NO-Classifier_NN-v2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# IC50 Classification Using Neural Network Models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#https://stackoverflow.com/questions/32419510/how-to-get-reproducible-results-in-keras\n",
    "import os\n",
    "import numpy as np\n",
    "import random as python_random\n",
    "#from mpl_toolkits.mplot3d import Axes3D\n",
    "#from sklearn.datasets import make_blobs\n",
    "from mlxtend.plotting import plot_decision_regions\n",
    "SEED=1234\n",
    "np.random.seed(SEED)\n",
    "python_random.seed(SEED)\n",
    "os.environ['PYTHONHASHSEED']=str(SEED)\n",
    "\n",
    "import tensorflow as tf\n",
    "import time\n",
    "from keras.preprocessing.text import Tokenizer\n",
    "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import train_test_split, StratifiedKFold #Split module from sklearn\n",
    "import pandas as pd\n",
    "from pandas import Series\n",
    "import csv\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn import metrics\n",
    "from sklearn import model_selection\n",
    "from math import log\n",
    "from sklearn.datasets import make_classification\n",
    "from sklearn.metrics import roc_curve, auc, roc_auc_score\n",
    "from matplotlib import pyplot\n",
    "from sklearn.metrics import make_scorer\n",
    "from keras import preprocessing\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Flatten, Embedding, SimpleRNN, LSTM\n",
    "from tensorflow.keras.optimizers import RMSprop\n",
    "from tensorflow.keras import layers\n",
    "from PIL import Image\n",
    "from tensorflow.keras.layers import Conv1D, GlobalMaxPooling1D, Flatten, Dense\n",
    "from sklearn.manifold import TSNE\n",
    "import keras\n",
    "from sklearn.metrics import matthews_corrcoef\n",
    "##https://github.com/mediumssharma23/avisualintroductiontonn/blob/master/visual_nn.ipynb##"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def readdata(filename):\n",
    "    text = []\n",
    "    y = []\n",
    "    with open(filename, encoding='latin1') as csvfile:\n",
    "        csvreader = csv.reader(csvfile)\n",
    "        for row in csvreader:\n",
    "            if row and len(row[2]) > 10: \n",
    "                text.append(row[2]) \n",
    "                y.append(row[7])\n",
    "    return [text, y]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def preprocessing(text, y):\n",
    "    tokenizer = Tokenizer(num_words=1024) # The number of Max Words in Document\n",
    "    tokenizer.fit_on_texts(text) #Word Index\n",
    "\n",
    "    X = tokenizer.texts_to_sequences(text) \n",
    "    print('lengths of 10 documents:', [len(doc) for doc in X[:10]])\n",
    "    print('num of tokens:', len(tokenizer.word_index))\n",
    "\n",
    "    maxlen = 212 #The number of Max Words in Document = 212\n",
    "    X = pad_sequences(X, maxlen=maxlen, truncating='pre') \n",
    "    # Labeling\n",
    "    label_dict = {'Active':0, 'Inactive':1}\n",
    "    y2 = [label_dict[c] for c in y]\n",
    "    X_train, X_test, y_train, y_test = train_test_split(X, y2, test_size=0.2, random_state=10)\n",
    "    \n",
    "    print('Train set count:', len(X_train))\n",
    "    print('Test set count:', len(X_test))\n",
    "    print('Test samples:', y_test[:20])\n",
    "    print(X_train.shape, X_test.shape, X.shape)\n",
    "    return [X_train, X_test, y_train, y_test, X, y2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def preprocessing_3rd(text, y):\n",
    "    max_words = 212 #Using Words\n",
    "    \n",
    "    tokenizer = Tokenizer(num_words=1024) # The number of Max Words in Document\n",
    "    tokenizer.fit_on_texts(text) #Word Index\n",
    "\n",
    "    X = tokenizer.texts_to_sequences(text) \n",
    "    print('lengths of 10 documents:', [len(doc) for doc in X[:10]])\n",
    "    print('num of tokens:', len(tokenizer.word_index))\n",
    "\n",
    "    maxlen = 212 #The number of Max Words in Document\n",
    "    X = pad_sequences(X, maxlen=maxlen, truncating='pre') \n",
    "    # Labeling\n",
    "    label_dict = {'Active':0, 'Inactive':1}\n",
    "    y2 = [label_dict[c] for c in y]\n",
    "\n",
    "    print('Train set count:', len(X))\n",
    "    print('Test samples:', y2[:20])\n",
    "    return [X, y2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lengths of 10 documents: [66, 66, 68, 66, 67, 70, 66, 69, 65, 66]\n",
      "num of tokens: 1024\n",
      "Train set count: 48220\n",
      "Test set count: 12055\n",
      "Test samples: [1, 0, 1, 0, 1, 1, 1, 0, 1, 0, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1]\n",
      "(48220, 212) (12055, 212) (60275, 212)\n",
      "[[  0   0   0 ...  54  23 562]\n",
      " [  0   0   0 ... 306 465 282]\n",
      " [  0   0   0 ... 819 309 950]\n",
      " ...\n",
      " [  0   0   0 ... 599 217 183]\n",
      " [  0   0   0 ... 166  26 104]\n",
      " [  0   0   0 ...  23 465  93]]\n"
     ]
    }
   ],
   "source": [
    "Classifier_input = preprocessing(readdata('dataset.csv')[0],readdata('dataset.csv')[1])\n",
    "X_train=Classifier_input[0]\n",
    "X_train = np.array(X_train)\n",
    "X_train= np.asarray(X_train)\n",
    "print(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "np.random.seed(42)\n",
    "tf.random.set_seed(42)\n",
    "def Loc_MLP(dataset,dataset_3rd):\n",
    "    maxlen = 212\n",
    "    X_train=dataset[0]\n",
    "    X_train = np.array(X_train)\n",
    "    X_test=dataset[1]\n",
    "    X_test= np.array(X_test)\n",
    "    y_train=dataset[2]\n",
    "    y_test=dataset[3]\n",
    "    X1 = dataset[4]\n",
    "    y1= dataset[5]\n",
    "    X = np.array(X1)\n",
    "    y = np.array(y1)\n",
    "    X_3rd=dataset_3rd[0]\n",
    "    y_3rd=dataset_3rd[1]\n",
    "    y_3rd = np.array(y_3rd)\n",
    "\n",
    "    model = keras.models.load_model('Embedding.h5')\n",
    "    model.compile(optimizer='rmsprop', loss='binary_crossentropy', metrics=['acc'])\n",
    "    model.summary()\n",
    "    history = model.fit(np.asarray(X_train), np.asarray(y_train), \n",
    "                    epochs=15,\n",
    "                    batch_size=256,\n",
    "                    validation_split=0.2)\n",
    "    history_dict = history.history\n",
    "    loss = history_dict['loss']\n",
    "    val_loss = history_dict['val_loss']\n",
    "        \n",
    "    y_pred_keras = model.predict(X_test).ravel()\n",
    "\n",
    "    y_pred_probs = model.predict(X_test)\n",
    "    y_pred_probs_train = model.predict(X_train)\n",
    "    y_pred_probs_3rd = model.predict(X_3rd)\n",
    "    \n",
    "    # Convert probabilities to class labels based on a threshold\n",
    "    threshold = 0.5  # Adjust the threshold as needed\n",
    "    y_pred_classes = (y_pred_probs > threshold).astype(int)\n",
    "    y_pred_classes_train = (y_pred_probs_train > threshold).astype(int)\n",
    "    y_pred_classes_3rd = (y_pred_probs_3rd > threshold).astype(int)\n",
    "\n",
    "    fpr_keras, tpr_keras, thresholds_keras = roc_curve(y_test, y_pred_keras)\n",
    "    \n",
    "    auc_keras = auc(fpr_keras, tpr_keras)\n",
    "    \n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    epochs = range(1, len(loss) + 1)\n",
    "\n",
    "    plt.plot(epochs, loss, 'bo', label='Training loss')\n",
    "    plt.plot(epochs, val_loss, 'r', label='Validation loss')\n",
    "    plt.title('Training and validation loss')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Loss')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "   #pyplot.show()\n",
    "    pyplot.savefig(\"MLP_LOSS.png\",dpi=300) \n",
    "    \n",
    "    #clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    history_dict = history.history\n",
    "    acc = history_dict['acc']\n",
    "    val_acc = history_dict['val_acc']\n",
    "\n",
    "    epochs = range(1, len(acc) + 1)\n",
    "\n",
    "    plt.plot(epochs, acc, 'bo', label='Training acc')\n",
    "    plt.plot(epochs, val_acc, 'r', label='Validation acc')\n",
    "    plt.title('Training and validation acc')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Accuracy')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "    #pyplot.show()\n",
    "    pyplot.savefig(\"MLP_ACC.png\",dpi=300)\n",
    "    score = model.evaluate(np.asarray(X_test),np.asarray(y_test))\n",
    "    score2 = model.evaluate(np.asarray(classifier_3rd[0]),np.asarray(classifier_3rd[1]))\n",
    "    print(score)\n",
    "    print(score2)\n",
    "    \n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    # generate a no skill prediction (majority class)\n",
    "    ns_probs = [0 for _ in range(len(y_test))]\n",
    "    ns_probs_2 = [0 for _ in range(len(y_train))]\n",
    "    ns_probs_3 = [0 for _ in range(len(y_3rd))]\n",
    "    # predict probabilities\n",
    "    lr_probs = model.predict(X_test)\n",
    "    lr_probs_2 = model.predict(X_train)\n",
    "    lr_probs_3 = model.predict(X_3rd)\n",
    "\n",
    "\n",
    "    # calculate roc curves\n",
    "    ns_fpr, ns_tpr, _ = roc_curve(y_test, ns_probs)\n",
    "    lr_fpr, lr_tpr, _ = roc_curve(y_test, lr_probs)\n",
    "  \n",
    "    ns_fpr_2, ns_tpr_2, _ = roc_curve(y_train, ns_probs_2)\n",
    "    lr_fpr_2, lr_tpr_2, _ = roc_curve(y_train, lr_probs_2)\n",
    "    \n",
    "    ns_fpr_3, ns_tpr_3, _ = roc_curve(y_3rd, ns_probs_3)\n",
    "    lr_fpr_3, lr_tpr_3, _ = roc_curve(y_3rd, lr_probs_3)\n",
    "    # calculate scores\n",
    "    ns_auc = roc_auc_score(y_test, ns_probs)\n",
    "    lr_auc = roc_auc_score(y_test, lr_probs)\n",
    "    \n",
    "    ns_auc_2 = roc_auc_score(y_train, ns_probs_2)\n",
    "    lr_auc_2 = roc_auc_score(y_train, lr_probs_2)\n",
    "    \n",
    "    ns_auc_3 = roc_auc_score(y_3rd, ns_probs_3)\n",
    "    lr_auc_3 = roc_auc_score(y_3rd, lr_probs_3)\n",
    "    \n",
    "    # plot the roc curve for the model\n",
    "    pyplot.plot(ns_fpr, ns_tpr, linestyle='--')\n",
    "    pyplot.plot(lr_fpr, lr_tpr, marker='.', label='MLP_Test, AUC=%.3f' % (lr_auc))\n",
    "    pyplot.plot(lr_fpr_2, lr_tpr_2, color='green', marker='.', label='MLP_Training, AUC=%.3f' % (lr_auc_2))\n",
    "    pyplot.plot(lr_fpr_3, lr_tpr_3, color='red', marker='.', label='MLP_3rd, AUC=%.3f' % (lr_auc_3))\n",
    "    \n",
    "    # axis labels\n",
    "    pyplot.xlabel('False Positive Rate')\n",
    "    pyplot.ylabel('True Positive Rate')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "    # show the plot\n",
    "    pyplot.savefig(\"MLP.png\",dpi=300)  \n",
    "    # summarize scores\n",
    "    print('MLP_Test: ROC AUC=%.3f' % (lr_auc))\n",
    "    print('MLP_Training: ROC AUC=%.3f' % (lr_auc_2))\n",
    "    print('MLP_3rd: ROC AUC=%.3f' % (lr_auc_3))\n",
    "\n",
    "    # confusion matrix\n",
    "    matrix = confusion_matrix(y_test,y_pred_classes)\n",
    "    matrix2= confusion_matrix(y_train, y_pred_classes_train)\n",
    "    matrix3= confusion_matrix(y_3rd, y_pred_classes_3rd)\n",
    "    print(matrix)\n",
    "    print(matrix2)\n",
    "    print(matrix3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Loc_MLP_simp(dataset,dataset_3rd):\n",
    "    maxlen = 212\n",
    "    X_train=dataset[0]\n",
    "    X_test=dataset[1]\n",
    "    y_train=dataset[2]\n",
    "    y_test=dataset[3]\n",
    "    X1 = dataset[4]\n",
    "    y1 = dataset[5]\n",
    "    X = np.array(X1)\n",
    "    y = np.array(y1)\n",
    "    X_3rd=dataset_3rd[0]\n",
    "    y_3rd=dataset_3rd[1]\n",
    "\n",
    "    model = keras.models.load_model('Embedding.h5')\n",
    "    model.compile(optimizer='rmsprop', loss='binary_crossentropy', metrics=['acc'])\n",
    "    model.summary()\n",
    "\n",
    "    history = model.fit(np.asarray(X_train), np.asarray(y_train), epochs=15, batch_size=256, validation_split=0.2)\n",
    "    history_dict = history.history\n",
    "    loss = history_dict['loss']\n",
    "    val_loss = history_dict['val_loss']    \n",
    "    epochs = range(1, len(loss) + 1)\n",
    "\n",
    "    plt.plot(epochs, loss, 'bo', label='Training loss')\n",
    "    plt.plot(epochs, val_loss, 'r', label='Validation loss')\n",
    "    plt.title('Training and validation loss')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Loss')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "    pyplot.savefig(\"MLP_LOSS.png\",dpi=300) \n",
    "    \n",
    "    #clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    history_dict = history.history\n",
    "    acc = history_dict['acc']\n",
    "    val_acc = history_dict['val_acc']\n",
    "\n",
    "    epochs = range(1, len(acc) + 1)\n",
    "    \n",
    "    plt.plot(epochs, acc, 'bo', label='Training acc')\n",
    "    plt.plot(epochs, val_acc, 'r', label='Validation acc')\n",
    "    plt.title('Training and validation acc')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Accuracy')\n",
    "# show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "    pyplot.savefig(\"MLP_ACC.png\",dpi=300)\n",
    "    score = model.evaluate(np.asarray(X_test),np.asarray(y_test))\n",
    "    score2 = model.evaluate(np.asarray(classifier_3rd[0]),np.asarray(classifier_3rd[1]))\n",
    "    print(score)\n",
    "    print(score2)\n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    # Load your Keras model\n",
    "    model = keras.models.load_model('Embedding.h5')\n",
    "    \n",
    "    activities = model.predict(X_3rd)\n",
    "\n",
    "    # Determine the classification of the compounds based on their predicted activity\n",
    "    classifications = np.where(activities > 0.5, \"Active\", \"Inactive\")\n",
    "\n",
    "  \n",
    "    if np.count_nonzero(classifications == \"Active\") > np.count_nonzero(classifications == \"Inactive\"):\n",
    "        print(\"The compound is classified as Active.\")\n",
    "    else:\n",
    "        print(\"The compound is classified as Inactive.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(42)\n",
    "tf.random.set_seed(42)\n",
    "\n",
    "def Loc_RNN(dataset,dataset_3rd):\n",
    "    X_train=dataset[0]\n",
    "    X_test=dataset[1]\n",
    "    y_train=dataset[2]\n",
    "    y_test=dataset[3]\n",
    "    X1 = dataset[4]\n",
    "    y1= dataset[5]\n",
    "    X = np.array(X1)\n",
    "    y = np.array(y1)\n",
    "    X_3rd=dataset_3rd[0]\n",
    "    y_3rd=dataset_3rd[1]\n",
    "    y_3rd = np.array(y_3rd)\n",
    "\n",
    "    model = keras.models.load_model('RNN.h5')\n",
    "    model.summary()\n",
    "    rmsprop = RMSprop(learning_rate=0.001) #Set the Learning Rate\n",
    "\n",
    "    model.compile(optimizer='rmsprop', loss='binary_crossentropy', metrics=['acc'])\n",
    "    model.summary()\n",
    "\n",
    "    history = model.fit(np.asarray(X_train), np.asarray(y_train), \n",
    "                    epochs=15,\n",
    "                    batch_size=256,\n",
    "                    validation_split=0.2)\n",
    "    history_dict = history.history\n",
    "    loss = history_dict['loss']\n",
    "    val_loss = history_dict['val_loss']\n",
    "        \n",
    "    y_pred_keras = model.predict(X_test).ravel()\n",
    "    \n",
    "    y_pred_classes = model.predict(X_test)\n",
    "    y_pred_classes= y_pred_classes.flatten()\n",
    "    y_pred_classes_train = model.predict(X_train)\n",
    "    y_pred_classes_train= y_pred_classes_train.flatten()\n",
    "    y_pred_classes_3rd = model.predict(X_3rd)\n",
    "    y_pred_classes_3rd= y_pred_classes_3rd.flatten()\n",
    "\n",
    "    # extract the predicted class labels\n",
    "    y_pred_classes = np.where(y_pred_classes > 0.5, 1, 0)\n",
    "    y_pred_classes_train = np.where(y_pred_classes_train > 0.5, 1, 0)\n",
    "    y_pred_classes_3rd = np.where(y_pred_classes_3rd > 0.5, 1, 0)\n",
    "\n",
    "    fpr_keras, tpr_keras, thresholds_keras = roc_curve(y_test, y_pred_keras)\n",
    "    \n",
    "    auc_keras = auc(fpr_keras, tpr_keras)\n",
    "    \n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    epochs = range(1, len(loss) + 1)\n",
    "\n",
    "    plt.plot(epochs, loss, 'bo', label='Training loss')\n",
    "    plt.plot(epochs, val_loss, 'r', label='Validation loss')\n",
    "    plt.title('Training and validation loss')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Loss')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "   #pyplot.show()\n",
    "    pyplot.savefig(\"RNN_LOSS.png\",dpi=300) \n",
    "    \n",
    "    #clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    history_dict = history.history\n",
    "    acc = history_dict['acc']\n",
    "    val_acc = history_dict['val_acc']\n",
    "\n",
    "    epochs = range(1, len(acc) + 1)\n",
    "\n",
    "    plt.plot(epochs, acc, 'bo', label='Training acc')\n",
    "    plt.plot(epochs, val_acc, 'r', label='Validation acc')\n",
    "    plt.title('Training and validation acc')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Accuracy')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "    #pyplot.show()\n",
    "    pyplot.savefig(\"RNN_ACC.png\",dpi=300)\n",
    "    score = model.evaluate(np.asarray(X_test),np.asarray(y_test))\n",
    "    score2 = model.evaluate(np.asarray(classifier_3rd[0]),np.asarray(classifier_3rd[1]))\n",
    "    print(score)\n",
    "    print(score2)\n",
    "    \n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    # generate a no skill prediction (majority class)\n",
    "    ns_probs = [0 for _ in range(len(y_test))]\n",
    "    ns_probs_2 = [0 for _ in range(len(y_train))]\n",
    "    ns_probs_3 = [0 for _ in range(len(y_3rd))]\n",
    "    # predict probabilities\n",
    "    lr_probs = model.predict(X_test)\n",
    "    lr_probs_2 = model.predict(X_train)\n",
    "    lr_probs_3 = model.predict(X_3rd)\n",
    "\n",
    "\n",
    "    # calculate roc curves\n",
    "    ns_fpr, ns_tpr, _ = roc_curve(y_test, ns_probs)\n",
    "    lr_fpr, lr_tpr, _ = roc_curve(y_test, lr_probs)\n",
    "  \n",
    "    ns_fpr_2, ns_tpr_2, _ = roc_curve(y_train, ns_probs_2)\n",
    "    lr_fpr_2, lr_tpr_2, _ = roc_curve(y_train, lr_probs_2)\n",
    "    \n",
    "    ns_fpr_3, ns_tpr_3, _ = roc_curve(y_3rd, ns_probs_3)\n",
    "    lr_fpr_3, lr_tpr_3, _ = roc_curve(y_3rd, lr_probs_3)\n",
    "    # calculate scores\n",
    "    ns_auc = roc_auc_score(y_test, ns_probs)\n",
    "    lr_auc = roc_auc_score(y_test, lr_probs)\n",
    "    \n",
    "    ns_auc_2 = roc_auc_score(y_train, ns_probs_2)\n",
    "    lr_auc_2 = roc_auc_score(y_train, lr_probs_2)\n",
    "    \n",
    "    ns_auc_3 = roc_auc_score(y_3rd, ns_probs_3)\n",
    "    lr_auc_3 = roc_auc_score(y_3rd, lr_probs_3)\n",
    "    \n",
    "    # plot the roc curve for the model\n",
    "    pyplot.plot(ns_fpr, ns_tpr, linestyle='--')\n",
    "    pyplot.plot(lr_fpr, lr_tpr, marker='.', label='RNN_Test, AUC=%.3f' % (lr_auc))\n",
    "    pyplot.plot(lr_fpr_2, lr_tpr_2, color='green', marker='.', label='RNN_Training, AUC=%.3f' % (lr_auc_2))\n",
    "    pyplot.plot(lr_fpr_3, lr_tpr_3, color='red', marker='.', label='RNN_3rd, AUC=%.3f' % (lr_auc_3))\n",
    "    \n",
    "    # axis labels\n",
    "    pyplot.xlabel('False Positive Rate')\n",
    "    pyplot.ylabel('True Positive Rate')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "    # show the plot\n",
    "    pyplot.savefig(\"RNN.png\",dpi=300)  \n",
    "    # summarize scores\n",
    "    print('RNN_Test: ROC AUC=%.3f' % (lr_auc))\n",
    "    print('RNN_Training: ROC AUC=%.3f' % (lr_auc_2))\n",
    "    print('RNN_3rd: ROC AUC=%.3f' % (lr_auc_3))\n",
    "\n",
    "    # confusion matrix\n",
    "    matrix = confusion_matrix(y_test,y_pred_classes)\n",
    "    matrix2= confusion_matrix(y_train, y_pred_classes_train)\n",
    "    matrix3= confusion_matrix(y_3rd, y_pred_classes_3rd)\n",
    "    print(matrix)\n",
    "    print(matrix2)\n",
    "    print(matrix3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Loc_RNN_simp(dataset,dataset_3rd):\n",
    "    X_train=dataset[0]\n",
    "    X_test=dataset[1]\n",
    "    y_train=dataset[2]\n",
    "    y_test=dataset[3]\n",
    "    X1 = dataset[4]\n",
    "    y1= dataset[5]\n",
    "    X = np.array(X1)\n",
    "    y = np.array(y1)\n",
    "    X_3rd=dataset_3rd[0]\n",
    "    y_3rd=dataset_3rd[1]    \n",
    "   \n",
    "    model = keras.models.load_model('RNN.h5')\n",
    "    model.summary()\n",
    "    rmsprop = RMSprop(learning_rate=0.001) #Set the Learning Rate\n",
    "\n",
    "    model.compile(optimizer='rmsprop', loss='binary_crossentropy', metrics=['acc'])\n",
    "    model.summary()\n",
    "\n",
    "    history = model.fit(np.asarray(X_train), np.asarray(y_train), \n",
    "                    epochs=15,\n",
    "                    batch_size=256,\n",
    "                    validation_split=0.2)\n",
    "    history_dict = history.history\n",
    "    loss = history_dict['loss']\n",
    "    val_loss = history_dict['val_loss']\n",
    "        \n",
    "    y_pred_keras = model.predict(X_test).ravel()\n",
    "    y_pred_classes0 = model.predict(X_test, verbose=0)\n",
    "    y_pred_classes = np.argmax(y_pred_classes0, axis=-1)\n",
    "    \n",
    "    y_pred_classes_train0= model.predict(X_train, verbose=0)\n",
    "    y_pred_classes_train=np.argmax(y_pred_classes_train0, axis=-1)\n",
    "    \n",
    "    y_pred_classes_3rd0= model.predict(X_3rd, verbose=0)\n",
    "    y_pred_classes_3rd = np.argmax(y_pred_classes_3rd0, axis=-1)\n",
    "\n",
    "    fpr_keras, tpr_keras, thresholds_keras = roc_curve(y_test, y_pred_keras)\n",
    "    \n",
    "    auc_keras = auc(fpr_keras, tpr_keras)\n",
    "    \n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    epochs = range(1, len(loss) + 1)\n",
    "\n",
    "    plt.plot(epochs, loss, 'bo', label='Training loss')\n",
    "    plt.plot(epochs, val_loss, 'r', label='Validation loss')\n",
    "    plt.title('Training and validation loss')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Loss')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "   #pyplot.show()\n",
    "    pyplot.savefig(\"RNN_LOSS.png\",dpi=300) \n",
    "    \n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "\n",
    "    history_dict = history.history\n",
    "    acc = history_dict['acc']\n",
    "    val_acc = history_dict['val_acc']\n",
    "\n",
    "    epochs = range(1, len(acc) + 1)\n",
    "\n",
    "    plt.plot(epochs, acc, 'bo', label='Training acc')\n",
    "    plt.plot(epochs, val_acc, 'r', label='Validation acc')\n",
    "    plt.title('Training and validation acc')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Accuracy')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "    #pyplot.show()\n",
    "    pyplot.savefig(\"RNN_ACC.png\",dpi=300)\n",
    "    score = model.evaluate(np.asarray(X_test),np.asarray( y_test))\n",
    "    score2 = model.evaluate(np.asarray(classifier_3rd[0]),np.asarray(classifier_3rd[1]))\n",
    "    print(score)\n",
    "    print(score2)\n",
    "    \n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "    # Load your Keras model\n",
    "    model = keras.models.load_model('RNN.h5')\n",
    "    \n",
    "    activities = model.predict(X_3rd)\n",
    "\n",
    "    # Determine the classification of the compounds based on their predicted activity\n",
    "    classifications = np.where(activities > 0.5, \"Active\", \"Inactive\")\n",
    "\n",
    "    if np.count_nonzero(classifications == \"Active\") > np.count_nonzero(classifications == \"Inactive\"):\n",
    "        print(\"The compound is classified as Active.\")\n",
    "    else:\n",
    "        print(\"The compound is classified as Inactive.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(42)\n",
    "tf.random.set_seed(42)\n",
    "\n",
    "def Loc_CNN(dataset,dataset_3rd):\n",
    "    X_train=dataset[0]\n",
    "    X_train = np.array(X_train)\n",
