4_Modelling_and_Evaluations_DNN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "70c3d8bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.impute import SimpleImputer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "cbd54bb3",
   "metadata": {},
   "outputs": [],
   "source": [
    "from cust_funcs import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "2ef7a987",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test,weights,y_test=prepare_dataset(remove_phi_cols(pd.read_csv('Data/testing.csv',index_col=0)))\n",
    "X_train,_,y_train=prepare_dataset(remove_phi_cols(pd.read_csv('Data/training.csv')))\n",
    "\n",
    "\n",
    "X_test=prepare_training_features(X_test,X_dep=X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e402ea9c",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train=X_train.values\n",
    "y_train=y_train.values\n",
    "#y_test=y_test.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "832dcc1e",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "a7f58d7e",
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler=MinMaxScaler().fit(X_train)\n",
    "X_train=scaler.transform(X_train)\n",
    "X_test=scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "72175fbe",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test,X_val,y_test,y_val=train_test_split(X_test,y_test,test_size=0.2,stratify=y_test,random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "7ac71b7f",
   "metadata": {},
   "outputs": [],
   "source": [
    "weights=weights.iloc[y_test.index]\n",
    "y_test=y_test.values"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "083e7b2f",
   "metadata": {},
   "source": [
    "## Training DNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "fb70f7a6",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-05-21 17:16:58.286281: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 AVX_VNNI FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2023-05-21 17:16:58.413097: I tensorflow/core/util/util.cc:169] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.\n",
      "2023-05-21 17:16:58.434638: E tensorflow/stream_executor/cuda/cuda_blas.cc:2981] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered\n",
      "2023-05-21 17:16:58.843905: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: :/home/philip/miniconda3/envs/tf/lib/:/home/philip/miniconda3/envs/tf/lib/python3.9/site-packages/nvidia/cudnn/lib\n",
      "2023-05-21 17:16:58.844032: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: :/home/philip/miniconda3/envs/tf/lib/:/home/philip/miniconda3/envs/tf/lib/python3.9/site-packages/nvidia/cudnn/lib\n",
      "2023-05-21 17:16:58.844041: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.\n"
