test sample 25 : Correlation  [ 0.54293275]  P-value:  [ 0.]\n",
      "Test sample 26 : Correlation  [ 0.53558528]  P-value:  [ 0.]\n",
      "Test sample 27 : Correlation  [ 0.53789842]  P-value:  [ 0.]\n",
      "Test sample 28 : Correlation  [ 0.53352827]  P-value:  [ 0.]\n",
      "Test sample 29 : Correlation  [ 0.52134794]  P-value:  [ 0.]\n",
      "Test sample 30 : Correlation  [ 0.54645187]  P-value:  [ 0.]\n",
      "Test sample 31 : Correlation  [ 0.56984901]  P-value:  [ 0.]\n",
      "Test sample 32 : Correlation  [ 0.59516382]  P-value:  [ 0.]\n",
      "Test sample 33 : Correlation  [ 0.60342544]  P-value:  [ 0.]\n",
      "Test sample 34 : Correlation  [ 0.61570615]  P-value:  [ 0.]\n",
      "Test sample 35 : Correlation  [ 0.58069795]  P-value:  [ 0.]\n",
      "Test sample 36 : Correlation  [ 0.46330675]  P-value:  [  8.12753109e-44]\n",
      "Test sample 37 : Correlation  [ 0.43580019]  P-value:  [  2.04871475e-38]\n",
      "Test sample 38 : Correlation  [ 0.43198213]  P-value:  [  1.05327747e-37]\n",
      "Test sample 39 : Correlation  [ 0.41893601]  P-value:  [  2.42474466e-35]\n",
      "Test sample 40 : Correlation  [ 0.43251744]  P-value:  [  8.38293511e-38]\n",
      "Test sample 41 : Correlation  [ 0.50586134]  P-value:  [ 0.]\n",
      "Test sample 42 : Correlation  [ 0.57291168]  P-value:  [ 0.]\n",
      "Test sample 43 : Correlation  [ 0.57107955]  P-value:  [ 0.]\n",
      "Test sample 44 : Correlation  [ 0.59281212]  P-value:  [ 0.]\n",
      "Test sample 45 : Correlation  [ 0.54976439]  P-value:  [ 0.]\n",
      "Test sample 46 : Correlation  [ 0.58347356]  P-value:  [ 0.]\n",
      "Test sample 47 : Correlation  [ 0.64239514]  P-value:  [ 0.]\n",
      "Test sample 48 : Correlation  [ 0.60914195]  P-value:  [ 0.]\n",
      "Test sample 49 : Correlation  [ 0.5774318]  P-value:  [ 0.]\n",
      "Test sample 50 : Correlation  [ 0.52893788]  P-value:  [ 0.]\n",
      "Test sample 51 : Correlation  [ 0.49263346]  P-value:  [ 0.]\n",
      "Test sample 52 : Correlation  [ 0.50055963]  P-value:  [ 0.]\n",
      "Test sample 53 : Correlation  [ 0.50195843]  P-value:  [ 0.]\n",
      "Test sample 54 : Correlation  [ 0.48660323]  P-value:  [ 0.]\n",
      "Test sample 55 : Correlation  [ 0.50188589]  P-value:  [ 0.]\n",
      "Test sample 56 : Correlation  [ 0.53112209]  P-value:  [ 0.]\n",
      "Test sample 57 : Correlation  [ 0.56193274]  P-value:  [ 0.]\n",
      "Test sample 58 : Correlation  [ 0.73322201]  P-value:  [ 0.]\n",
      "Test sample 59 : Correlation  [ 0.75946045]  P-value:  [ 0.]\n",
      "Test sample 60 : Correlation  [ 0.76898402]  P-value:  [ 0.]\n",
      "Test sample 61 : Correlation  [ 0.77808362]  P-value:  [ 0.]\n",
      "Test sample 62 : Correlation  [ 0.74547553]  P-value:  [ 0.]\n",
      "Test sample 63 : Correlation  [ 0.6540069]  P-value:  [ 0.]\n",
      "Test sample 64 : Correlation  [ 0.66231847]  P-value:  [ 0.]\n",
      "Test sample 65 : Correlation  [ 0.66483212]  P-value:  [ 0.]\n",
      "Test sample 66 : Correlation  [ 0.66246521]  P-value:  [ 0.]\n",
      "Test sample 67 : Correlation  [ 0.67726582]  P-value:  [ 0.]\n",
      "Test sample 68 : Correlation  [ 0.59428388]  P-value:  [ 0.]\n",
      "average correlation =  0.567271\n",
      "Pearson's R: [ 0.59428388] p-value [ 0.]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f291d6f7e48>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "reconstruct_training_sample = False\n",
    "chan = 13\n",
    "samp = 36\n",
    "\n",
    "from scipy.stats import pearsonr\n",
    "# now let's see how the model performs:\n",
    "r_accumulator = []\n",
    "\n",
    "if (reconstruct_training_sample==True):\n",
    "    reconst = decoded_train[samp,chan,:,:].reshape(leng,1)\n",
    "    orig = x_train[samp,chan,:,:].reshape(leng,1)\n",
    "    R,p = pearsonr(reconst, orig)\n",
    "    title = \"Reconstructed input based on TRAINING data\" + \"\\n\" + \"Corr: \" + str(R) + \"P-value \" + str(p)\n",
    "    for samp in range(y_train.shape[0]):\n",
    "        reconst = decoded_train[samp,chan,:,:].reshape(leng,1)\n",
    "        orig = x_train[samp,chan,:,:].reshape(leng,1)\n",
    "        R,p = pearsonr(reconst, orig)\n",
    "        r_accumulator.append(R)\n",
    "        print(\"Training sample\", samp, \": Correlation \", R, \" P-value: \", p)\n",
    "    print('average correlation = ', np.mean(r_accumulator))\n",
    "else:\n",
    "    reconst = decoded_test[samp,chan,:,:].reshape(leng,1)\n",
    "    orig = x_test[samp,chan,:,:].reshape(leng,1)\n",
    "    R,p = pearsonr(reconst, orig)\n",
    "    title = \"Reconstructed input based on TESTING data\" + \"\\n\" + \"Corr: \" + str(R) + \"P-value \" + str(p)\n",
    "    for samp in range(y_test.shape[0]):\n",
    "        reconst = decoded_test[samp,chan,:,:].reshape(leng,1)\n",
    "        orig = x_test[samp,chan,:,:].reshape(leng,1)\n",
    "        R,p = pearsonr(reconst, orig)\n",
    "        r_accumulator.append(R)\n",
    "        print(\"Test sample\", samp, \": Correlation \", R, \" P-value: \", p)\n",
    "    print('average correlation = ', np.mean(r_accumulator))\n",
    "\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.set_title(title)\n",
    "ax.grid()\n",
    "plt.subplot(212)\n",
    "plt.plot(orig)\n",
    "ax = plt.gca()\n",
    "plt.tight_layout()\n",
    "ax.set_title(\"Original\")\n",
    "ax.grid()"
   ]
  },
  {
   "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
}