    "    X_test=dataset[1]\n",
    "    X_test= np.array(X_test)\n",
    "    y_train=dataset[2]\n",
    "    y_test=dataset[3]\n",
    "    X1 = dataset[4]\n",
    "    y1= dataset[5]\n",
    "    X = np.array(X1)\n",
    "    y = np.array(y1)\n",
    "    X_3rd=dataset_3rd[0]\n",
    "    y_3rd=dataset_3rd[1]\n",
    "    y_3rd = np.array(y_3rd)\n",
    "\n",
    "    model = keras.models.load_model('CNN.h5')\n",
    "    model.summary()\n",
    "   \n",
    "    rmsprop = RMSprop(learning_rate=0.001) #Learning Rate\n",
    "\n",
    "    model.compile(optimizer=rmsprop, loss='binary_crossentropy', metrics=['acc'])\n",
    "\n",
    "    history = model.fit(np.asarray(X_train), np.asarray(y_train), \n",
    "                    epochs=15,\n",
    "                    batch_size=256,\n",
    "                    validation_split=0.2)\n",
    "    history_dict = history.history\n",
    "    loss = history_dict['loss']\n",
    "    val_loss = history_dict['val_loss']\n",
    "    \n",
    "    y_pred_keras = model.predict(X_test).ravel()\n",
    "\n",
    "    y_pred_classes = model.predict(X_test)\n",
    "    y_pred_classes= y_pred_classes.flatten()\n",
    "    \n",
    "    y_pred_classes_train = model.predict(X_train)\n",
    "    y_pred_classes_train= y_pred_classes_train.flatten()\n",
    "    y_pred_classes_3rd = model.predict(X_3rd)\n",
    "    y_pred_classes_3rd= y_pred_classes_3rd.flatten()\n",
    "\n",
    "    # extract the predicted class labels\n",
    "    y_pred_classes = np.where(y_pred_classes > 0.5, 1, 0)\n",
    "    y_pred_classes_train = np.where(y_pred_classes_train > 0.5, 1, 0)\n",
    "    y_pred_classes_3rd = np.where(y_pred_classes_3rd > 0.5, 1, 0)\n",
    "    \n",
    "    fpr_keras, tpr_keras, thresholds_keras = roc_curve(y_test, y_pred_keras)\n",
    "    \n",
    "    auc_keras = auc(fpr_keras, tpr_keras)\n",
    "    \n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    epochs = range(1, len(loss) + 1)\n",
    "\n",
    "    plt.plot(epochs, loss, 'bo', label='Training loss')\n",
    "    plt.plot(epochs, val_loss, 'r', label='Validation loss')\n",
    "    plt.title('Training and validation loss')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Loss')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "   #pyplot.show()\n",
    "    pyplot.savefig(\"CNN_LOSS.png\",dpi=300) \n",
    "    \n",
    "    #clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    history_dict = history.history\n",
    "    acc = history_dict['acc']\n",
    "    val_acc = history_dict['val_acc']\n",
    "\n",
    "    epochs = range(1, len(acc) + 1)\n",
    "\n",
    "    plt.plot(epochs, acc, 'bo', label='Training acc')\n",
    "    plt.plot(epochs, val_acc, 'r', label='Validation acc')\n",
    "    plt.title('Training and validation acc')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Accuracy')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "    #pyplot.show()\n",
    "    pyplot.savefig(\"CNN_ACC.png\",dpi=300)\n",
    "    score = model.evaluate(np.asarray(X_test),np.asarray(y_test))\n",
    "    score2 = model.evaluate(np.asarray(classifier_3rd[0]),np.asarray(classifier_3rd[1]))\n",
    "    print(score)\n",
    "    print(score2)\n",
    "    \n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    # generate a no skill prediction (majority class)\n",
    "    ns_probs = [0 for _ in range(len(y_test))]\n",
    "    ns_probs_2 = [0 for _ in range(len(y_train))]\n",
    "    ns_probs_3 = [0 for _ in range(len(y_3rd))]\n",
    "    # predict probabilities\n",
    "    lr_probs = model.predict(X_test)\n",
    "    lr_probs_2 = model.predict(X_train)\n",
    "    lr_probs_3 = model.predict(X_3rd)\n",
    "\n",
    "\n",
    "    # calculate roc curves\n",
    "    ns_fpr, ns_tpr, _ = roc_curve(y_test, ns_probs)\n",
    "    lr_fpr, lr_tpr, _ = roc_curve(y_test, lr_probs)\n",
    "  \n",
    "    ns_fpr_2, ns_tpr_2, _ = roc_curve(y_train, ns_probs_2)\n",
    "    lr_fpr_2, lr_tpr_2, _ = roc_curve(y_train, lr_probs_2)\n",
    "    \n",
    "    ns_fpr_3, ns_tpr_3, _ = roc_curve(y_3rd, ns_probs_3)\n",
    "    lr_fpr_3, lr_tpr_3, _ = roc_curve(y_3rd, lr_probs_3)\n",
    "    # calculate scores\n",
    "    ns_auc = roc_auc_score(y_test, ns_probs)\n",
    "    lr_auc = roc_auc_score(y_test, lr_probs)\n",
    "    \n",
    "    ns_auc_2 = roc_auc_score(y_train, ns_probs_2)\n",
    "    lr_auc_2 = roc_auc_score(y_train, lr_probs_2)\n",
    "    \n",
    "    ns_auc_3 = roc_auc_score(y_3rd, ns_probs_3)\n",
    "    lr_auc_3 = roc_auc_score(y_3rd, lr_probs_3)\n",
    "    \n",
    "    # plot the roc curve for the model\n",
    "    pyplot.plot(ns_fpr, ns_tpr, linestyle='--')\n",
    "    pyplot.plot(lr_fpr, lr_tpr, marker='.', label='CNN_Test, AUC=%.3f' % (lr_auc))\n",
    "    pyplot.plot(lr_fpr_2, lr_tpr_2, color='green', marker='.', label='CNN_Training, AUC=%.3f' % (lr_auc_2))\n",
    "    pyplot.plot(lr_fpr_3, lr_tpr_3, color='red', marker='.', label='CNN_3rd, AUC=%.3f' % (lr_auc_3))\n",
    "    \n",
    "    # axis labels\n",
    "    pyplot.xlabel('False Positive Rate')\n",
    "    pyplot.ylabel('True Positive Rate')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "    # show the plot\n",
    "    pyplot.savefig(\"CNN.png\",dpi=300)  \n",
    "    # summarize scores\n",
    "    print('CNN_Test: ROC AUC=%.3f' % (lr_auc))\n",
    "    print('CNN_Training: ROC AUC=%.3f' % (lr_auc_2))\n",
    "    print('CNN_3rd: ROC AUC=%.3f' % (lr_auc_3))\n",
    "\n",
    "    # confusion matrix\n",
    "    matrix = confusion_matrix(y_test,y_pred_classes)\n",
    "    matrix2= confusion_matrix(y_train, y_pred_classes_train)\n",
    "    matrix3= confusion_matrix(y_3rd, y_pred_classes_3rd)\n",
    "    print(matrix)\n",
    "    print(matrix2)\n",
    "    print(matrix3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Loc_CNN_simp(dataset,dataset_3rd):\n",
    "    maxlen = 212\n",
    "    X_train=dataset[0]\n",
    "    X_test=dataset[1]\n",
    "    y_train=dataset[2]\n",
    "    y_test=dataset[3]\n",
    "    X1 = dataset[4]\n",
    "    y1 = dataset[5]\n",
    "    X = np.array(X1)\n",
    "    y = np.array(y1)\n",
    "    \n",
    "    X_3rd=dataset_3rd[0]\n",
    "    y_3rd=dataset_3rd[1]\n",
    "    \n",
    "    model = keras.models.load_model('CNN.h5')\n",
    "    model.summary()\n",
    "   \n",
    "    rmsprop = RMSprop(learning_rate=0.001) #Learning Rate\n",
    "\n",
    "    model.compile(optimizer=rmsprop, loss='binary_crossentropy', metrics=['acc'])\n",
    "\n",
    "    history = model.fit(np.asarray(X_train), np.asarray(y_train), \n",
    "                    epochs=15,\n",
    "                    batch_size=256,\n",
    "                    validation_split=0.2)\n",
    "    history_dict = history.history\n",
    "    loss = history_dict['loss']\n",
    "    val_loss = history_dict['val_loss']\n",
    "        \n",
    "    y_pred_keras = model.predict(X_test).ravel()\n",
    "    y_pred_classes0 = model.predict(X_test, verbose=0)\n",
    "    y_pred_classes = np.argmax(y_pred_classes0, axis=-1)\n",
    "    \n",
    "    y_pred_classes_train0= model.predict(X_train, verbose=0)\n",
    "    y_pred_classes_train=np.argmax(y_pred_classes_train0, axis=-1)\n",
    "    \n",
    "    y_pred_classes_3rd0= model.predict(X_3rd, verbose=0)\n",
    "    y_pred_classes_3rd=np.argmax(y_pred_classes_3rd0, axis=-1)\n",
    "    \n",
    "    fpr_keras, tpr_keras, thresholds_keras = roc_curve(y_test, y_pred_keras)\n",
    "    \n",
    "    auc_keras = auc(fpr_keras, tpr_keras)\n",
    "    \n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    epochs = range(1, len(loss) + 1)\n",
    "\n",
    "    plt.plot(epochs, loss, 'bo', label='Training loss')\n",
    "    plt.plot(epochs, val_loss, 'r', label='Validation loss')\n",
    "    plt.title('Training and validation loss')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Loss')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "    #pyplot.show()\n",
    "    pyplot.savefig(\"CNN_LOSS.png\",dpi=300)\n",
    "    \n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "    \n",
    "    history_dict = history.history\n",
    "    acc = history_dict['acc']\n",
    "    val_acc = history_dict['val_acc']\n",
    "\n",
    "    epochs = range(1, len(acc) + 1)\n",
    "\n",
    "    plt.plot(epochs, acc, 'bo', label='Training acc')\n",
    "    plt.plot(epochs, val_acc, 'r', label='Validation acc')\n",
    "    plt.title('Training and validation acc')\n",
    "    plt.xlabel('Epochs')\n",
    "    plt.ylabel('Accuracy')\n",
    "    # show the legend\n",
    "    plt.legend(fontsize=12, loc='best')\n",
    "    plt.tight_layout()\n",
    "    #pyplot.show()\n",
    "    pyplot.savefig(\"CNN_ACC.png\",dpi=300)\n",
    "  \n",
    "    score = model.evaluate(np.asarray(X_test),np.asarray(y_test))\n",
    "    score2 = model.evaluate(np.asarray(classifier_3rd[0]),np.asarray(classifier_3rd[1]))\n",
    "    print(score)\n",
    "    print(score2)\n",
    "    # clear the plot\n",
    "    pyplot.clf()\n",
    "\n",
    "    # Load your Keras model\n",
    "    model = keras.models.load_model('CNN.h5')\n",
    "    \n",
    "    activities = model.predict(X_3rd)\n",
    "\n",
    "    # Determine the classification of the compounds based on their predicted activity\n",
    "    classifications = np.where(activities > 0.5, \"Active\", \"Inactive\")\n",
    "\n",
    "    if np.count_nonzero(classifications == \"Active\") > np.count_nonzero(classifications == \"Inactive\"):\n",
    "        print(\"The compound is classified as Active.\")\n",
    "    else:\n",
    "        print(\"The compound is classified as Inactive.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lengths of 10 documents: [66, 66, 68, 66, 67, 70, 66, 69, 65, 66]\n",
      "num of tokens: 1024\n",
      "Train set count: 48220\n",
      "Test set count: 12055\n",
      "Test samples: [1, 0, 1, 0, 1, 1, 1, 0, 1, 0, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1]\n",
      "(48220, 212) (12055, 212) (60275, 212)\n",
      "lengths of 10 documents: [37, 37, 36, 36, 37, 36, 37, 36, 37, 37]\n",
      "num of tokens: 595\n",
      "Train set count: 200\n",
      "Test samples: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n"
     ]
    }
   ],
   "source": [
    "Classifier_input = preprocessing(readdata('dataset.csv')[0],readdata('dataset.csv')[1])\n",
    "classifier_3rd = preprocessing_3rd(readdata('lignan.csv')[0],readdata('lignan.csv')[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_177\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_177 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " flatten_177 (Flatten)       (None, 6784)              0         \n",
      "                                                                 \n",
      " dense_177 (Dense)           (None, 1)                 6785      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 39,553\n",
      "Trainable params: 39,553\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.2868 - acc: 0.8842 - val_loss: 0.1992 - val_acc: 0.9172\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 1s 9ms/step - loss: 0.1456 - acc: 0.9424 - val_loss: 0.1426 - val_acc: 0.9420\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.1029 - acc: 0.9608 - val_loss: 0.1200 - val_acc: 0.9521\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0811 - acc: 0.9702 - val_loss: 0.1088 - val_acc: 0.9546\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0677 - acc: 0.9758 - val_loss: 0.1042 - val_acc: 0.9581\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0592 - acc: 0.9788 - val_loss: 0.1018 - val_acc: 0.9589\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0531 - acc: 0.9804 - val_loss: 0.1036 - val_acc: 0.9575\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0482 - acc: 0.9827 - val_loss: 0.1012 - val_acc: 0.9577\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0446 - acc: 0.9834 - val_loss: 0.1020 - val_acc: 0.9590\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0416 - acc: 0.9844 - val_loss: 0.1032 - val_acc: 0.9582\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 1s 9ms/step - loss: 0.0391 - acc: 0.9849 - val_loss: 0.1031 - val_acc: 0.9587\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0368 - acc: 0.9854 - val_loss: 0.1065 - val_acc: 0.9591\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0351 - acc: 0.9853 - val_loss: 0.1074 - val_acc: 0.9584\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0334 - acc: 0.9863 - val_loss: 0.1112 - val_acc: 0.9574\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0323 - acc: 0.9863 - val_loss: 0.1136 - val_acc: 0.9585\n",
      "377/377 [==============================] - 1s 2ms/step - loss: 0.1046 - acc: 0.9611\n",
      "7/7 [==============================] - 0s 1ms/step - loss: 1.0264 - acc: 0.5900\n",
      "[0.10461772233247757, 0.9610949754714966]\n",
      "[1.0264185667037964, 0.5899999737739563]\n",
      "7/7 [==============================] - 0s 1ms/step\n",
      "The compound is classified as Inactive.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_MLP_simp(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_527\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_527 (Embedding)   (None, None, 212)         217088    \n",
      "                                                                 \n",
      " global_max_pooling1d_160 (G  (None, 212)              0         \n",
      " lobalMaxPooling1D)                                              \n",
      "                                                                 \n",
      " dense_687 (Dense)           (None, 5)                 1065      \n",
      "                                                                 \n",
      " dense_688 (Dense)           (None, 1)                 6         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 218,159\n",
      "Trainable params: 218,159\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 11s 71ms/step - loss: 0.3698 - acc: 0.8642 - val_loss: 0.2373 - val_acc: 0.9082\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 11s 74ms/step - loss: 0.1758 - acc: 0.9321 - val_loss: 0.1657 - val_acc: 0.9329\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 10s 67ms/step - loss: 0.1227 - acc: 0.9549 - val_loss: 0.1308 - val_acc: 0.9489\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 10s 67ms/step - loss: 0.0968 - acc: 0.9640 - val_loss: 0.1155 - val_acc: 0.9538\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 11s 70ms/step - loss: 0.0802 - acc: 0.9715 - val_loss: 0.1056 - val_acc: 0.9586\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.0690 - acc: 0.9757 - val_loss: 0.0982 - val_acc: 0.9603\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 11s 74ms/step - loss: 0.0611 - acc: 0.9778 - val_loss: 0.0995 - val_acc: 0.9612\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 11s 72ms/step - loss: 0.0553 - acc: 0.9791 - val_loss: 0.0951 - val_acc: 0.9615\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 11s 74ms/step - loss: 0.0502 - acc: 0.9811 - val_loss: 0.0903 - val_acc: 0.9626\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 11s 71ms/step - loss: 0.0463 - acc: 0.9822 - val_loss: 0.0884 - val_acc: 0.9638\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0433 - acc: 0.9827 - val_loss: 0.0876 - val_acc: 0.9641\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 11s 71ms/step - loss: 0.0402 - acc: 0.9834 - val_loss: 0.0858 - val_acc: 0.9642\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.0383 - acc: 0.9841 - val_loss: 0.0893 - val_acc: 0.9633\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.0360 - acc: 0.9844 - val_loss: 0.0867 - val_acc: 0.9646\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.0349 - acc: 0.9850 - val_loss: 0.0884 - val_acc: 0.9658\n",
      "377/377 [==============================] - 1s 3ms/step\n",
      "377/377 [==============================] - 2s 4ms/step - loss: 0.0796 - acc: 0.9679\n",
      "7/7 [==============================] - 0s 3ms/step - loss: 6.7455 - acc: 0.1650\n",
      "[0.07958674430847168, 0.9678971171379089]\n",
      "[6.74549674987793, 0.16500000655651093]\n",
      "7/7 [==============================] - 0s 3ms/step\n",
      "The compound is classified as Active.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_CNN_simp(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 26s 162ms/step - loss: 0.3344 - acc: 0.8540 - val_loss: 0.2250 - val_acc: 0.9064\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 25s 164ms/step - loss: 0.1564 - acc: 0.9359 - val_loss: 0.1574 - val_acc: 0.9367\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 25s 168ms/step - loss: 0.1057 - acc: 0.9570 - val_loss: 0.1205 - val_acc: 0.9517\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 26s 173ms/step - loss: 0.0834 - acc: 0.9660 - val_loss: 0.1004 - val_acc: 0.9581\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 26s 169ms/step - loss: 0.0679 - acc: 0.9711 - val_loss: 0.1268 - val_acc: 0.9517\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 28s 183ms/step - loss: 0.0612 - acc: 0.9741 - val_loss: 0.1050 - val_acc: 0.9583\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 28s 184ms/step - loss: 0.0546 - acc: 0.9764 - val_loss: 0.1036 - val_acc: 0.9600\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 26s 175ms/step - loss: 0.0496 - acc: 0.9778 - val_loss: 0.1005 - val_acc: 0.9599\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 25s 163ms/step - loss: 0.0455 - acc: 0.9797 - val_loss: 0.0946 - val_acc: 0.9611\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 25s 163ms/step - loss: 0.0422 - acc: 0.9808 - val_loss: 0.0945 - val_acc: 0.9641\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 24s 160ms/step - loss: 0.0405 - acc: 0.9806 - val_loss: 0.1051 - val_acc: 0.9599\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 24s 161ms/step - loss: 0.0393 - acc: 0.9814 - val_loss: 0.1008 - val_acc: 0.9598\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 24s 162ms/step - loss: 0.0370 - acc: 0.9820 - val_loss: 0.1034 - val_acc: 0.9612\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 24s 161ms/step - loss: 0.0344 - acc: 0.9830 - val_loss: 0.1034 - val_acc: 0.9632\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 25s 163ms/step - loss: 0.0350 - acc: 0.9829 - val_loss: 0.1210 - val_acc: 0.9579\n",
      "377/377 [==============================] - 5s 12ms/step\n",
      "377/377 [==============================] - 5s 13ms/step - loss: 0.1126 - acc: 0.9619\n",
      "7/7 [==============================] - 0s 13ms/step - loss: 0.5122 - acc: 0.7600\n",
      "[0.11263934522867203, 0.9619244933128357]\n",
      "[0.5122488141059875, 0.7599999904632568]\n",
      "7/7 [==============================] - 0s 11ms/step\n",
      "The compound is classified as Inactive.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_RNN_simp(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lengths of 10 documents: [66]\n",
      "num of tokens: 66\n",
      "Train set count: 1\n",
      "Test samples: [0]\n"
     ]
    }
   ],
   "source": [
    "classifier_3rd = preprocessing_3rd(readdata('Steroid.csv')[0],readdata('Steroid.csv')[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_177\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_177 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " flatten_177 (Flatten)       (None, 6784)              0         \n",
      "                                                                 \n",
      " dense_177 (Dense)           (None, 1)                 6785      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 39,553\n",
      "Trainable params: 39,553\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.2868 - acc: 0.8842 - val_loss: 0.1992 - val_acc: 0.9172\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.1456 - acc: 0.9424 - val_loss: 0.1426 - val_acc: 0.9420\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.1029 - acc: 0.9608 - val_loss: 0.1200 - val_acc: 0.9521\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0811 - acc: 0.9702 - val_loss: 0.1088 - val_acc: 0.9546\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0677 - acc: 0.9758 - val_loss: 0.1042 - val_acc: 0.9581\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0592 - acc: 0.9788 - val_loss: 0.1018 - val_acc: 0.9589\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0531 - acc: 0.9804 - val_loss: 0.1036 - val_acc: 0.9575\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0482 - acc: 0.9827 - val_loss: 0.1012 - val_acc: 0.9577\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0446 - acc: 0.9834 - val_loss: 0.1020 - val_acc: 0.9590\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0416 - acc: 0.9844 - val_loss: 0.1032 - val_acc: 0.9582\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 1s 9ms/step - loss: 0.0391 - acc: 0.9849 - val_loss: 0.1031 - val_acc: 0.9587\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 1s 9ms/step - loss: 0.0368 - acc: 0.9854 - val_loss: 0.1065 - val_acc: 0.9591\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0351 - acc: 0.9853 - val_loss: 0.1074 - val_acc: 0.9584\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0334 - acc: 0.9863 - val_loss: 0.1112 - val_acc: 0.9574\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0323 - acc: 0.9863 - val_loss: 0.1136 - val_acc: 0.9585\n",
      "377/377 [==============================] - 0s 965us/step - loss: 0.1046 - acc: 0.9611\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 4.9920 - acc: 0.0000e+00\n",
      "[0.10461772233247757, 0.9610949754714966]\n",
      "[4.99196195602417, 0.0]\n",
      "1/1 [==============================] - 0s 34ms/step\n",
      "The compound is classified as Active.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_MLP_simp(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_527\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_527 (Embedding)   (None, None, 212)         217088    \n",
      "                                                                 \n",
      " global_max_pooling1d_160 (G  (None, 212)              0         \n",
      " lobalMaxPooling1D)                                              \n",
      "                                                                 \n",
      " dense_687 (Dense)           (None, 5)                 1065      \n",
      "                                                                 \n",
      " dense_688 (Dense)           (None, 1)                 6         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 218,159\n",
      "Trainable params: 218,159\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 11s 68ms/step - loss: 0.3698 - acc: 0.8642 - val_loss: 0.2373 - val_acc: 0.9082\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 10s 67ms/step - loss: 0.1758 - acc: 0.9321 - val_loss: 0.1657 - val_acc: 0.9329\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.1227 - acc: 0.9549 - val_loss: 0.1308 - val_acc: 0.9489\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 10s 66ms/step - loss: 0.0968 - acc: 0.9640 - val_loss: 0.1155 - val_acc: 0.9538\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.0802 - acc: 0.9715 - val_loss: 0.1056 - val_acc: 0.9586\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 10s 67ms/step - loss: 0.0690 - acc: 0.9757 - val_loss: 0.0982 - val_acc: 0.9603\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.0611 - acc: 0.9778 - val_loss: 0.0995 - val_acc: 0.9612\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0553 - acc: 0.9791 - val_loss: 0.0951 - val_acc: 0.9615\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 10s 66ms/step - loss: 0.0502 - acc: 0.9811 - val_loss: 0.0903 - val_acc: 0.9626\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0463 - acc: 0.9822 - val_loss: 0.0884 - val_acc: 0.9638\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.0433 - acc: 0.9827 - val_loss: 0.0876 - val_acc: 0.9641\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 10s 70ms/step - loss: 0.0402 - acc: 0.9834 - val_loss: 0.0858 - val_acc: 0.9642\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 11s 70ms/step - loss: 0.0383 - acc: 0.9841 - val_loss: 0.0893 - val_acc: 0.9633\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 11s 70ms/step - loss: 0.0360 - acc: 0.9844 - val_loss: 0.0867 - val_acc: 0.9646\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.0349 - acc: 0.9850 - val_loss: 0.0884 - val_acc: 0.9658\n",