     ]
    }
   ],
   "source": [
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense,Dropout,BatchNormalization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "23f8aa70",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.optimizers import Adam"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "7ab1af72",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "model=Sequential()\n",
    "\n",
    "model.add(Dense(30,activation='relu'))\n",
    "model.add(Dropout(0.6))\n",
    "model.add(Dense(15,activation='relu'))\n",
    "model.add(Dropout(0.6))\n",
    "model.add(Dense(1,activation='sigmoid'))\n",
    "\n",
    "model.compile(optimizer='adam',loss='binary_crossentropy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "4f008aa8",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.callbacks import EarlyStopping"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "90afd661",
   "metadata": {},
   "outputs": [],
   "source": [
    "early_stop = EarlyStopping(monitor='val_loss', mode='min', verbose=1, patience=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "d440114e",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/600\n",
      "123/123 [==============================] - 1s 9ms/step - loss: 0.6627 - val_loss: 0.6140\n",
      "Epoch 2/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.6062 - val_loss: 0.5866\n",
      "Epoch 3/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.5841 - val_loss: 0.5666\n",
      "Epoch 4/600\n",
      "123/123 [==============================] - 1s 7ms/step - loss: 0.5678 - val_loss: 0.5445\n",
      "Epoch 5/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.5558 - val_loss: 0.5300\n",
      "Epoch 6/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.5452 - val_loss: 0.5157\n",
      "Epoch 7/600\n",
      "123/123 [==============================] - 1s 7ms/step - loss: 0.5363 - val_loss: 0.5079\n",
      "Epoch 8/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.5283 - val_loss: 0.4937\n",
      "Epoch 9/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.5202 - val_loss: 0.4939\n",
      "Epoch 10/600\n",
      "123/123 [==============================] - 1s 7ms/step - loss: 0.5154 - val_loss: 0.4866\n",
      "Epoch 11/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.5084 - val_loss: 0.4813\n",
      "Epoch 12/600\n",
      "123/123 [==============================] - 1s 7ms/step - loss: 0.5034 - val_loss: 0.4746\n",
      "Epoch 13/600\n",
      "123/123 [==============================] - 1s 8ms/step - loss: 0.4991 - val_loss: 0.4743\n",
      "Epoch 14/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4944 - val_loss: 0.4678\n",