      "377/377 [==============================] - 1s 3ms/step\n",
      "377/377 [==============================] - 1s 3ms/step - loss: 0.0796 - acc: 0.9679\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 9.5084 - acc: 0.0000e+00\n",
      "[0.07958674430847168, 0.9678971171379089]\n",
      "[9.508434295654297, 0.0]\n",
      "1/1 [==============================] - 0s 40ms/step\n",
      "The compound is classified as Active.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_CNN_simp(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 29s 180ms/step - loss: 0.3344 - acc: 0.8540 - val_loss: 0.2250 - val_acc: 0.9064\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 26s 175ms/step - loss: 0.1564 - acc: 0.9359 - val_loss: 0.1574 - val_acc: 0.9367\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 26s 173ms/step - loss: 0.1057 - acc: 0.9570 - val_loss: 0.1205 - val_acc: 0.9517\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 27s 176ms/step - loss: 0.0834 - acc: 0.9660 - val_loss: 0.1004 - val_acc: 0.9581\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 26s 173ms/step - loss: 0.0679 - acc: 0.9711 - val_loss: 0.1268 - val_acc: 0.9517\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 26s 173ms/step - loss: 0.0612 - acc: 0.9741 - val_loss: 0.1050 - val_acc: 0.9583\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 26s 172ms/step - loss: 0.0546 - acc: 0.9764 - val_loss: 0.1036 - val_acc: 0.9600\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 26s 174ms/step - loss: 0.0496 - acc: 0.9778 - val_loss: 0.1005 - val_acc: 0.9599\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 27s 176ms/step - loss: 0.0455 - acc: 0.9797 - val_loss: 0.0946 - val_acc: 0.9611\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 26s 171ms/step - loss: 0.0422 - acc: 0.9808 - val_loss: 0.0945 - val_acc: 0.9641\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 26s 170ms/step - loss: 0.0405 - acc: 0.9806 - val_loss: 0.1051 - val_acc: 0.9599\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 26s 174ms/step - loss: 0.0393 - acc: 0.9814 - val_loss: 0.1008 - val_acc: 0.9598\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 26s 172ms/step - loss: 0.0370 - acc: 0.9820 - val_loss: 0.1034 - val_acc: 0.9612\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 27s 178ms/step - loss: 0.0344 - acc: 0.9830 - val_loss: 0.1034 - val_acc: 0.9632\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 27s 178ms/step - loss: 0.0350 - acc: 0.9829 - val_loss: 0.1210 - val_acc: 0.9579\n",
      "377/377 [==============================] - 5s 12ms/step\n",
      "377/377 [==============================] - 5s 12ms/step - loss: 0.1126 - acc: 0.9619\n",
      "1/1 [==============================] - 0s 23ms/step - loss: 8.1926 - acc: 0.0000e+00\n",
      "[0.11263934522867203, 0.9619244933128357]\n",
      "[8.192581176757812, 0.0]\n",
      "1/1 [==============================] - 0s 341ms/step\n",
      "The compound is classified as Active.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_RNN_simp(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lengths of 10 documents: [52, 51, 51, 53, 52, 52, 52, 53, 52, 55]\n",
      "num of tokens: 614\n",
      "Train set count: 130\n",
      "Test samples: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n"
     ]
    }
   ],
   "source": [
    "classifier_3rd = preprocessing_3rd(readdata('phenylpropanoid.csv')[0],readdata('phenylpropanoid.csv')[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_177\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_177 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " flatten_177 (Flatten)       (None, 6784)              0         \n",
      "                                                                 \n",
      " dense_177 (Dense)           (None, 1)                 6785      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 39,553\n",
      "Trainable params: 39,553\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 2s 13ms/step - loss: 0.2868 - acc: 0.8842 - val_loss: 0.1992 - val_acc: 0.9172\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.1456 - acc: 0.9424 - val_loss: 0.1426 - val_acc: 0.9420\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.1029 - acc: 0.9608 - val_loss: 0.1200 - val_acc: 0.9521\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0811 - acc: 0.9702 - val_loss: 0.1088 - val_acc: 0.9546\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0677 - acc: 0.9758 - val_loss: 0.1042 - val_acc: 0.9581\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0592 - acc: 0.9788 - val_loss: 0.1018 - val_acc: 0.9589\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0531 - acc: 0.9804 - val_loss: 0.1036 - val_acc: 0.9575\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0482 - acc: 0.9827 - val_loss: 0.1012 - val_acc: 0.9577\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0446 - acc: 0.9834 - val_loss: 0.1020 - val_acc: 0.9590\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0416 - acc: 0.9844 - val_loss: 0.1032 - val_acc: 0.9582\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0391 - acc: 0.9849 - val_loss: 0.1031 - val_acc: 0.9587\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 1s 9ms/step - loss: 0.0368 - acc: 0.9854 - val_loss: 0.1065 - val_acc: 0.9591\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0351 - acc: 0.9853 - val_loss: 0.1074 - val_acc: 0.9584\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0334 - acc: 0.9863 - val_loss: 0.1112 - val_acc: 0.9574\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0323 - acc: 0.9863 - val_loss: 0.1136 - val_acc: 0.9585\n",
      "377/377 [==============================] - 0s 1ms/step - loss: 0.1046 - acc: 0.9611\n",
      "5/5 [==============================] - 0s 2ms/step - loss: 4.4715 - acc: 0.0615\n",
      "[0.10461772233247757, 0.9610949754714966]\n",
      "[4.471503257751465, 0.0615384615957737]\n",
      "5/5 [==============================] - 0s 2ms/step\n",
      "The compound is classified as Active.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_MLP_simp(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_527\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_527 (Embedding)   (None, None, 212)         217088    \n",
      "                                                                 \n",
      " global_max_pooling1d_160 (G  (None, 212)              0         \n",
      " lobalMaxPooling1D)                                              \n",
      "                                                                 \n",
      " dense_687 (Dense)           (None, 5)                 1065      \n",
      "                                                                 \n",
      " dense_688 (Dense)           (None, 1)                 6         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 218,159\n",
      "Trainable params: 218,159\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 11s 69ms/step - loss: 0.3698 - acc: 0.8642 - val_loss: 0.2373 - val_acc: 0.9082\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.1758 - acc: 0.9321 - val_loss: 0.1657 - val_acc: 0.9329\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.1227 - acc: 0.9549 - val_loss: 0.1308 - val_acc: 0.9489\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.0968 - acc: 0.9640 - val_loss: 0.1155 - val_acc: 0.9538\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 10s 66ms/step - loss: 0.0802 - acc: 0.9715 - val_loss: 0.1056 - val_acc: 0.9586\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 10s 67ms/step - loss: 0.0690 - acc: 0.9757 - val_loss: 0.0982 - val_acc: 0.9603\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0611 - acc: 0.9778 - val_loss: 0.0995 - val_acc: 0.9612\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0553 - acc: 0.9791 - val_loss: 0.0951 - val_acc: 0.9615\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.0502 - acc: 0.9811 - val_loss: 0.0903 - val_acc: 0.9626\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 11s 70ms/step - loss: 0.0463 - acc: 0.9822 - val_loss: 0.0884 - val_acc: 0.9638\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 11s 73ms/step - loss: 0.0433 - acc: 0.9827 - val_loss: 0.0876 - val_acc: 0.9641\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 11s 76ms/step - loss: 0.0402 - acc: 0.9834 - val_loss: 0.0858 - val_acc: 0.9642\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 11s 75ms/step - loss: 0.0383 - acc: 0.9841 - val_loss: 0.0893 - val_acc: 0.9633\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 11s 72ms/step - loss: 0.0360 - acc: 0.9844 - val_loss: 0.0867 - val_acc: 0.9646\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.0349 - acc: 0.9850 - val_loss: 0.0884 - val_acc: 0.9658\n",
      "377/377 [==============================] - 1s 3ms/step\n",
      "377/377 [==============================] - 1s 3ms/step - loss: 0.0796 - acc: 0.9679\n",
      "5/5 [==============================] - 0s 3ms/step - loss: 9.7857 - acc: 0.0231\n",
      "[0.07958674430847168, 0.9678971171379089]\n",
      "[9.785733222961426, 0.023076923564076424]\n",
      "5/5 [==============================] - 0s 3ms/step\n",
      "The compound is classified as Active.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_CNN_simp(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 29s 179ms/step - loss: 0.3344 - acc: 0.8540 - val_loss: 0.2250 - val_acc: 0.9064\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 27s 178ms/step - loss: 0.1564 - acc: 0.9359 - val_loss: 0.1574 - val_acc: 0.9367\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 26s 174ms/step - loss: 0.1057 - acc: 0.9570 - val_loss: 0.1205 - val_acc: 0.9517\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 26s 173ms/step - loss: 0.0834 - acc: 0.9660 - val_loss: 0.1004 - val_acc: 0.9581\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 26s 174ms/step - loss: 0.0679 - acc: 0.9711 - val_loss: 0.1268 - val_acc: 0.9517\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 26s 171ms/step - loss: 0.0612 - acc: 0.9741 - val_loss: 0.1050 - val_acc: 0.9583\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 27s 182ms/step - loss: 0.0546 - acc: 0.9764 - val_loss: 0.1036 - val_acc: 0.9600\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 26s 175ms/step - loss: 0.0496 - acc: 0.9778 - val_loss: 0.1005 - val_acc: 0.9599\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 26s 172ms/step - loss: 0.0455 - acc: 0.9797 - val_loss: 0.0946 - val_acc: 0.9611\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 26s 174ms/step - loss: 0.0422 - acc: 0.9808 - val_loss: 0.0945 - val_acc: 0.9641\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 26s 173ms/step - loss: 0.0405 - acc: 0.9806 - val_loss: 0.1051 - val_acc: 0.9599\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 26s 172ms/step - loss: 0.0393 - acc: 0.9814 - val_loss: 0.1008 - val_acc: 0.9598\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 26s 174ms/step - loss: 0.0370 - acc: 0.9820 - val_loss: 0.1034 - val_acc: 0.9612\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 26s 173ms/step - loss: 0.0344 - acc: 0.9830 - val_loss: 0.1034 - val_acc: 0.9632\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 26s 173ms/step - loss: 0.0350 - acc: 0.9829 - val_loss: 0.1210 - val_acc: 0.9579\n",
      "377/377 [==============================] - 5s 12ms/step\n",
      "377/377 [==============================] - 5s 12ms/step - loss: 0.1126 - acc: 0.9619\n",
      "5/5 [==============================] - 0s 12ms/step - loss: 2.6536 - acc: 0.1308\n",
      "[0.11263934522867203, 0.9619244933128357]\n",
      "[2.6536483764648438, 0.13076923787593842]\n",
      "5/5 [==============================] - 0s 12ms/step\n",
      "The compound is classified as Active.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_RNN_simp(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lengths of 10 documents: [32, 33, 31, 33, 32]\n",
      "num of tokens: 67\n",
      "Train set count: 5\n",
      "Test samples: [0, 0, 0, 0, 0]\n"
     ]
    }
   ],
   "source": [
    "classifier_3rd = preprocessing_3rd(readdata('Iridoid.csv')[0],readdata('Iridoid.csv')[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_177\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_177 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " flatten_177 (Flatten)       (None, 6784)              0         \n",
      "                                                                 \n",
      " dense_177 (Dense)           (None, 1)                 6785      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 39,553\n",
      "Trainable params: 39,553\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.2868 - acc: 0.8842 - val_loss: 0.1992 - val_acc: 0.9172\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.1456 - acc: 0.9424 - val_loss: 0.1426 - val_acc: 0.9420\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.1029 - acc: 0.9608 - val_loss: 0.1200 - val_acc: 0.9521\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0811 - acc: 0.9702 - val_loss: 0.1088 - val_acc: 0.9546\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0677 - acc: 0.9758 - val_loss: 0.1042 - val_acc: 0.9581\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0592 - acc: 0.9788 - val_loss: 0.1018 - val_acc: 0.9589\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0531 - acc: 0.9804 - val_loss: 0.1036 - val_acc: 0.9575\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0482 - acc: 0.9827 - val_loss: 0.1012 - val_acc: 0.9577\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0446 - acc: 0.9834 - val_loss: 0.1020 - val_acc: 0.9590\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0416 - acc: 0.9844 - val_loss: 0.1032 - val_acc: 0.9582\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0391 - acc: 0.9849 - val_loss: 0.1031 - val_acc: 0.9587\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0368 - acc: 0.9854 - val_loss: 0.1065 - val_acc: 0.9591\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0351 - acc: 0.9853 - val_loss: 0.1074 - val_acc: 0.9584\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0334 - acc: 0.9863 - val_loss: 0.1112 - val_acc: 0.9574\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0323 - acc: 0.9863 - val_loss: 0.1136 - val_acc: 0.9585\n",
      "377/377 [==============================] - 0s 1ms/step - loss: 0.1046 - acc: 0.9611\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 1.0993 - acc: 0.4000\n",
      "[0.10461772233247757, 0.9610949754714966]\n",
      "[1.0992711782455444, 0.4000000059604645]\n",
      "1/1 [==============================] - 0s 36ms/step\n",
      "The compound is classified as Active.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_MLP_simp(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_527\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_527 (Embedding)   (None, None, 212)         217088    \n",
      "                                                                 \n",
      " global_max_pooling1d_160 (G  (None, 212)              0         \n",
      " lobalMaxPooling1D)                                              \n",
      "                                                                 \n",
      " dense_687 (Dense)           (None, 5)                 1065      \n",
      "                                                                 \n",
      " dense_688 (Dense)           (None, 1)                 6         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 218,159\n",
      "Trainable params: 218,159\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 11s 70ms/step - loss: 0.3698 - acc: 0.8642 - val_loss: 0.2373 - val_acc: 0.9082\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.1758 - acc: 0.9321 - val_loss: 0.1657 - val_acc: 0.9329\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.1227 - acc: 0.9549 - val_loss: 0.1308 - val_acc: 0.9489\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0968 - acc: 0.9640 - val_loss: 0.1155 - val_acc: 0.9538\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 10s 68ms/step - loss: 0.0802 - acc: 0.9715 - val_loss: 0.1056 - val_acc: 0.9586\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0690 - acc: 0.9757 - val_loss: 0.0982 - val_acc: 0.9603\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0611 - acc: 0.9778 - val_loss: 0.0995 - val_acc: 0.9612\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 11s 70ms/step - loss: 0.0553 - acc: 0.9791 - val_loss: 0.0951 - val_acc: 0.9615\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0502 - acc: 0.9811 - val_loss: 0.0903 - val_acc: 0.9626\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 11s 70ms/step - loss: 0.0463 - acc: 0.9822 - val_loss: 0.0884 - val_acc: 0.9638\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0433 - acc: 0.9827 - val_loss: 0.0876 - val_acc: 0.9641\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 11s 72ms/step - loss: 0.0402 - acc: 0.9834 - val_loss: 0.0858 - val_acc: 0.9642\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0383 - acc: 0.9841 - val_loss: 0.0893 - val_acc: 0.9633\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 11s 71ms/step - loss: 0.0360 - acc: 0.9844 - val_loss: 0.0867 - val_acc: 0.9646\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 11s 72ms/step - loss: 0.0349 - acc: 0.9850 - val_loss: 0.0884 - val_acc: 0.9658\n",
      "377/377 [==============================] - 1s 3ms/step\n",
      "377/377 [==============================] - 1s 3ms/step - loss: 0.0796 - acc: 0.9679\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 8.9588 - acc: 0.0000e+00\n",
      "[0.07958674430847168, 0.9678971171379089]\n",
      "[8.958808898925781, 0.0]\n",
      "1/1 [==============================] - 0s 45ms/step\n",
      "The compound is classified as Active.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_CNN_simp(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 28s 178ms/step - loss: 0.3344 - acc: 0.8540 - val_loss: 0.2250 - val_acc: 0.9064\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 28s 185ms/step - loss: 0.1564 - acc: 0.9359 - val_loss: 0.1574 - val_acc: 0.9367\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 29s 190ms/step - loss: 0.1057 - acc: 0.9570 - val_loss: 0.1205 - val_acc: 0.9517\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 26s 173ms/step - loss: 0.0834 - acc: 0.9660 - val_loss: 0.1004 - val_acc: 0.9581\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 26s 170ms/step - loss: 0.0679 - acc: 0.9711 - val_loss: 0.1268 - val_acc: 0.9517\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 25s 169ms/step - loss: 0.0612 - acc: 0.9741 - val_loss: 0.1050 - val_acc: 0.9583\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 26s 173ms/step - loss: 0.0546 - acc: 0.9764 - val_loss: 0.1036 - val_acc: 0.9600\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 26s 171ms/step - loss: 0.0496 - acc: 0.9778 - val_loss: 0.1005 - val_acc: 0.9599\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 27s 176ms/step - loss: 0.0455 - acc: 0.9797 - val_loss: 0.0946 - val_acc: 0.9611\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 28s 184ms/step - loss: 0.0422 - acc: 0.9808 - val_loss: 0.0945 - val_acc: 0.9641\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 28s 183ms/step - loss: 0.0405 - acc: 0.9806 - val_loss: 0.1051 - val_acc: 0.9599\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 27s 179ms/step - loss: 0.0393 - acc: 0.9814 - val_loss: 0.1008 - val_acc: 0.9598\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 27s 178ms/step - loss: 0.0370 - acc: 0.9820 - val_loss: 0.1034 - val_acc: 0.9612\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 27s 179ms/step - loss: 0.0344 - acc: 0.9830 - val_loss: 0.1034 - val_acc: 0.9632\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 27s 178ms/step - loss: 0.0350 - acc: 0.9829 - val_loss: 0.1210 - val_acc: 0.9579\n",
      "377/377 [==============================] - 5s 12ms/step\n",
      "377/377 [==============================] - 5s 12ms/step - loss: 0.1126 - acc: 0.9619\n",
      "1/1 [==============================] - 0s 24ms/step - loss: 1.3130 - acc: 0.4000\n",
      "[0.11263934522867203, 0.9619244933128357]\n",
      "[1.3129587173461914, 0.4000000059604645]\n",
      "1/1 [==============================] - 0s 329ms/step\n",
      "The compound is classified as Active.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_RNN_simp(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lengths of 10 documents: [133, 139, 133, 135, 137, 133, 138, 134, 137, 131]\n",
      "num of tokens: 1012\n",
      "Train set count: 275\n",
      "Test samples: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n"
     ]
    }
   ],
   "source": [
    "classifier_3rd = preprocessing_3rd(readdata('Saponins.csv')[0],readdata('Saponins.csv')[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_527\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_527 (Embedding)   (None, None, 212)         217088    \n",
      "                                                                 \n",
      " global_max_pooling1d_160 (G  (None, 212)              0         \n",
      " lobalMaxPooling1D)                                              \n",
      "                                                                 \n",
      " dense_687 (Dense)           (None, 5)                 1065      \n",
      "                                                                 \n",
      " dense_688 (Dense)           (None, 1)                 6         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 218,159\n",
      "Trainable params: 218,159\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 11s 66ms/step - loss: 0.3698 - acc: 0.8642 - val_loss: 0.2373 - val_acc: 0.9082\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 10s 65ms/step - loss: 0.1758 - acc: 0.9321 - val_loss: 0.1657 - val_acc: 0.9329\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 10s 66ms/step - loss: 0.1227 - acc: 0.9549 - val_loss: 0.1308 - val_acc: 0.9489\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 10s 66ms/step - loss: 0.0968 - acc: 0.9640 - val_loss: 0.1155 - val_acc: 0.9538\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 11s 70ms/step - loss: 0.0802 - acc: 0.9715 - val_loss: 0.1056 - val_acc: 0.9586\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 11s 72ms/step - loss: 0.0690 - acc: 0.9757 - val_loss: 0.0982 - val_acc: 0.9603\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 11s 74ms/step - loss: 0.0611 - acc: 0.9778 - val_loss: 0.0995 - val_acc: 0.9612\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0553 - acc: 0.9791 - val_loss: 0.0951 - val_acc: 0.9615\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 11s 70ms/step - loss: 0.0502 - acc: 0.9811 - val_loss: 0.0903 - val_acc: 0.9626\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 11s 70ms/step - loss: 0.0463 - acc: 0.9822 - val_loss: 0.0884 - val_acc: 0.9638\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 11s 71ms/step - loss: 0.0433 - acc: 0.9827 - val_loss: 0.0876 - val_acc: 0.9641\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 11s 71ms/step - loss: 0.0402 - acc: 0.9834 - val_loss: 0.0858 - val_acc: 0.9642\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 11s 72ms/step - loss: 0.0383 - acc: 0.9841 - val_loss: 0.0893 - val_acc: 0.9633\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 10s 69ms/step - loss: 0.0360 - acc: 0.9844 - val_loss: 0.0867 - val_acc: 0.9646\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 11s 70ms/step - loss: 0.0349 - acc: 0.9850 - val_loss: 0.0884 - val_acc: 0.9658\n",