      "Epoch 15/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4919 - val_loss: 0.4927\n",
      "Epoch 16/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4884 - val_loss: 0.4788\n",
      "Epoch 17/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4865 - val_loss: 0.4654\n",
      "Epoch 18/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4840 - val_loss: 0.4599\n",
      "Epoch 19/600\n",
      "123/123 [==============================] - 1s 7ms/step - loss: 0.4818 - val_loss: 0.4610\n",
      "Epoch 20/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4798 - val_loss: 0.4561\n",
      "Epoch 21/600\n",
      "123/123 [==============================] - 1s 7ms/step - loss: 0.4795 - val_loss: 0.4553\n",
      "Epoch 22/600\n",
      "123/123 [==============================] - 1s 8ms/step - loss: 0.4776 - val_loss: 0.4476\n",
      "Epoch 23/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4774 - val_loss: 0.4524\n",
      "Epoch 24/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4764 - val_loss: 0.4540\n",
      "Epoch 25/600\n",
      "123/123 [==============================] - 1s 8ms/step - loss: 0.4749 - val_loss: 0.4495\n",
      "Epoch 26/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4737 - val_loss: 0.4494\n",
      "Epoch 27/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4739 - val_loss: 0.4456\n",
      "Epoch 28/600\n",
      "123/123 [==============================] - 1s 8ms/step - loss: 0.4718 - val_loss: 0.4541\n",
      "Epoch 29/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4715 - val_loss: 0.4554\n",
      "Epoch 30/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4702 - val_loss: 0.4511\n",
      "Epoch 31/600\n",
      "123/123 [==============================] - 1s 7ms/step - loss: 0.4692 - val_loss: 0.4372\n",
      "Epoch 32/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4685 - val_loss: 0.4440\n",
      "Epoch 33/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4689 - val_loss: 0.4423\n",
      "Epoch 34/600\n",
      "123/123 [==============================] - 1s 8ms/step - loss: 0.4679 - val_loss: 0.4510\n",
      "Epoch 35/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4682 - val_loss: 0.4432\n",
      "Epoch 36/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4671 - val_loss: 0.4494\n",
      "Epoch 37/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4670 - val_loss: 0.4354\n",
      "Epoch 38/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4661 - val_loss: 0.4398\n",
      "Epoch 39/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4678 - val_loss: 0.4380\n",
      "Epoch 40/600\n",