      "377/377 [==============================] - 1s 3ms/step\n",
      "377/377 [==============================] - 1s 4ms/step\n",
      "1507/1507 [==============================] - 5s 4ms/step\n",
      "9/9 [==============================] - 0s 4ms/step\n",
      "377/377 [==============================] - 1s 4ms/step - loss: 0.0796 - acc: 0.9679\n",
      "9/9 [==============================] - 0s 4ms/step - loss: 0.5983 - acc: 0.7018\n",
      "[0.07958674430847168, 0.9678971171379089]\n",
      "[0.5983113050460815, 0.7018181681632996]\n",
      "377/377 [==============================] - 1s 4ms/step\n",
      "1507/1507 [==============================] - 5s 3ms/step\n",
      "9/9 [==============================] - 0s 3ms/step\n",
      "CNN_Test: ROC AUC=0.996\n",
      "CNN_Training: ROC AUC=0.999\n",
      "CNN_3rd: ROC AUC=0.760\n",
      "[[4597  212]\n",
      " [ 175 7071]]\n",
      "[[19122   527]\n",
      " [  288 28283]]\n",
      "[[151  49]\n",
      " [ 33  42]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_CNN(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_177\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_177 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " flatten_177 (Flatten)       (None, 6784)              0         \n",
      "                                                                 \n",
      " dense_177 (Dense)           (None, 1)                 6785      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 39,553\n",
      "Trainable params: 39,553\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.2868 - acc: 0.8842 - val_loss: 0.1992 - val_acc: 0.9172\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.1456 - acc: 0.9424 - val_loss: 0.1426 - val_acc: 0.9420\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.1029 - acc: 0.9608 - val_loss: 0.1200 - val_acc: 0.9521\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0811 - acc: 0.9702 - val_loss: 0.1088 - val_acc: 0.9546\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0677 - acc: 0.9758 - val_loss: 0.1042 - val_acc: 0.9581\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0592 - acc: 0.9788 - val_loss: 0.1018 - val_acc: 0.9589\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0531 - acc: 0.9804 - val_loss: 0.1036 - val_acc: 0.9575\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0482 - acc: 0.9827 - val_loss: 0.1012 - val_acc: 0.9577\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0446 - acc: 0.9834 - val_loss: 0.1020 - val_acc: 0.9590\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 2s 12ms/step - loss: 0.0416 - acc: 0.9844 - val_loss: 0.1032 - val_acc: 0.9582\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 3s 18ms/step - loss: 0.0391 - acc: 0.9849 - val_loss: 0.1031 - val_acc: 0.9587\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 2s 13ms/step - loss: 0.0368 - acc: 0.9854 - val_loss: 0.1065 - val_acc: 0.9591\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0351 - acc: 0.9853 - val_loss: 0.1074 - val_acc: 0.9584\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0334 - acc: 0.9863 - val_loss: 0.1112 - val_acc: 0.9574\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0323 - acc: 0.9863 - val_loss: 0.1136 - val_acc: 0.9585\n",
      "377/377 [==============================] - 0s 988us/step\n",
      "377/377 [==============================] - 0s 920us/step\n",
      "1507/1507 [==============================] - 1s 948us/step\n",
      "9/9 [==============================] - 0s 1ms/step\n",
      "377/377 [==============================] - 0s 1ms/step - loss: 0.1046 - acc: 0.9611\n",
      "9/9 [==============================] - 0s 1ms/step - loss: 3.8679 - acc: 0.2509\n",
      "[0.10461772233247757, 0.9610949754714966]\n",
      "[3.8678905963897705, 0.25090909004211426]\n",
      "377/377 [==============================] - 0s 928us/step\n",
      "1507/1507 [==============================] - 1s 918us/step\n",
      "9/9 [==============================] - 0s 1ms/step\n",
      "MLP_Test: ROC AUC=0.994\n",
      "MLP_Training: ROC AUC=0.999\n",
      "MLP_3rd: ROC AUC=0.251\n",
      "[[4540  269]\n",
      " [ 200 7046]]\n",
      "[[19080   569]\n",
      " [  255 28316]]\n",
      "[[ 29 171]\n",
      " [ 35  40]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_MLP(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 38s 239ms/step - loss: 0.3344 - acc: 0.8540 - val_loss: 0.2250 - val_acc: 0.9064\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 37s 246ms/step - loss: 0.1564 - acc: 0.9359 - val_loss: 0.1574 - val_acc: 0.9367\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 36s 238ms/step - loss: 0.1057 - acc: 0.9570 - val_loss: 0.1205 - val_acc: 0.9517\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 37s 242ms/step - loss: 0.0834 - acc: 0.9660 - val_loss: 0.1004 - val_acc: 0.9581\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 34s 227ms/step - loss: 0.0679 - acc: 0.9711 - val_loss: 0.1268 - val_acc: 0.9517\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 35s 230ms/step - loss: 0.0612 - acc: 0.9741 - val_loss: 0.1050 - val_acc: 0.9583\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 35s 229ms/step - loss: 0.0546 - acc: 0.9764 - val_loss: 0.1036 - val_acc: 0.9600\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 40s 263ms/step - loss: 0.0496 - acc: 0.9778 - val_loss: 0.1005 - val_acc: 0.9599\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 41s 272ms/step - loss: 0.0455 - acc: 0.9797 - val_loss: 0.0946 - val_acc: 0.9611\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 39s 258ms/step - loss: 0.0422 - acc: 0.9808 - val_loss: 0.0945 - val_acc: 0.9641\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 38s 252ms/step - loss: 0.0405 - acc: 0.9806 - val_loss: 0.1051 - val_acc: 0.9599\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 39s 257ms/step - loss: 0.0393 - acc: 0.9814 - val_loss: 0.1008 - val_acc: 0.9598\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 34s 228ms/step - loss: 0.0370 - acc: 0.9820 - val_loss: 0.1034 - val_acc: 0.9612\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 34s 225ms/step - loss: 0.0344 - acc: 0.9830 - val_loss: 0.1034 - val_acc: 0.9632\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 35s 231ms/step - loss: 0.0350 - acc: 0.9829 - val_loss: 0.1210 - val_acc: 0.9579\n",
      "377/377 [==============================] - 5s 13ms/step\n",
      "377/377 [==============================] - 5s 13ms/step\n",
      "1507/1507 [==============================] - 20s 13ms/step\n",
      "9/9 [==============================] - 0s 13ms/step\n",
      "377/377 [==============================] - 5s 13ms/step - loss: 0.1126 - acc: 0.9619\n",
      "9/9 [==============================] - 0s 13ms/step - loss: 0.5805 - acc: 0.7782\n",
      "[0.11263934522867203, 0.9619244933128357]\n",
      "[0.5805054903030396, 0.778181791305542]\n",
      "377/377 [==============================] - 5s 13ms/step\n",
      "1507/1507 [==============================] - 20s 13ms/step\n",
      "9/9 [==============================] - 0s 13ms/step\n",
      "RNN_Test: ROC AUC=0.994\n",
      "RNN_Training: ROC AUC=0.998\n",
      "RNN_3rd: ROC AUC=0.688\n",
      "[[4602  207]\n",
      " [ 252 6994]]\n",
      "[[19121   528]\n",
      " [  479 28092]]\n",
      "[[193   7]\n",
      " [ 54  21]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_RNN(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lengths of 10 documents: [41, 43, 42, 42, 67, 67, 68, 68, 69, 69]\n",
      "num of tokens: 1024\n",
      "Train set count: 484\n",
      "Test samples: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n"
     ]
    }
   ],
   "source": [
    "classifier_3rd = preprocessing_3rd(readdata('Flavonoid.csv')[0],readdata('Flavonoid.csv')[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_177\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_177 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " flatten_177 (Flatten)       (None, 6784)              0         \n",
      "                                                                 \n",
      " dense_177 (Dense)           (None, 1)                 6785      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 39,553\n",
      "Trainable params: 39,553\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 2s 13ms/step - loss: 0.2868 - acc: 0.8842 - val_loss: 0.1992 - val_acc: 0.9172\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.1456 - acc: 0.9424 - val_loss: 0.1426 - val_acc: 0.9420\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 2s 12ms/step - loss: 0.1029 - acc: 0.9608 - val_loss: 0.1200 - val_acc: 0.9521\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 2s 13ms/step - loss: 0.0811 - acc: 0.9702 - val_loss: 0.1088 - val_acc: 0.9546\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 2s 12ms/step - loss: 0.0677 - acc: 0.9758 - val_loss: 0.1042 - val_acc: 0.9581\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0592 - acc: 0.9788 - val_loss: 0.1018 - val_acc: 0.9589\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0531 - acc: 0.9804 - val_loss: 0.1036 - val_acc: 0.9575\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0482 - acc: 0.9827 - val_loss: 0.1012 - val_acc: 0.9577\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0446 - acc: 0.9834 - val_loss: 0.1020 - val_acc: 0.9590\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0416 - acc: 0.9844 - val_loss: 0.1032 - val_acc: 0.9582\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0391 - acc: 0.9849 - val_loss: 0.1031 - val_acc: 0.9587\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0368 - acc: 0.9854 - val_loss: 0.1065 - val_acc: 0.9591\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 2s 12ms/step - loss: 0.0351 - acc: 0.9853 - val_loss: 0.1074 - val_acc: 0.9584\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 2s 12ms/step - loss: 0.0334 - acc: 0.9863 - val_loss: 0.1112 - val_acc: 0.9574\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0323 - acc: 0.9863 - val_loss: 0.1136 - val_acc: 0.9585\n",
      "377/377 [==============================] - 0s 989us/step\n",
      "377/377 [==============================] - 0s 1ms/step\n",
      "1507/1507 [==============================] - 2s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "377/377 [==============================] - 0s 1ms/step - loss: 0.1046 - acc: 0.9611\n",
      "16/16 [==============================] - 0s 1ms/step - loss: 1.8366 - acc: 0.4711\n",
      "[0.10461772233247757, 0.9610949754714966]\n",
      "[1.836582899093628, 0.4710743725299835]\n",
      "377/377 [==============================] - 0s 1ms/step\n",
      "1507/1507 [==============================] - 1s 911us/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "MLP_Test: ROC AUC=0.994\n",
      "MLP_Training: ROC AUC=0.999\n",
      "MLP_3rd: ROC AUC=0.700\n",
      "[[4540  269]\n",
      " [ 200 7046]]\n",
      "[[19080   569]\n",
      " [  255 28316]]\n",
      "[[ 67  13]\n",
      " [243 161]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_MLP(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_527\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_527 (Embedding)   (None, None, 212)         217088    \n",
      "                                                                 \n",
      " global_max_pooling1d_160 (G  (None, 212)              0         \n",
      " lobalMaxPooling1D)                                              \n",
      "                                                                 \n",
      " dense_687 (Dense)           (None, 5)                 1065      \n",
      "                                                                 \n",
      " dense_688 (Dense)           (None, 1)                 6         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 218,159\n",
      "Trainable params: 218,159\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 12s 78ms/step - loss: 0.3698 - acc: 0.8642 - val_loss: 0.2373 - val_acc: 0.9082\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 11s 73ms/step - loss: 0.1758 - acc: 0.9321 - val_loss: 0.1657 - val_acc: 0.9329\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 11s 75ms/step - loss: 0.1227 - acc: 0.9549 - val_loss: 0.1308 - val_acc: 0.9489\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 11s 73ms/step - loss: 0.0968 - acc: 0.9640 - val_loss: 0.1155 - val_acc: 0.9538\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 11s 75ms/step - loss: 0.0802 - acc: 0.9715 - val_loss: 0.1056 - val_acc: 0.9586\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 11s 75ms/step - loss: 0.0690 - acc: 0.9757 - val_loss: 0.0982 - val_acc: 0.9603\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 11s 74ms/step - loss: 0.0611 - acc: 0.9778 - val_loss: 0.0995 - val_acc: 0.9612\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 11s 75ms/step - loss: 0.0553 - acc: 0.9791 - val_loss: 0.0951 - val_acc: 0.9615\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 11s 72ms/step - loss: 0.0502 - acc: 0.9811 - val_loss: 0.0903 - val_acc: 0.9626\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 11s 73ms/step - loss: 0.0463 - acc: 0.9822 - val_loss: 0.0884 - val_acc: 0.9638\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 13s 84ms/step - loss: 0.0433 - acc: 0.9827 - val_loss: 0.0876 - val_acc: 0.9641\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 12s 77ms/step - loss: 0.0402 - acc: 0.9834 - val_loss: 0.0858 - val_acc: 0.9642\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 11s 74ms/step - loss: 0.0383 - acc: 0.9841 - val_loss: 0.0893 - val_acc: 0.9633\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 11s 72ms/step - loss: 0.0360 - acc: 0.9844 - val_loss: 0.0867 - val_acc: 0.9646\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 11s 70ms/step - loss: 0.0349 - acc: 0.9850 - val_loss: 0.0884 - val_acc: 0.9658\n",
      "377/377 [==============================] - 1s 3ms/step\n",
      "377/377 [==============================] - 1s 3ms/step\n",
      "1507/1507 [==============================] - 5s 3ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "377/377 [==============================] - 1s 3ms/step - loss: 0.0796 - acc: 0.9679\n",
      "16/16 [==============================] - 0s 3ms/step - loss: 0.3032 - acc: 0.8967\n",
      "[0.07958674430847168, 0.9678971171379089]\n",
      "[0.30316346883773804, 0.8966942429542542]\n",
      "377/377 [==============================] - 1s 3ms/step\n",
      "1507/1507 [==============================] - 5s 3ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "CNN_Test: ROC AUC=0.996\n",
      "CNN_Training: ROC AUC=0.999\n",
      "CNN_3rd: ROC AUC=0.947\n",
      "[[4597  212]\n",
      " [ 175 7071]]\n",
      "[[19122   527]\n",
      " [  288 28283]]\n",
      "[[ 71   9]\n",
      " [ 41 363]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_CNN(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 32s 202ms/step - loss: 0.3344 - acc: 0.8540 - val_loss: 0.2250 - val_acc: 0.9064\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 29s 195ms/step - loss: 0.1564 - acc: 0.9359 - val_loss: 0.1574 - val_acc: 0.9367\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 30s 198ms/step - loss: 0.1057 - acc: 0.9570 - val_loss: 0.1205 - val_acc: 0.9517\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 30s 196ms/step - loss: 0.0834 - acc: 0.9660 - val_loss: 0.1004 - val_acc: 0.9581\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 29s 195ms/step - loss: 0.0679 - acc: 0.9711 - val_loss: 0.1268 - val_acc: 0.9517\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 29s 191ms/step - loss: 0.0612 - acc: 0.9741 - val_loss: 0.1050 - val_acc: 0.9583\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 29s 194ms/step - loss: 0.0546 - acc: 0.9764 - val_loss: 0.1036 - val_acc: 0.9600\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 29s 195ms/step - loss: 0.0496 - acc: 0.9778 - val_loss: 0.1005 - val_acc: 0.9599\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 30s 196ms/step - loss: 0.0455 - acc: 0.9797 - val_loss: 0.0946 - val_acc: 0.9611\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 30s 200ms/step - loss: 0.0422 - acc: 0.9808 - val_loss: 0.0945 - val_acc: 0.9641\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 32s 209ms/step - loss: 0.0405 - acc: 0.9806 - val_loss: 0.1051 - val_acc: 0.9599\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 31s 208ms/step - loss: 0.0393 - acc: 0.9814 - val_loss: 0.1008 - val_acc: 0.9598\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 32s 210ms/step - loss: 0.0370 - acc: 0.9820 - val_loss: 0.1034 - val_acc: 0.9612\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 37s 243ms/step - loss: 0.0344 - acc: 0.9830 - val_loss: 0.1034 - val_acc: 0.9632\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 34s 225ms/step - loss: 0.0350 - acc: 0.9829 - val_loss: 0.1210 - val_acc: 0.9579\n",
      "377/377 [==============================] - 5s 13ms/step\n",
      "377/377 [==============================] - 5s 14ms/step\n",
      "1507/1507 [==============================] - 20s 13ms/step\n",
      "16/16 [==============================] - 0s 13ms/step\n",
      "377/377 [==============================] - 5s 13ms/step - loss: 0.1126 - acc: 0.9619\n",
      "16/16 [==============================] - 0s 13ms/step - loss: 1.2600 - acc: 0.4607\n",
      "[0.11263934522867203, 0.9619244933128357]\n",
      "[1.2599860429763794, 0.46074381470680237]\n",
      "377/377 [==============================] - 5s 13ms/step\n",
      "1507/1507 [==============================] - 20s 13ms/step\n",
      "16/16 [==============================] - 0s 13ms/step\n",
      "RNN_Test: ROC AUC=0.994\n",
      "RNN_Training: ROC AUC=0.998\n",
      "RNN_3rd: ROC AUC=0.739\n",
      "[[4602  207]\n",
      " [ 252 6994]]\n",
      "[[19121   528]\n",
      " [  479 28092]]\n",
      "[[ 62  18]\n",
      " [243 161]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_RNN(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lengths of 10 documents: [58, 55, 62, 59, 60, 58, 62, 59, 58, 62]\n",
      "num of tokens: 1024\n",
      "Train set count: 1074\n",
      "Test samples: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]\n"
     ]
    }
   ],
   "source": [
    "classifier_3rd = preprocessing_3rd(readdata('Coumarin.csv')[0],readdata('Coumarin.csv')[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_177\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_177 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " flatten_177 (Flatten)       (None, 6784)              0         \n",
      "                                                                 \n",
      " dense_177 (Dense)           (None, 1)                 6785      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 39,553\n",
      "Trainable params: 39,553\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.2868 - acc: 0.8842 - val_loss: 0.1992 - val_acc: 0.9172\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.1456 - acc: 0.9424 - val_loss: 0.1426 - val_acc: 0.9420\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.1029 - acc: 0.9608 - val_loss: 0.1200 - val_acc: 0.9521\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 2s 12ms/step - loss: 0.0811 - acc: 0.9702 - val_loss: 0.1088 - val_acc: 0.9546\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 2s 13ms/step - loss: 0.0677 - acc: 0.9758 - val_loss: 0.1042 - val_acc: 0.9581\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 2s 13ms/step - loss: 0.0592 - acc: 0.9788 - val_loss: 0.1018 - val_acc: 0.9589\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0531 - acc: 0.9804 - val_loss: 0.1036 - val_acc: 0.9575\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0482 - acc: 0.9827 - val_loss: 0.1012 - val_acc: 0.9577\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0446 - acc: 0.9834 - val_loss: 0.1020 - val_acc: 0.9590\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0416 - acc: 0.9844 - val_loss: 0.1032 - val_acc: 0.9582\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 1s 10ms/step - loss: 0.0391 - acc: 0.9849 - val_loss: 0.1031 - val_acc: 0.9587\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 1s 9ms/step - loss: 0.0368 - acc: 0.9854 - val_loss: 0.1065 - val_acc: 0.9591\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 2s 11ms/step - loss: 0.0351 - acc: 0.9853 - val_loss: 0.1074 - val_acc: 0.9584\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 2s 12ms/step - loss: 0.0334 - acc: 0.9863 - val_loss: 0.1112 - val_acc: 0.9574\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 2s 10ms/step - loss: 0.0323 - acc: 0.9863 - val_loss: 0.1136 - val_acc: 0.9585\n",
      "377/377 [==============================] - 0s 797us/step\n",
      "377/377 [==============================] - 0s 874us/step\n",
      "1507/1507 [==============================] - 1s 862us/step\n",
      "34/34 [==============================] - 0s 885us/step\n",
      "377/377 [==============================] - 0s 1ms/step - loss: 0.1046 - acc: 0.9611\n",
      "34/34 [==============================] - 0s 1ms/step - loss: 1.5843 - acc: 0.5149\n",
      "[0.10461772233247757, 0.9610949754714966]\n",
      "[1.5842934846878052, 0.5148975849151611]\n",
      "377/377 [==============================] - 0s 995us/step\n",
      "1507/1507 [==============================] - 1s 887us/step\n",
      "34/34 [==============================] - 0s 907us/step\n",
      "MLP_Test: ROC AUC=0.994\n",
      "MLP_Training: ROC AUC=0.999\n",
      "MLP_3rd: ROC AUC=0.598\n",
      "[[4540  269]\n",
      " [ 200 7046]]\n",
      "[[19080   569]\n",
      " [  255 28316]]\n",
      "[[525 503]\n",
      " [ 18  28]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAnYAAAHWCAYAAAD6oMSKAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAAClkElEQVR4nOzdd1zV9ffA8dfnXvZURBygOBD3CjVHztyjYamlFY5SyzSzaf2+ldY3bWhmpmmhthyVWllmWV81Zy6c5Bb14kYB2XDv5/fH9V7WBe6FC5dxno/HfQDv+xnnXhEO73HeiqqqKkIIIYQQotzTODoAIYQQQghhH5LYCSGEEEJUEJLYCSGEEEJUEJLYCSGEEEJUEJLYCSGEEEJUEJLYCSGEEEJUEJLYCSGEEEJUEJLYCSGEEEJUEE6ODqC0GQwGLl26hLe3N4qiODocIYQQQogCqarK7du3qV27NhpNwX1ylS6xu3TpEnXq1HF0GEIIIYQQNrl48SJBQUEFHlPpEjtvb2/A+Ob4+Pg4OBohhBBCiIIlJCRQp04dcw5TkEqX2JmGX318fCSxE0IIIUS5Yc0UMlk8IYQQQghRQUhiJ4QQQghRQUhiJ4QQQghRQUhiJ4QQQghRQUhiJ4QQQghRQUhiJ4QQQghRQUhiJ4QQQghRQUhiJ4QQQghRQUhiJ4QQQghRQUhiJ4QQQghRQUhiJ4QQQghRQTg0sfv7778ZMmQItWvXRlEUfvzxx0LP2bp1K2FhYbi5udGgQQM+++yzkg9UCCGEEKIccGhil5SUROvWrVmwYIFVx587d46BAwfStWtXIiMjee2115gyZQpr1qwp4UiFECKLLkHH5nOb0SXoSuV+e2P2MnfXXPbG7JVrF1d8DJz72/hRVAw6HWzebPxYDEX9fr289xciF0zi8t5finV/e1FUVVUdHQSAoiisW7eOBx54IN9jXnnlFX7++Wf+/fdfc9vEiRM5dOgQu3btsuo+CQkJ+Pr6Eh8fj4+PT3HDLhOmbpzK5/s+J12fjpuzGxmZGRhUA1Xcq5CmT0NRFBpXa0ynOp3wcfVhy7kt1PapTWi1UA5dOcSDTR/k7K2zfHPoG0KqhdA/pD87Luwg05BJ/ar18XH14dCVQ/h7+HMj+Qata7amilsV/Nz9OHjlIID5GNNzXet2pX1ge/bG7GXbhW1ciL/A/87+jypuVbir9l2MajkKgG0XttG1blfWn1zP2qi1DG02lJk9Z+Z4fXtj9vLtkW8BGNVyFO0D2+f7XuyN2cucrW9zKe48Y9uE09zdn23Hf6RrjTa0d/Njb+pNtl09SNcmD9C+enP2/vUm226exq9BNw66egDQpmYb8+tqU7MNN1Numl8Puv1wYRfU7cRexcC2wyvpiob2LR+FoDDrY9btZ++RlXybeAm8A4zHqBrztQkKy3Gv7NfO/XpN72FB70uB8ruPbj8c+d74ecth7FUMWa+pxl20T0nIN7bixrU38ivjv1uTB2jf9gljOAk6dl7YCQp0rtOZIJ8gy+faeu/cr7+g9+PCLiJSrjJ+x7sYVAMaRcOSwUsYd9c4m1+jtUb/OJovD31p/jq8dTjLH1heqa9dZAe+gvXPgWoARQNDPoa7nnBsTKJ4IiJg/HgwGECjgU8+gfBwmy8zfv14vj2ywvz1qJYjWTJkSaHn7ZjWm46f70argl6Bnc/cQ9cF22y+f2FsyV3KVWLXrVs32rZty8cff2xuW7duHcOHDyc5ORlnZ+c856SlpZGWlmb+OiEhgTp16pSbxE6XoKPBRw3IIMPRodgs0DuQmNu2/1XshsLXfq1RA5rwqW47WxNz/hXW1bcB42vehRp3EdU3CENGCmpqPF/EnWJnyrWsA03f2Yrxc3/gRravqwGx2b5GKTiuu9z8GZ5yGxX4ngwOKAbzfdqg4aGaYdB0MGv+XcvBq4dynNs6oCX3NxqEqqpw6nd+vHaII6g57tlCVRjEne9hVx9IS8h60q8h1Lsn24uCX6O3cPTW6azzq4YwqG7XXC8eyPFfXM3xFBd3w63orK+r1oOg9nBxD8Sdz7oXGRxVsl8TWnAnXt8gqNUm69jL+ziacCErLp86DKwZlnVifu7EueHcXxzNSDL/u7Rw9iTIL4SN13K+p/0DWtHKp26O6264epCj2b7nWngFMrBG67z3Nn16LQpuX85qd3KDzNSsr70CwD8UbpxETbzGbVQ+IyPP98rE2p3wdnLL/7UVyvL7cjktnm+uRuZpf6xGW2q5+hZyyYJ/tF9OS+Cba3mvPSqgDbVci/ez8XJaAt9eO2jdta3+DVT4gaqFz8wy0+Fy3pio1RqcXK24fUn8qrTiNanWHVcS9y7KodZdzpYL5n+s7610/jP9IJoykcUYZSpw/Z/11Go/2K7XrbCJXWhoKKNHj+a1114zt+3cuZMuXbpw6dIlatWqleect956ixkzZuRpL8uJ3bifxrH04FJHhyGEEEKUWT3OweYvCz+utB1c8CxtJn1i12vaktg52fXOpUBRcv6pbMpLc7ebTJ8+nWnTppm/NvXYlUUdFndg75UyNBfFgdxUqIXCOcXy3x2hqkI9NCgYO0+iMXA8n2Ptqb1qnJa619Rbl03HO8/ttvAcQBdVQxu0HETPjnyO6a5qCMvvv6VGCxpjj94+Qxp/q2l5DumuuBHm5J6t5c7/izz/PxTQp0Fm3mugcQJDpvnLfWTyd2HxunqDixf70hL4O+NWnuN6uFQlzLVKjpDyxAPsS7rKVn2SxXtZMtwzkDpOxiH0fWm32Jp6I++93fwJc6uW996p8ZB8s/CbuHhBeiIAJ9Dzi6LPc8hgr2CauFXJ2VhID7A1B17OSOLbuFN52kdVaUQtZ09rb2DTtR+rEkotl9zXtvrFGK+dnsQ3cScsXLvxnWtbeT3bbnvnlHxOSomD+At526vUBfcq9rl5gYr+mvN9Tfa8r9WXs8/1lAK+spaPexoGZX+OHjuDAnNntiGhqhW9sHfuezk9kRU3juZ5dqR/C2q5eOVpN6hwJSEFp+vxfPpxNNps989UoEaHfra+FLsqV4ldzZo1uXLlSo62a9eu4eTkRLVq1Sye4+rqiqurNf/AjjN311xe+OMFR4dhf1YMb+Z33ks4MwRnOqjJea+hwje40z7bt+9eMi0fW1A8tsanwqcYh9vy3EuF+bjBgA/psPEZi6d/NHoz7dGyd3nffF/XB7jleF05jP2fec7X3pi9dPiiQ55DPhj3t/Vz2nT74YteeduHLISfsl5Dvu9t9ngf/x2CwvKN6/0nfrcqrr2RX9Hhp/A8762i5B2Q0Spa5ozfbZ5rl++9H9tg+d75vf7cBsw3vx86DPyqJqJmi09BYdFT2/Od81dcTiU4V60kr60ta3Ps4mPgo+bk/E5SYMwW8A10UFCi2LwWwqRJxs+1WjSLF/PiONvnvGZY+H793ML36/nYJKasjORQQjy4wsYxb9Nv2T84qcakbtcz99DVzsOwtipXdew6derEpk2bcrT98ccftGvXzuL8urJub8xelBmK45O63L8x8/s6v4/5PNcQxfKxpq9Vy597ATPvJG7hOOW5TzhOeZKf/I4tKJ4C47Mg3C+U9nfuHZ79/qaYWj9B+7ufJrx13om74a3DaV+vG9TrQvvWT+SNlVyvq2r9nE+2HpljIn/7wPZ57hPeOty2hQpBYcbr5r5P21E52vO8Xsj575AttuLG1b7tE4T7heb4dwn3C+XzIV/k6LXQKBoWD16cI5my+d6WXr+l9z3b+xGEhs9xM//g1KDh8yGfl1hSB7D8geXseXIPH/X7iD1P7rFrclRer10kvoFw33zI9q/HffMlqSvvsi+UiIqCIiR1YN3364+RMQyav51Dunh83Z1Z/HgYgyJ2c/2f9Rxc8CzX/1lfIgsnbOXQOXaJiYmcPm2c/N22bVvmzp1Lz5498fPzo27dukyfPp2YmBi++uorwFjupEWLFkyYMIGnnnqKXbt2MXHiRFauXMlDDz1k1T3LyqpY93fcSdWnFn6gjVwVY++kaVVsuj4dFGhcrTFd6nTBSzXw96EV1E69RQgKRzBwP06cxcC3ZNAQDf1xYhd6MjDQACe8gCMY8ANuAi3R4IeGKsBhjEN0pmNMz3VBS3uc2EsmO9ATjZ7NZOILtMOZR+8sEtiBni5oWU8G68jkQZyYiXuO17SXTFaSAS4+PJqekn+P1p1jPyKNGFTG4EzzKiHsqNGYLjVa0969GntTYtlx9RBdmtxvXhW749YZqtTvymE3L1ChVc1WHL562Px5XGocXep0yVoVe3E31OnIXsXAjiMrja+1xSN5VsWuPLoSVHi05aOWV8UeXcXK2zHgXdN4jKoxX9u8OjP715Zeb8xedlzckRVfUeR3H91+OPo9qAq0fJi9iiHrNdVoQ/vU2/nGVty49kZ+xY7jPxn/nbKtit110bj6vVOdTgWuirXp3rlff0Hvx512nU8NTt88TYhfSIkmdaIExMfAzbPg10CSuoogKQm87gyXJiaCZ/GmKFi8RVom//npKGsPGBdmdajnx7xH2lC7inshZ9pPuVk8sWXLFnr27JmnPTw8nOXLlzN69Giio6PZsmWL+bmtW7fy/PPPc+zYMWrXrs0rr7zCxIkTrb6noxO7rhFd2a7bbvfr+rv5E/l0ZN5fMvEx8NMUOLcV1DK6sta1CvR/F6o3MSYSV/4FQwbU6wLpt82JRY5fuilxcPUo1GhhPCbhzmrYm2chQWf8gd1jOjTu78hXJoQQoiSVQmI3ZtkeNp+4jkaB5+4N5dleIWg19p6DWbByk9g5giMTO2VG0b8RnHEm/c106w7euQC2fgBp8ZTMMvlCVKkH3V+G47+AezVIiTV+vBwJ6Sng4gYpt8GnFnSdJsmXEEIURqeDU6egUSMIkl5is+yJ3YkTEBpq91sc0cXz7MoDfPBwazrU97P79a1RoVfFlldFSerq+dTj3PPnrD/h2xFwaqPN9ykSzxrGGe0BTcEnCM5uBk//nL1kbUeVTixCCFGR2akIb4W0JFsR4aZNjV8XcZ6dyY3ENCIvxNGnWQ0AWgb58te07jhpy8eyBOmxKwXub7uTarB+Pp2Xkxe3X79d+IE7F8DfcyA1Acgs9HDbKdDtJeNQ55V/jSUf6t8DHcbL3BQhhCgNOh0EBxuTOlE4rRaio4vcq7n91A2e/+4g8ckZrH2mMy0CCykEXkqkx64M2Ruz1+qkzgknMt60Yh7czgXwx+vFjMyCbi/DlcMQp4M2j0LnZ+1/DyGEENY7dUqSOlvo9XD6tM2JXYbewJw/TrL47zOoKoTW8MLFqXz00OUmiV0Js1RXyxL1TSs7TmfVybnVVHG5+hrnw0kSJ4QQZU+jRsbh1+zJnVZrLO0RWMlHTmJijMOvud+bkBCbLnMhNpnJqyI5dDEOgFF31+X/BjXD3UVrx2BLjyR2JeiXk78Uekz7mu3ZM2GPdRf8b23IsL4qv0WKM9RqBQM/yLeEhhBCiDIiKMg4py5bEV4WLy6RRQLlTmiocU7dhAnGnjrTe2NDb93Phy7x2tojJKZl4uPmxPsPt6J/i7zbk5YnMseuBFWdVZW49Lh8n7e6lw7grWKM87tXg/bjoFcJDN8KIYQoWaWw8rNc0+mMw68hITYPwS7acob3Nh6nfb2qzHukLYGlWJvOFjLHrowoKKnz0HpYf6GiJHXOntDzNRliFUKIiqSyD79aEhRkU0KXqTeYV7hO6NYAfy8XHmwbWG5WvRZGEjsHSfo/K4dUbUnqqgRD8wdl1aoQQpSm0qwxFxMjPXZFpKoqy3ZEs+aAjh8mdsbdRYtGozCsXR1Hh2ZXktiVZdYmdXc/DQNml2wsQggh8iqNGnMlUKutsolNTOOlHw7zv+PGXYp+OKDj8Y7BDo6qZMgcuxLS6ONGnI47ne/zhc6ve8sfsKL0yVvxtgUmhBDCPhxVY66Ytdoqm52nbzB19UGu3U7DxUnDfwY15bGOwShK6W4LVhwyx64MKCipK3R+3Ts1kKROCCHKOEfVmCtirbbKJkNv4KNNJ1m01VibLiTAiwUj29KkZunvE1+aJLFzgLfvfTv/J+c2h8xCChq7eMFrMfYNSgghhG1Ko8acnWq1VUbvbviXZTuiAXi0Q13eGFx+a9PZQhI7B5jWaZrlJ05shARdwSc3exCGL7d7TEIIIWxUGjXm7FCrrbKa0K0hm49f4+X+TRjYsnzXprOFzLErIZoZGlTyvrUKCoY38+m6L2yxRJ2OMO53O0QnhBDCLkqrxlwxarVVFsnpmWyKusr9bbJ6S7OXNinPZI5dGWApqSuo3aoVsJLUCSFE2VWSNeZsrNVW2Ry7FM/klZGcvZ6Eu7OWvs1rAlSIpM5Wle8VlxJNPm+txfaZAYVfUBZKCCGE4+h0sHmz8WN+YmTuc2kz1qY7x4Of7uTs9SRq+Lji4+7s6LAcShK7EmLA8nBrnvadC8CQVvDF+v7XTlEJIYSwWUSEsaxJr17GjwsXGodgk5Ly1piLiHBcnJXMzaR0nvpqHzPWR5GuN9C7aQC/PdeNjg2qOTo0h5I5diVEmZF/fZwcNewKG4J19YHpF+0UlRBCCJvYWqtOasyVit1nY3luVSRXE9Jw0Wp4bWATwjvXK1e16Wwhc+wcTFfAylZnsnURv1vINiauvjD9gp2iEkIIYTNba9VJjblSEZecztWENBpW9+STR++iWe2KXZvOFpLYlYDxP4/P97mWNVsaPzmxEdITCr6QJHVCCOFYBdWqA6kxV4qyr3Dt36IW80a0oW/zGni4SCqTncyxKwGbozfn+9yMnjOMn3xXyF6Cj662Y0RCCCGKxFSrziR7rTpTjTmtNudz0ltnd78evkzvuVu5mpBVwP+BtoGS1FkgiV0JyNDnvx3Y4NDBxt46fQG7S3jXhsb9SyAyIYQQNgvP9od4VBSMG5f19bhxxjl1mzcbP2Z/ThRbSrqe6WsPM2nFAaJjk1m89ayjQyrzJNUtAfmtiFW4M6lzy7sFX+CFf+0ckRBCCLuwVKtOasyViH8vJzB5ZSSnryWiKPBMj4ZM7V1CBaArEEnsSoCL4kKamreEiYviYvzk8qH8T7776RKKSgghRA46nXFxRKNG1idmMTElt7uEAIy16b7efZ53fv2X9EwDAd6uzBvRhs4h/o4OrVyQodgSUNe3rsX24CrBxmHYggyYXQIRCSGEyKGg2nS5H1KrrlR9s/s8b/x0jPRMA72aBPDbc10lqbOB1LErAXXm1EGXmLfkSZBXEBddakLsScsnhvaHkbJoQgghSpSttelyk1p1JSo5PZOHFu1iWFgQY7pU3Np0tpA6dg5mKakzt6sFlDjp9nIJRSSEEMLM1tp0uUmtOrvK1BtYFxnDQ3cFodEoeLg4sf7ZLpVyn1d7kMSurFCcICjM0VEIIUTFV1BtutyLI2JipFZdCYqJS+G5lZHsO3+Lm0npTOjeEECSumKQd64E+Ln5WW4v6O32lb/8hBCiVBRUm87TM+dDatWVmI1HLzNg3t/sO38LL1cnaldxd3RIFYL02JUATT4JnEYtoOu/w1MlFI0QQog8wsNh0iTj51FRBa90HTcO+vUzDr+GhEhSV0ypGXre/iWKb/8x7q7Uuk4VPnmkLXWreTg4sopBErsS4ObkZrm9oJM6P1sisQghhCiEpdp0uUmtOrs4dfU2k1Yc4OTVRAAmdm/IC31DcZahV7uRd7IE1PKuZbE93x8dNduUVChCCCEsiYmx/LkoUWmZBs7dSKK6tytfj+vAqwOaSFJnZ/JuloDjN45bbI/K74Se00ssFiGEELlERECTJllfN2kitelKUKY+axpSi0BfFoy8i9+e60rXRtUdGFXFJYldCbidcdtye34nyL6wQghROnQ6eOopyF7CVVVhwgTjc8Ku9py7Se+5WzkaE29u69e8Jv5erg6MqmKTxM7RandwdARCCFF5nDqVM6kzMdWmE3ahN6jM+/MkjyzZRXRsMnP+OOHokCoNWTzhaCOWOzoCIYSoPBo1AkXJm9xJbTq7uRSXwtTVB9lz7iYAD90VxIz7mzs4qspDeuxKwD1B9+RtVCFPq4sX+FqxGksIIYR9BAXBggU52zQaqU1nJxuPXmHAx9vYc+4mni5a5o1ow5zhrfFylX6k0iKJXQnoVq+bxfa55KrREzq4FKIRQgiRQ3h41udffQXnzxtr1Yli2XbqOhO/2U98Sgatgnz5dUpXHmgrnRelTVJoO9Ml6Hh3+7t5n1DgRTWFrXhntfV5o/QCE0IIkdfQocYdJkSxdWnoT7fQ6jSp6c2LfRvj4iR9R44giZ2dvfbXa/k+d5Jsczpq3SXDsEIIUZp0OuPiiewFiWNiCt51QuRLVVV+PnSJPs1q4OHihEajsDS8nezz6mDy7tvZlnNbLD+hwojsebSTLPUWQohSExEBwcHQq1fOGnZNm0oNuyKIT85g0ooDPLfqIDPXZ1VplaTO8eRfwM5qeta02K4A87LPsbtxqnQCEkKIyk6ng/HjwXCnUG72FbEGg9Sws9G+6JsMnL+NDUeu4KRRaFDdE9VSCRnhEDIUa2eKRrHY3j7PgZaPE0IIYWenTmUldZaYatjJqtgC6Q0qCzefZt5fp9AbVIKreTD/kba0rlPF0aGJbCSxs7OL8Rcttuf5W7DdmBKPRQghBMbadRpN/smd1LAr1NWEVJ5bFcnus8badA+0qc3bD7TA283ZwZGJ3GQo1s6CvC3/xVcn+xdO7tDr9VKJRwghKr2gIPjkk6yvFcWY6IExqZMadlY5eTURDxctc4e3Zt4jbSWpK6Okx87OfN19LbeTbeh10JxSikYIIQRgrF03aZLx8+PHwcPDOPwaEiJJXT4y9QbzYogaPm4sHHUXNXzcqO8v5WHKMumxs7PouGjL7dlLnUR+WzrBCCGEyCsw0JjM9eghSV0+Tl+7zZAFO9h49Iq5rWODapLUlQOS2NmZolpeFJFjvVBmcqnEIoQQFZJOB5s3F30la0yMfeOpQFRVZfXeCwz5ZAf/Xk7g/d+PozfIitfyRBI7O6viXsViu1/2L/T60ghFCCEqnuz16IKDYeFCSEoq/LFkSdY1pHadRfEpGUxeGckra46QkqHnnhB/Vj3VEW0+1R5E2SRz7OzMzcnNcnv2OXb+svpKCCFslrsencFgnDdnmjtnLVPtun79ZCj2jv3nb/Hcqkh0t1Jw0ii82K8x47s2QCNJXbkjPXZ21qN+j7yNKnTLnkO3GlFa4QghRMVRWD06W5hq1wnOxyYxYvEudLdSqOPnzvcTOzGxe0NJ6sop6bGzs4S0hLyNCiRmn6LgIpNPhRDCZpbq0Wm1EBWVc//X3GJijMOvuc+T2nUABFfzZOTddYlLzuCdB1vgI2VMyjXpsbOzXRd3WWzfQcadzxTwa1B6AQkhREWRux6dqQZdaCh4eub/CA01zrHTanOeV4mHYf93/CpX4lPNX78xuBkfP9JGkroKQBI7Oyt054kaLcC3gL8shRBC5C88POvzqCgYN86688aNg+ho42ra6Gjrz6tg0jL1vPXzMcYu38fU1ZHmFa9OWg2KbHVZIchQrJ25OrlabHcxfdLx6VKLRQghKrSChl8tCQqq1L10Z64nMnlFJFGXjVOGmtbyQW9QZdVrBSOJnZ2lZaZZbE8v5TiEEKJC0emMiyeyJ3MxMcZhVlEgVVX5fr+ON386RkqGHj9PFz54uBX3Nq3h6NBECZChWDtTsVzI0Txld/dnpRaLEEJUCNlr1zVpktUu9egKlZiWyZRVB3n5h8OkZOjp3LAavz3XVZK6CkwSOzu7mXzTcrv5Mzst1RdCiMogd+06Ndsfz6Z6dEXdgaIS0CoK/15OQKtReLl/Y74edzc1fCzXWxUVgwzF2pmqWu6xM7d2fKbUYhFCiHKvsNp1pnp0lXjuXG6GOwsiNBoFdxctC0a2JTldz111qzo4MlEapMfOztrUbGOx/S4AFy9oO6o0wxFCiPLNVLsuP1KPLodrCak8sXQPi/8+a25rUtNHkrpKRBI7O+sf2j9vowp9cYb0RNDtL/2ghBCivMpdu05RshI9qUeXw+bj1+j/8Ta2n77Bwi2niU/OKPwkUeFIYmdnP5/4OW+jAr+Qafz84u7SDUgIIcq77LXrjh+H8+crfT267NIy9bz9SxRjlu/lZlI6TWv5sO6ZLvh6SLHhykjm2NlZclqyxfZE0yy7Oh1LMRohhKhgAgONu0lILx0AZ68nMmVVJEdjjLXpRneux6sDmuDmrHVwZMJRHN5jt3DhQurXr4+bmxthYWFs27atwOO//fZbWrdujYeHB7Vq1WLMmDHExsaWUrSFC64SbLG9PgpUrQ9BYaUckRBClAE6nbGXrbgrWGNi7BNPBZCYlsnQRTs5GpNAVQ9nvniiHW/d11ySukrOoYnd6tWrmTp1Kq+//jqRkZF07dqVAQMGcOHCBYvHb9++nSeeeIJx48Zx7Ngxvv/+e/bu3cuTTz5ZypHn71LiJYvtOlS4dU7m2AkhKp/sdeiCg2HhQkhKsv6xZEnWtaR2nZmXqxNTejWiYwM/fnuuG72bSW06AYqaX32OUnD33Xdz1113sWjRInNb06ZNeeCBB5g1a1ae4z/88EMWLVrEmTNnzG2ffPIJ77//PhcvWt6jNbeEhAR8fX2Jj4/Hx8en+C8iF893PEnW5x2O9VAhCR/o9y50mmT3+wohRJmk0xmTuYJKlthKqzXOr6uEw7GHLsah1Si0CPQFjCW2DCqyLVgFZ0vu4rAeu/T0dPbv30/fvn1ztPft25edO3daPKdz587odDo2bNiAqqpcvXqVH374gUGDBuV7n7S0NBISEnI8SpKns6fldtMnfg1L9P5CCFGmFFaHrihMtesqEYNBZfHWMzy0aCfPrjhAYppxQZ6iKJLUiRwcltjduHEDvV5PjRo5u45r1KjBlStXLJ7TuXNnvv32W0aMGIGLiws1a9akSpUqfJJ9KXwus2bNwtfX1/yoU6eOXV9Hbq1rtrbY3pY7//FcLCd+QghRIVmqQ6fVwokTkJhY+OPECcvnV6LadddupxK+bA+zfjtOpkGleW1fDI4bbBNlnMMXTyhKzr80VFXN02YSFRXFlClTeOONN9i/fz8bN27k3LlzTJw4Md/rT58+nfj4ePPD2iHboipwjh2AX4MSvb8QQpQpuevQmWrPhYYaV7cW9ggNNc6x02pznl9JhmG3nrzOwI+3se3UDdycNcwe2pIFI9vi4yalTIRlDit34u/vj1arzdM7d+3atTy9eCazZs2iS5cuvPTSSwC0atUKT09PunbtyjvvvEOtWrXynOPq6oqrq6v9X0A+LiZYThx1AAEtwDew1GIRQogyITwcJt2ZWxwVZUzWbDFuHPTrZxx+DQmpFEldht7AB7+fYMmdHSSa1PRmwci2hAR4OzgyUdY5rMfOxcWFsLAwNm3alKN906ZNdO7c2eI5ycnJaHJ1yWvv/BXnwDUgOVRzq2ax3Q9A1ZdqLEIIUeYEFvGP26Ag6NGjUiR1AFpF4fiV2wA80SmYHyd1kaROWMWhBYqnTZvG448/Trt27ejUqRNLlizhwoUL5qHV6dOnExMTw1dffQXAkCFDeOqpp1i0aBH9+vXj8uXLTJ06lQ4dOlC7dm1HvhSzOr51iE6IztMejAKB7Uo/ICGEKCk6nXFxRKNG1idcMTG299hVIpl6A05aDRqNwpxhrTl4MY4+UsZE2MChid2IESOIjY1l5syZXL58mRYtWrBhwwaCg41Ffi9fvpyjpt3o0aO5ffs2CxYs4IUXXqBKlSr06tWL9957z1EvIY8e9Xuw7WKuIssqdMMJXOWvLSFEBRERAePHG1e8ajTGeXTZt/7KLncduiVLZCuwXBLTMnnjx6M4azW893ArAKp7u0pSJ2zm0Dp2jlDSdewe+eERVh9bnaf9UVXLio4vQv/Zdr+nEEKUquLWpqvEdegsOayLY8rKSKJjk9Eo8Mfz3WTYVeRgS+4ie8Xa2d/Rf1ts34oearQs5WiEEKIEFLc2nakOXSVP7AwGlYjt53j/9+Nk6FVq+7rx8aOyQEIUjyR2dhabbHnf2liA1LjSDEUIIUqGqTZd9uROqzWueM29OCImxjj8mvvYSlSHzpLrt9N48ftDbD15HYD+zWvy3kOt8PWQMiaieBxex66yUAHcqjg4CiGEsANbatNV8jp0lhgMKo9H/MPWk9dxddLw3wdbsOixuySpE3YhiZ2duTm7WW4H6bETQlQc2RdKREUVvBhi3DjjnLrNm40fK/nCCY1G4ZX+TWhS05v1k+9h1N3B+RbmF8JWMhRrZ/p8atXpQXrshBAVkzW16YKCKnUv3fnYJC7eTOGeRv4A9GwSQNdG/jhppX9F2Jd8R9mZm9Zyj507SI+dEKL80OmMPWw6XeHHxsSUfDzl2E8HYxg0fzvPfLufmLgUc7skdaIkyHeVnXWuY2HXDBU6oUiPnRCifIiIMJYz6dXL+HHhQkhKyvnIXZsuIsJx8ZZRSWmZvPDdIZ5bdZDEtEwa1/RGIyOuooRJHTs7q/tedS6m3sjTXkeFC93+C71es/s9hRDCbopao05q0+VwNCaeySsjOXcjCY0Ck3s1YnKvEOmlE0Uidewc6IqFpA7gCoCXVBAXQpRxRa1RJ7XpzCK2n2P2b/+SoVep5evGvBFtuLuB5X3EhbA3+dPBzrRYXq6uBWjcv1RjEUIIm5lq1GWn1cKJE5CYaHycOGH5mEpem87kQmwSGXqVfs1r8NtzXSWpE6VKEjs7c3N2tdyucQVfK1aOCSGEI1lTo05q0+WRqc/q5Zw+sCnzRrThs8fCqOLh4sCoRGUkc+zszGuGM0lk5mn3xInENzPsfj8hhLC7pCTw8jJ+fuKEMZGzRKczDr+GhFTapC5Db2DOHyc5dDGOb568G62sjhAlQObYOZABy3Xs8msXQogyraAadZW8Nt3Fm8lMXhnJwYtxAGw9eY1eTWQutXAsGYq1M+d8cuX82oUQokhsqTNXHFKjzqKfD11i4MfbOHgxDh83JxaOukuSOlEmSLZhZ5kaLRjyDrlmarQOiEYIUSFFRMD48cbVqxqNcU5c9i2+iit3jbolSyr9NmAmyemZvPnTMb7fb0yo2wVXZd4jbQiq6uHgyIQwkjl2duY6w5l0C3PsXHAm7c10u99PCFHJFLXOXHFIjTqziV/vZ+OxKygKTO4ZwpR7G0ltOlHiZI6dA/k6uXI9M29i5+tkebWsEELYpKh15opDatSZPd8nlONXEpg1tBWdGkoZE1H2SGJnZ/U8q3M9PilPewPP6g6IRghR4ZjqzGVP7rRaiIoqeKGDtWJijMOvua9fSWvU3UxK55+zsQxoWQuAxjW9+XNad+mlE2WWJHZ2di7xmsX2M4lXSzkSIUSFZKozN2mS8evsdebswVSjbsIEY09dJa5Rt/P0DaauPkhsUjrf+7pxV92qAJLUiTJNEjs7i9On2tQuhBA2Cw/PSuyiouyX1JmMGwf9+lXaGnUZegPz/jzJwi1nUFVoWN0TDxdZACfKB0ns7MzDyY2EzGSL7UIIYXf2GH61pJLWqLt4M5nnVkVy4EIcAI92qMN/BjfDw0V+XYryQb5T7cwVy1XH82sXQogcdDrjAolGjaxLrGJi7N9jV0n9duQyL685zO3UTLzdnJg1tCWDW9V2dFhC2EQmCtjZTQu9dQW1CyGEWUSEsZRJr17GjwsXGrf3yv3IXWcuIsJxMVcgVxNSuZ2aSdu6VdgwpaskdaJckjp2dqbMyL9nTn2zUr3VQghbFKc+ndSZK7JMvcG8GEJVVdYeiOG+NrVxlgUSogyxJXeR71whhCgLilOfzlRnTlhNVVW+3BnNwPnbuJ1q3C1IURQeCguSpE6Ua/LdK4QQZYGpPl12Wi2cOAGJiVmPEycsH1dJ68wVxa2kdJ76aj9v/nyMk1cTWb33oqNDEsJuJLGzs/wGYmXphBCiQKb6dCbZ69N5emY9THXmtNqcx8kwrFV2n41lwMfb+PPfq7hoNbw5pBnj7qnv6LCEsBuZY2dnbjNcSCMjT7srzqTKXrFCiIIkJYGXl/HzEycKXu2q01XaOnNFkak3MP+vU3yy+TSqCg2qe/LJo21pXtvX0aEJUSjZK9aBLCV1BbULIYRFhdWnq6R15opqzqaTLNpyBoDh7YJ4677mUptOVEjyXS2EEGWBTgeHD2d9LfXp7GrcPfX5/dgVnru3Efe3KaGizkKUATLHTgghHM1Uv27QoKw2qU9XLCnpetbs15m/9vdy5Y+p3SSpExWe9NgJIYQj6XQwfnzeUicGA0yYYNyzVYZcbXL8SgKTV0Ry6loiTlrFnMw5SRkTUQnId7mdueJsU7sQopIrqH6d1KeziaqqfL37PPcv2MGpa4lU93bF38vV0WEJUaqkx87ONFon0OddKKHRylsthLDAVL/OUnIn9emsFpeczss/HOaPqKsA9GxcnQ+HtaaaJHaikpEeOztL1afa1C6EqORy168zkfp0VtsbfZMBH2/jj6irOGsV/m9QUyLC20tSJyol6UayMy1aMsm02C6EEBaFh8OkScbPt2419t5JfTqrJaVlcjk+lfr+xtp0LQKlNp2ovCSxszONogELJZ81inSOCiGsEBZm3GFCFChTbzAvhujROIBPHm1LzyYBeLnKrzVRuUm2YWfpquXdJfJrF0IIYZuNR6/Qc84WLt5MNrcNaV1bkjohkMROCCFEOZGaoef/fjzCxG/2c/FmCp9tPePokIQoc+TPGyGEEGXeyau3eXbFAU5eTQRgQvcGvNCnsYOjEqLskcROCCFEmaWqKiv2XGDm+ijSMg34e7kyd3hruoVWd3RoQpRJktgJIYQos77fp+P1dUcB6B5anTnDW0vRYSEKIImdEEKIMuu+NrX5ds8FhrSqxdgu9dFoFEeHJESZVqTFE5mZmfz5558sXryY27dvA3Dp0iUSExPtGpwQQojKRW9Q+W7vRTL1xp043Jy1rH26M092bSBJnRBWsLnH7vz58/Tv358LFy6QlpZGnz598Pb25v333yc1NZXPPvusJOIUQghRwV2OT2HqqoP8c+4ml+JTmNo7FACtJHRCWM3mHrvnnnuOdu3acevWLdzd3c3tDz74IH/99ZddgxNCiHJHp4PNm40fiyImxr7xlBN/HLvCgI+38c+5m3i6aAmu5uHokIQol2zusdu+fTs7duzAxcUlR3twcDAxlfQHkhBCABARAePHG7cE02iMe8CGhxd+3pIlWZ83bWr8ety4kouzDEnN0PPuhn/5atd5AFoG+jL/0bbU95fdN4QoCpsTO4PBgF6vz9Ou0+nw9va2S1BCCFHu6HRZSR0YP06alLUHrLUMBpgwAfr1q/B7xZ6+lsizKw5w/Ipxrvb4bg14sW9jXJykdr4QRWXz/54+ffowb94889eKopCYmMibb77JwIED7RmbEEKUH6dOZSV1xaXXw+nT9rlWGWZQVc7dSMLfy4Uvx3bgtYFNJakTophs7rH76KOP6NmzJ82aNSM1NZWRI0dy6tQp/P39WblyZUnEKIQQZV+jRsbh1+zJnVYLUVEQGJj/eTExxuHX3OeFhJRcrA6UqTfgpDUmb6E1vFn02F20CPQlwNvNwZEJUTHY/KdR7dq1OXjwIC+99BITJkygbdu2zJ49m8jISAICAkoiRiGEKPuCgoxz6ky0Wli8GEJDwdMz/0doqHFOnVab87wKOAy7//xN7p27lf3nb5nbejWpIUmdEHakqKqq2nLC33//TefOnXFyytnZl5mZyc6dO+nWrZtdA7S3hIQEfH19iY+Px8fHx+7XV2bkvyxffdOmt1oIUd4kJYGXl/HzEyeMSZu1dDrj8GtISIVL6vQGlUVbTvPRn6fQG1S6hFTj2yc7OjosIcoNW3IXm4die/bsyeXLl/P0zsXHx9OzZ0+LCyuEEKLSKWj41ZKgoAqX0AFciU/l+dUH2XU2FoD729TmnQdaODgqISoumxM7VVVRlLy9UrGxsXh6yvJ0IUQlpdPB4cNZX8fE2NZjVwH9GXWVl344xK3kDDxctMy8vwUP3RVo8XeIEMI+rE7shg4dChhXwY4ePRpX16xNmPV6PYcPH6Zz5872j1AIIcq67PXrTCpZPbrc/jkby5Nf7QOgeW0fPnm0LQ2qezk4KiEqPqsTO19fX8DYY+ft7Z1j1wkXFxc6duzIU089Zf8IhRCiLMtdv86kEtWjs6RDfT/6NqtBXT8PXurfGFcnraNDEqJSsDqxW7ZsGQD16tXjxRdflGFXIYSAguvXmerRVYLETlVVfjp4iXubBuDt5oyiKCx6LEz2eRWilNlc7uTNN9+UpE4IIUxM9essqcD16LJLSM1g8spIpq4+yP/9eBRTsQVJ6oQofTYvngD44Ycf+O6777hw4QLp6ek5njtw4IBdAhNCiHLBVL8u99ZhFbgeXXYHLtxiyspIdLdScNIoNKlp/zJSQgjr2dxjN3/+fMaMGUNAQACRkZF06NCBatWqcfbsWQYMGFASMQohRNkWHp71+datsHkzREdX6IUTBoPKp5tPM+yzXehupVDHz53vJ3bi6R4NZdWrEA5kc4/dwoULWbJkCY8++ihffvklL7/8Mg0aNOCNN97g5s2bJRGjEEKUH2Fhxh0lKrBrt4216XacNtamG9K6Nv99sAU+bs4OjkwIYXOP3YULF8xlTdzd3bl9+zYAjz/+uOwVK4Qov3Q6Y0+bTle868TE2CeeMkyrKJy8moi7s5b3H27F/EfaSFInRBlhc2JXs2ZNYmONf6UFBweze/duAM6dO4eNu5MJIUTZEBEBwcHQq5fx48KFxu3BrH0sWZJ1raZNjderYDL1WSt/q3m5smjUXayffA/D29WRoVchyhCbE7tevXqxfv16AMaNG8fzzz9Pnz59GDFiBA8++KDNASxcuJD69evj5uZGWFgY27ZtK/D4tLQ0Xn/9dYKDg3F1daVhw4YsXbrU5vsKIQSQtw6dwWBcCOHlZf1j2rSs65nq1xW3568MOXs9kfs/3cGPkVm9ke3q+RESIAWHhShrbJ5jt2TJEgx3fgBOnDgRPz8/tm/fzpAhQ5g4caJN11q9ejVTp05l4cKFdOnShcWLFzNgwACioqKoW7euxXOGDx/O1atXiYiIICQkhGvXrpGZmWnryxBCCKOC6tAVVQWpX6eqKj/s1/Hmz8dITtczZ9MJBrWqhbPW5j4BIUQpUVQ7jp/GxMQQaMPG13fffTd33XUXixYtMrc1bdqUBx54gFmzZuU5fuPGjTzyyCOcPXsWPz+/IsWYkJCAr68v8fHx+PjYf1m+MiP/IQn1TRmqFqLM0emMw6/ZkzutFqKiwJqfZzExxuHX3OdHR5frxO52agavrzvKz4cuAdCpQTXmPdKGGj5uDo5MiMrHltzFLn92XblyhcmTJxNiQyHO9PR09u/fT9++fXO09+3bl507d1o85+eff6Zdu3a8//77BAYGEhoayosvvkhKSkq+90lLSyMhISHHQwghzEx16ExM9edCQ42rWwt7hIYa59hptTnPL8dJ3cGLcQyav52fD11Cq1F4qV9jvnnybknqhCgHrE7s4uLiGDVqFNWrV6d27drMnz8fg8HAG2+8QYMGDdi9e7dNc91u3LiBXq+nRo0aOdpr1KjBlStXLJ5z9uxZtm/fztGjR1m3bh3z5s3jhx9+YFLuwqDZzJo1C19fX/OjTp06VscohKgkstehi4qyvf7cuHHGHroKUL8uJi6F4Z/t4sLNZAKruPPdhE5M6hkiu0gIUU5YPcfutdde4++//yY8PJyNGzfy/PPPs3HjRlJTU/ntt9/o3r17kQLIvZpKVdV8V1gZDAYUReHbb7/F19cXgLlz5/Lwww/z6aef4u7unuec6dOnMy3bxOaEhARJ7oQQ+bNhOkkOQUHlupfOJLCKO6O71CPmVgrvDm2Jr7uUMRGiPLE6sfv1119ZtmwZvXv35plnniEkJITQ0FDmzZtXpBv7+/uj1Wrz9M5du3YtTy+eSa1atQgMDDQndWCck6eqKjqdjkaNGuU5x9XVFVdX1yLFKISogHQ644KJRo0sJ2IxMcbh1Upky4lrNKzuRR0/DwBe6d8EjZL3D28hRNln9VDspUuXaNasGQANGjTAzc2NJ598ssg3dnFxISwsjE2bNuVo37Rpk7kAcm5dunTh0qVLJCYmmttOnjyJRqMhqAL8pSyEKGH51aurBHXoLEnL1PPOL1GMXraXKasiybhTq06rUSSpE6KcsjqxMxgMODtndclrtVo8i7ltzrRp0/jiiy9YunQp//77L88//zwXLlwwl02ZPn06TzzxhPn4kSNHUq1aNcaMGUNUVBR///03L730EmPHjrU4DCuEEGYF1aur4HXoLDl3I4mHFu3ki+3nAGgV6ItBiswLUe5ZPRSrqiqjR482D2umpqYyceLEPMnd2rVrrb75iBEjiI2NZebMmVy+fJkWLVqwYcMGgoODAbh8+TIXLlwwH+/l5cWmTZuYPHky7dq1o1q1agwfPpx33nnH6nsKISopW+rVVZA6dPlZs1/Hf346SnK6nioeznzwcGv6NLM8BUYIUb5YXcduzJgxVl1w2bJlxQqopEkdOyEqqfzq1f3vf9CzZ4WrQ2dJUlom//fjUdbd2UHi7vp+zHukDbV8ZcRDiLLMltzF6h67sp6wCSFEgUz16kzlkUz15rp1M86xmzDB2FNXAerQ5cdJq3Dy6m00CkztHSplTISogOy680R5ID12QlRiSUnGOXUAJ07kXP2q0xmHX0NCKlRSZzCoqGBO4M5eT+RmUjrt6hVt9x4hROkrkR47IYSoUHLXq6sgdeiyu347jRe+P0SbOlWY1seYxDao7kWD6g4OTAhRYmQnZyGE4+h0xt0aHLECNSam9O9Ziv4+eZ0BH//N3yev88W2s8Qmpjk6JCFEKZAeOyGEY0REZJUf0WiM89+yb+1VEnLXq1uypFxv/2VJeqaBOX+cYPHfZwFoXMObT0a2pZqXFGoXojKQOXZ2JnPshLCCpRWqjlDBVr9G30hiyqpIDuviAXi8YzCvD2qKm7PWwZEJIYrDltylSEOxX3/9NV26dKF27dqcP38egHnz5vHTTz8V5XJCiMrGlppyJclUr64CSM3Q8/Bnuzisi8fX3ZnFj4fx9gMtJKkTopKxObFbtGgR06ZNY+DAgcTFxaHX6wGoUqVKkfeNFUJUMo0aGYdfs9NqjStVExNL5nHihOV7hoSU3usuQW7OWl7sG0qHen789lxX+jWv6eiQhBAOYHNi98knn/D555/z+uuvo9Vm/SXYrl07jhw5YtfghBAVlKmmnImpdlxoKHh6lswjNNQ4p870c6sC1Ks7GhNP5IVb5q9HtK/DyvEdqV1FCg4LUVnZvHji3LlztG3bNk+7q6srSUlJdglKCFEJhIdnFQuOispZU66kjBsH/fqV+3p1BoPK0h3neG/jcQK83dgwpSu+Hs4oioJW6g0LUanZnNjVr1+fgwcPmvdzNfntt99o1qyZ3QITQlQiuWvKlaRyXq/uRmIaL35/iC0nrgPQItD+i8CEEOWXzYndSy+9xKRJk0hNTUVVVfbs2cPKlSuZNWsWX3zxRUnEKIQoz3Q642KJRo3yT6hiYkqnx66c237qBs9/d5Drt9NwddLwn8HNGHV3XRRFuumEEEY2J3ZjxowhMzOTl19+meTkZEaOHElgYCAff/wxjzzySEnEKIQorwqqVVcJasrZi96g8sHvJ1j89xlUFUJrePHJo3fRuKa3o0MTQpQxxapjd+PGDQwGAwEBAfaMqURJHTshSomtteoqWE05e1JVlYnf7Of3Y1cZeXdd/jOoGe4uUsZEiMqiROvYzZgxgzNnzgDg7+9frpI6IUQpsrVWXQWqKWcvGXrj+6coCu891Iolj4fx7oMtJakTQuTL5sRuzZo1hIaG0rFjRxYsWMD169dLIi4hRHlXUK26Cl5TrriS0jJ56ftDPL/6IKZBlSoeLvSV2nRCiELYnNgdPnyYw4cP06tXL+bOnUtgYCADBw5kxYoVJCcnl0SMQojyqKBadRWwppy9HI2JZ8gn2/l+v45fj1zm2KUER4ckhChHir1X7I4dO1ixYgXff/89qampJCSU7R9CMsdOiFKUlAReXsbPT5zIu/JVpyv3NeXsRVVVlu2IZvZvx0nXG6jp48ZHI9rQqWE1R4cmhHAwW3IXm1fF5ubp6Ym7uzsuLi7cvn27uJcTQlRUlmrVlfOacvYSm5jGSz8c5n/HrwHQp1kN3n+oFVU9XRwcmRCivLF5KBaMu0/897//pVmzZrRr144DBw7w1ltvceXKFXvHJ4Qor3Q62Lo16+uYGMfFUoapqsrY5Xv53/FruDhpmHl/c5Y8HiZJnRCiSGzusevUqRN79uyhZcuWjBkzxlzHTgghzLLXrzORWnUWKYrCKwOa8NbPx/j4kbY0rSU7SQghis7mOXavvfYao0aNonnz5iUVU4mSOXZClLCC6tdJrToALt5M5vS1RHo2ySoXpTeoaDWyg4QQIq8SnWP37rvvFjkwIUQlUFD9OlOtukqc2K0/dInX1h5Br6r8OqUr9f09ASSpE0LYhVWJ3bRp03j77bfx9PRk2rRpBR47d+5cuwQmhCinTPXr8uuxq6S16pLTM5nxcxSr910EICy4Ks5aSeaEEPZlVWIXGRlJRkaG+XMhhMiXqX7dpEk52ytxrbqoSwlMXnmAM9eTUBR4tmcIz93bCCdtkdavCSFEvopdx668kTl2QpSC7PXrtm419t5V0lp1X+2K5p1f/iVdb6CGjysfjWhD54b+jg5LCFGOlOhesWPHjrVYry4pKYmxY8faejkhREUXFgY9elTKpA4gJi6FdL2B3k0D+O25bpLUCSFKlM2J3ZdffklKSkqe9pSUFL766iu7BCWEKEN0Oti82fixKCph/boMfdb8whf6NGbeiDZ8/kQ7/KQ2nRCihFmd2CUkJBAfH4+qqty+fZuEhATz49atW2zYsIGAgIDCLySEKD8iIoylS3r1Mn5cuNA4zFrYY8mSrGs0bWq8TiWQqTcw548TjFi8y5zcuThpeKBtIIoiCyWEECXP6nInVapUQVEUFEUhNPd+jxiLbM6YMcOuwQkhHEiny1lk2GAwLojIvSiiMAYDTJgA/fpV6OFY3a1knlt1kP3nbwGwKeoqA1vWcnBUQojKxurEbvPmzaiqSq9evVizZg1+fn7m51xcXAgODqZ27dolEqQQwgEKqkdnqwpev27Dkcu8suYwt1Mz8XZ14t2hLSWpE0I4hNWJXffu3QHjPrF169aVYQUhKjpL9ei0WoiKgoK2EYyJMQ6/5j6vAtavS0nXM/OXKFbuuQBA27pVmP9IW+r4eTg4MiFEZWVVYnf48GFatGiBRqMhPj6eI0eO5Htsq1at7BacEMKBctejM9WhszAVI4fQUOMcuwkTjD11Fbh+3StrDvPzoUsoCjzToyFTe4fiLLXphBAOZFUdO41Gw5UrVwgICECj0aAoCpZOUxQFvV5fIoHai9SxE8IG2evRnThReFKXnU5nHH6twPXrzl5PZNyX+3jngRZ0CZEyJkKIkmH3vWLPnTtH9erVzZ8LISqhgoZfLQkKqnAJXVxyOn+fusF9rY3ziRtU9+LPad1ln1chRJlhVWIXHBxs8XMhRAWm08Hhw1lfx8TY1mNXwfxzNpapqw9yJSGV6l6udGpYDUCSOiFEmVKkAsW//vqr+euXX36ZKlWq0LlzZ86fP2/X4IQQDmKqXzdoUFZbJapHl12m3sDcTSd59PPdXI5PpX41T7zdrF53JoQQpcrmxO7dd9/F3d0dgF27drFgwQLef/99/P39ef755+0eoBCilOWuX2diqkdX1B0oyqGYuBQe/Xw38/86hUGFYWFBrJ98Dy0CfR0dmhBCWGTzn50XL14k5E7Zgh9//JGHH36Y8ePH06VLF3r06GHv+IQQpa2g+nUVvB5ddr8fu8JL3x8iITUTL1cn/vtgC+5vY+M8QyGEKGU299h5eXkRGxsLwB9//EHv3r0BcHNzs7iHrBCinDHVr7OkgtajsyQuOZ2E1ExaB/ny65R7JKkTQpQLNvfY9enThyeffJK2bdty8uRJBt2Zg3Ps2DHq1atn7/iEEKUtd/06kwpcj84kQ28w16Eb3q4Ork5aBrashYuT1KYTQpQPNv+0+vTTT+nUqRPXr19nzZo1VKtmXBm2f/9+Hn30UbsHKIRwgPDwrM+3boXNmyE6GsaNc1hIJUlVVb7ZfZ5+8/4mPjkDMNblfKBtoCR1QohyxaoCxRWJFCgWwgrZCxMnJoKnp2PjKUFxyem8uuYIG49dAeClfo2Z1LNyDDcLIcoHuxcozi0uLo6IiAj+/fdfFEWhadOmjBs3Dl9fWSkmRIWzfz906+boKErEnnM3mboqkkvxqThrFV7p34SxXeo7OiwhhCgym8cY9u3bR8OGDfnoo4+4efMmN27c4KOPPqJhw4YcOHCgJGIUQpS2CROyPu/eHUaPdlgoJUFvUJn350keWbKLS/Gp1Kvmwdqnu/Bk1wZopOCwEKIcs3kotmvXroSEhPD555/j5GTs8MvMzOTJJ5/k7Nmz/P333yUSqL3IUKwQhdi7Fzp0yNu+Zw+0b1/68ZSAjzad5OO/TgHw0F1BzLi/OV6uUnRYCFE2lehQ7L59+3IkdQBOTk68/PLLtGvXzvZohRBly7Ztltt37Kgwid3YLvX57ehlnukRwgNtpYyJEKLisHko1sfHhwsXLuRpv3jxIt7e3nYJSgjhQF27Wm7v0qV047Cj1Aw9q/dewDRA4evhzG/PdZOkTghR4dic2I0YMYJx48axevVqLl68iE6nY9WqVTz55JNS7kSIiqB9exg1KmdbeHi57a07efU29y/YwStrjvDdvovmdq3MpRNCVEA2D8V++OGHKIrCE088QWZmJgDOzs48/fTTzJ492+4BCiEcYPFi+PZb4+dbt5bLVbGqqrJyz0Vm/nKM1AwD/l4u1PJ1d3RYQghRoopcxy45OZkzZ86gqiohISF4eHjYO7YSIYsnhLBCOa9jF5+cwfR1h9lwxFibrmsjf+YOb0N1b1cHRyaEELazJXexeig2OTmZSZMmERgYSEBAAE8++SS1atWiVatW5SapKxX55W6S04nSptMZd4zQ6Yp3nZgY+8RTSvafv8XA+dvYcOQKThqF1wY24csxHSSpE0JUClYndm+++SbLly9n0KBBPPLII2zatImnn366JGMrn/LrsJPpPKI0RURAcDD06mX8uHChsRfO2seSJVnXatrUeL1yIj3TwKX4FIKrebDm6c6M79ZQatMJISoNq4diGzZsyH//+18eeeQRAPbs2UOXLl1ITU1Fq9WWaJD2JEOxosLT6YzJnMFgv2tqtca9YoOC7HdNO8rQG3DWZv2duvHoZbqE+OPt5uzAqIQQwj5KZCj24sWLdM1WBqFDhw44OTlx6dKlokcqhLC/U6fsm9QB6PVw+rR9r2knm6Ku0uODLZy9nmhu69+iliR1QohKyerETq/X4+LikqPNycnJvDJWCFFGNGoEmlz/tbVaOHHCuBCisMeJE5bPDwkpvddghdQMPW/+dJSnvtpHTFwKC7eccXRIQgjhcFaXO1FVldGjR+PqmjUBOTU1lYkTJ+KZbcXc2rVr7RuhEMI2QUHwyScwaZLxa63WWL4kNNS680NDjXPsJkww9tSZzi9Dw7Cnr93m2RWRHL9yG4An76nPy/2bODgqIYRwPKvn2I0ZM8aqCy5btqxYAZU0mWMnKoXs5UpOnLA+qctOpzMOv4aElJmkTlVVvtt3kbd+jiIlQ081Txc+HN6ano0DHB2aEEKUmBLZK7asJ2xCiHwEFnHbrKCgMpPQmfx86BKvrDkCwD0h/swd0ZoAbzcHRyWEEGWHzTtPCCGEowxsWYtvd1+gV9MAxndtIGVMhBAiF5v3ihVClAG2FB8uZwWGs9MbVFbtuUB6pnGVr7NWw8rxHZnYXWrTCSGEJZLYCVHeWFN8uBwXGDa5mpDK4xH/8OraI8zZdMLcrpWETggh8iVDsUKUJzodjB+fVafOYDCufjWtgLXEYDCucO3Xr8zNmcvPX/9e5cXvD3ErOQMPFy0h1b0cHZIQQpQLDu+xW7hwIfXr18fNzY2wsDC2bdtm1Xk7duzAycmJNm3alGyAQpQlRS0+XIYLDGeXlqlnxvpjjPtyH7eSM2he24f1k+9hWLs6jg5NCCHKhSIldl9//TVdunShdu3anD9/HoB58+bx008/2XSd1atXM3XqVF5//XUiIyPp2rUrAwYM4MKFCwWeFx8fzxNPPMG9995blPCFKL+sKT5cTgoM53buRhIPfrqTZTuiARjTpR5rn+lMQ+mtE0IIq9mc2C1atIhp06YxcOBA4uLi0Ov1AFSpUoV58+bZdK25c+cybtw4nnzySZo2bcq8efOoU6cOixYtKvC8CRMmMHLkSDp16mRr+EKUb6biwybZiw97ehofpgLDpj2cy2CBYUsU4HxsEn6eLkSEt+PNIc1xdSo/+1ALIURZYHNi98knn/D555/z+uuvo9Vm/dBt164dR44csfo66enp7N+/n759++Zo79u3Lzt37sz3vGXLlnHmzBnefPNNq+6TlpZGQkJCjocQ5Vp4eNbnUVEwblzeY8aNg+ho48rZ6GjLx5QBGfqsYeV6/p589ngYvz3XlXub1nBgVEIIUX7ZnNidO3eOtm3b5ml3dXUlKSnJ6uvcuHEDvV5PjRo5f4DXqFGDK1euWDzn1KlTvPrqq3z77bc4OVm37mPWrFn4+vqaH3XqyFwdUYEUVHw4KAh69CizPXWRF27Re+5Wdp6+YW7r2qg6NXyk4LAQQhSVzYld/fr1OXjwYJ723377jWbNmtkcgKLkLF2gqmqeNgC9Xs/IkSOZMWMGoTZsjzR9+nTi4+PNj4sXL9ocoxClzto6deWwRp3BoLJwy2mGfbaL87HJzN10Eit3NhRCCFEIm8udvPTSS0yaNInU1FRUVWXPnj2sXLmSWbNm8cUXX1h9HX9/f7RabZ7euWvXruXpxQO4ffs2+/btIzIykmeffRYAg8GAqqo4OTnxxx9/0KtXrzznubq64urqauOrFMKBIiKySppoNMY5ddmHX3PXqFuypMwOteZ2LSGVad8dYvudXrpBrWrx7oMtLf4xJ4QQwnaKWoQ/lT///HPeeecdc+9XYGAgb731FuNs/OVy9913ExYWxsKFC81tzZo14/7772fWrFk5jjUYDERFReVoW7hwIf/73//44YcfqF+/Pp6enoXe05aNdItCmZH/Lyj1TemVEIXQ6YxFh20paaLVGufRldEhV5PNx6/xwveHuJmUjruzlrfua8bwdnUkqRNCiELYkrsUqUDxU089xVNPPcWNGzcwGAwEBAQUKdBp06bx+OOP065dOzp16sSSJUu4cOECEydOBIzDqDExMXz11VdoNBpatGiR4/yAgADc3NzytDuUinF5n6V2IQpTlDp1php1ZTixO3QxjjHL9wLQpKY3C0a2JSTA28FRCSFExVOsnSf8/f2LdfMRI0YQGxvLzJkzuXz5Mi1atGDDhg0EBwcDcPny5UJr2glRoZjq1GVP7rRa4+rXwEDjnLqmTfM+X8Zr1LUK8uW+1rXx83Th1QFNcHOWMiZCCFESbB6KrV+/foFDJ2fPni12UCWpxIdi31Ly7bFT35JuO2GFhQuztggz1aDLPs0hIsK4RZheb/n5MkBVVX4+dIlujapT1dMFAL1BlX1ehRCiCEp0KHbq1Kk5vs7IyCAyMpKNGzfy0ksv2Xq5iie/31vy+0xYKzw8K7GLijIWHM5u3Djjvq+nTxt76srYEOzt1Az+8+NRfjx4ib7NarD48TAURZGkTgghSoHNid1zzz1nsf3TTz9l3759xQ5ICJFNfnXqgoLKXEIHxrl0U1ZFcj42Ga1GoVWQL6oKsj5CCCFKR5H2irVkwIABrFmzxl6XE6LyMdWuy16brpzUqTMYVBZvPcNDi3ZyPjaZwCrufDehI8/2aoRGeuqEEKLUFGvxRHY//PADfn5+9rpc+SWrYkVRZK9dl717qxzUqYtNTGPq6oNsO3WnNl3LWrw7tCW+7s4OjkwIISofmxO7tm3b5lg8oaoqV65c4fr16znq0QkhrKTTZSV1ANnXMxkMxoUS/fqVyaFXACethrPXk3Bz1vDWkOaMaC+16YQQwlFsTuweeOCBHF9rNBqqV69Ojx49aNKkib3iEqLyKKx2XRmsU5ehN+CkUVAUBV93ZxaOugsPFy2NakhtOiGEcCSbErvMzEzq1atHv379qFmzZknFJETlYql2XXZlrE5d9I0kpqyK5NEOdXm0Q10AWtep4tighBBCADYunnBycuLpp58mLS2tpOIRovIJCjLuB2uiKMZED7Lq1JWR3rp1kToGzd/GYV088/86RWqG3tEhCSGEyMbmodi7776byMhI8+4QQgg7yF677vhx8PAoU3XqEtMyeePHo6yNNK7S7VDfj3kj2sgOEkIIUcbYnNg988wzvPDCC+h0OsLCwvD09MzxfKtWrewWXLkkBYpFcQUGgqdnmUjoAA7r4piyMpLo2GQ0CkztHcqkniFScFgIIcogqxO7sWPHMm/ePEaMGAHAlClTzM8pioKqqiiKgl4vQzNC5EunMy6WaNQo/8QtJibvbhMOcu12KsM+20VapoHavm58/Ghb2teTskZCCFFWWZ3Yffnll8yePZtz586VZDxCVFzZa9VpNMZ5deHhxueWLMk6rgzVrgvwdmNC94acunqb2UNb4eshtemEEKIsU1RVtap0rkaj4cqVKwQEBJR0TCXKlo10i0J5S8m3QLH6llQprrR0OggOLrisSXZaLURHO2Q4dtup69Su4k7D6l6AcVcJRUFq0wkhhIPYkrvYtCpWfrALUUSF1arLzVS7rhSlZxqY9du/PB6xh8krIknLNE6r0NypVyeEEKLss2nxRGhoaKE/4G/evFmsgMo9WTwhLLFUq06rhago4+dNm+Z9rhRr152PTWLKykgO6eIBaFu3Ctb15QshhChLbErsZsyYga+vb0nFIkTFZapVZyppYqpPZ1oksWSJceswvb7Ua9f9dDCG19cdJTEtEx83J95/uBX9W9QqlXsLIYSwL5ljZ2fKjPy75tQ3pQukUktKAi/jvDVOnMi78lWnK9XadSnpev7vx6OsOaADoH29qsx7pC2BVdxL/N5CCCGsZ0vuYnWPncyxsZJKvosnhDALDMzbFhRUqoslnLQKZ28kolFgcq9GTO4VgpPWpmm3QgghyhirEzsrO/aEEGW4Vp2qqugNKk5aDc5aDfMfaculuBTublCt1GMRQghhf1b/eW4wGMr9MKwQJS4iwljWpFcv48eFC41DsElJeWvVRUSUamg3EtMYu3wv7/9+wtxWx89DkjohhKhArJ5jV1HIHDtRYspwrbrtp27w/HcHuX47DTdnDVtf6kkNH7cSv68QQojiK5E5dsJKMseu8ipqrboSTOwy9AbmbjrJZ1vPoKrQKMCLT0a2laROCCEqKEnshLCXMlar7uLNZCavjOTgxTgARt5dl/8Maoa7i7bE7imEEMKxZAmcEPZiqlVnkr1WXWiocY6dVpvzuRLqrUvPNDB88S4OXozDx82JhaPu4t0HW0pSJ4QQFZzMsbMzmWNXyZWhWnU/Rsbwze7zzHukDUFVPUr0XkIIIUqOzLEToiwo5Vp1xy7Fk5Smp0N9PwAeaBvIkNa10WqkBqUQQlQWMhRrb/l1yklnXcWh08HmzcaPBYmJKZVwVFVl2Y5zPPjpTiatOEBsYpr5OUnqhBCicpHETghbFFSnzgG16m4mpfPkl/uYsT6KdL2B1kFV0MguMUIIUWnJHDs7076lYLDwe1Wjgv6tSvVWVzy21qmDEq1Vt/PMDaauOsi122m4OGl4fWBTnugULNv/CSFEBSNz7Bwov1/5NqQCoqyytU4dlEitOoNBZe6mk3y65TSqCg2re/LJo3fRrLb9/1ARQghRvshQrBDWMtWpy06rNa5+TUw0frT0vJ1r1SkKXLyVjKrCI+3rsH7yPZLUCSGEACSxsztnG9tFOVJQnTpPzxKvVZehN/YWKorCOw+0YPHjYcx+qBUeLtLxLoQQwkjm2NmZ61sK6RamOLmokCZz7Mq/wurUgd1r1SWnZzJzfRQ3EtP5/IkwmUMnhBCVjMyxcyBLSV1B7aIcs1SnDuxaqy7qUgKTVx7gzPUkFAUiL8ZxV92qdrm2EEKIikeGYu1N6thVHIXVqyvBOnWqqvLlzmgeWLiDM9eTCPB25dtxd0tSJ4QQokCS2NlZfh1z8kaXM/nVqyuFOnW3ktJ56qv9vPnzMdIzDdzbJICNU7vROcTf7vcSQghRscgcOztzmqGgt9CuBTJlr9jywZZ6dSVQp27YZzvZG30LF62G6QObMLpzPZlXJ4QQlZgtuYt0JNmZ1LGrAGypV2eqU2dHrw5oSmgNL9Y+05kxXepLUieEEMJqktjZmfTJVQD51avburVE6tTpbiXzx7Er5q/Dgqvy23PdaBHoW6zrCiGEqHwksbOz/N5Q6XMpR/KrV9etm93r1P125DIDP97G5JWRnLx6O+uWGvmOEUIIYTspd2Jn+f06ll/T5Ux4OEyaZPw8KiqrXt24cdCvX7Hr1KWk65n5SxQr91wAoE2dKrg7a+0RuRBCiEpMEjs7cwWS82kX5VTuenXFrFN34sptJq88wMmriSgKTOzekGl9QnHWSge6EEKI4pHEzs7y+9Usv7LLOJ3OuGiiUaO8SVtMjOUdJorg23/OM3N9FGmZBqp7u/LR8Dbc00jKmAghhLAPSezsLNPGdlEGRETA+PHGlbAajXF+XVpa1vNNmxrn1o0bV+xb3bidTlqmgZ6Nq/PhsNZU85K+XCGEEPYjdezszHeGQoKFdh8gXurYlT3W1qwrRr26DL3BPMyqN6j8cvgSQ1rVRiMLJIQQQlhB6tg5kJuN7cLBrK1ZV4R6dZl6A/P+PMnQhTtJzTCWrdZqFO5vEyhJnRBCiBIhQ7F2lm5ju3AwU8267MmdqVZd9jYb69Vdikth6qqD7Im+CcDGo1d4oG1gIWcJIYQQxSM9dnaWYWO7cDBLNeuWLClWvbqNR68w4ONt7Im+iaeLlnkj2khSJ4QQolRIj52dSR27cii/mnU21qtLzdDzzq9RfLPbWJuudZAv8x9tS3A1z5KKXAghhMhBEjs709vYLsqY7DXrbKxX98ZPR/lunw6ACd0b8EKfxrg4Sae4EEKI0iO/dewsv3Wvsh7WwXQ62LzZ+LEgMTFFvsWUexvRoLonX43twPQBTSWpE0IIUerkN4+dyVBsGRQRYSxp0quX8ePChZCUlPVYsiTr2KZNjcdbIT45g7UHshLFoKoebHq+O91Cq9v7FQghhBBWkTp2diZ17MoYa+vUZWdFzbq90Td5bmUkl+JTWTamPT0bBxQ/ViGEEMICW3IXmWNnZ04oWBp4dZI+O8ewtk5ddqaadRYSO71BZcH/TvPxXycxqFCvmgf+nrJ7hBBCiLJBEjs780DDTQtLJTxk1NsxLNWp02qNq18DA41z6po2tapm3eV4Y226f84Za9MNbRvIzAda4OUq/42EEEKUDZJt2JlHPj1zXtJj5xiW6tQtXmwsaeLpafxoRc26v/69yoCPt/HPOWNturnDWzN3RBtJ6oQQQpQp8lvJzvLbYSItn3ZRCvKrU2cyblyhNesS0zKJS86gZaCxNl19f6lNJ4QQouyRxM7OMm1sF6UsMJ8dICzUrEvPNJhLltzfJhCNotCveU0pYyKEEKLMkt9QdpbfIgmtDMU6hk4HW7dmfW1FnTpVVVm55wL3zt3C9dtZfa1DWteWpE4IIUSZJr+l7Cy/ATqvUo1CAFn16wYNymorpE5dfEoGz66IZPraI1y8mcLXu8+XQqBCCCGEfUhiZ3eW31JF3urSpdPB+PF5S50YDDBhgsUdKPafv8nAj7fx65HLOGkUpg9owtR7G5VSwEIIIUTxyRw7O0vPZ/OwVNlUrHQVVL8uV506vUFl0ZbTfPTnKfQGlbp+Hsx/tC1t6lQpvXiFEEIIO5DEzs7ym0mnLdUohMX6dSa56tQt+fssH/5xEoD729TmnQda4O3mXFqRCiGEEHbj8PHBhQsXUr9+fdzc3AgLC2Pbtm35Hrt27Vr69OlD9erV8fHxoVOnTvz++++lGG3hpNxJGZG7fp2JhTp1j3cKplktHz4c1pp5I9pIUieEEKLccmhit3r1aqZOncrrr79OZGQkXbt2ZcCAAVy4cMHi8X///Td9+vRhw4YN7N+/n549ezJkyBAiIyNLOXLbyUCsA4SHZ32+dSts3gzR0aSFj2bFPxcwbZPs5erEL5Pv4eGwIBRFVi8LIYQovxTV9NvNAe6++27uuusuFi1aZG5r2rQpDzzwALNmzbLqGs2bN2fEiBG88cYbVh1vy0a6RXHXf72JzEzM2+7kxf7Xb9v9fqIASUngdWc9cmIieHpy+loiU1ZGEnU5gTeHNGNMl/qOjVEIIYQohC25i8Pm2KWnp7N//35effXVHO19+/Zl586dVl3DYDBw+/Zt/Pz88j0mLS2NtLSsgdCEhISiBWwlf43lYbz82kUR6XTGBRKNGlncKSI3Vafj+wQP3vz5GCkZeqp5ulCvmuweIYQQomJx2FDsjRs30Ov11KhRI0d7jRo1uHLlilXXmDNnDklJSQwfPjzfY2bNmoWvr6/5UadOnWLFXZgMvd6mdlEEpvp0vXoZPy5caOydy/1YssR8itq0Gfve+JCUDD1dQqrx23Nd6dkkwIEvQgghhLA/hy+eyD2nSVVVq+Y5rVy5krfeeovVq1cTEJD/L+jp06cTHx9vfly8eLHYMRfEW2N5/atPPu3CRrnr0xkMxn1gvbzyPqZNM5+mUQ28+/sCZoT58PXYuwnwcXPQCxBCCCFKjsOGYv39/dFqtXl6565du5anFy+31atXM27cOL7//nt69+5d4LGurq64uroWO15r5ZuTypx8+yioPl0hnFQD4dX1oJF/DCGEEBWTw3rsXFxcCAsLY9OmTTnaN23aROfOnfM9b+XKlYwePZoVK1YwKPtWUWVEgj7TpnZhI1N9uuy0WjhxwrhAIjGR9LgE49eWjstWv04IIYSoaBw6FDtt2jS++OILli5dyr///svzzz/PhQsXmDhxImAcRn3iiSfMx69cuZInnniCOXPm0LFjR65cucKVK1eIj4931EvIwzmfNcb5tQsb5a5PZ6pLFxoKnp5svphEj4V7OO5T0zjHTqvNeZwVCy2EEEKI8sqhO0+MGDGC2NhYZs6cyeXLl2nRogUbNmwgODgYgMuXL+eoabd48WIyMzOZNGkSkyZNMreHh4ezfPny0g7fotsaBSyMFN6W4T/7CQ83zqsDiIqC0FDSMvW899sJlu44B8Cnm8/wybhx0K+fcfuwkBBJ6kSJ0ev1ZGRkODoMIUQ54+TkhFartWsNVYfWsXOEkq5j1+ndKuzOyNuD2MnZl52vxdn9fpVSrvp0Z5NVJq+M5NglYymbMV3q8Ur/Jrg5y4IVUbJUVeXKlSvExcU5OhQhRDml1WoJCAjA19c33wSvXNSxq6iqa93BQmIXoHV3QDRliI1156z128Z9vHAoheR0PVU9nPlwWGvubVrw4hsh7MWU1AUEBODh4SE7lwghrKaqKpmZmSQkJHD58mVSUlKoVatWsa8riZ2deTi7Qqql9kpcXiMiIqtEiUZjnCOXfbsvW2WrT9d3WE8G95vMxQcfZd4jbaghZUxEKdHr9eakrlq1ao4ORwhRTnl7e+Pq6sqNGzcICAhAqy3eaJMkdqJk5Vd3LtscyeLQqiqz/1iAuvhFtJLUiVJkmlPn4eHh4EiEEOWdp6cn169fJyMjo9iJncMLFFc4Sj65slJJ53sVo+6ctTQGA9qzZ0r0HkLkR4ZfhRDFZc+fI9JjZ2/afIoh59de0ZnqzmVP7rRa42rWwECrL3PtdiqvrT3C6ciT/C/iaTRqrutJfTohhBBCeuzsL7/eqZLttSqzCqk7Z81jsy6JAZ8f4M8LSVypWYd9r82S+nRCCCGEBZLY2V1+b2klfquzL5SIioJx46w6LS1Tzzu/RDFm2V5ik9JpUtObXybfQ4d3XoboaNi82fjRyusJIay3fPlyFEVBURS2bNmS53lVVQkJCUFRFHr06JHjOUVRePbZZwu8fo8ePczXVxQFd3d3Wrduzbx58zBYOX2jXr16Oa6R38NedU7fffddfvzxR7tcC2Do0KEFvlemf4N9+/ZZfH7w4MHUq1cvT3tCQgL//e9/adeuHT4+Pri6ulKvXj3Gjh3LgQMH7BL7tWvXGD16NP7+/nh4eNCpUyf++usvq8//9ttvadu2LW5ubvj7+zNy5EiLe7nfvn2bKVOmEBgYiKurK6Ghobz//vvo9foCr//FF1+gKApeptJYlYgMxdqd9NgVyMrh1wuxyTyzYj9HY4y16cI7BTN9YNOs2nRBQdJLJ0Qp8Pb2JiIiIk/ytnXrVs6cOYO3t3eRr92gQQO+/fZbwJgofPbZZzz//PNcvnyZ9957r9Dz161bR1pamvnrL774goiICDZu3Iivr6+5vWHDhkWOMbt3332Xhx9+mAceeKDY17p27Rq//PILYExyPvzwQ9zcir8A7MyZM/Tt25dr164xceJEZsyYgZeXF9HR0Xz33XeEhYURFxeX4/2xVVpaGvfeey9xcXF8/PHHBAQE8Omnn9K/f3/+/PNPunfvXuD5n3zyCVOmTOHJJ59k9uzZ6HQ6/vOf/9C1a1ciIyOpWrUqAJmZmfTp04eTJ0/y9ttvExoaysaNG3n11VfR6XTMnz/f4vVjYmJ48cUXqV27dpnamaq0SGJnd5Wsx87W+nQxMcZh2EJoNMbkroqHMx883Jo+zaQ2nRCOMGLECL799ls+/fTTHIVRIyIi6NSpEwkJCUW+tru7Ox07djR/PWDAAJo0acKCBQt45513cHZ2LvD8tm3b5vh648aNAISFheHv71/kuErDV199RUZGBoMGDeLXX39l7dq1jBw5sljX1Ov1PPjgg9y4cYNdu3bRokUL83Pdu3cnPDyc3377rdD3tTAREREcPXqUnTt30qlTJwB69uxJ69atefnll/nnn3/yPTctLY3//Oc/DBkyhM8//9zc3qxZMzp37syHH37If//7XwB++OEH/vnnH9asWcPQoUMB6NOnD4mJiXz66adMmjSJxo0b57nHxIkT6datG35+fvzwww/Feq3lUQXNNhzHM59lyvm1l2sRERAcDL16GT8uXGjcFSL3I1vdOZo2NZ5nQXpmVq9mUFUPFj/ejo3PdZOkTlRO8TFw7m/jRwd69NFHAeNe3Sbx8fGsWbOGsWPH2vVezs7OhIWFkZyczPXr1+1yTVVVWbhwIW3atMHd3Z2qVavy8MMPc/bs2RzHRUZGMnjwYAICAnB1daV27doMGjQInU4HGIeXk5KS+PLLL81DvLl7MW2xdOlSatSowZdffom7uztLly4tzssE4Mcff+TIkSNMnz49R1KX3YABA4pdomfdunU0btzYnNSBcWusxx57jD179hATk//37NGjR4mPj2fgwIE52jt16oSfnx9r1qwxt+3YsQNFURgwYECOYwcPHozBYGDdunV5rv/NN9+wdetWFi5cWNSXV+5JYmdn/hrLfwlV07iUciQlLL/6dF5eeR/TpmWdZzDAhAnG87M5dDGOvh9t5X/Hr5rbOjWsRk1fqU0nyhlVhfSk4j32fA7zWsCXQ4wf93xe9GsVc9dIHx8fHn744RyJx8qVK9FoNIwYMaK471YeZ86cwcnJyTwcV1wTJkxg6tSp9O7dmx9//JGFCxdy7NgxOnfuzNWrxp83SUlJ9OnTh6tXr/Lpp5+yadMm5s2bR926dbl9+zYAu3btwt3dnYEDB7Jr1y527dpV5ORh586d/PvvvzzxxBNUq1aNhx56iP/973+cO3euWK/1jz/+ALB6qNi084E1j+yOHj1Kq1at8lzP1Hbs2LF875meng6Aq2veShGurq6cOnWK1NRU87EajSZPD6Pp3MOHD+dov3btGlOnTmX27NkEVeKpOjIUa2c3UuMstsem3irdQEpacerT6fVw+jQEBWEwqHyx/SzvbzxBpkHlo02n6Nk4QGqDifIrIxnerW2/66kG2PCi8VEUr10CF89ihTB27Fh69uzJsWPHaN68OUuXLmXYsGHFml9nYkoarl+/zvz58zlw4ADDhg3D3b342zDu3r2bzz//nDlz5jAt2x+YXbt2JTQ0lLlz5/Lee+9x/PhxYmNjiYiI4P777zcfN3z4cPPnHTt2RKPRUL169RzDx0URcWfUwtTjOW7cOL755huWLVvGzJkzi3zdCxcuAFC/fn2rjt+6dSs9e/a06thz586ZF2rExsbi5+eX5xhTW2xsbL7Xady4MRqNhh07djBmzBhz+5kzZ7h8+TIAt27dolatWjRr1gy9Xs/u3bu55557zMdu377d4n2eeeYZGjduzNNPP23Va6qoJLGzs6R8ChEn5Ve4uLyytj5dTIxx+DX3cSEhXL+dxgvfH+Lvk8YhlwEtajJ7aCtJ6oQoY7p3707Dhg1ZunQpo0ePZu/evcyZM6fY1z127FiO3hhnZ2dGjRrFp59+WuxrA/zyyy8oisJjjz2Wo9epZs2atG7d2rzaNyQkhKpVq/LKK69w+fJlunXrRrNmzewSQ26JiYl89913dO7cmSZNmgBZ7+/y5ct566230GhKZzAtLCyMvXv3WnVs7do5/1gp6Od0Qc/5+fkxatQovvrqK9q3b8+wYcPQ6XSMHz8erVaLXq83v/5Ro0Yxc+ZMxo8fz7Jly2jcuDG//fabedFE9vdpzZo1rF+/nsjIyEr/O6SCZRuO55nPnrCezhWsQLGpPp1pa7Ds9emyCw01zrGbMMHYU3fnuL+TXZm2/G9uJKbj6qThzSHNebRDnUr/H1JUAM4exl6yokq4BJ92MPbUmShamPQP+BShJ9C5+FueKYrCmDFjmD9/PqmpqYSGhtK1a9diX7dhw4asWrUKRVFwc3Ojfv36dt2i7erVq6iqSo0alufpNmjQAABfX1+2bt3Kf//7X1577TVzj9FTTz3F//3f/xV7sUF2q1evJjExkeHDhxMXF2duHz58OLNmzWLTpk3069cPMM5bA/It7ZGZmZkjtrp16wLG3jVT0lgQLy8v2rRpY1XcplgAqlWrZrFX7ubNmwAWe/OyW7RoEaqq8swzzzBx4kQ0Gg2PP/44NWrU4Pfffzfvvezv78/GjRsJDw8395JWq1aNuXPnMm7cOALvdCIkJiYyadIkJk+eTO3atc3vq2nYNy4uDmdnZzw9i9dzXV5IYmdn/k6Wv3GqOVXAWjrh4VmJXVRU/qtdx42Dfv2Mw68hIfyr9eGJj7cB0LiGN5+MbEtojeIP6QhRJihK8YY+/RvBkI9h/VRQ9cakbsg8Y7sDjR49mjfeeIPPPvvMvGqxuNzc3GjXrp1drmWJv78/iqKwbdu2fOd0mbRs2ZJVq1ahqiqHDx9m+fLlzJw5E3d3d1599VW7xWQahp06dSpTp061+LwpsTMlpPktRoiJicmRtPbr148lS5bw448/WhVzUYdiW7ZsyZEjR/IcY2rLb+GGiaenJ19//TXz58/n4sWL1K5dG39/f5o0aULnzp1zJJHt27cnKiqK6OhokpKSaNSoEfv37wegW7duANy4cYOrV68yZ84ciz3JVatW5f7777drDcKyTBI7O0tTMyy2p6vppRxJKSusPl22unNNgRHt6uDipOH1Qdlq0wkhjO56AhreCzfPgl8D8LV++72SEhgYyEsvvcTx48cJz150vAwbPHgws2fPJiYmJsd8uYIoikLr1q356KOPWL58eY6Cvq6urqSkpBQ5nn///Zddu3bx0EMPWSxK/M477/DTTz8RGxtLtWrV6NixI15eXqxevdpc7sMkKiqKY8eO8cYbb5jb7r//flq2bMmsWbMYPHiwxQTr999/p2vXrnh4eBR5KPbBBx/kmWee4Z9//uHuu+8GjL2H33zzDXfffXeeYdv8VK1a1bxI5ueff+bEiRP51i80JZWqqjJnzhxq167NsGHDAOPQ+ubNm/OcM3v2bLZu3cpvv/1W5svf2JMkdnYW4Gy556l6Pu3lgjW16gqpT/fzoUt0alCN6t7Gv5BnDW2JRiPDrkLkyzewTCR02c2ePdvqY8+cOWOxhlizZs1KbP5abl26dGH8+PGMGTOGffv20a1bNzw9Pbl8+TLbt2+nZcuWPP300/zyyy8sXLiQBx54gAYNGqCqKmvXriUuLo4+ffqYr9eyZUu2bNnC+vXrqVWrFt7e3uY6aqbEIzo6Ot94TL11L7/8Mh06dMjz/O3bt/nrr7/45ptveO655/D29mbGjBm88MILGAwGRowYQdWqVTly5AjvvvsuwcHBTJkyxXy+Vqtl3bp19O3bl06dOvH000/Ts2dPPD09OX/+PD/88APr16/n1i3jYj5vb+8i9ZiOHTuWTz/9lGHDhjF79mwCAgJYuHAhJ06c4M8//8xx7L333svWrVtzzHFcs2YNly5domnTpqSmprJlyxY+/vhjJk6cmGPxCsDrr79Oy5YtqVWrFhcuXGDp0qX8888//Prrr+YFNm5ubhZLzyxfvhytVlussjTlklrJxMfHq4AaHx9fItdfuKyPylvkeSxa1qdE7lfivvhCVTUaVQXjx08/VdXERONj7lxju+m5L77Ic/rt1Az1+dWRavArv6iPR/yj6vUGB7wIIewvJSVFjYqKUlNSUhwdSolYtmyZCqh79+4t8LjmzZur3bt3z9EG5Pt48803VVVV1e7du6vNmze3a8xvvvmmCqjXr1/P0b506VL17rvvVj09PVV3d3e1YcOG6hNPPKHu27dPVVVVPX78uProo4+qDRs2VN3d3VVfX1+1Q4cO6vLly3Nc5+DBg2qXLl1UDw8PFcjxuv39/dWOHTvmG1t6eroaEBCgtmnTJt9jMjMz1aCgILVly5Y52r/77jv1nnvuUb29vVUnJye1bt266tNPP61euXLF4nXi4uLUt99+W73rrrtULy8v1dnZWa1bt6762GOPqTt27Mj3/ra4cuWK+sQTT6h+fn6qm5ub2rFjR3XTpk15juvevbuaO9VYt26d2qZNG/O/R7t27dSIiAjVYMj7++Hpp59W69atq7q4uKj+/v7qQw89pB4+fNiqGMPDw1VPT8+ivcBSVtjPE1tyF0VVi1nkqJxJSEjA19eX+Pj4HFXU7WXRt0N55nTeoomLQh5k4qi1dr9fidLpjIWHrS1rotUa926906t3RBfPlFWRnLuRhEaB5+4N5dleIWilp05UAKmpqZw7d4769evbZSsoUX5FRUXRvHlzfvnlFwYNGuTocEQ5VNjPE1tyFxmKtTfnfGov2WFlWqmztVbdnfp0htqBLN1xjvc2HidDr1Lb1415j7SlQ/2CV0oJIUR5tHnzZjp16iRJnSgTJLGzt4x8JtZmJJduHPZQUK06sFifLq52XaZ+uZctJ4y16fo1r8F7D7WiikcF23lDCFGiVFXNt8yHiVarLRMlkiZNmsQkU4UAIRxMthSzt4xE29rLMlOtOpPstepM9elMe+Deec4puC7RN5JwddLwzgMt+OyxMEnqhBA2+/LLL3F2di7wsXXrVkeHKUSZIz129uacT726/NrLuoJq1d2pT5d54iTa0EYodergBSwcFYZWo9C4ZjleCSyEcKghQ4YUWorDtCJVCJFFEjt7q0hz7HKzUKvugrsfk4+7cr9rJmPrGNua1bb/ohQhROVSrVo18w4EQgjrSWJnb675VJzPr70s0+ng8OGsr3PVqvvpYAyvrztKYlomMbdSeKRDHTxc5FtKCCGEcBSZY2dv7vms/MyvvayKiDCWOsm+yqtpU4iIICktk5e+P8Rzqw6SmJZJu+Cq/PRsF0nqhBBCCAeT38R2Vk1veeuwagbLW42VSTodjB+ft9SJwYA6YQJjon3Yk+GBRoFnezViSq8QnLTyN4IQQgjhaJLY2Vl9jZuxvnr2Ffgq1FPybkBdZhVQv07R69GcOU3NFh2Y90gbOjaQOTBCCCFEWSHdLHZ2zpCWM6kDUCDakOaQeIrEVL/OAoNGQ90Orfjtua6S1AkhhBBljCR2dhardbapvUzKVb/OvOecVgufLea95wZS1VNq0wkhhBBljSR2dlYh5tgBGY89bv78+afnkfbHnxAdjeapJ8tEpXchRMlavnw5iqKgKApbtmzJ87yqqoSEhKAoCj169MjxnKIoPPvsswVev0ePHubrK4qCu7s7rVu3Zt68eRis3MqwXr16Oa6R32P58uVWvmrLTO9FdHS0zedGR0fbJYaSctddd6EoCh9++KHF59966y0UReHGjRsWn2/RokWef3+Aq1ev8uqrr9KyZUu8vLxwc3OjUaNGPPfcc5w6dcousZ89e5ahQ4dSpUoVvLy86NOnDwcOHLDqXFVVmT9/Pk2aNMHV1ZVatWrx9NNPc+vWrTzHXr58mdGjRxMQEICbmxutWrUiIiLC4nV///13unTpgru7O76+vgwZMoRjx44V63XaSubY2VlFmGN38WYyL335D6vufO1zz90YurcHF61D4xJClD5vb28iIiLy/PLeunUrZ86cwdu76IXIGzRowLfffgvAtWvX+Oyzz3j++ee5fPky7733XqHnr1u3jrS0rGkuX3zxBREREWzcuBFfX19ze8OGDYscI8CgQYPYtWsXtWrVsvncWrVqsWvXrmLHUBIOHjxIZGQkABEREbz44ot2ue6ePXsYPHgwqqry7LPP0qlTJ1xcXDhx4gTffPMNHTp0sJhA2eL69et07dqVqlWrsnTpUtzc3Jg1axY9evRg7969hRavfvHFF5k3bx4vvvgivXv3JioqijfeeIO9e/eya9cunJ2No2zx8fHcc889pKen8/7771OrVi1WrlzJk08+SXx8PNOmTTNf86effuLBBx/k/vvvZ82aNcTHxzNjxgy6du3K3r17S+97QK1k4uPjVUCNj48vkeuv3jBN5S3yPL7b8EKJ3M/e/vh9rzrm8ffU7k8uVlUwPk6ccHRYQpQ5KSkpalRUlJqSkuLoUErEsmXLVEB98sknVXd39zw/Mx977DG1U6dOavPmzdXu3bvneA5QJ02aVOD1u3fvrjZv3jxHW3p6utqgQQPVw8NDTU9PtznmN998UwXU69evF3hcUlKSzdeuiCZNmqQC6qBBg1RA3bFjR55jCntPc//7x8fHqzVr1lTr1KmjXrx40eI533//fbFjf+mll1RnZ2c1Ojo6x739/f3V4cOHF3iuTqdTtVqtOnny5BztK1asUAF1yZIl5rZZs2apgLpv374cx/bt21f19PRUb926ZW5r3Lix2qpVK9VgMJjboqOjVRcXF3XkyJEFxlTYzxNbchcZirU3j+qW2z39SzcOG6Vm6Pnh6bfo1b8DS79+hf99MSHryTv164QQpUeXoGPzuc3oEnQOjePRRx8FYOXKlea2+Ph41qxZw9ixY+16L2dnZ8LCwkhOTub69et2uebo0aPx8vLiyJEj9O3bF29vb+69914ANm3axP33309QUBBubm6EhIQwYcKEPMOOloZie/ToQYsWLdi7dy9du3bFw8ODBg0aMHv27BxDyZaGYk3Dm8eOHePRRx/F19eXGjVqMHbsWOLj43PcOy4ujnHjxuHn54eXlxeDBg3i7NmzKIrCW2+9VeT3JTU1lRUrVhAWFsZHH30EwNKlS4t8PZPPP/+cK1eu8P777xMUFGTxmIcffrjY91m3bh29evUiODjY3Obj48PQoUNZv349mZmZ+Z67e/du9Ho9AwcOzNE+ePBgANasWWNu27FjBzVq1CAsLCzPsUlJSWzcuBGA2NhYTpw4wYABA3JMVwoODqZFixb8+OOP6PX6or9gG0hiZ2f13atmW21whwr13Mp2gWLny5cYuvhttKox+BzfGAYDTJhgrG8nhCiQqqokpScV67Fw70KC5wXT66teBM8LZuHehUW+lqrm/oFkGx8fHx5++OEcv/RXrlyJRqNhxIgRxX278jhz5gxOTk5UrVrVbtdMT0/nvvvuo1evXvz000/MmDHDfK9OnTqxaNEi/vjjD9544w3++ecf7rnnHjIyCp8XfeXKFUaNGsVjjz3Gzz//zIABA5g+fTrffPONVXE99NBDhIaGsmbNGl599VVWrFjB888/b37eYDAwZMgQVqxYwSuvvMK6deu4++676d+/f9HeiGzWrl3LrVu3GDt2LI0aNeKee+5h9erVJCYmFuu6f/zxB1qtliFDhlh1vKqqZGZmWvUwSUlJ4cyZM7Rq1SrP9Vq1akVKSgpnz57N957p6ca58K6uOadIOTs7oygKh7PtuJSenp7nuOznmo7N75qmtuTkZM6cOZNvTPYkc+zsLNGjqsVyJ0ke9vshZS+qqpJpUHHWatCeOQ1qAROW9Xo4fdq4YlYIka/kjGS8ZnnZ7XoG1cCkDZOYtGFSkc5PnJ6Ip0vxtjQcO3YsPXv25NixYzRv3pylS5cybNiwYs2vMzH9wr5+/Trz58/nwIEDDBs2DHf3fPbdLoKMjAzeeOMNxowZk6N94sSJ5s9VVaVz58706NGD4OBgfvvtN+67774CrxsbG8uGDRvo0KEDAL1792bLli2sWLGCJ554otC4xo0bx0svvWQ+9/Tp0yxdupSIiAgURWHjxo1s376dRYsWmWPt06cPLi4uTJ8+3ab3ILeIiAjc3NwYOXKkOZYxY8bw3XffFasn9sKFC1SvXh1PT+u+57788ss8/y75Mf2RcuvWLVRVxc8vb4eJqS02Njbf6zRr1gww9sb17NnT3L5z505UVc1xbrNmzfjzzz+5cOECdevWNbdv3749x31q1KiBn58fO3bsyHGvuLg4jh49WmhM9iQ9dnbm5exhscfO09l+P6Ts4WZSOk99tY+Z66OMDQXUrgOMpU5CQkonOCFEmdK9e3caNmzI0qVLOXLkCHv37rXLMOyxY8dwdnbG2dmZ2rVrM2fOHEaNGsXnn39uh6hzeuihh/K0Xbt2jYkTJ1KnTh2cnJxwdnY2D+39+++/hV6zZs2a5qTOpFWrVpw/f96qmHInjq1atSI1NZVr164BxgUqAMOHD89xnGl4vKjOnTvH5s2bzStKAXOibo/hWFsMGTKEvXv3WvXIraAKDQU917p1a7p168YHH3zA999/T1xcHDt37mTixIlotVo02X4Xjh8/HmdnZ0aNGsWxY8eIjY3l008/ZfXq1QDmYzUaDZMmTeKvv/7i7bff5tq1a5w+fZrHHnuM5OTkHMeWNOmxs7PEm6ct99jdzL9buLTtOhPL1NWRXE1Iw8VJw/huDahjql036U6vgKIYHwaDMalbvFh664SwgoezB4nTiz6cFZMQQ9OFTTFk60HXKlqinoki0CewSPEUl6IojBkzhvnz55OamkpoaChdu3Yt9nUbNmzIqlWrUBQFNzc36tevj4dH8ePNzcPDAx8fnxxtBoOBvn37cunSJf7zn//QsmVLPD09MRgMdOzYkZSUlEKvW61a3iLtrq6uVp1r6XzTMJ7p/NjYWJycnPL0TNWoUcOq6+dn6dKlqKrKww8/TFxcnLn9vvvu49tvv+X48eM0adIEACcnY5qQ3/ywzMxM8wpSgLp163Lq1CmSkpKs6rXz8/PLsYLZGlWrVkVRFIs9YDdv3jRftyDff/89o0ePNifNLi4uPP/88/z555853pOmTZuybt06JkyYQIsWLQCoU6cOc+bMYfLkyQQGZv2ffOONN0hMTOSdd97hjTfeAIwrqseMGcMXX3yR49iSJD12dubl5me5x86tiiPCySFTb2DOHycY+cVuriak0aC6J+ue6Uwdvzs/SMPDsw4+fhzOn4fNmyE6GsaNc0jMQpQ3iqLg6eJZ5EeofyhLBi9BqxjLC2kVLYsHLybUP7RI17NX3cnRo0dz48YNPvvsM6uHzgrj5uZGu3btCAsLo3nz5iWS1IHl3pujR49y6NAhPvjgAyZPnkyPHj1o3769xWTNUapVq0ZmZqY5WTG5cuVKka9pMBjMCzmGDh1K1apVzQ9T6ZnsvXamJDImJibPtVRV5fLlyzkSzX79+qHX61m/fr1V8Xz55ZfmXtvCHibu7u6EhIRw5MiRPNc7cuQI7u7uNGjQoMD7BgQEsGHDBq5evcqhQ4e4du0aM2fO5OTJk3Tr1i3HsQMGDOD8+fOcPHmSqKgozp07Z/4+yX6sk5MTc+fOJTY2lsOHD3Pp0iV++eUXLly4QP369fNdTGJv0mNnZ4mpNy332KXGOSIcM92tZJ5bdZD95421g0a0q8Ob9zXDwyWfb4HAQPD0lF46IRxg3F3j6BfSj9M3TxPiF0KQj+P/HwYGBvLSSy9x/PhxwrP/EVhOmZK93JPdFy9e7IhwLOrevTvvv/8+q1ev5umnnza3r1q1qoCzCvb777+j0+mYNGmSxdWpzz77LF999RXvvvsuTk5O9OrVC0VRWL16NXfddVeOYzdu3EhCQgK9e/c2t40bN44PPviAl19+ma5du1rspVq7di1Dhw4FsoZibfXggw8yb948Ll68SJ06dQC4ffs2a9eu5b777jP3NBYmICCAgIAAAObPn09SUpLF4tqKotCoUSPAuFDi448/pk2bNnmSQAAvLy9atmwJwIEDB/jrr7+YM2eOza+xqCSxs7NGdbqgqKBmS+40QEidTg6LKVNvYOTn/3DhZjLerk68O7QlQ1rXzntg9r/IYmIgNLT0ghRC5BDkE1QmErrsZs+ebfWxZ86c4YcffsjT3qxZM/PkdUdq0qQJDRs25NVXXzVPxF+/fj2bNm1ydGhm/fv3p0uXLrzwwgskJCQQFhbGrl27+Oqrr4C8c7YURaF79+4WdwoxiYiIwMnJiddee43atfP+HpgwYQJTpkzh119/5f7776dhw4Y8++yzfPDBB8TFxTFw4EDc3d3Zu3cvs2fPpl27duYFGAC+vr789NNPDB48mLZt2+YoUHzq1Cm++eYbDh06ZE7sqlWrVqRe0hdffJGvv/6aQYMGMXPmTFxdXZk9ezapqal5ysCE3Jkffvr0aXObaR5nw4YNiYuL47fffiMiIoJ33303TwJr6tGtVq0aZ8+eZf78+eh0OvMcSJMtW7awd+9eWrVqhaqq7Nmzh/fee4/+/fsXuhOLPUliVwqKWW2g2Jy0Gl4f1JTPtp5h/iNts4Zes4uIgKeeyvq6SRP4/HMZghVCFMnGjRvNNb6ye/PNN4tVf81enJ2dWb9+Pc899xwTJkzAycmJ3r178+eff+ZY/ehIGo2G9evX88ILLzB79mzS09Pp0qUL33zzDR07djQvfADMZUoK2h3jxo0brF+/nsGDB1tM6gAef/xxXnnlFSIiIrj//vsB+Pjjj2nWrBkRERF88803ZGZmEhwczKRJk/i///s/XFxy7h3eoUMHjhw5wkcffcR3333He++9h16vp06dOtx7770sWLCgmO8MVK9enW3btvHiiy8SHh5OZmYmnTp1YsuWLeb5gSaWatqpqsq8efM4f/48Go2Gtm3bsm7dOvNrzu7ixYtMnjyZGzduUK1aNfr3789PP/2Uo4YeGOfprVmzhnfeeYe0tDQaNWrEzJkzmTJlClpt6e3cpKjFLXJUziQkJODr60t8fHyeybT2sHnXPHr98Xze9r4f06PTFLvfLz/HryQQm5hOl5CswsgGg4pGY2G+jU4HdevmzUC1WuP8OhmOFSKP1NRUzp07R/369XFzc3N0OKISWbFiBaNGjWLHjh107twZgA0bNjB48GAOHTpkHgYU5UdhP09syV2kx87OGtXpgkYFQ7b8SUvpDcWqqso3u8/z9q//4umiZePUbtTwMX6TWEzqAE6dstytKLXrhBDCoVauXElMTAwtW7ZEo9Gwe/duPvjgA7p162ZO6gA2b97MI488IkmdkMTO3oK8a9MXLRu5szRchcdwJsjbcre3PcUlp/PyD4f5I+oqAF0aVsMpv2Quu0aNjKVNLPXYSe06IYQDqKpa6BZMWq3Wbqt+yypvb29WrVrFO++8Q1JSErVq1WL06NG88847OY774IMPHBShKGuk3Imd6S7u4g+y/TBS4Bsy0F3cXaL3/edsLAM+3sYfUVdx1ir8Z3Azlo5uTzWvvNub5BEUBLnnPGg0UrtOCOEw1pTByD15vSIaPHgw+/btIy4ujoyMDC5cuMD8+fNLZCqRqBikx87OTmHIMQwLoFfgNAZKIkVSVZV5f57ik/+dwqBCA39P5j/alhaBthV8JDw8qzjxV19Bz56S1AkhHMaaMhiNGzcupWiEKD8ksbOzRnU7o0HBkK1KsVbREFK3ZObYKYrCtdupGFQYFhbEW/c1x9O1mP+sQ4caa9gJIYSDFLUMhhCVnSR2dhbkE8QLtTryweVdAGhVWNx6tN3rUaVHX8Dl3Blo1Ig3BjenR+MA+jWvaZ+LSw07IYQQolySOXb2Fh8Dl/ebvzQAHFplbLeD1Aw9ayfNwKlBfejVC4KDcf9iMf3qeUNSUtEfS5Zk3aRpU2NdOyGEEEKUK9JjZ2e6i7uYo6abtxVTFZigJtPv4m6CfB8q1rVPXLnNjEUb+XrRDDSmFawGg3FunGl+nD0YDDBhAvTrJ/PshBBCiHJEeuzsrKDFE0Vlqk1334LtqCdPoy2NmtKmGnZCCCGEKDekx87OjIsnyJHGFWfxRFxyOq+uOcLGY1cAqBHWAvU7DYoh2x20WoiKAgubLVslJsY4/Jr7mlLDTgghhChXJLGzsyCfIPr4NuD3+LPGBhUeC+pc5MUTz66IZPvpGzhrFV7p34SxXeqjeHySNfSq1RrrzRVnsUNoqHGO3YQJxp460zVlGFYIIYQoV2Qo1s50ur1sMiV1YCxQrNuOTldwPab8vDqgCaE1vFj7dBee7NrAuC1YeHjWAVFRMG5cMaPGeI3oaNi82fjRHtcUQgghRKmSxM7OTl3ckWc2nR44fXGXVedfjk9hw5HL5q9bBPqy8blutAzKp+BwUYdfLQkKgh49pKdOCMHy5ctRFAVFUdiyZUue51VVJSQkBEVR6NGjR47nFEXh2WefLfD6PXr0MF9fURTc3d1p3bo18+bNw2Cwfk7yxYsXefDBB2nQoAGenp74+vrStm1bFixYQGZmptXXsSQ6OhpFUVi+fHmxrpORkUHNmjVRFIUffvjB4jGjR4/Gy8sr32t4eXkxevToPO1nz57l2WefJTQ0FHd3dzw8PGjevDn/93//R0yMfaoxHDhwgN69e+Pl5UWVKlUYOnQoZ8+eLfxE8v47mx79+/fPc+zJkyd56KGHqFq1Kh4eHtx99938/PPPFq+7Zs0aunTpgp+fH1WqVKFDhw58/fXXxXqdFYUkdnbWqE6XPG+qFgipU/gcu9+PXWHAx9t4blUkR2Pize2aSzHGnjSdLu9JdvqPK4QQlnh7exNhofzR1q1bOXPmDN7e3kW+doMGDdi1axe7du1i9erVBAYG8vzzzzN9+nSrr5GUlISPjw//+c9/+Pnnn1m1ahX33HMPkydPZuLEiUWOzZ5++eUXrl417uFt6b0sznVbtWrFL7/8wvjx4/nll1/Mn69fv57BgwcX+x7Hjx+nR48epKen891337F06VJOnjxJ165duX79ulXXyP7vbHrMmzcvxzHR0dF06tSJEydO8Nlnn/H9999TvXp1HnjgAdasWZPj2KVLl/Lwww9Tq1Ytvv32W1atWkXDhg154okn+Oijj4r9mss9tZKJj49XATU+Pr7E7vHi4o4qb6HyFqr2LdQv1o0t8PiU9Ez1/9YdUYNf+UUNfuUXdcgn29ToG4nGJ7/4QlU1GlUF48dPP1XVuXONX5vavviixF6LEMKylJQUNSoqSk1JSXF0KCVi2bJlKqA++eSTqru7e56fmY899pjaqVMntXnz5mr37t1zPAeokyZNKvD63bt3V5s3b56jLT09XW3QoIHq4eGhpqenFyv+4cOHq05OTmpqamqBxyUlJeX73Llz51RAXbZsWbFiGTRokOri4qL26dNH1Wg06sWLF/McEx4ernp6euZ7DU9PTzU8PNz89dmzZ1VPT0+1bdu2alxcXJ7jDQaDumbNmmLFraqqOmzYMNXf3z/Hv390dLTq7Oysvvzyy4Web+nf2ZIJEyaobm5uqk6nM7dlZmaqTZs2VevUqaPq9Xpze5cuXdTg4OAcbQaDQW3SpInaqlUra19amVLYzxNbchfpsSsBg1qNBKCu4kT0uD2MeyD/v9BOXb3NA5/u4Ovd5wGY0K0BP0zsTHA1T2MP3fjxWatVTTXrpk3LuoCp5pyl3jwhRPml0+XfU1+KHn30UQBWrlxpbouPj2fNmjWMHTvWrvdydnYmLCyM5ORkq3uD8lO9enU0Gg1ardbc1qNHD1q0aMHff/9N586d8fDwML+GS5cuMXz4cLy9vfH19WXEiBFcuXKlWDGYrrtx40aGDBnCSy+9hMFgKPbQLsDcuXNJSkpi4cKF+PrmnaqjKApDhw4t1j0yMzP55ZdfeOihh/Dx8TG3BwcH07NnT9atW1es62e3Y8cOWrduTWC26UVarZYBAwZw8eJF9uzZY253dnbGy8sLjSYrhVEUBR8fH9zc3OwWU3kliV1JuFNnzlPjRFBQ+3wPW733AkMWbOf4ldv4e7nw1dgOTB/YFBenO/8sp07lLEGSH6k5J0TZoarF2wUmKQkWLoTgYPPuMixcWPRrFbPupY+PDw8//DBLly41t61cuRKNRsOIESOK+27lcebMGZycnKhatapN56mqSmZmJrdu3WL16tUsX76cF154ASennMUfLl++zGOPPcbIkSPZsGEDzzzzDCkpKfTu3Zs//viDWbNm8f3331OzZk27vL7ly5ej1+sZO3YsvXv3Jjg4mKVLl6IW89/ljz/+oEaNGnTs2NGq4w0GA5mZmYU+9Hq9+ZwzZ86QkpJCq1at8lyvVatWnD59mtTU1ELvfebMGfz8/HBycqJhw4a8/vrrpKSk5DgmPT0dV1fXPOea2g4fPmxumzx5Mv/++y///e9/uX79Ojdu3ODDDz9k//79vPjii1a9HxWZlDtxoLjkDFIzDHRt5M/c4W2o7p3rm7pRI9BociZ3pr9QpOacEGVTcjIUMAneZsXdXSYxETw9ixXC2LFj6dmzJ8eOHaN58+YsXbqUYcOGFWt+nYlpgcP169eZP38+Bw4cYNiwYbi7u9t0nffee888N09RFF577TXeeeedPMfdvHmT77//nl69epnbPvvsM/79919++ukn7rvvPgD69u1LSkoKn3/+eVFfGqqqsmzZMgIDA+nXrx+KojB69GhmzJjB5s2bc8RgqwsXLtCmTRurj585cyYzZswo9Ljg4GCio6MBiI2NBcDPzy/PcX5+fqiqyq1bt6hVq1a+17vnnnsYMWIETZo0ISUlhd9++43333+f7du3s3nzZnOvW7NmzdiyZQuJiYk5FpFs3749RywAQ4cOZe3atYSHh/N///d/ALi7u/Pll18ybNiwQl9jRSeJXSlLzzSYe+Se6tqA2lXcGdSylrGMSW5BQfCJhZp1IDXnhBClpnv37jRs2JClS5cyevRo9u7dy5w5c4p93WPHjuHs7Gz+2tnZmVGjRvHpp5/afK3Ro0fTu3dvbt68yf/+9z8++OAD4uPj+eSTT3IcV7Vq1TwJ1ebNm/H29jYndSYjR44sVmK3detWTp8+zWuvvWYeEh4zZgwzZ85k6dKlxUrsbDV+/HirFlNY6jVTFAu/n6x4DsiTXA8cOJB69erx4osv8tNPP/Hggw8C8Oyzz/LTTz/xxBNP8OGHH+Lp6cmCBQvYuXMnQI5h140bN/LYY48xbNgwhg8fjpOTEz///DOjR48mPT2dMWPGFPo6KzJJ7EqJ3qCycPNpfjl8mXWTOuPh4oRGozCkde2CTwwPz0rsoqKyChH362ccfg0JkaROiLLEw8PYS1ZU+e0EU9TdZTw8ih7LHYqiMGbMGObPn09qaiqhoaF07dq12Ndt2LAhq1atQlEU3NzcqF+/Ph5FjLdmzZrUrFkTMPa2Va1alVdffZWxY8fStm1b83GWepdiY2OpUaOGxWsWh2kF7IMPPkhcXBwAvr6+3HPPPaxZs4YFCxZQpUoVAJycnHIMg+aWmZmZIwmuW7cu586dszqWmjVrEhAQUOhx2RO1atWqATl7y0xu3ryJoijm+G3x2GOP8eKLL7J7925zYnfvvfeybNkyXnjhBRo2bAgYe/HefvttXnvtNfPcO1VVGTt2LN26dcsxPaB3797Ex8czefJkhg8fjmcxe6nLM4fPsVu4cCH169fHzc2NsLAwtm3bVuDxW7duJSwsDDc3Nxo0aMBnn31WSpEW3ZX4VEZ9sZs5m05y4upt1h+6VLQLZf+hLjXnhCibFMU49FnUh2knGNOk/+y7yxTleoX0qFhr9OjR3Lhxg88++8xuPSJubm60a9eOsLAwmjdvXuSkzpIOHToAxtpo2VnqYapWrZq5HEl2xVk8YVpgAtC+fXuqVq1qfmzbto3U1FRWrFhhPr5GjRqkpqZy8+bNPNeKjY0lLS0tR/LZr18/rl69yu7du62KZ+bMmTg7Oxf6MCVVYEy83d3dOXLkSJ7rHTlyhJCQkGItVsjeCwcQHh7OlStXiIqK4tSpUxw7dgww/puZ/pC4evUqly9fNv/7Zte+fXuSkpLMQ8mVlUMTu9WrVzN16lRef/11IiMj6dq1KwMGDODChQsWjz937hwDBw6ka9euREZG8tprrzFlypQ8NW7KgnY6eGR7On8tmU//j/9m99mbeLhomTu8NSPa1y3aRaVmnRCVQxncCSYwMJCXXnqJIUOGEJ5995syavPmzQCEWDH/uGfPnty+fTtPMdzsiZetVqxYQUpKCm+//TabN2/O8/D398/T4wTG34u5fffddzmOAXj++efx9PTkmWeeIT4+Ps85qqrmWLU6fvx49u7dW+hj/fr15nOcnJwYMmQIa9eu5fbt2+b2CxcusHnz5iKvuv3yyy8BLC78cHJyomnTpoSEhBAfH8+SJUu4//77CQ4OBoxD6W5ubhYT2l27dqHRaAqc81cp2K8Ki+06dOigTpw4MUdbk/9v716DojrPOID/l12uC4uJMVwERdQVNcYLFBVHHQ1RByvRqLGVMepoAjGKkRqKk46QtthJrcR4bx0L0cF6C1gTsUq8oKLTIEq8YBWVWC8g1URAvCDw9INl68qC7roX3P3/Zs6H8+57znmOj7vn4ex53w0JkaSkJIP9ExMTJSQkRK8tNjZWBgwY8MzHtMY8dt8NeEUa/jfPXAMg377mK28v3i2lP9wQuXPHuIVz1hG1So4yj11BQUGL/Zqbx27UqFGydevWJsuZM2dE5NnnN3uahQsXSmxsrGRmZsqBAwdk+/btEhcXJ0qlUiZOnKjXt7lj1tTUiFarFW9vb1mxYoXs3r1b5s6dKx06dGgyj13j3HaPzylnSGhoqLz00kvN/v9ISEgQAFJUVKRri46O1s0Pt3PnTvnmm28kMTFRnJ2dJTo6usk+vv76a/Hw8JCgoCD505/+JHv37pW9e/fK8uXLpW/fvtKnT58WY3wWZ8+eFU9PTxkyZIjk5ORIVlaWvPbaa+Lv7y8VFRV6fZVKpQwfPly3fvDgQRk5cqSsWbNG9uzZIzt27JAPPvhA1+/xeehu3LghiYmJ8ve//1327dsnq1atkqCgIAkODpZr164Z/LebMmWKfPPNN7Jr1y6JjY0VADJjxoznPmdbMOc8djYr7B48eCBKpVKysrL02uPj42XIkCEGtxk8eLDEx8frtWVlZYlKpWp2Msv79+9LZWWlbrly5YpFC7vyr9N1RZ1FFqVSxMDklkRkXSzsHmmusGtuSU5OFhHzFXY7duyQyMhI8fHxEZVKJZ6enhIeHi7Lli2Thw8f6vVt6ZhXr16V8ePHi6enp3h5ecn48ePlyJEjTQq7U6dOCYBmb0CIiHz//fcCQD766KNm+/zrX/8SADJnzhxdW21trSxatEh69uwprq6u4urqKj179pRFixY1e427ePGizJo1S7p06SKurq7i7u4uPXr0kISEBCktLW32+MY4duyYvPHGG+Lh4SEajUbGjh0rFy5caNIPgN7/hZKSEomKipL27duLq6uruLm5Sa9evSQ1NbXJxNG3bt2SESNGSLt27cTZ2Vk6dOggc+bMkf/85z9NjlNfXy9r166VsLAwadOmjWg0Gunbt6+sWLHiuSe2thW7KOyuXbsmACQ/P1+vPTU1VbRarcFtunbtKqmpqXpt+fn5AkCuX79ucJvk5GSDHy6WKuzOz3nLckVd47J/v0ViJ6JnZ++FHRm2cuVKUavVUl5ebutQyI7Y1S9PPPkgq4gYNbRa/jfJY3PbLFiwAJWVlbrlypUrzxlxyzQjxuLJaScFwI2tax6NlDNmOXfu//PWNeKcdURENrN//37Ex8cbHEVL1BrYbLqTV155BUqlssmoo4qKimbfML6+vgb7q1Qq3bDsJ7m6uhqcl8dSfH4+DeeHpaLr/gtQ4FFRVzKsC7QTYo3fWePoOM5ZR0QORkRanP4DePSTU0+bR83ctm7datXjERnLZnfsXFxcEBoaitzcXL323NxcREREGNxm4MCBTfrv2bMHYWFhevP72Jp2Xwkqvk7HhfixqPg6Hdp9JabvrBWOjiMisrQvv/zyqVNz5OXl2TpMolbHphMUJyQkYMqUKQgLC8PAgQPxl7/8Bf/+978RFxcH4NHXqNeuXcP69esBAHFxcVixYgUSEhLw3nvv4ejRo1i3bp3ej1O3Fj4/nwafn08zz84CAniXjogcypgxY1BQUNBin27dulkpGqIXh00Lu0mTJuHWrVv47W9/i7KyMrz22mvIycnRzVdTVlamN6ddp06dkJOTg3nz5mHlypXw9/fHsmXLMH78eFudAhERWUDbtm2bfcSGiJqnkMbRBw6iqqoK3t7eqKyshEajsXU4RPSCun//PkpLS3W/nENEZKqnfZ4YU7vYfFQsEdGLzMH+NiYiCzDn5wgLOyIiEzQO2Lp7966NIyGiF11NTQ0UCoVZBoLa9Bk7IqIXlVKpRJs2bVBRUQEA8PDwsPrUG0T04hIR1NXVoaqqClVVVWjTpg2USuVz75eFHRGRiXx9fQFAV9wRERlLqVTCz88P3t7eZtkfCzsiIhMpFAr4+fnh1VdfxcOHD20dDhG9YFQqldkn2mZhR0T0nJRKpVm+QiEiel4cPEFERERkJ1jYEREREdkJFnZEREREdoKFHREREZGdYGFHREREZCccblRs4892VFVV2TgSIiIioqdrrFme5afHHK6wq66uBgAEBgbaOBIiIiKiZ1ddXf3UiYwV4mC/YN3Q0IDr16/Dy8vLoj//U1VVhcDAQFy5cgUajcZix6Fnx5y0TsxL68OctE7MS+tkjbyICKqrq+Hv7w8np5afonO4O3ZOTk4ICAiw2vE0Gg3fgK0Mc9I6MS+tD3PSOjEvrZOl8/KsPznGwRNEREREdoKFHREREZGdYGFnIa6urkhOToarq6utQ6H/YU5aJ+al9WFOWifmpXVqbXlxuMETRERERPaKd+yIiIiI7AQLOyIiIiI7wcKOiIiIyE6wsHsOq1atQqdOneDm5obQ0FAcOnSoxf55eXkIDQ2Fm5sbgoODsWbNGitF6jiMyUlWVhbefPNNtGvXDhqNBgMHDsTu3butGK3jMPa90ig/Px8qlQp9+vSxbIAOyNicPHjwAJ988gk6duwIV1dXdO7cGX/961+tFK3jMDYvmZmZ6N27Nzw8PODn54fp06fj1q1bVorW/h08eBBjxoyBv78/FAoFtm/f/tRtbH6tFzLJpk2bxNnZWdauXSvFxcUyd+5cUavVcvnyZYP9L126JB4eHjJ37lwpLi6WtWvXirOzs2zbts3KkdsvY3Myd+5c+eyzz+S7776T8+fPy4IFC8TZ2VmOHz9u5cjtm7F5aXT79m0JDg6WESNGSO/eva0TrIMwJSfR0dHSv39/yc3NldLSUvnnP/8p+fn5Voza/hmbl0OHDomTk5N88cUXcunSJTl06JD07NlTxo4da+XI7VdOTo588skn8tVXXwkAyc7ObrF/a7jWs7AzUXh4uMTFxem1hYSESFJSksH+iYmJEhISotcWGxsrAwYMsFiMjsbYnBjSo0cP+fTTT80dmkMzNS+TJk2S3/zmN5KcnMzCzsyMzcmuXbvE29tbbt26ZY3wHJaxeVm8eLEEBwfrtS1btkwCAgIsFqMje5bCrjVc6/lVrAlqa2tRWFiIESNG6LWPGDECR44cMbjN0aNHm/QfOXIkjh07hocPH1osVkdhSk6e1NDQgOrqarz88suWCNEhmZqX9PR0XLx4EcnJyZYO0eGYkpMdO3YgLCwMf/zjH9G+fXtotVrMnz8f9+7ds0bIDsGUvERERODq1avIycmBiODGjRvYtm0bRo8ebY2QyYDWcK13uN+KNYebN2+ivr4ePj4+eu0+Pj4oLy83uE15ebnB/nV1dbh58yb8/PwsFq8jMCUnT1qyZAlqamrwzjvvWCJEh2RKXkpKSpCUlIRDhw5BpeJHlLmZkpNLly7h8OHDcHNzQ3Z2Nm7evIlZs2bhxx9/5HN2ZmJKXiIiIpCZmYlJkybh/v37qKurQ3R0NJYvX26NkMmA1nCt5x2756BQKPTWRaRJ29P6G2on0xmbk0Z/+9vfkJKSgs2bN+PVV1+1VHgO61nzUl9fj8mTJ+PTTz+FVqu1VngOyZj3SkNDAxQKBTIzMxEeHo6oqCikpaUhIyODd+3MzJi8FBcXIz4+HgsXLkRhYSH+8Y9/oLS0FHFxcdYIlZph62s9/xw2wSuvvAKlUtnkr6iKioomlXojX19fg/1VKhXatm1rsVgdhSk5abR582bMmDEDW7duRWRkpCXDdDjG5qW6uhrHjh3DiRMnMHv2bACPigoRgUqlwp49ezB8+HCrxG6vTHmv+Pn5oX379vD29ta1de/eHSKCq1evomvXrhaN2RGYkpc//OEPGDRoED7++GMAwOuvvw61Wo3Bgwfj97//Pb8JsoHWcK3nHTsTuLi4IDQ0FLm5uXrtubm5iIiIMLjNwIEDm/Tfs2cPwsLC4OzsbLFYHYUpOQEe3ambNm0aNm7cyOdSLMDYvGg0Gpw6dQpFRUW6JS4uDt26dUNRURH69+9vrdDtlinvlUGDBuH69eu4c+eOru38+fNwcnJCQECAReN1FKbk5e7du3By0r+MK5VKAP+/S0TW1Squ9VYbpmFnGoelr1u3ToqLi+Wjjz4StVotP/zwg4iIJCUlyZQpU3T9G4dAz5s3T4qLi2XdunWc7sTMjM3Jxo0bRaVSycqVK6WsrEy33L5921anYJeMzcuTOCrW/IzNSXV1tQQEBMiECRPkzJkzkpeXJ127dpWZM2fa6hTskrF5SU9PF5VKJatWrZKLFy/K4cOHJSwsTMLDw211CnanurpaTpw4ISdOnBAAkpaWJidOnNBNQdMar/Us7J7DypUrpWPHjuLi4iL9+vWTvLw83WtTp06VoUOH6vU/cOCA9O3bV1xcXCQoKEhWr15t5YjtnzE5GTp0qABoskydOtX6gds5Y98rj2NhZxnG5uTs2bMSGRkp7u7uEhAQIAkJCXL37l0rR23/jM3LsmXLpEePHuLu7i5+fn4SExMjV69etXLU9mv//v0tXida47VeIcL7tURERET2gM/YEREREdkJFnZEREREdoKFHREREZGdYGFHREREZCdY2BERERHZCRZ2RERERHaChR0RERGRnWBhR0RERGQnWNgRUauRkZGBNm3a2DoMkwUFBWHp0qUt9klJSUGfPn2sEg8ROR4WdkRkVtOmTYNCoWiyXLhwwdahISMjQy8mPz8/vPPOOygtLTXL/gsKCvD+++/r1hUKBbZv367XZ/78+di7d69ZjtecJ8/Tx8cHY8aMwZkzZ4zez4tcaBM5IhZ2RGR2o0aNQllZmd7SqVMnW4cFANBoNCgrK8P169exceNGFBUVITo6GvX19c+973bt2sHDw6PFPp6enmjbtu1zH+tpHj/PnTt3oqamBqNHj0Ztba3Fj01EtsPCjojMztXVFb6+vnqLUqlEWloaevXqBbVajcDAQMyaNQt37txpdj/ff/89hg0bBi8vL2g0GoSGhuLYsWO6148cOYIhQ4bA3d0dgYGBiI+PR01NTYuxKRQK+Pr6ws/PD8OGDUNycjJOnz6tu6O4evVqdO7cGS4uLujWrRs2bNigt31KSgo6dOgAV1dX+Pv7Iz4+Xvfa41/FBgUFAQDGjRsHhUKhW3/8q9jdu3fDzc0Nt2/f1jtGfHw8hg4darbzDAsLw7x583D58mWcO3dO16elfBw4cADTp09HZWWl7s5fSkoKAKC2thaJiYlo37491Go1+vfvjwMHDrQYDxFZBws7IrIaJycnLFu2DKdPn8aXX36Jffv2ITExsdn+MTExCAgIQEFBAQoLC5GUlARnZ2cAwKlTpzBy5Ei8/fbbOHnyJDZv3ozDhw9j9uzZRsXk7u4OAHj48CGys7Mxd+5c/OpXv8Lp06cRGxuL6dOnY//+/QCAbdu24fPPP8ef//xnlJSUYPv27ejVq5fB/RYUFAAA0tPTUVZWplt/XGRkJNq0aYOvvvpK11ZfX48tW7YgJibGbOd5+/ZtbNy4EQB0/35Ay/mIiIjA0qVLdXf+ysrKMH/+fADA9OnTkZ+fj02bNuHkyZOYOHEiRo0ahZKSkmeOiYgsRIiIzGjq1KmiVCpFrVbrlgkTJhjsu2XLFmnbtq1uPT09Xby9vXXrXl5ekpGRYXDbKVOmyPvvv6/XdujQIXFycpJ79+4Z3ObJ/V+5ckUGDBggAQEB8uDBA4mIiJD33ntPb5uJEydKVFSUiIgsWbJEtFqt1NbWGtx/x44d5fPPP9etA5Ds7Gy9PsnJydK7d2/denx8vAwfPly3vnv3bnFxcZEff/zxuc4TgKjVavHw8BAAAkCio6MN9m/0tHyIiFy4cEEUCoVcu3ZNr/2NN96QBQsWtLh/IrI8lW3LSiKyR8OGDcPq1at162q1GgCwf/9+LFq0CMXFxaiqqkJdXR3u37+PmpoaXZ/HJSQkYObMmdiwYQMiIyMxceJEdO7cGQBQWFiICxcuIDMzU9dfRNDQ0IDS0lJ0797dYGyVlZXw9PSEiODu3bvo168fsrKy4OLigrNnz+oNfgCAQYMG4YsvvgAATJw4EUuXLkVwcDBGjRqFqKgojBkzBiqV6R+lMTExGDhwIK5fvw5/f39kZmYiKioKL7300nOdp5eXF44fP466ujrk5eVh8eLFWLNmjV4fY/MBAMePH4eIQKvV6rU/ePDAKs8OElHLWNgRkdmp1Wp06dJFr+3y5cuIiopCXFwcfve73+Hll1/G4cOHMWPGDDx8+NDgflJSUjB58mTs3LkTu3btQnJyMjZt2oRx48ahoaEBsbGxes+4NerQoUOzsTUWPE5OTvDx8WlSwCgUCr11EdG1BQYG4ty5c8jNzcW3336LWbNmYfHixcjLy9P7itMY4eHh6Ny5MzZt2oQPPvgA2dnZSE9P171u6nk6OTnpchASEoLy8nJMmjQJBw8eBGBaPhrjUSqVKCwshFKp1HvN09PTqHMnIvNjYUdEVnHs2DHU1dVhyZIlcHJ69Hjvli1bnrqdVquFVqvFvHnz8Mtf/hLp6ekYN24c+vXrhzNnzjQpIJ/m8YLnSd27d8fhw4fx7rvv6tqOHDmid1fM3d0d0dHRiI6OxocffoiQkBCcOnUK/fr1a7I/Z2fnZxptO3nyZGRmZiIgIABOTk4YPXq07jVTz/NJ8+bNQ1paGrKzszFu3LhnyoeLi0uT+Pv27Yv6+npUVFRg8ODBzxUTEZkfB08QkVV07twZdXV1WL58OS5duoQNGzY0+Wrwcffu3cPs2bNx4MABXL58Gfn5+SgoKNAVWb/+9a9x9OhRfPjhhygqKkJJSQl27NiBOXPmmBzjxx9/jIyMDKxZswYlJSVIS0tDVlaWbtBARkYG1q1bh9OnT+vOwd3dHR07djS4v6CgIOzduxfl5eX46aefmj1uTEwMjh8/jtTUVEyYMAFubm6618x1nhqNBjNnzkRycjJE5JnyERQUhDt37mDv3r24efMm7t69C61Wi5iYGLz77rvIyspCaWkpCgoK8NlnnyEnJ8eomIjIAmz5gB8R2Z+pU6fKW2+9ZfC1tLQ08fPzE3d3dxk5cqSsX79eAMhPP/0kIvoP6z948EB+8YtfSGBgoLi4uIi/v7/Mnj1bb8DAd999J2+++aZ4enqKWq2W119/XVJTU5uNzdBggCetWrVKgoODxdnZWbRaraxfv173WnZ2tvTv3180Go2o1WoZMGCAfPvtt7rXnxw8sWPHDunSpYuoVCrp2LGjiDQdPNHoZz/7mQCQffv2NXnNXOd5+fJlUalUsnnzZhF5ej5EROLi4qRt27YCQJKTk0VEpLa2VhYuXChBQUHi7Owsvr6+Mm7cODl58mSzMRGRdShERGxbWhIRERGROfCrWCIiIiI7wcKOiIiIyE6wsCMiIiKyEyzsiIiIiOwECzsiIiIiO8HCjoiIiMhOsLAjIiIishMs7IiIiIjsBAs7IiIiIjvBwo6IiIjITrCwIyIiIrITLOyIiIiI7MR/ATQD3WAsKpiGAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_MLP(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_527\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_527 (Embedding)   (None, None, 212)         217088    \n",
      "                                                                 \n",
      " global_max_pooling1d_160 (G  (None, 212)              0         \n",
      " lobalMaxPooling1D)                                              \n",
      "                                                                 \n",
      " dense_687 (Dense)           (None, 5)                 1065      \n",
      "                                                                 \n",
      " dense_688 (Dense)           (None, 1)                 6         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 218,159\n",
      "Trainable params: 218,159\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 12s 74ms/step - loss: 0.3698 - acc: 0.8642 - val_loss: 0.2373 - val_acc: 0.9082\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 11s 71ms/step - loss: 0.1758 - acc: 0.9321 - val_loss: 0.1657 - val_acc: 0.9329\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 11s 71ms/step - loss: 0.1227 - acc: 0.9549 - val_loss: 0.1308 - val_acc: 0.9489\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 11s 75ms/step - loss: 0.0968 - acc: 0.9640 - val_loss: 0.1155 - val_acc: 0.9538\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 11s 73ms/step - loss: 0.0802 - acc: 0.9715 - val_loss: 0.1056 - val_acc: 0.9586\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 11s 75ms/step - loss: 0.0690 - acc: 0.9757 - val_loss: 0.0982 - val_acc: 0.9603\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 11s 76ms/step - loss: 0.0611 - acc: 0.9778 - val_loss: 0.0995 - val_acc: 0.9612\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 11s 76ms/step - loss: 0.0553 - acc: 0.9791 - val_loss: 0.0951 - val_acc: 0.9615\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 12s 80ms/step - loss: 0.0502 - acc: 0.9811 - val_loss: 0.0903 - val_acc: 0.9626\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 12s 78ms/step - loss: 0.0463 - acc: 0.9822 - val_loss: 0.0884 - val_acc: 0.9638\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 12s 78ms/step - loss: 0.0433 - acc: 0.9827 - val_loss: 0.0876 - val_acc: 0.9641\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 14s 92ms/step - loss: 0.0402 - acc: 0.9834 - val_loss: 0.0858 - val_acc: 0.9642\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 12s 77ms/step - loss: 0.0383 - acc: 0.9841 - val_loss: 0.0893 - val_acc: 0.9633\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 12s 82ms/step - loss: 0.0360 - acc: 0.9844 - val_loss: 0.0867 - val_acc: 0.9646\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 14s 96ms/step - loss: 0.0349 - acc: 0.9850 - val_loss: 0.0884 - val_acc: 0.9658\n",
      "377/377 [==============================] - 2s 4ms/step\n",
      "377/377 [==============================] - 2s 4ms/step\n",
      "1507/1507 [==============================] - 6s 4ms/step\n",
      "34/34 [==============================] - 0s 4ms/step\n",
      "377/377 [==============================] - 2s 4ms/step - loss: 0.0796 - acc: 0.9679\n",
      "34/34 [==============================] - 0s 4ms/step - loss: 2.9118 - acc: 0.3864\n",
      "[0.07958674430847168, 0.9678971171379089]\n",
      "[2.9118075370788574, 0.38640594482421875]\n",
      "377/377 [==============================] - 2s 4ms/step\n",
      "1507/1507 [==============================] - 6s 4ms/step\n",
      "34/34 [==============================] - 0s 3ms/step\n",
      "CNN_Test: ROC AUC=0.996\n",
      "CNN_Training: ROC AUC=0.999\n",
      "CNN_3rd: ROC AUC=0.913\n",
      "[[4597  212]\n",
      " [ 175 7071]]\n",
      "[[19122   527]\n",
      " [  288 28283]]\n",
      "[[369 659]\n",
      " [  0  46]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_CNN(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Model: \"sequential_365\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding_365 (Embedding)   (None, 212, 32)           32768     \n",
      "                                                                 \n",
      " lstm_187 (LSTM)             (None, 32)                8320      \n",
      "                                                                 \n",
      " dense_365 (Dense)           (None, 1)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 41,121\n",
      "Trainable params: 41,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/15\n",
      "151/151 [==============================] - 34s 215ms/step - loss: 0.3344 - acc: 0.8540 - val_loss: 0.2250 - val_acc: 0.9064\n",
      "Epoch 2/15\n",
      "151/151 [==============================] - 33s 220ms/step - loss: 0.1564 - acc: 0.9359 - val_loss: 0.1574 - val_acc: 0.9367\n",
      "Epoch 3/15\n",
      "151/151 [==============================] - 31s 206ms/step - loss: 0.1057 - acc: 0.9570 - val_loss: 0.1205 - val_acc: 0.9517\n",
      "Epoch 4/15\n",
      "151/151 [==============================] - 31s 204ms/step - loss: 0.0834 - acc: 0.9660 - val_loss: 0.1004 - val_acc: 0.9581\n",
      "Epoch 5/15\n",
      "151/151 [==============================] - 30s 201ms/step - loss: 0.0679 - acc: 0.9711 - val_loss: 0.1268 - val_acc: 0.9517\n",
      "Epoch 6/15\n",
      "151/151 [==============================] - 31s 205ms/step - loss: 0.0612 - acc: 0.9741 - val_loss: 0.1050 - val_acc: 0.9583\n",
      "Epoch 7/15\n",
      "151/151 [==============================] - 31s 203ms/step - loss: 0.0546 - acc: 0.9764 - val_loss: 0.1036 - val_acc: 0.9600\n",
      "Epoch 8/15\n",
      "151/151 [==============================] - 31s 202ms/step - loss: 0.0496 - acc: 0.9778 - val_loss: 0.1005 - val_acc: 0.9599\n",
      "Epoch 9/15\n",
      "151/151 [==============================] - 31s 204ms/step - loss: 0.0455 - acc: 0.9797 - val_loss: 0.0946 - val_acc: 0.9611\n",
      "Epoch 10/15\n",
      "151/151 [==============================] - 31s 202ms/step - loss: 0.0422 - acc: 0.9808 - val_loss: 0.0945 - val_acc: 0.9641\n",
      "Epoch 11/15\n",
      "151/151 [==============================] - 31s 204ms/step - loss: 0.0405 - acc: 0.9806 - val_loss: 0.1051 - val_acc: 0.9599\n",
      "Epoch 12/15\n",
      "151/151 [==============================] - 31s 202ms/step - loss: 0.0393 - acc: 0.9814 - val_loss: 0.1008 - val_acc: 0.9598\n",
      "Epoch 13/15\n",
      "151/151 [==============================] - 31s 203ms/step - loss: 0.0370 - acc: 0.9820 - val_loss: 0.1034 - val_acc: 0.9612\n",
      "Epoch 14/15\n",
      "151/151 [==============================] - 31s 203ms/step - loss: 0.0344 - acc: 0.9830 - val_loss: 0.1034 - val_acc: 0.9632\n",
      "Epoch 15/15\n",
      "151/151 [==============================] - 31s 203ms/step - loss: 0.0350 - acc: 0.9829 - val_loss: 0.1210 - val_acc: 0.9579\n",
      "377/377 [==============================] - 5s 13ms/step\n",
      "377/377 [==============================] - 5s 13ms/step\n",
      "1507/1507 [==============================] - 20s 13ms/step\n",
      "34/34 [==============================] - 0s 13ms/step\n",
      "377/377 [==============================] - 5s 13ms/step - loss: 0.1126 - acc: 0.9619\n",
      "34/34 [==============================] - 0s 13ms/step - loss: 1.6853 - acc: 0.3063\n",
      "[0.11263934522867203, 0.9619244933128357]\n",
      "[1.6853364706039429, 0.30633148550987244]\n",
      "377/377 [==============================] - 5s 13ms/step\n",
      "1507/1507 [==============================] - 20s 13ms/step\n",
      "34/34 [==============================] - 0s 13ms/step\n",
      "RNN_Test: ROC AUC=0.994\n",
      "RNN_Training: ROC AUC=0.998\n",
      "RNN_3rd: ROC AUC=0.717\n",
      "[[4602  207]\n",
      " [ 252 6994]]\n",
      "[[19121   528]\n",
      " [  479 28092]]\n",
      "[[283 745]\n",
      " [  0  46]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Loc_RNN(Classifier_input,classifier_3rd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAArgAAAGMCAYAAAAxw9EJAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy81sbWrAAAACXBIWXMAAA9hAAAPYQGoP6dpAABcKklEQVR4nO3deViN6f8H8Pdp3xSVUkoLUqYSY98TQ5ZBBslWGEv2fQnZS9ZmhuyViSxDdmbIPshuRJahspStKKHt9Pz+8Ot8nWlRp3JyvF/X1XU593Pf9/N5+nam9/fuPs8jEgRBABERERGRglCSdwFERERERKWJAZeIiIiIFAoDLhEREREpFAZcIiIiIlIoDLhEREREpFAYcImIiIhIoTDgEhEREZFCYcAlIiIiIoXCgEtERERECoUBl4jKraioKIhEIohEIvj5+RXaN7dfXFxcgX3mzJkDkUgET0/PQs/5888/w8bGBhUqVICGhgYsLS3Rq1cvREREICcnp0i1h4SESGrK/dLW1oapqSlat26NqVOn4tatW58dP2fOnCKdr6xYWlpCJBJJtcXFxUEkEqF169byKeoTuf+bhoSEyLsUIipHGHCJqNz6/fff8/13WcjKysLQoUPRuHFjbNiwASKRCG3btkXXrl1hZGSEXbt2wc3NDe3atSvWvNWrV8fAgQMxcOBAdO3aFfb29rh16xYCAgJgb2+Pfv36ITU1tUyuqXXr1p8N/eWdIlwDEX15KvIugIgoP1lZWdi+fTtEIhGMjY0RExODq1evol69emVyPi8vL2zZsgU2NjYIDg5G06ZNpY4nJCRg3rx5+Ouvv4o1b/PmzfOsLgqCgIMHD2L06NHYsmULnjx5gqNHj0JVVVXSp3v37mjcuDEMDQ1lvqbSEBkZiaysLLnWUJhRo0bB3d0dJiYm8i6FiMoRruASUbl0+PBhvHr1Ci1btsTQoUMBlN0q7q5du7BlyxYYGxvj9OnTecItAJiammLNmjWlUoNIJELnzp0RFRUFU1NTnDp1CkFBQVJ99PT0YGtrK/eAW716ddja2sq1hsIYGhrC1tYWenp68i6FiMoRBlwiKpdyg2S/fv3Qr18/AEB4eDjEYnGpn2vJkiUAPu7nNDY2LrRvs2bNSu28RkZGmDdvHgDgl19+kTpW0B7crKwsrF27Fg0bNoShoSG0tLRgaWmJzp07Y9u2bQD+t0f21KlTAAArKyupvcC5PD09IRKJcPLkSfz5559wdnZGxYoVIRKJ8ObNGwD578H9VGpqKsaOHQtzc3NoaGjAzs4OK1asyHevskgkgqWlZb7z/Pd6i3oNhe3BTUpKwuTJk1GzZk1oaGhAX18fHTp0KHAVPrc+sViMgIAA2NjYQF1dHebm5pg6dSoyMjIK/D4QUfnCLQpEVO6kpKTgwIEDUFdXx08//YSKFSuiYcOGuHjxIo4ePYoOHTqU2rlevXqFixcvQiQSwd3dvdTmLapevXph6NChePDgAR4/fgxzc/NC+/fv3x/bt2+HoaEhmjZtCi0tLTx9+hRnzpxBWloa3N3doaOjg4EDB+LIkSN4/vw5evToAR0dnQLn3Lp1KzZs2ID69evD1dUVDx48KDTU5srIyECbNm3w4MEDtGnTBpmZmYiMjMSECRPwzz//IDg4uNjfj1zFvYb/evr0KVq2bImHDx+iWrVq6NatG16+fIljx47hzz//xPLlyzF+/Ph8x/bt2xcHDhxAw4YNUatWLZw5cwYBAQF4+vQpwsLCZL4mIvpyGHCJqNzZsWMH0tPT0aNHD1SsWBHAx5XcixcvIiwsrFQD7vXr1yEIAqpXry4515dUoUIFWFtb499//0VMTEyhATcuLg7bt29HgwYNcPr0aWhoaEiOffjwAdevXwfw8c/2ISEhaN26NZ4/f46lS5cWuHIKAOvXr8e2bdvQu3fvYtV+4cIFODo64v79+5KtFA8ePEDLli0REhKC7t2748cffyzWnLmKew3/NXz4cDx8+BD9+/fHxo0bJfubz549i/bt22Py5MlwcXGBo6Oj1Lj4+HhoaWkhOjpacr7Y2Fh8//332LJlC+bOnYvq1avLdE1E9OVwiwIRlTufbk/I5e7uDhUVFURERCAtLa3UzpWUlAQAqFy5cqnNWVy54fD169eF9nvx4gUAoGnTplLhFgA0NTXRpEkTmc7fqVOnYofbXEuXLpXaJ1y9enXMmjULALBq1SqZ5iyphw8f4sCBA9DV1cUvv/wi9eG95s2bY/jw4RCLxVi9enW+43/99VepMG1lZSX5WTxz5kyZ1k5EpYMBl4jKlbi4OJw9exb6+vro2LGjpL1y5cpo37493r9/j4iIiFI7nyAIpTZXSWv43LYAW1tbaGtrIzg4GOvXr5eE85KSdZVVX18/39umeXh4AADOnTsnl+/v2bNnAQAdO3bMd1W+f//+APIPq6qqqvne39fGxgYAkJiYWHqFElGZYcAlonIlLCwMgiCgV69eUFNTkzqWu4om650M8guSuauPL1++lGnO0vDq1SsAHwNjYXR1dbF+/Xrk5ORg6NChqFy5Muzs7ODt7Y0LFy7IfP5q1arJNM7CwqLAOitWrIi0tLQyu8dvYRISEgCgwC0Nue25/T5lYmICZWXlPO25+3/5QTOirwP34BJRuZL7IZ7IyEg0b95c6lhuuIiMjERiYqLUvU81NTXx4cMHvHv3rsC5379/DwDQ1taWtDk5OQH4+GftN2/efPF9uKmpqXj48CEAoHbt2p/t36dPH7Rt2xZ79+7FX3/9JbnFWFBQECZPnoyAgIBi1/Df7Q6lobgrt0V9QlxxFLQintue3/GifLiOiMo/ruASUblx8eJF3L17FwBw//59/P3331Jfly9fBvAxDG3dulVqbO6Hs3LDYn5yj5mZmUnaDA0N0bBhQwiCILnN1pe0Y8cOCIIAGxsbmJqaFmlM5cqVMWTIEOzYsQPPnj3D4cOHoauriyVLluD27dtlXPH/PHr0KN/21NRUpKSkQFtbG7q6upJ2VVXVAvdPP378uNTqyv0+xsbG5ns896lofDgEkeJiwCWiciN368HkyZMhCEK+X7n3MP3v7ZpatmwJANi/f3++c7979w7Hjx+X6ptr0qRJAD7eUzX3g1wFOXfuXDGvqmAvXryAr68vAGDs2LEyzSESidChQwd06tQJABAdHS05lrvFIzs7u4SV5i8pKQnHjh3L0x4eHg7g44fhPl0RNTExQVJSEpKTk/OMKejetLJcQ+7K/8GDByX38/1U7s9OixYtijwnEX1dGHCJqFzIzs7G9u3bAXz8M3xB2rRpAyMjI1y/fl0qzI0cORLKysoIDg7Gnj17pMakp6dj+PDhePPmDerXr5/nSWU9e/aEu7s7nj9/jpYtW+L8+fN5zvvs2TOMGjVK6s4OshIEAYcOHULjxo2RkJCANm3aSJ7WVphr165h9+7deR6d+/r1a0RFRQGQ3k+bu5KZuypeFiZPniz1YbfY2FjMnz8fAODt7S3Vt1WrVgAgOQ58/F74+fkV+H8cZLkGa2trdOrUCW/fvsXYsWOlvl/nz59HUFAQlJWV89RHRIqDe3CJqFw4fPgwXr58iVq1aqFu3boF9lNWVsZPP/2E1atXIywsDP7+/gA+7qUNDAzEmDFj0L17dzg5OcHOzg4fPnzAuXPn8OLFC5iZmeXZ2pBr8+bN0NLSwqZNm9C0aVPY2tqidu3aUFVVRVxcHC5fvgyxWJzvXQMKc/bsWXh6egIAMjMzkZSUhKtXr0o+WNa/f3+sWrUKKiqf/89xfHw8evToAT09PdSvXx9VqlTBmzdvcObMGaSmpqJ79+5o3LixpP+PP/6I0NBQeHh44IcffpA8znbDhg3FuoaCNG7cGJmZmahZs6bUgx7ev3+Pfv36oVu3blL9p06dij/++AMrV67EyZMnUb16ddy8eROPHz+Gt7d3vrftkvUa1q5dixYtWmDz5s04deoUmjRpgpcvX+LkyZMQi8VYtmxZnnvgEpECEYiIyoGePXsKAARfX9/P9j1z5owAQDAzMxPEYrHUsQsXLgju7u6CmZmZoKamJujo6Ah16tQRfHx8hFevXn127nPnzgmDBg0SatSoIWhpaQnq6uqChYWF0KtXL2Hv3r1CTk5Oka4nODhYACD1pampKZiYmAgtW7YUpkyZIkRHR392/Kffj8TERGHBggVCmzZtJNdnbGwsNG/eXAgNDRWysrLyzLNixQqhdu3agrq6uqSOXAMHDhQACCdOnCiwDgsLC+G/vypiY2MFAEKrVq2EN2/eCN7e3oKpqamgpqYm1KpVS1i6dKmQnZ2d73znz58XWrduLWhpaQm6urqCq6urcP369XyvtyjX4OvrKwAQgoOD84x79eqVMHHiRKF69eqCmpqaULFiReGHH34Q/vzzz3xrAyBYWFjke6yw+oio/BEJQjm4CSQRERERUSnhHlwiIiIiUigMuERERESkUBhwiYiIiEihMOASERERkUJhwCUiIiIihcKAS0REREQKhQ96+H85OTlISEhAhQoVpB4tSURERETlgyAIePv2LUxNTaGkVPA6LQPu/0tISIC5ubm8yyAiIiKiz3j8+DHMzMwKPC73gPv27VvMnz8f169fx7Vr1/Dq1Sv4+vpizpw5RRr/4sULTJkyBQcOHMD79+9Rp04dLFiwAC4uLsWqo0KFCgA+fsN0dXWLexlEREREVMZSU1Nhbm4uyW0FkXvATUpKwrp161CnTh1069atWM9Iz8jIgIuLC968eYPAwEAYGRlh1apV6NChA44dO4ZWrVoVea7cbQm6uroMuERERETl2Oe2k8o94FpYWOD169cQiUR49epVsQLuxo0bER0djXPnzqFJkyYAAGdnZ9SpUwdTpkxBVFRUWZVNREREROWU3O+iIBKJZP5QV0REBGrVqiUJtwCgoqKCfv364eLFi3j69GlplUlEREREXwm5B9ySiI6OhqOjY5723LZbt2596ZKIiIiISM7kvkWhJJKSkqCvr5+nPbctKSmpwLEZGRnIyMiQvE5NTS39AomIiKhAYrEYWVlZ8i6DyhFlZWWoqKiU+JatX3XABQrfZFzYMT8/P8ydO7csSiIiIqLPSEtLw5MnTyAIgrxLoXJGS0sLJiYmUFNTk3mOrzrgGhgY5LtKm5ycDAD5ru7mmj59OiZMmCB5nXvbCSIiIipbYrEYT548gZaWFipXrswHLBGAjw9xyMzMxMuXLxEbG4uaNWsW+jCHwnzVAdfBwQE3b97M057bZm9vX+BYdXV1qKurl1ltRERElL+srCwIgoDKlStDU1NT3uVQOaKpqQlVVVXEx8cjMzMTGhoaMs3zVX/IrHv37rhz547U7cCys7MRFhaGRo0awdTUVI7VERERUWG4ckv5kXXV9lPlYgX38OHDePfuHd6+fQsAuH37Nv744w8AQMeOHaGlpYXBgwcjNDQUDx48gIWFBQBg0KBBWLVqFXr27Al/f38YGRlh9erVuHv3Lo4dOya36yEiIiIi+SkXAXfEiBGIj4+XvN65cyd27twJAIiNjYWlpSXEYjHEYrHUZnR1dXVERkZiypQpGD16NN6/fw8nJyccPny4WE8xIyIiIiLFIRL48UUAHz9kpqenh5SUFD6ql4iIqAylp6cjNjYWVlZWMu+xJMVV2M9HUfNauVjBJSIiIrKcdvCLni+mn0Wxxwz18UHYvn0YNmwY1qxZI3XM29sbQUFBGDhwIEJCQuDp6Yk3b95gz549kj7PHtyX/LtBK2c8+f+nrmpoaMDC3ByDBvTHgD7uec5bpXrNYtf6LfuqP2RGRERE9KWZm5tj27Zt+PDhg6QtPT0d4eHhqFatWrHmmjxuLG6c/xvHD+xHh3ZtMXXWbOw9+GWDviJiwCUiIiIqhnr16qFatWrYvXu3pG337t0wNzdH3bp1izWXjrY2jCpXhpWlBaZNGA9rS0scOcoPypcUAy4RERFRMWS8z8ZPXftgbdB6vIhPxYv4VKxZvQ49u3kg43020tOy8CI+FelpWch4ny3p8yI+9bNzq6urIysr+wtchWJjwCUiIiIqpp5u7rh46QIePY7H4yePcOlyFH7q3lvm+bKzs7F9127E3L2LFk2blGKl3yZ+yIyIiIiomAz0DdC2zQ/YsSscgiCgbZsfYKBvUOx5FgYsweIVK5GZmQlVVVV4/zwE/fP5kBkVDwMuERERkQz69OqP6bMnAQD85y+TaY4RPw9Bbzc3aGpqwNjIiE93KyUMuEREREQyaNOqLbKysgAAzi1dZJpDv1IlWFkW/3ZlVDgGXCIiIiIZKCsr4+yxi5J/5yf1bSqib/0jeZ0jTkLFihVhZmr6RWr8VjHgEhGVAw6hDjKNuznwZilXQiQ/cf6dvti5PkRHl8o8FSoU/vTTcxfOwKVTC6m2Xm7dERiwuFTOT/njo3r/Hx/VS0TyxIBL35Ly8KjekgTctxWK9zCHT+VkP5dp3Lf0JLPSeFQvbxNGRERERAqFAZeIiIiIFAoDLhEREREpFAZcIiIiIlIoDLhEREREpFAYcImIiIhIoTDgEhEREZFCYcAlIiIiIoXCJ5lRqZL1ZvUAb1hPREREpYMBl4iIiMqHOXpf7FSaAD789PcXOx99WdyiQERERFREQ318YGypB2NLPZhW10e9pt9his94vEl5LelTv5kDjC31cPnqJamxM+dOg5tHP8nrpYG/wKSGDabMmi3VL/r2bZjUsMHjJ0/K9mIUGAMuERERUTG0adUWNy/ew+WzN7Hc/xf8FXkEU2dOlOqjoa6BBf6+n51LQ10d4Tv/wIPY2LIq95vEgEtERERUDGpq6jAyMoapSVW0bumCrp3dcOrMcak+/T28cOX6JRw78Vehc1W3tkLTRo2wePnKMqz428OAS0RERCSjuEexOHHqGFRUVKXazc2qYYDHICxcPBc5OTmFzuEzeRIO/vknrv/zT1mW+k1hwCUiIiIqhqPHj8CqtiksahmjUUsn3L1/B6OGj8vTb/zoyXj0JB679uwodD5H++/wY0dXLFyytIwq/vYw4BIREREVQ7MmLXD80Bkc2hOJwZ7D4NzSBUM8h+XpZ2hgCO+fR2Px8oXIzMwsdM6p48cj6vIVnDxztqzK/qYw4BIREREVg5amNqwsq+M7O3ssmhOAjMxMLA30z7fv8CEjkZ6ejuDfNxQ6p6VFNfTt1RMLlyyFIJRF1d8WBlwiIiKiEpg0diqC1v2KZ88T8xzT1tbBhDFTELhqKd6mvS10ngmjR+FhXBz2HjxYVqV+MxhwiYiIiEqgWZMWqGVji8BVy/I93r+PJypU0MWefX8UOk9lQ0MMG+SFjaGby6LMbwqfZEZERETlw5yUL3aqD9HRpTrfsCGjMG6yd74fNlNVVcXUCTMxYuzgz87jPWQwQrdsRXpGRqnW960RCQJ3egBAamoq9PT0kJKSAl1dXXmX89VyCHWQeezNgTdLsRKir4us7x2+b+hrlJ6ejtjYWFhZWUFDQ0MuNZQk4L6tUE3msTnZz2UaV6V6TZnP+bUp7OejqHmNWxSIiIiISKFwiwIRERXbst6dZR47cfuBUqyEiCgvruASERERkULhCi4RERFROZf5pPBbjBVGzaxCKVbydeAKLhEREREpFAZcIiIiIlIoDLhEREREpFAYcImIiIhIoTDgEhEREZFC4V0UiIiIqFwoydMwZXHx+/Avej76criCS0RERFQML148x3TfyWjQwhHmNpVRt0lt9BvcG6f/PgkAqN/MAcaWerh89ZLUuJlzp8HNo5/k9dLAX2BSwwZTZs2W6hd9+zZMatjg8ZMnZX4tiooBl4iIiKiI4p8+RbsurfD3udOYPX0eTh45j/DQXWjepAWmz5ok6aehroEF/r6fnU9DXR3hO//Ag9jYsiz7m8MtCkREX7EYWzuZx9rdiSnFSoi+DWMXLIBIJMLhvcehraUtabe1sUOfnv9bne3v4YXQLRtx7MRfaOv8Q4HzVbe2goG+ARYvX4l1vwaWae3fEq7gEhERERVBckoKjv79N7z6D5EKt7n09CpK/m1uVg0DPAZh4eK5yMnJKXRen8mTcPDPP3H9n39Ku+RvFgMuERERURE8ePQIgiCgZnWbIvUfP3oyHj2Jx649Owrt52j/HX7s6IqFS5aWRpkEBlwiIiKiIhEE4eM/RKIi9Tc0MIT3z6OxePlCZGZmFtp36vjxiLp8BSfPnC1pmYRysgc3LS0NM2fOxI4dO5CcnAxbW1tMmzYN7u7unx174sQJLFq0CDdu3MD79+9hbW2NIUOGYOTIkVBWVv4C1RMRfZ1WDT8u7xKIvio1LCwgEolw/9+7QPvORRozfMhIBP++AcG/byi0n6VFNfTt1RMLlyzFcr9FpVHuN61crOC6ubkhNDQUvr6+OHz4MBo0aIA+ffpg69athY47duwY2rZti+zsbKxfvx579uxB69atMXbsWEyYMOELVU9ERETfAn09PbRt2hTBv2/Au/fv8hxPSXmTp01bWwcTxkxB4KqleJv2ttD5J4wehYdxcdh78GBplfzNknvAPXToEI4ePYrVq1dj2LBhcHZ2xvr169GuXTtMnjwZYrG4wLEhISFQVVXFgQMH0LVrV7Rt2xa//PILfvjhB4SEhHy5iyAiIqJvQuDMmRCLxXDt2gYHDu/Fw9gHuPfvXawPXoNObu3yHdO/jycqVNDFnn1/FDp3ZUNDDBvkhY2hm8ui9G+K3LcoREREQEdHBz179pRq9/LygoeHB6KiotC0adN8x6qqqkJNTQ2amppS7RUrVoSGhkaZ1UxERESl7+bAm1/sXB+io2UaZ2lmhmMHT2Plb0sxZ8FMPH/5DAb6hnC0d8LiBcvzHaOqqoqpE2ZixNjBn53fe8hghG7ZivSMDJnqo4/kHnCjo6NhZ2cHFRXpUhwdHSXHCwq4w4cPR3h4OMaMGYMZM2ZAS0sL+/fvR0REBPz8/Ao9b0ZGBjI++eFJTU0t4ZUQERHRt8DYqAr85i2F37z873pw+e+8Qd2t609w6/oTcrKfS9omjR2DSWPHSPXT0dHBrUtRpVvwN0juWxSSkpKgr6+fpz23LSkpqcCxjRo1wvHjxxEREYGqVauiUqVK8PLywsKFCzFx4sRCz+vn5wc9PT3Jl7m5eckuhIiIiIjKBbkHXAAQFXK7jcKOXblyBd27d8f333+P/fv34/jx45g+fTpmzpyJ+fPnF3rO6dOnIyUlRfL1+PFjmesnIiIiovJD7lsUDAwM8l2lTU5OBoB8V3dzjRw5EsbGxoiIiJDcEszZ2RlKSkqYM2cO+vbtC2tr63zHqqurQ11dvRSugIiIiIjKE7mv4Do4OCAmJgbZ2dlS7Tdvfty/Ym9vX+DY69ev4/vvv89zv9sGDRogJycHMTF8zjoRERHRt0buAbd79+5IS0vDrl27pNpDQ0NhamqKRo0aFTjW1NQUly9fznMrsfPnzwMAzMzMSr9gIiIiIirX5L5FwdXVFe3atcOIESOQmpqKGjVqIDw8HEeOHEFYWJhkdXbw4MEIDQ3FgwcPYGFhAQAYP348xowZgy5dumDYsGHQ0tJCZGQkli1bhrZt26JOnTryvDQiIiIikgO5B1wA2L17N3x8fDB79mzJo3rDw8OlHtUrFoshFov/9xxoAKNHj0bVqlWxYsUKDBkyBB8+fIClpSV8fX0xfvx4eVwKEREREclZuQi4Ojo6CAwMRGBgYIF9QkJC8n06mZubG9zc3MqwOiIiIiL6msh9Dy4RERERUWkqFyu4RERERDG2dl/0fJZ/7Pyi5yuKBq2c8bPnQAz18pRp/OYdWzBp7jS8uPVt39+fK7hERERERTTUxwcDf/Yos/kP796Ffu69i9TXpok9ftmwSqqtZxc3RJ+6WhalfVW4gktERERUThgaFPyAq6LQ1NSEpqZmKVXz9eIKLhEREVEpOHfhLNp3dYa5TWU4NLDBfH9fqQdZpaW9hfeEibB2qIM6TZph7aZguHn0w6wFCyV9GrRyxrrgEMnrpYG/4PsWrVChuiEsv7fB+NmTAQDtenZE/JNHmDx3OtTNdaFurgvg4xYFo+/Mperat28f6tevDw0NDRgaGn4TH85nwCUiIiIqocRnCfDw6om6jvVw/NDfWLxgOcJ3/I4Vvy6R9Jm9YAYuXbmK0LVB2B4ajKjLl3Hz1q0C5zxw+AjWBYcgYME83Dp9DTs3bIW97XcAgO3rwmBmUhW+E30Qf+U+4q/cz3eOQ5FH4Obmhk6dOuHatWuIjIxE/fr1S/fiyyFuUSAiIiIqoeDfN6CqSVX4zVsKkUiEmjVs8PxFIub7z8HEsVPx/v077NgVjtXLl6FF06YAgJWL/eHUtHmBcz5JSIBR5cpo2bQpjHXMUa2qORrU/RhO9SvpQ1lZGTo6OqhiZFzgHIt/XQp3d3fMnTtX0vYtPAiLK7hEREREJXT/37uoX68hRCKRpK3h943x7l0aEhKfIv5RHLKysuBUx1FyXLdCBVS3sipwzi6urviQno7Gzi4YMWU09h7eL7XloShu3LoJFxeX4l/QV44Bl4iIiKiEBAHAJ+H2Y9vHp6+KRKL//RuifAbmr6qpCc4e/ROL5vhCQ0MDY2ZOgMtPHZCVlVXkujQ1NIrcV5Ew4BIRERGVkE3NWrh8JUoSZAHg0tUo6OhUgEkVU1haWEFVVRXX/vlHcvzt2zQ8jI8vdF5NDQ20b+uCFfOW4K8dh3DhykVE3/m4b1dVVRVicU6h4x3s7BEZGVmCK/s6cQ8uEVFpmaMn+1iraqVXBxGVqdS3qYi+9Y9UW/8+nli3KQgzfCdj0ICh+PfhfSxZ4Yfhg0dCSUkJOjoV0KtHH8xfvBiV9PRgaGCAJYG/QEkkyruq+/+279oNsViMek51UEX3A7buCoemhiaqmX28S4KFWTWcjfobvX7sAXV1dRjqG+SZw2f8NHRw74Lq1avD3d0d2dnZOHz4MKZMmVL635hyhAGXiIiIygW7OzHF6n/rVcF3IPgc62cFbw34nHMXzsClUwuptt49PLA1eCfm+s1CWMdmqKhXCX169cf40ZMlfebNXIRJM0ag/9BhqKCjA++fhyDhWSLU1dXyPY+ubgX8tmYd5izyQ06OAHvb2tgdvB0GlT4GWd9JPhg5bRzsWtRBRkYGMh6n5pmjVZMW2LlzJ+bPnw9/f3/o6uqiZcuWMl/710IkCIVs/viGpKamQk9PDykpKdDV1ZV3OV8th1AHmcfeHHizFCshkoMSrOA6yLiCu8OveB84+dTx1qs+36kA6a+Xyzx24vYDMo8lxZCeno7Y2FhYWVlBowR7ROUVcN9WkP0vLjnZzyX/fv/+Peo2awHf6dPg0atnoeP01avIfE41swoyj5WHwn4+iprXuIJLRERE9AXcjL6Be/cvo66jI1LfvsXy334DALRv++3d5aCsMeASERERfSFBGzbiQWws1FRV4Wj/HfZs2woD/ZI9npfyYsAlIiIi+gIc7Ovgr70R8i7jm8DbhBERERGRQuEKLuXFWx0RURl6Mu2MTOPM/Ft8vhMREbiCS0REREQKhgGXiIiIiBQKAy4RERERKRQGXCIiIiJSKPyQGRHRJyynHZR5bJzsD2QiIgCrhh//Yuc6CWDQKKMymdvYUg/Ba7egY/vO+R5//OQJGrZug6P79sC+du0yqeFbxxVcIiIioiIa6uMDY0s9GFvqwbS6Puo1/Q5TfMbjTcprSZ+bF+/BpXU7OVZJXMElIiIiKoY2rdoicMlqZIuzce/+HYybMgopqSlY++smAICRkbGcKyQGXPqmLeud/5+PimLi9gOlWAkREX0t1NTUJSHW1KQqunZ2w/Y/tkiO/3eLwtXrVzB5xljc//ceatnUxFjv4Xnm/PNYJOb6+SPx2TPUc3JC7x7dMXbKNNy5ehl6uroAgPOXozDTzxeXb1yFob4BfuzQGQumzYG2lvYXuOqvCwMuffW+5J4tIiIqAwnX5F2BzOIexeLEqWNQUVHN9/i79+/Qb3AvNG/SEqtWrkd83A3Mmr9Aqs/jJ0/w8+gxGDJwADx69UT07duY57dYqk90zC107tcdvpN8sGbpb3iVlIRxsyZh3MxJWL88qMyu72vFgEtERERUDEePH4FVbVPkiMVIz0gHAMyduSjfvrv27ECOWIyVS1ZBS1MLNtb6SHj2DNNm+0r6bN4ajupWVpg9bSoAoIa1Ne7cu4/A1f8LrsvXBqJ3t54YM2QkAKCmVQ0snxuAtj1d8euiFdDQ4KdcP8WAS0RERFQMzZq0QMCC5Xj/4QO2bN+Mhw//xRDPYfn2vf/vPdS2s4eWppakrX5dJ6k+/8bGwsnRQaqtrqOj1OurN6/jQdxDbIvYIWkTBAE5OTmIfRwPu5q1SnhVioUBl4iIiKgYtDS1YWVZHQCwaE4Aurt3xtJAf0ybODNPX0EQPjufIAgQQVTouJycHAzp64WRXnn371aral6c8r8JvE0YERERUQlMGjsVQet+xbPniXmO2dSshdsx0fiQ/kHSduX6Dak+Na2tcf3mTam2G9HRUq/r2tfB7Xt3UMOqep4vNTW1UrwaxcCAS0RERFQCzZq0QC0bWwSuWpbnWI+uPaGkpITxU0bh7v07iDx5Ems2bJTq07+PO/59+BALApbgQWws9h08hB27dgMARKKPK7uTvMcj6spFjPGZgBu3/sH92H+x/69DGDdrUtlf4FeIWxSIiIioXBg5u1Kx+t8qwcql9bPPbx0ojmFDRmHcZG+MGj5Oql1bWwebN2zHFJ/xaNupBWyqV4fPlEkYMnK0pE81c3Os//UXzPXzx4aQUHxfty7GeI/AtNm+ktVZBzt7HNt5CLMD5qFNjw4QBAHWFlbo2cWtVK9DUTDgEhERERXRuoUL8bZCtTztPbr2RI+uPQEAz+NSpI7Vr9cAxw+fBQDkZD8HACT+e0+qT/u2Lmjf1kXyeuXqIJhWqQINdfX/zeP0PQ5t3Vs6F6LgGHCJiIiI5CwkbAucHB1QqWJFXLxyFUHrN8Crfz95l/XVYsAlIiIikrOHcfFYuToIb968QVVTUwwbPAhjhud/6zH6PAZcIiIiIjmbN3MG5s2cIe8yFAbvokBERERECoUBl4iIiIgUCgMuERERESkUBlwiIiIiUigMuERERESkUHgXBSIZPZl2RuaxZv4tSrESIiIi+hQDLhEREZULy8bP+qLn857l/0XPR18OtygQERERFdFQHx8M/NlD3mXQZzDgEhEREZFC4RYFKjdibO1kG9h6VekWQkREJIM1G35D+M4tiH8Uh0oVK+EHlw6YPX0etLV1AACPnzzC9FljcPHKFWRmZcG8alXMnjYFLq1b401KCmbMnYdTZ87i/fv3MKlSBWNGDIf7Tz0AANExtzBxzlRcuHIRWpqa6NaxK5bMXgSd/5+bpJWLFdy0tDSMGzcOpqam0NDQgJOTE7Zt21bk8Xv37kWrVq2gq6sLbW1tfPfdd1i3bl0ZVkxEREQkTSRSwkLfxTj113n8siwIZ8+dxjy/2ZLj02ZPQkZmJiLCt+DEwf2YOWUytLS0AQABK1bi/v1/sWXTBpz+8zD8582BfqVKAID3Hz6gS383VNSriHMHTmLrms04fvYkxs2cJI/L/CqUixVcNzc3XLp0Cf7+/rCxscHWrVvRp08f5OTkwMOj8H0u/v7+8PHxwfDhwzF9+nSoqqrizp07yMzM/ELVExEREQHDBntL/m1hbompE30wdeYELF6wHADwNOEJOv3QBna1an3sU62apP/ThETYf1cbTg4OAABzMzPJsd179+FD+gdsWrkW2lra+A7AyvlL4ObVGwtnzINxZaMvcHVfF7kH3EOHDuHo0aOSUAsAzs7OiI+Px+TJk9G7d28oKyvnO/bKlSvw8fGBn58fpkyZIml3cXH5IrUTERER5Tp77jQCVy/Dvft38TbtLcTZ2UjPSMe79++graWNIZ7DMHXmBJw6+zdaNG2CTh3ao7atLQBgQN8+GDJyNG7euoVWzZujQ7u2aFCvHgDg/oMHcKztAO3/X+0FgKb1GyMnJwf3HtxnwM2H3LcoREREQEdHBz179pRq9/LyQkJCAqKiogoc+9tvv0FdXR2jR48u6zKJiIiICvT4ySP09eoJW5va2Bj0O47uPwW/+UsBANlZWQCAfu4DceF4JHp064qYe/fQoXsPbNy8GQDg0qoVLp06iZ89PfHs+Qv06j8Qc/0+3sZMEACRSJTveQtq/9bJPeBGR0fDzs4OKirSi8mOjo6S4wU5ffo07OzssGvXLtSqVQvKysowMzPDtGnTPrtFISMjA6mpqVJfRERERLK4cfMassXZmDtzIerXa4Dq1jXw7PmzPP2qmppgoEcfbFq9CsMGeWHL9p2SY4YG+ujdww2rli/F3Jk+CNu2HQBgU6M6btz6B+/ev5P0PXf5ApSUlFDTukbZX9xXSO5bFJKSkmBtbZ2nXV9fX3K8IE+fPsXLly8xZswYzJ8/H7Vr10ZkZCT8/f3x+PFjbNmypcCxfn5+mDt3bskvgIiIiL4pqW9TEX3rH6k2A31DZGdnY0PIWvzQtgMuXY7C5i2bpPrMnDsNzi3qo7qVJd6kpOLvCxdQs/rHDBSwMhCO332HWjVrIjMzE8eOn0DNGtUBAG5df8TyX1Zh8PjhmDV+Ol4mv8L4WZPR182d2xMKIPeACxS+vF7YsZycHLx9+xbh4eFwd3cH8HH/7rt377By5UrMnTsXNWrk//9spk+fjgkTJkhep6amwtzcXMYrKH8spx2UeWycRikWQkREVEQTV8wvVv9bamoyn8v6mSDz2HMXzsClk/Qj13v38MDcmYvw25qVWBQwF40bNoXPFF+MmjBM0kecI8aMOXOR+OwZdHR04NyyBeb6zAAAqKqqYtHSZXj89Ck0NTTQsP73WLNyBQBAS1MTB8IiMHHOVDTt3FrqNmGUP7kHXAMDg3xXaZOTkwH8byW3oLHPnj1D+/btpdpdXV2xcuVKXL16tcCAq66uDnV19RJUTkRERN+adQsXYtkvvxd4fPiQkVKve7q5S/7tN3cJcmblf2uv8SO9MX6kd77HAMDe7jv8uf1AMav9dsl9D66DgwNiYmKQnZ0t1X7z5k0AgL29fYFjc/fp/pcgfPx/ZUpKcr88IiIiIvrC5J4Au3fvjrS0NOzatUuqPTQ0FKampmjUqFGBY3v0+Ph0j8OHD0u1Hzp0CEpKSmjQoEHpF0xERERE5Zrctyi4urqiXbt2GDFiBFJTU1GjRg2Eh4fjyJEjCAsLk9wDd/DgwQgNDcWDBw9gYWEB4OOtxNauXQtvb2+8evUKtWvXxrFjx7Bq1Sp4e3tL+hERERHRt0PuARcAdu/eDR8fH8yePRvJycmwtbWV+uAYAIjFYojFYsn2A+DjhuyjR49ixowZWLRoEZKTk2FlZQV/f3+pD5ARERER0bejXARcHR0dBAYGIjAwsMA+ISEhCAkJydOur6+PNWvWYM2aNWVYIRERERF9LeS+B5eIiIiIqDQx4BIRERGRQmHAJSIiIiKFwoBLRERE9IW5efTDrAUL5V2GwirSh8wEQcCBAwdgZWVV4IMXbt68ibi4OHTp0qVUCyQiIqJvw5Nf0orVX68E50oCYNCvUrHGDPXxQdi+ffCZ4osx3v+7W9OhPw/Aa1hfPI9LKfJcG1f/BlWVsv2s/5Dxw/H7H1slr/X19dGgQQMEBAQU+LAsRVGkFdyDBw+id+/e0NTULLCPtrY23N3dsWPHjlIrjoiIiKg80VBXx29rVuJNyusSzVOpYkXo6OiUUlUF+6F1WyQmJiIxMRGRkZFQUVFB586dy/y88lakgLt27VoMGDAA1atXL7CPtbU1Bg4ciM2bN5dacURERETliXPjxjCqbIxfVi8vsE/y62QMGz0ITo3tYGlbBa3aN8HuvX9I9fl0i8LCJUvRqUfPPPO06dQFS1b+7xaqodvD4OhcH7o1KsOh9fdYE7r+s/Wqq6mjSpUqqFKlCpycnDB16lQ8fvwYL1++lPSZOnUqbGxsoKWlBWtra8yaNQtZWVkAgLi4OCgpKeHy5ctS8/7666+wsLCQPJ/g9u3b6NixI3R0dGBsbIz+/fvj1atXkv5//PEHHBwcoKmpCQMDA7Rt2xbv3r37bP2yKlLAvXjxIjp06PDZfj/88EOebwARERGRolBWUsKMybOxMWQdEhKf5tsnIyMddRycELZxO07+dR79+3hi1IShuHIt/4zUo+uPuHrjBuLiH0na7t67j5i7d+HW9UcAwMatIfANmId5U2bjxvFLmD/VF3OXLsDvO7cUufa0tDRs2bIFNWrUgIGBgaS9QoUKCAkJwe3btxEYGIj169djxYoVAABLS0u0bdsWwcHBUnMFBwfD09MTIpEIiYmJaNWqFZycnHD58mUcOXIEz58/R69evQAAiYmJ6NOnDwYNGoSYmBicPHkSbm5uUg/vKm1F2vzx+vVrVK5c+bP9DA0N8fp1yZbsiYiIiMqzjh264LvaDghYsQgrA1blOW5SxRTeQ8dIXg/xHIYTp45h/6EI1HUYk6e/rY0NatvWwu79+zFh1EgAwK59++Dk6IDqVlYAAL/AACyetRDdXD8GXqtqloi5dwcbtgSjf8++BdZ6KPKIZCvEu3fvYGJiggMHDkBJ6X9rnDNnzpT829LSEhMnTsT27dsxZcqUj/UPGYLhw4dj+fLlUFdXx40bN3D9+nXs3r0bABAUFIR69eph0aJFknk2bdoEc3Nz3Lt3D2lpacjOzoabmxssLCwAAA4ODoV9i0usSCu4enp6ePbs2Wf7PX/+HLq6uiUuioiIiKg8mzVtLnbsCsfd+3fyHBOLxVjx2xK07tAUtk6WsKptipNnjuNpwpMC53P78Ufs3rcPwMcP9+85cBBuP34Ms6+SkvE44QmGTR4F/Vomki+/X5fgYXxsoXW2atoS169fx/Xr1xEVFYUffvgBrq6uiI+Pl/T5448/0Lx5c1SpUgU6OjqYNWsWHj3632pyt27doKKigoiICAAfw6uzszMsLS0BAFeuXMGJEyego6Mj+bK1tQUAPHjwAHXq1IGLiwscHBzQs2dPrF+/vswXRIsUcJ2cnCQXVZiIiAg4OTmVtCYiIiKicq1Jo2ZwbumCRQHz8hwLWv8r1m1cjVHDxmLX1v04fugMWrd0QWZmVoHzde/SBQ9j4/BP9C1cunoVCYmJ6Na5EwBAEHI+zhvwCy4eOSv5unr0Ak7vjSy0Tm1NLdSoUQM1atRAw4YNsXHjRrx79w7r13/cv3vhwgW4u7vD1dUVBw4cwLVr1+Dj44PMzEzJHGpqaujfvz+Cg4ORmZmJrVu3YtCgQZLjOTk56NKliyRI537dv38fLVu2hLKyMo4ePYrDhw+jdu3a+PXXX1GrVi3ExhYezkuiSFsUBgwYAC8vL7i6uqJv3/yXwX///Xds27YNISEhpVkfERERUbnkM3UOXDo2R3Vr6Q/hX7h0Hu3bdcRP3XsD+BgAY+MeoGb1WgXOZWpSBY0bNMDuffuQnp6BFk2borKhIQCgsqEhqlYxRWx8HPr8/5yyEolEUFJSwocPHwAAf//9NywsLODj4yPp8+nqbq4hQ4bA3t4eq1evRlZWFtzc3CTH6tWrh127dsHS0hIqBdz6TCQSoVmzZmjWrBlmz54NCwsLREREYMKECfn2L6kiBdx+/fphy5YtGDBgALZu3YquXbvC6v/3hMTGxmLPnj34888/0b59+wIDMBEREZEiqW37HXp064WNIeuk2q0srHHwyD5cuhIFPb2KWLPhN7x4+aLQgAsAbj92wbJff0VmZhbm+syQOjZzwnRMmD0FFSpUQHvndsjMyMSVf67hdcobjBs6qsA5MzIzJNtMX79+jd9++w1paWmS5xbUqFEDjx49wrZt29CgQQMcPHgw37/a29nZoXHjxpg6dSoGDRokdevYkSNHYv369ejTpw8mT54MQ0ND/Pvvv9i2bRvWr1+Py5cvIzIyEj/88AOMjIwQFRWFly9fws7OrvBvcAkUKeCKRCLs3bsX48ePx8aNG3H48GGIRCIAH/eJqKqqSjYf57YTERERKbqpE3yw74B0IJwwZjIePY5H7wFu0NTURP8+nnBt1wmpb1MLnatLR1fMnDcfSsrKcG3XVurYoD4DoaWhieVrf8GMRbOhrakFe9vvMHqwd6Fz/nXyGExMTAB8vFuCra0tdu7cidatWwMAunbtivHjx2PUqFHIyMhAp06dMGvWLMyZMyfPXIMHD8a5c+ekticAgKmpKf7++29MnToV7du3R0ZGBiwsLNChQwcoKSlBV1cXp0+fxsqVK5GamgoLCwssW7YMrq6uhdZeEiKhmPdoeP78OU6cOCHZfFytWjU4OzvD2Ni4TAr8UlJTU6Gnp4eUlBSF+KCc5bSDMo+N0/CQeayDVTWZx+7wy5Zp3PHWeT/BWlTprwu+j+Hn9LaaKvNYM/8WMo+lsvW1vXdkfd8AX997h+8bxZGeno7Y2FhYWVlBQ0MDSLgm0zy31NRkrsH6mey3qHpbQfbfdTnZz2Uap69eReZzqplVkHnspxYuXIht27bh5s2bpTJfQfL8fHyiqHmt2M+IMzY2hru7e/GrJSIiIqKvTlpaGmJiYvDrr79i/vz58i6nSIoUcD+9VcR/qaqqwsDAAGol+H9RRERERFQ+jRo1CuHh4ejWrVue7QnlVZECrqWlZaF7a5WVldGiRQv4+fmhYcOGpVYcEREREclXSEjIV3eXrCIF3ICAgAIDbnZ2Np4+fYp9+/ahdevWOHfuHO+FS0RERERyU6SAO2nSpM/2CQgIQNOmTbFw4ULs3LmzxIUREREREcmiSE8yKwoNDQ2MHTsWp0+fLq0piYiISIEV80ZO9I3Iyckp8RzFvotCYSwsLMr82cJERET0dVNVVYVIJMLLly9RuXJliLJlC7o5ItmDUEaO7OE6Kzvz850KkCMWyzQuvSTnTE+XeeyXJAgCMjMz8fLlSygpKZXoBgalGnAfPXqESpUqleaUREREpGCUlZVhZmaGJ0+eIC4uDnjzUqZ5XhTwWNiiEAp/5kKh0lNlC6kAIOTIduI3Kh9kPqfKB43PdypHtLS0UK1aNSgpyb7RoNQCbmZmJlauXIkWLXgjbiIiIiqcjo4OatasiaysLOC3njLNMbaqqcznX7FO9oek3Gw4W+axmakhMo1zNRsi8zmrTCy7R+KWNmVlZaioqJT4ybhFCrjLlxf8xBqxWCy5i8KzZ8+wcePGEhVERERE3wZlZWUoKysDaY9lGp+YKXsIUkqUPeCmp8i+NSL99SuZxqnoyb6l4r9PA/sWlMpdFJSVldGsWTNs3boVdevWLZXCiIiIiIhkUaSAGxsbW/AEKiowNDSEuro6AEg2jBMRERERyUORAq6FhUWhxwVBwKFDh7Bx40YcOHAAGRkZpVIcEREREVFxlehDZg8ePMCmTZsQGhqKxMREqKmpoUePHqVVGxERERFRsRU74Kanp2Pnzp3YuHEjzpw5A0EQIBKJMGHCBEybNg0GBgZlUScRERERUZEU+QZjly5dwvDhw1GlShV4enri6tWr8PT0xIEDByAIArp06cJwS0RERERyV6QVXEdHR9y6dQsA0KRJEwwaNAi9e/eGtrY2UlJSyrRAIiIiIqLiKFLAjY6OhkgkQqdOneDv74/atWuXdV1ERERERDIp0haFlStXwtHREQcOHICDgwOaNGmCDRs24O3bt2VdHxERERFRsRQp4I4ZMwbXrl3DxYsXMXToUNy5cwdDhw6FiYkJhg4dCpFIVOJHqhERERERlYYif8gMAOrXr4+goCAkJiYiNDQU9evXxx9//AFBEDB48GAsW7YMSUlJZVUrEREREdFnFSvg5tLQ0ED//v1x8uRJ3Lt3D9OmTcP79+8xefJkmJubl3aNRERERERFJlPA/VT16tWxaNEiPHr0CPv27UOHDh1Koy4iIiIiIpmU6Elmn1JSUkLnzp3RuXPn0pqSiIiIiKjYSryCS0RERERUnjDgEhEREZFCYcAlIiIiIoXCgEtERERECoUBl4iIiIgUCgMuERERESkUBlwiIiIiUigMuERERESkUMpFwE1LS8O4ceNgamoKDQ0NODk5Ydu2bcWeZ+bMmRCJRLC3ty+DKomIiIjoa1BqTzIrCTc3N1y6dAn+/v6wsbHB1q1b0adPH+Tk5MDDw6NIc1y/fh1Lly6FsbFxGVdLREREROWZ3APuoUOHcPToUUmoBQBnZ2fEx8dj8uTJ6N27N5SVlQudIzs7G15eXhg2bBhu3LiBV69efYnSiYiIiKgckvsWhYiICOjo6KBnz55S7V5eXkhISEBUVNRn5/D390dycjIWLlxYVmUSERER0VdC7gE3OjoadnZ2UFGRXkx2dHSUHC/M7du3sWDBAgQFBUFHR6fI583IyEBqaqrUFxERERF9/eQecJOSkqCvr5+nPbctKSmpwLE5OTkYNGgQ3Nzc0LFjx2Kd18/PD3p6epIvc3Pz4hVOREREROWS3AMuAIhEIpmOLV++HPfv38fKlSuLfc7p06cjJSVF8vX48eNiz0FERERE5Y/cP2RmYGCQ7yptcnIyAOS7ugsAjx49wuzZs+Hv7w81NTW8efMGwMcPnOXk5ODNmzdQV1eHpqZmvuPV1dWhrq5eOhdBREREROWG3FdwHRwcEBMTg+zsbKn2mzdvAkCB97R9+PAhPnz4gLFjx6JSpUqSr7///hsxMTGoVKkSpk+fXub1ExEREVH5IvcV3O7du2P9+vXYtWsXevfuLWkPDQ2FqakpGjVqlO84JycnnDhxIk/7uHHjkJKSguDgYJiZmZVZ3URERERUPsk94Lq6uqJdu3YYMWIEUlNTUaNGDYSHh+PIkSMICwuT3AN38ODBCA0NxYMHD2BhYYGKFSuidevWeearWLEisrOz8z1GRERERIpP7gEXAHbv3g0fHx/Mnj0bycnJsLW1RXh4ONzd3SV9xGIxxGIxBEGQY6VEREREVN6Vi4Cro6ODwMBABAYGFtgnJCQEISEhn53r5MmTpVcYEREREX115P4hMyIiIiKi0sSAS0REREQKhQGXiIiIiBQKAy4RERERKRQGXCIiIiJSKAy4RERERKRQGHCJiIiISKEw4BIRERGRQmHAJSIiIiKFwoBLRERERAqFAZeIiIiIFAoDLhEREREpFAZcIiIiIlIoDLhEREREpFAYcImIiIhIoTDgEhEREZFCYcAlIiIiIoXCgEtERERECoUBl4iIiIgUCgMuERERESkUBlwiIiIiUigMuERERESkUBhwiYiIiEihMOASERERkUJhwCUiIiIihcKAS0REREQKhQGXiIiIiBQKAy4RERERKRQGXCIiIiJSKAy4RERERKRQGHCJiIiISKEw4BIRERGRQmHAJSIiIiKFwoBLRERERAqFAZeIiIiIFAoDLhEREREpFAZcIiIiIlIoKvIugIiIiL5+ltMOyjw2TqMUCyECV3CJiIiISMEw4BIRERGRQmHAJSIiIiKFwoBLRERERAqFAZeIiIiIFAoDLhEREREpFAZcIiIiIlIoDLhEREREpFAYcImIiIhIoZSLgJuWloZx48bB1NQUGhoacHJywrZt2z47bvfu3ejTpw9q1KgBTU1NWFpaom/fvrh///4XqJqIiIiIyqNy8aheNzc3XLp0Cf7+/rCxscHWrVvRp08f5OTkwMPDo8BxixcvRpUqVeDj4wNra2s8fvwYixYtQr169XDhwgV89913X/AqiIiIiKg8kHvAPXToEI4ePSoJtQDg7OyM+Ph4TJ48Gb1794aysnK+Y/fv3w8jIyOptjZt2sDS0hIrVqzAhg0byrx+IiIiIipf5L5FISIiAjo6OujZs6dUu5eXFxISEhAVFVXg2P+GWwAwNTWFmZkZHj9+XOq1EhEREVH5J/eAGx0dDTs7O6ioSC8mOzo6So4Xx8OHDxEfH8/tCURERETfKLlvUUhKSoK1tXWedn19fcnxosrOzsbgwYOho6OD8ePHF9o3IyMDGRkZktepqalFPg8RERERlV9yX8EFAJFIJNOxTwmCgMGDB+PMmTPYvHkzzM3NC+3v5+cHPT09ydfn+hMRERHR10HuAdfAwCDfVdrk5GQA/1vJLYwgCBgyZAjCwsIQEhKCrl27fnbM9OnTkZKSIvninl0iIiIixSD3gOvg4ICYmBhkZ2dLtd+8eRMAYG9vX+j43HAbHByMDRs2oF+/fkU6r7q6OnR1daW+iIiIiOjrJ/eA2717d6SlpWHXrl1S7aGhoTA1NUWjRo0KHCsIAn7++WcEBwdj7dq18PLyKutyiYiIiKick/uHzFxdXdGuXTuMGDECqampqFGjBsLDw3HkyBGEhYVJ7oE7ePBghIaG4sGDB7CwsAAAjBkzBhs3bsSgQYPg4OCACxcuSOZVV1dH3bp15XJNRERERCQ/cg+4wMdH7vr4+GD27NlITk6Gra0twsPD4e7uLukjFoshFoshCIKkbf/+/QCATZs2YdOmTVJzWlhYIC4u7ovUT0RERETlR7kIuDo6OggMDERgYGCBfUJCQhASEiLVxgBLRERERP8l9z24RERERESliQGXiIiIiBQKAy4RERERKRQGXCIiIiJSKAy4RERERKRQGHCJiIiISKEw4BIRERGRQmHAJSIiIiKFwoBLRERERAqFAZeIiIiIFAoDLhEREREpFAZcIiIiIlIoDLhEREREpFAYcImIiIhIoTDgEhEREZFCYcAlIiIiIoXCgEtERERECoUBl4iIiIgUCgMuERERESkUBlwiIiIiUigMuERERESkUBhwiYiIiEihMOASERERkUJhwCUiIiIihcKAS0REREQKhQGXiIiIiBQKAy4RERERKRQGXCIiIiJSKAy4RERERKRQGHCJiIiISKEw4BIRERGRQmHAJSIiIiKFwoBLRERERAqFAZeIiIiIFAoDLhEREREpFAZcIiIiIlIoDLhEREREpFAYcImIiIhIoTDgEhEREZFCYcAlIiIiIoXCgEtERERECoUBl4iIiIgUCgMuERERESkUBlwiIiIiUigMuERERESkUBhwiYiIiEihMOASERERkUIpFwE3LS0N48aNg6mpKTQ0NODk5IRt27YVaeyLFy/g6ekJQ0NDaGlpoUmTJoiMjCzjiomIiIiovFKRdwEA4ObmhkuXLsHf3x82NjbYunUr+vTpg5ycHHh4eBQ4LiMjAy4uLnjz5g0CAwNhZGSEVatWoUOHDjh27BhatWr1Ba+CiIiIiMoDuQfcQ4cO4ejRo5JQCwDOzs6Ij4/H5MmT0bt3bygrK+c7duPGjYiOjsa5c+fQpEkTydg6depgypQpiIqK+mLXQURERETlg9y3KEREREBHRwc9e/aUavfy8kJCQkKhITUiIgK1atWShFsAUFFRQb9+/XDx4kU8ffq0zOomIiIiovJJ7gE3OjoadnZ2UFGRXkx2dHSUHC9sbG6//MbeunWrFCslIiIioq+B3LcoJCUlwdraOk+7vr6+5HhhY3P7FXdsRkYGMjIyJK9TUlIAAKmpqUUrvJzLyXgv89hUkSDzWPEHscxj08Syjf2Q+U7mc2ZkZck89m2G7OdVlJ8zRfS1vXdkfd8AX997h++b8k0e7x15/M4B5PPe4e+cj3KvRRAK/5mRe8AFAJFIJNOxkoz18/PD3Llz87Sbm5sXer5vgV6JRsfIPLKhrAP//VHmc5bETPwl++CVpVYGlSPyeO/I/L4Bvr73zspSLYPKEdnfO3L4nQPI5b3D3znS3r59Cz29gn9y5B5wDQwM8l1pTU5OBoB8V2hLY+z06dMxYcIEyeucnBwkJyfDwMDgs6GavqzU1FSYm5vj8ePH0NXVlXc5RF8NvneIio/vm/JNEAS8ffsWpqamhfaTe8B1cHBAeHg4srOzpfbh3rx5EwBgb29f6Njcfp8qylh1dXWoq6tLtVWsWLE4pdMXpqury//YEMmA7x2i4uP7pvwqbOU2l9w/ZNa9e3ekpaVh165dUu2hoaEwNTVFo0aNCh17584dqTstZGdnIywsDI0aNfpsuiciIiIixSP3FVxXV1e0a9cOI0aMQGpqKmrUqIHw8HAcOXIEYWFhknvgDh48GKGhoXjw4AEsLCwAAIMGDcKqVavQs2dP+Pv7w8jICKtXr8bdu3dx7NgxeV4WEREREcmJ3AMuAOzevRs+Pj6YPXs2kpOTYWtri/DwcLi7u0v6iMViiMViqU/NqaurIzIyElOmTMHo0aPx/v17ODk54fDhw3yKmQJRV1eHr69vni0lRFQ4vneIio/vG8UgEj53nwUiIiIioq+I3PfgEhERERGVJgZcIiIiIlIoDLhEREREpFAYcClf//zzD7y8vGBlZQUNDQ3o6OigXr16CAgIkDxIQ1G1bt0arVu3lncZ9BUJCQmBSCTK92vSpEkAAEtLS3h6esq30DJy8uRJiEQinDx58rN9PT09YWlpWeY10dcvKioK3bt3R7Vq1aCurg5jY2M0adIEEydOlHdpxVKc9weVnnJxFwUqX9avXw9vb2/UqlULkydPRu3atZGVlYXLly9jzZo1OH/+PCIiIuRdZplZvXq1vEugr1RwcDBsbW2l2r6F+3HXq1cP58+fR+3ateVdCimIgwcP4scff0Tr1q0REBAAExMTJCYm4vLly9i2bRuWLVsm7xKLjO8P+eBdFEjK+fPn0aJFC7Rr1w579uzJc5uUzMxMHDlyBD/+KJ9n2Jel9+/fQ0tLS95l0FcoJCQEXl5euHTpEurXr59vH0tLS7Ru3RohISFftrhyxtPTEydPnkRcXJy8S6FyrFWrVnj69Cnu3Lkj9ZRTAMjJyYGSEv8ATYXjTwhJWbRoEUQiEdatW5fvPQDV1NQk4TYnJwcBAQGwtbWFuro6jIyMMGDAADx58kRqTEF/mv3vVoDcP+Ns3boVU6dOhYmJCXR0dNClSxc8f/4cb9++xdChQ2FoaAhDQ0N4eXkhLS1Nas5Vq1ahZcuWMDIygra2NhwcHBAQEICsrKw857a3t8fp06fRtGlTaGlpYdCgQfnWFRcXB5FIhKVLl2L58uWwsrKCjo4OmjRpggsXLhTn20skkZ6ejokTJ8LJyQl6enrQ19dHkyZNsHfvXql+devWRYsWLfKMF4vFqFq1Ktzc3CRtycnJ8Pb2RtWqVaGmpgZra2v4+PggIyNDaqxIJMKoUaPw+++/w87ODlpaWqhTpw4OHDiQ5zxnz56Fi4sLKlSoAC0tLTRt2hQHDx6U6lPQn2BDQkJQq1YtqKurw87ODps3by7ut4m+UUlJSTA0NMwTbgFIhdvt27fjhx9+gImJCTQ1NWFnZ4dp06bh3bt3UmM8PT2ho6ODW7duwcXFBdra2qhcuTJGjRqF9+/fS/VNT0/H9OnTYWVlBTU1NVStWhUjR47EmzdvpPpZWlqic+fOOHLkCOrVqwdNTU3Y2tpi06ZNUv3ye3/k1vPvv/+iY8eO0NHRgbm5OSZOnJjn/UqyYcAlCbFYjOPHj+P777+Hubn5Z/uPGDECU6dORbt27bBv3z7Mnz8fR44cQdOmTfHq1SuZ65gxYwZevHiBkJAQLFu2DCdPnkSfPn3Qo0cP6OnpITw8HFOmTMHvv/+OGTNmSI198OABPDw88Pvvv+PAgQMYPHgwlixZgmHDhuU5T2JiIvr16wcPDw8cOnQI3t7ehda1atUqHD16FCtXrsSWLVvw7t07dOzYESkpKTJfKykWsViM7Oxsqa+CZGRkIDk5GZMmTcKePXsQHh6O5s2bw83NTSoIenl54ezZs7h//77U+L/++gsJCQnw8vIC8PGXsrOzMzZv3owJEybg4MGD6NevHwICAqRCcK6DBw/it99+w7x587Br1y7o6+uje/fuePjwoaTPqVOn0KZNG6SkpGDjxo0IDw9HhQoV0KVLF2zfvr3Q70XuqradnR127dqFmTNnYv78+Th+/HiRvpf0bWvSpAmioqIwZswYREVF5VmkyHX//n107NgRGzduxJEjRzBu3Djs2LEDXbp0ydM3KysLHTt2hIuLC/bs2YNRo0Zh7dq16N27t6SPIAjo1q0bli5div79++PgwYOYMGECQkND0aZNmzzh88aNG5g4cSLGjx+PvXv3wtHREYMHD8bp06c/e41ZWVn48ccf4eLigr1792LQoEFYsWIFFi9eXMzvFuVLIPp/z549EwAI7u7un+0bExMjABC8vb2l2qOiogQAwowZMyRtFhYWwsCBA/PM0apVK6FVq1aS1ydOnBAACF26dJHqN27cOAGAMGbMGKn2bt26Cfr6+gXWKBaLhaysLGHz5s2CsrKykJycLHVuAEJkZORn64qNjRUACA4ODkJ2drak/eLFiwIAITw8vMAa6NsQHBwsAMj3KysrSxCEgt8HubKzs4WsrCxh8ODBQt26dSXtr169EtTU1KTeU4IgCL169RKMjY0l869Zs0YAIOzYsUOq3+LFiwUAwl9//SVpAyAYGxsLqampkrZnz54JSkpKgp+fn6StcePGgpGRkfD27VupOu3t7QUzMzMhJydHEIT/vXdPnDghCMLH956pqalQr149SR9BEIS4uDhBVVVVsLCwKOzbSSS8evVKaN68ueR9pKqqKjRt2lTw8/OT+nn8VE5OjpCVlSWcOnVKACDcuHFDcmzgwIECACEwMFBqzMKFCwUAwtmzZwVBEIQjR44IAISAgACpftu3bxcACOvWrZO0WVhYCBoaGkJ8fLyk7cOHD4K+vr4wbNgwSdt/3x+f1vPf92vHjh2FWrVqFfG7RIXhCi7J5MSJEwCQZ+tBw4YNYWdnh8jISJnn7ty5s9RrOzs7AECnTp3ytCcnJ0ttU7h27Rp+/PFHGBgYQFlZGaqqqhgwYADEYjHu3bsnNb5SpUpo06ZNkevq1KkTlJWVJa8dHR0BAPHx8UWegxTb5s2bcenSJamv/P7Emmvnzp1o1qwZdHR0oKKiAlVVVWzcuBExMTGSPgYGBujSpQtCQ0ORk5MDAHj9+jX27t2LAQMGSOY/fvw4tLW18dNPP0mdI/c9+t/3pLOzMypUqCB5bWxsDCMjI8nP87t37xAVFYWffvoJOjo6kn7Kysro378/njx5grt37+Z7XXfv3kVCQgI8PDwgEokk7RYWFmjatGmB3w+iXAYGBjhz5gwuXboEf39/dO3aFffu3cP06dPh4OAg+Svhw4cP4eHhgSpVqkj+m9+qVSsAkHof5erbt6/Uaw8PDwD/+52W+xeG//5u69mzJ7S1tfO8j5ycnFCtWjXJaw0NDdjY2BTp94JIJMqz0uzo6MjfKaWEAZckDA0NoaWlhdjY2M/2TUpKAgCYmJjkOWZqaio5Lgt9fX2p12pqaoW2p6enAwAePXqEFi1a4OnTpwgMDJT8x3HVqlUAgA8fPkiNz6/2whgYGEi9zt2j/N956dtlZ2eH+vXrS30VZPfu3ejVqxeqVq2KsLAwnD9/HpcuXcKgQYMkP9O5Bg0ahKdPn+Lo0aMAgPDwcGRkZEj9Ek5KSkKVKlWkAiUAGBkZQUVFJc978r8/z8DHn+ncn+fXr19DEIQC3+O558xPbnuVKlXyHMuvjagg9evXx9SpU7Fz504kJCRg/PjxiIuLQ0BAANLS0tCiRQtERUVhwYIFOHnyJC5duoTdu3cDyPvfZhUVlTw/97k/j7k/s0lJSVBRUUHlypWl+olEIlSpUqXY76PCaGlpQUNDI8/Y/77/STa8TRhJKCsrw8XFBYcPH8aTJ09gZmZWYN/cN3ViYmKefgkJCTA0NJS81tDQyHfT/KtXr6T6ldSePXvw7t077N69GxYWFpL269ev59v/v0GA6EsKCwuDlZUVtm/fLvWzmN97pX379jA1NUVwcDDat2+P4OBgNGrUSOq2QwYGBoiKioIgCFLzvXjxAtnZ2cV+r1WqVAlKSkpITEzMcywhIQEACpwz978Pz549y3MsvzaiolBVVYWvry9WrFiB6OhoHD9+HAkJCTh58qRk1RZAng+D5crOzkZSUpJUKM39ecxtMzAwQHZ2Nl6+fCkVcgVBwLNnz9CgQYMyuDIqC1zBJSnTp0+HIAj4+eefkZmZmed4VlYW9u/fL/nTflhYmNTxS5cuISYmBi4uLpI2S0tL/PPPP1L97t27V+CfN2WV+0v907s/CIKA9evXl+p5iEqDSCSCmpqaVBh99uxZnrsoAP/bFrBnzx6cOXMGly9fltz1I5eLiwvS0tKwZ88eqfbcD6x9+p4sCm1tbTRq1Ai7d++WWo3KyclBWFgYzMzMYGNjk+/YWrVqwcTEBOHh4RA+uRNlfHw8zp07V6w66NuU3/+xAv637cDU1DTf/+YDwNq1awucd8uWLVKvt27dCgCSO+fkvk/++7tt165dePfuXbHfRyQ/XMElKU2aNEFQUBC8vb3x/fffY8SIEfjuu++QlZWFa9euYd26dbC3t0dERASGDh2KX3/9FUpKSnB1dUVcXBxmzZoFc3NzjB8/XjJn//790a9fP3h7e6NHjx6Ij49HQEBAnj8BlVS7du2gpqaGPn36YMqUKUhPT0dQUBBev35dquchKg2dO3fG7t274e3tjZ9++gmPHz/G/PnzYWJikueOCcDHbQqLFy+Gh4cHNDU1pT75DQADBgzAqlWrMHDgQMTFxcHBwQFnz57FokWL0LFjR7Rt27bYNfr5+aFdu3ZwdnbGpEmToKamhtWrVyM6Ohrh4eEF/hVESUkJ8+fPx5AhQ9C9e3f8/PPPePPmDebMmcMtClQk7du3h5mZGbp06QJbW1vk5OTg+vXrWLZsGXR0dDB27FiYmpqiUqVKGD58OHx9faGqqootW7bgxo0b+c6ppqaGZcuWIS0tDQ0aNMC5c+ewYMECuLq6onnz5gA+/h5p3749pk6ditTUVDRr1gz//PMPfH19UbduXfTv3/9LfhuoBBhwKY+ff/4ZDRs2lNyu5NmzZ1BVVYWNjQ08PDwwatQoAEBQUBCqV6+OjRs3YtWqVdDT00OHDh3g5+cn9ScgDw8PJCQkYM2aNQgODoa9vT2CgoIwd+7cUq3b1tZWcjsiNzc3GBgYwMPDAxMmTICrq2upnouopLy8vPDixQusWbMGmzZtgrW1NaZNm4YnT57k+96wsbFB06ZNce7cOfTt2xd6enpSxzU0NHDixAn4+PhgyZIlePnyJapWrYpJkybB19dXphpbtWqF48ePw9fXF56ensjJyUGdOnWwb9++PB8G/a/BgwcDABYvXgw3NzdYWlpixowZOHXqFB9ZSp81c+ZM7N27FytWrEBiYiIyMjJgYmKCtm3bYvr06ZIPHx88eBATJ05Ev379oK2tja5du2L79u2oV69enjlVVVVx4MABjBkzBgsWLICmpiZ+/vlnLFmyRNJHJBJhz549mDNnDoKDg7Fw4UIYGhqif//+WLRoUb73h6fyiU8yIyIiIoXm6emJP/74I8/DgUhxcQ8uERERESkUBlwiIiIiUijcokBERERECoUruERERESkUBhwiYiIiEihMOASERERkUJhwCUiIiIihcKAS0REREQKhQGXiIhKXVxcHEQiEa5fvy7vUojoG8SAS0RUQs+ePcPo0aNhbW0NdXV1mJubo0uXLoiMjJR3aURE3yQVeRdARPQ1i4uLQ7NmzVCxYkUEBATA0dERWVlZ+PPPPzFy5EjcuXNH3iUSEX1zuIJLRFQC3t7eEIlEuHjxIn766SfY2Njgu+++w4QJE3DhwgUAwKNHj9C1a1fo6OhAV1cXvXr1wvPnzyVzzJkzB05OTti0aROqVasGHR0djBgxAmKxGAEBAahSpQqMjIywcOFCqXOLRCIEBQXB1dUVmpqasLKyws6dO6X63Lx5E23atIGmpiYMDAwwdOhQpKWlSY63bt0a48aNkxrTrVs3eHp6Sl5bWlpi0aJFGDRoECpUqIBq1aph3bp1UmMuXryIunXrQkNDA/Xr18e1a9ekjr9+/Rp9+/ZF5cqVoampiZo1ayI4OLjY328ioqJgwCUiklFycjKOHDmCkSNHQltbO8/xihUrQhAEdOvWDcnJyTh16hSOHj2KBw8eoHfv3lJ9Hzx4gMOHD+PIkSMIDw/Hpk2b0KlTJzx58gSnTp3C4sWLMXPmTElozjVr1iz06NEDN27cQL9+/dCnTx/ExMQAAN6/f48OHTqgUqVKuHTpEnbu3Iljx45h1KhRxb7WZcuWSYKrt7c3RowYIVmdfvfuHTp37oxatWrhypUrmDNnDiZNmpSnztu3b+Pw4cOIiYlBUFAQDA0Ni10HEVGRCEREJJOoqCgBgLB79+4C+/z111+CsrKy8OjRI0nbrVu3BADCxYsXBUEQBF9fX0FLS0tITU2V9Gnfvr1gaWkpiMViSVutWrUEPz8/yWsAwvDhw6XO16hRI2HEiBGCIAjCunXrhEqVKglpaWmS4wcPHhSUlJSEZ8+eCYIgCK1atRLGjh0rNUfXrl2FgQMHSl5bWFgI/fr1k7zOyckRjIyMhKCgIEEQBGHt2rWCvr6+8O7dO0mfoKAgAYBw7do1QRAEoUuXLoKXl1eB3yciotLEFVwiIhkJggDg41aBgsTExMDc3Bzm5uaSttq1a6NixYqSlVbg4zaAChUqSF4bGxujdu3aUFJSkmp78eKF1PxNmjTJ8zp33piYGNSpU0dqdblZs2bIycnB3bt3i3OpcHR0lPxbJBKhSpUqklpyz6OlpVVgXSNGjMC2bdvg5OSEKVOm4Ny5c8U6PxFRcTDgEhHJqGbNmhCJRFJB9b8EQcg3AP+3XVVVVeq4SCTKty0nJ+ezdeXOW9C5P+2jpKQkCeq5srKy8vQvrJb/js+Pq6sr4uPjMW7cOCQkJMDFxSXPNgYiotLCgEtEJCN9fX20b98eq1atwrt37/Icf/PmDWrXro1Hjx7h8ePHkvbbt28jJSUFdnZ2Ja7hv3tyL1y4AFtbWwAfV4qvX78uVdvff/8NJSUl2NjYAAAqV66MxMREyXGxWIzo6Ohi1VC7dm3cuHEDHz58KLCu3HN5enoiLCwMK1euzPNBNSKi0sKAS0RUAqtXr4ZYLEbDhg2xa9cu3L9/HzExMfjll1/QpEkTtG3bFo6Ojujbty+uXr2KixcvYsCAAWjVqhXq169f4vPv3LkTmzZtwr179+Dr64uLFy9KPkTWt29faGhoYODAgYiOjsaJEycwevRo9O/fH8bGxgCANm3a4ODBgzh48CDu3LkDb29vvHnzplg1eHh4QElJCYMHD8bt27dx6NAhLF26VKrP7NmzsXfvXvz777+4desWDhw4UCoBn4goPwy4REQlYGVlhatXr8LZ2RkTJ06Evb092rVrh8jISAQFBUEkEmHPnj2oVKkSWrZsibZt28La2hrbt28vlfPPnTsX27Ztg6OjI0JDQ7FlyxbUrl0bAKClpYU///wTycnJaNCgAX766Se4uLjgt99+k4wfNGgQBg4cKAndVlZWcHZ2LlYNOjo62L9/P27fvo26devCx8cHixcvluqjpqaG6dOnw9HRES1btoSysjK2bdtW8m8AEVE+REJRNk8REVG5IxKJEBERgW7dusm7FCKicoUruERERESkUBhwiYiIiEihqMi7ACIikg13mBER5Y8ruERERESkUBhwiYiIiEihMOASERERkUJhwCUiIiIihcKAS0REREQKhQGXiIiIiBQKAy4RERERKRQGXCIiIiJSKP8HSOHk7PG1IusAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Make a data definition\n",
    "_data=[[\"Coumarin\",0.598, 0.717, 0.913, 0.620, 0.687, 0.691, 0.633],[\"Flavonoid\",0.700,0.739, 0.947, 0.829, 0.817, 0.867, 0.718],[\"Saponin\",0.251, 0.688, 0.760, 0.975, 0.967, 0.975, 0.861]]\n",
    "\n",
    "# Plot multiple columns bar chart\n",
    "_df = pd.DataFrame(_data, columns=[\"Compounds\", 'MLP', 'RNN', 'CNN', 'Logistic', 'Ridge','Lasso', 'Naive Bayes'], )\n",
    "_df.plot(x=\"Compounds\", y=['MLP', 'RNN', 'CNN','Logistic', 'Ridge','Lasso', 'Naive Bayes'], kind=\"bar\",rot=0,figsize=(8,4), fontsize=12)\n",
    "plt.ylabel(\"AUC\",fontsize=12)\n",
    "plt.title('AUC Distribution', fontsize=15)\n",
    "\n",
    "# Display the plot\n",
    "plt.savefig('AUC score based on compounds.jpg', dpi=600, transparent=False, bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}