      "123/123 [==============================] - 1s 7ms/step - loss: 0.4653 - val_loss: 0.4435\n",
      "Epoch 41/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4651 - val_loss: 0.4528\n",
      "Epoch 42/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4656 - val_loss: 0.4670\n",
      "Epoch 43/600\n",
      "123/123 [==============================] - 1s 8ms/step - loss: 0.4645 - val_loss: 0.4559\n",
      "Epoch 44/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4635 - val_loss: 0.4474\n",
      "Epoch 45/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4645 - val_loss: 0.4564\n",
      "Epoch 46/600\n",
      "123/123 [==============================] - 1s 7ms/step - loss: 0.4633 - val_loss: 0.4427\n",
      "Epoch 47/600\n",
      "123/123 [==============================] - 1s 5ms/step - loss: 0.4639 - val_loss: 0.4554\n",
      "Epoch 48/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4631 - val_loss: 0.4597\n",
      "Epoch 49/600\n",
      "123/123 [==============================] - 1s 7ms/step - loss: 0.4634 - val_loss: 0.4632\n",
      "Epoch 50/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4615 - val_loss: 0.4577\n",
      "Epoch 51/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4622 - val_loss: 0.4549\n",
      "Epoch 52/600\n",
      "123/123 [==============================] - 1s 7ms/step - loss: 0.4612 - val_loss: 0.4611\n",
      "Epoch 53/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4624 - val_loss: 0.4527\n",
      "Epoch 54/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4604 - val_loss: 0.4510\n",
      "Epoch 55/600\n",
      "123/123 [==============================] - 1s 7ms/step - loss: 0.4600 - val_loss: 0.4477\n",
      "Epoch 56/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4611 - val_loss: 0.4549\n",
      "Epoch 57/600\n",
      "123/123 [==============================] - 1s 6ms/step - loss: 0.4602 - val_loss: 0.4596\n",
      "Epoch 57: early stopping\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x7f245c23b8b0>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train,y_train,epochs=600,batch_size=2048,callbacks=[early_stop],validation_data=(X_val,y_val))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "9badb929",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkAAAAGwCAYAAABB4NqyAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABtAElEQVR4nO3dd3iUVdrH8e/MpFdKIAkQQui9hY5gQVEUFRuoiH0VFRXbruj6WndRd1V0FbtgQWUVcFVQwUITUEqC9A6hJIQA6T153j8OGYgppM4k5Pe5rrkyeeqZB3RuzrnPfWyWZVmIiIiINCB2dzdARERExNUUAImIiEiDowBIREREGhwFQCIiItLgKAASERGRBkcBkIiIiDQ4CoBERESkwfFwdwPqosLCQg4dOkRgYCA2m83dzREREZEKsCyLtLQ0WrRogd1efh+PAqBSHDp0iIiICHc3Q0RERKpg//79tGrVqtxjFACVIjAwEDAPMCgoyM2tERERkYpITU0lIiLC+T1eHgVApSga9goKClIAJCIiUs9UJH1FSdAiIiLS4CgAEhERkQZHAZCIiIg0OMoBEhERKUNBQQF5eXnuboacwsvL67RT3CtCAZCIiMifWJZFQkICycnJ7m6K/IndbicqKgovL69qXUcBkIiIyJ8UBT/NmzfHz89PRXHriKJCxfHx8bRu3bpafy4KgERERE5RUFDgDH6aNm3q7ubInzRr1oxDhw6Rn5+Pp6dnla+jJGgREZFTFOX8+Pn5ubklUpqioa+CgoJqXUcBkIiISCk07FU31dSfiwIgERERaXAUAImIiEiDowBIRETkDHHOOecwefJkdzejXlAA5EIFhRaJqdnsTcpwd1NEREQaNAVALrRy11EG/PMn7vh4jbubIiIi0qApAHKh5kHeACSm5bi5JSIiUhmWZZGZm+/yl2VZVW7z8ePHufHGG2ncuDF+fn6MGjWKHTt2OPfv27ePSy+9lMaNG+Pv70+3bt1YsGCB89zx48fTrFkzfH196dChAzNmzKj2c6xLVAjRhZoFmAAoOTOPnPwCvD0cbm6RiIhURFZeAV3/7weX33fzMxfi51W1r+qbb76ZHTt28PXXXxMUFMTf/vY3Lr74YjZv3oynpyf33HMPubm5LF26FH9/fzZv3kxAQAAATzzxBJs3b+a7774jJCSEnTt3kpWVVZMfze0UALlQIz9PvBx2cgsKOZKWQ6vGKrIlIiI1ryjw+fXXXxkyZAgAs2bNIiIigq+++oprrrmGuLg4rrrqKnr06AFA27ZtnefHxcXRp08f+vXrB0CbNm1c/hlqmwIgF7LZbDQL9OZgcpYCIBGResTX08HmZy50y32rYsuWLXh4eDBw4EDntqZNm9KpUye2bNkCwH333cddd93FwoULOf/887nqqqvo2bMnAHfddRdXXXUV69atY+TIkYwZM8YZSJ0plAPkYiGBygMSEalvbDYbfl4eLn9VtepxWblDlmU5r3n77beze/duJkyYwIYNG+jXrx//+c9/ABg1ahT79u1j8uTJHDp0iBEjRvDwww9X7eHVUQqAXKy5AiAREallXbt2JT8/n99++8257ejRo2zfvp0uXbo4t0VERDBx4kTmzp3LQw89xLvvvuvc16xZM26++WY++eQTpk2bxjvvvOPSz1DbNATmYkUB0BEFQCIiUks6dOjA5Zdfzl/+8hfefvttAgMDefTRR2nZsiWXX345AJMnT2bUqFF07NiR48eP8/PPPzuDo//7v/8jOjqabt26kZOTw7ffflsscDoTqAfIxZo5A6BsN7dERETOZDNmzCA6OprRo0czePBgLMtiwYIFeHp6AmY19XvuuYcuXbpw0UUX0alTJ6ZPnw6YFdenTJlCz549GT58OA6Hg88//9ydH6fG2azqFBk4Q6WmphIcHExKSgpBQUE1eu1Pf4vjsXkbGNG5Oe/f3L9Gry0iItWXnZ3Nnj17iIqKwsfHx93NkT8p78+nMt/f6gFyMecQWLqGwERERNxFAZCLOatBpyoAEhERcRcFQC5WlAOUlJ5DYaFGH0VERNxBAZCLhQR4Y7NBfqHF8cxcdzdHRESkQVIA5GKeDjtN/LwA1QISERFxFwVAbtBMxRBFRETcyu0B0PTp051T2aKjo1m2bFm5x+fk5PD4448TGRmJt7c37dq144MPPnDunzlzJjabrcQrO7vu1N1xBkCpdadNIiIiDYlbK0HPnj2byZMnM336dIYOHcrbb7/NqFGj2Lx5M61bty71nLFjx3L48GHef/992rdvT2JiIvn5+cWOCQoKYtu2bcW21aVaDs0DTVs0FV5ERMQ93BoAvfzyy9x2223cfvvtAEybNo0ffviBN998k6lTp5Y4/vvvv2fJkiXs3r2bJk2aANCmTZsSx9lsNsLCwmq17dVxsgdIAZCIiIg7uG0ILDc3l7Vr1zJy5Mhi20eOHMmKFStKPefrr7+mX79+vPjii7Rs2ZKOHTvy8MMPk5WVVey49PR0IiMjadWqFaNHjyYmJqbctuTk5JCamlrsVZu0HpiIiNRFbdq0Ydq0aRU61maz8dVXX9Vqe2qT23qAkpKSKCgoIDQ0tNj20NBQEhISSj1n9+7dLF++HB8fH+bNm0dSUhJ33303x44dc+YBde7cmZkzZ9KjRw9SU1N59dVXGTp0KOvXr6dDhw6lXnfq1Kk8/fTTNfsBy1FUDFEBkIiIiHu4PQnaZrMV+92yrBLbihQWFmKz2Zg1axYDBgzg4osv5uWXX2bmzJnOXqBBgwZxww030KtXL4YNG8Z///tfOnbsyH/+858y2zBlyhRSUlKcr/3799fcByxFs4CiWWBKghYREXEHtwVAISEhOByOEr09iYmJJXqFioSHh9OyZUuCg4Od27p06YJlWRw4cKDUc+x2O/3792fHjh1ltsXb25ugoKBir9rUPMgkQWsavIhIPWFZkJvh+lcl1it/++23admyJYWFhcW2X3bZZdx0003s2rWLyy+/nNDQUAICAujfvz8//vhjjT2iDRs2cN555+Hr60vTpk254447SE9Pd+5fvHgxAwYMwN/fn0aNGjF06FD27dsHwPr16zn33HMJDAwkKCiI6Oho1qxZU2NtK43bhsC8vLyIjo5m0aJFXHHFFc7tixYt4vLLLy/1nKFDh/LFF1+Qnp5OQEAAANu3b8dut9OqVatSz7Esi9jYWHr06FHzH6KKinKAMnMLyMjJx9/brbnoIiJyOnmZ8M8Wrr/vY4fAy79Ch15zzTXcd999/PLLL4wYMQKA48eP88MPP/DNN9+Qnp7OxRdfzHPPPYePjw8ffvghl156Kdu2bStz5nVFZWZmctFFFzFo0CBWr15NYmIit99+O5MmTWLmzJnk5+czZswY/vKXv/DZZ5+Rm5vL77//7hzxGT9+PH369OHNN9/E4XAQGxuLp6dntdp0Om795n3wwQeZMGEC/fr1Y/DgwbzzzjvExcUxceJEwAxNHTx4kI8++giA66+/nmeffZZbbrmFp59+mqSkJB555BFuvfVWfH19AXj66acZNGgQHTp0IDU1lddee43Y2FjeeOMNt33OP/P39sDfy0FGbgGJaTlEKQASEZFqatKkCRdddBGffvqpMwD64osvaNKkCSNGjMDhcNCrVy/n8c899xzz5s3j66+/ZtKkSdW696xZs8jKyuKjjz7C398EbK+//jqXXnopL7zwAp6enqSkpDB69GjatWsHmBGcInFxcTzyyCN07twZoMyc3Zrk1m/ecePGcfToUZ555hni4+Pp3r07CxYsIDIyEoD4+Hji4uKcxwcEBLBo0SLuvfde+vXrR9OmTRk7dizPPfec85jk5GTuuOMOEhISCA4Opk+fPixdupQBAwa4/POVp1mgNxlHM0lMzSYqpGLRvYiIuImnn+mNccd9K2H8+PHccccdTJ8+HW9vb2bNmsW1116Lw+EgIyODp59+mm+//ZZDhw6Rn59PVlZWse/ZqtqyZQu9evVyBj9gRm0KCwvZtm0bw4cP5+abb+bCCy/kggsu4Pzzz2fs2LGEh4cDpkPk9ttv5+OPP+b888/nmmuucQZKtcXtXQ933303d999d6n7Zs6cWWJb586dWbRoUZnXe+WVV3jllVdqqnm1pnmgD3uPZqoYoohIfWCzVXgoyp0uvfRSCgsLmT9/Pv3792fZsmW8/PLLADzyyCP88MMP/Pvf/6Z9+/b4+vpy9dVXk5tb/YW5y5vAVLR9xowZ3HfffXz//ffMnj2bv//97yxatIhBgwbx1FNPcf311zN//ny+++47nnzyST7//PNiKTI1ze2zwBqqZkEqhigiIjXL19eXK6+8klmzZvHZZ5/RsWNHoqOjAVi2bBk333wzV1xxBT169CAsLIy9e/fWyH27du1KbGwsGRkZzm2//vordrudjh07Orf16dOHKVOmsGLFCrp3786nn37q3NexY0ceeOABFi5cyJVXXsmMGTNqpG1lUQDkJienwisAEhGRmjN+/Hjmz5/PBx98wA033ODc3r59e+bOnUtsbCzr16/n+uuvLzFjrDr39PHx4aabbmLjxo388ssv3HvvvUyYMIHQ0FD27NnDlClTWLlyJfv27WPhwoVs376dLl26kJWVxaRJk1i8eDH79u3j119/ZfXq1cVyhGqD24fAGqqiYoiqBSQiIjXpvPPOo0mTJmzbto3rr7/euf2VV17h1ltvZciQIYSEhPC3v/2txlY+8PPz44cffuD++++nf//++Pn5cdVVVzmH3/z8/Ni6dSsffvghR48eJTw8nEmTJnHnnXeSn5/P0aNHufHGGzl8+DAhISFceeWVtV6g2GZZlSgy0ECkpqYSHBxMSkpKrdUE+nLtAR7+Yj3DOoTw8W0Da+UeIiJSednZ2ezZs4eoqKg6tZC2GOX9+VTm+1tDYG7STOuBiYiIuI0CIDcpKoaoHCAREalrZs2aRUBAQKmvbt26ubt5NUI5QG5SFAAdy8glr6AQT4diURERqRsuu+wyBg4sPT2jtis0u4oCIDdp7OeFh91GfqFFUnoO4cG+7m6SiIicoiGnyAYGBhIYGOjuZpSqpv5c1O3gJna7jZAA1QISEalrino4MjMz3dwSKU1R4UaHw1Gt66gHyI2aB3mTkJqtRGgRkTrE4XDQqFEjEhMTATOFu6wqx+JahYWFHDlyBD8/Pzw8qhfCKAByIyVCi4jUTWFhYQDOIEjqDrvdTuvWrasdlCoAcqNmgSqGKCJSF9lsNsLDw2nevDl5eXnubo6cwsvLC7u9+hk8CoDcqFmgKeCkITARkbrJ4XBUO9dE6iYlQbuRhsBERETcQwGQGzVTACQiIuIWCoDcqKgHKEkBkIiIiEspAHKj5kEnc4AacsEtERERV1MA5EYhAV4A5BYUkpypWQYiIiKuogDIjbw9HDTyMxVHlQckIiLiOgqA3KwoD0hT4UVERFxHAZCbqRiiiIiI6ykAcrPmJ4ohaghMRETEdRQAuZmGwERERFxPAZCbqRiiiIiI6ykAcjNnAJSqHCARERFXUQDkZkU5QEfS1QMkIiLiKgqA3Kx50IkcoFQFQCIiIq6iAMjNiobA0nLyycotcHNrREREGgYFQG4W6O2Bj6f5Y9BMMBEREddQAORmNpvtlFpASoQWERFxBQVAdYCmwouIiLiWAqA6oLmmwouIiLiUAqA6wFkNWlPhRUREXEIBUB1wshiiAiARERFXUABUB2hBVBEREddSAFQHNAvSgqgiIiKupACoDmiuWWAiIiIupQCoDijKATqakUN+QaGbWyMiInLmUwBUBzT198ZuA8uCYxm57m6OiIjIGU8BUB3gsNsICdAwmIiIiKsoAKojTlaDVjFEERGR2qYAqI5wFkNUD5CIiEitUwBURzhrAakYooiISK1TAFRHaEFUERER11EA5Er5uRC/HvYuL7GreZBygERERFxFAZAr7fwR3h4OC/5aYpdygERERFxHAZArtehtfh7ZCnlZxXZpCExERMR1FAC5UmA4+DcDqwASNhbbdeqCqJZluaN1IiIiDYYCIFey2SC8t3kfH1tsV1EPUG5+IanZ+a5tl4iISAOjAMjViobBDsUW2+zj6SDIxwOAI0qEFhERqVUKgFzN2QO0vsQuZx6QagGJiIjUKgVAruZMhN4CecV7eorygI6kKwASERGpTQqAXC2oJfiFQGE+HN5UbJezFpB6gERERGqVAiBXs9kgvJd5Hx9TbFezABVDFBERcQUFQO5QRiJ0UQ+QiiGKiIjULgVA7lDGVPhTawGJiIhI7VEA5A5FPUCJWyD/ZLCjatAiIiKuoQDIHYIjwLdJiUTosGDTA3TgeCZ5BYXuap2IiMgZTwGQOxRLhI51bo5q6k8jP0+y8wrZcDDFPW0TERFpABQAuUspidB2u43+bZoA8NvuY65vk4iISAOhAMhdykiEHhh1IgDac9S17REREWlAFAC5S1EP0OHNkJ/r3DyobVMA1uw9Tr7ygERERGqFAiB3aRQJPo2gMA8SNzs3dwkPItDHg/ScfDbHp7qvfSIiImcwtwdA06dPJyoqCh8fH6Kjo1m2bFm5x+fk5PD4448TGRmJt7c37dq144MPPih2zJw5c+jatSve3t507dqVefPm1eZHqJoyEqEdp+QB/b5HeUAiIiK1wa0B0OzZs5k8eTKPP/44MTExDBs2jFGjRhEXF1fmOWPHjuWnn37i/fffZ9u2bXz22Wd07tzZuX/lypWMGzeOCRMmsH79eiZMmMDYsWP57bffXPGRKqeMitBFeUCrlAgtIiJSK2yWZVnuuvnAgQPp27cvb775pnNbly5dGDNmDFOnTi1x/Pfff8+1117L7t27adKkSanXHDduHKmpqXz33XfObRdddBGNGzfms88+q1C7UlNTCQ4OJiUlhaCgoEp+qkrYOBe+vAVa9IE7Fjs3x+5PZswbvxLs60nMExdgt9tqrw0iIiJniMp8f7utByg3N5e1a9cycuTIYttHjhzJihUrSj3n66+/pl+/frz44ou0bNmSjh078vDDD5OVleU8ZuXKlSWueeGFF5Z5TTDDaqmpqcVeLuFMhN4EBXnOzd1bBOHv5SAlK4+tCWmuaYuIiEgD4rYAKCkpiYKCAkJDQ4ttDw0NJSEhodRzdu/ezfLly9m4cSPz5s1j2rRpfPnll9xzzz3OYxISEip1TYCpU6cSHBzsfEVERFTjk1VC4yjwDoaCXLMsxgkeDjvRbTQdXkREpLa4PQnaZis+vGNZVoltRQoLC7HZbMyaNYsBAwZw8cUX8/LLLzNz5sxivUCVuSbAlClTSElJcb72799fjU9UCTYbhPc078uqB6Q8IBERkRrntgAoJCQEh8NRomcmMTGxRA9OkfDwcFq2bElwcLBzW5cuXbAsiwMHDgAQFhZWqWsCeHt7ExQUVOzlMmUkQg9qe2Im2N5juDFNS0RE5IzktgDIy8uL6OhoFi1aVGz7okWLGDJkSKnnDB06lEOHDpGenu7ctn37dux2O61atQJg8ODBJa65cOHCMq/pdmVUhO7RshE+nnaOZeSyIzG9xGkiIiJSdW4dAnvwwQd57733+OCDD9iyZQsPPPAAcXFxTJw4ETBDUzfeeKPz+Ouvv56mTZtyyy23sHnzZpYuXcojjzzCrbfeiq+vLwD3338/Cxcu5IUXXmDr1q288MIL/Pjjj0yePNkdH/H0WvQxPxM2FkuE9vKw07d1YwB+2608IBERkZrk1gBo3LhxTJs2jWeeeYbevXuzdOlSFixYQGRkJADx8fHFagIFBASwaNEikpOT6devH+PHj+fSSy/ltddecx4zZMgQPv/8c2bMmEHPnj2ZOXMms2fPZuDAgS7/fBXSOAq8g6AgB45sK7ZrYJRZFmOVCiKKiIjUKLfWAaqrXFYHqMiMS2Dfcrj8Dehzg3Pzqt1HufadVYQEeLP68RHlJnKLiIg0dPWiDpCcooxE6N4RjfDysJOUnsPupAyXN0tERORMpQCoLigjEdrH00HviEaApsOLiIjUJAVAdUFRD1DCRijIL7ZrUJQKIoqIiNQ0BUB1QZN24BUA+VmQtL3YroFtTSL0b7tVD0hERKSmKACqC+x2CCu9InTf1o3xdNhISM1m/7GskueKiIhIpSkAqivKSIT29XLQs1UjAFZpGExERKRGKACqK8pIhAatCyYiIlLTFADVFc5E6A1QWFBslzMPSD1AIiIiNUIBUF3RtD14+kNeZolE6OjIxjjsNg4cz+JgsvKAREREqksBUF1hd0BYD/M+fn2xXQHeHnRvGQxoXTAREZGaoACoLikjERqUByQiIlKTFADVJRVJhFYekIiISLUpAKpLWvQxPw/FQn5OsV392jTBZoO9RzM5nJrt+raJiIicQRQA1SXNOkFAqKkIHbeq2K5gX0+6hpuVbVcpD0hERKRaFADVJTYbtD3XvN/9S4ndA6OKpsMrD0hERKQ6FADVNe1OBEC7fi6xa2DbokRo9QCJiIhUhwKguqbtOeZn/HrISCq2a2BUExx2G7uOZLAnKcP1bRMRETlDKACqawLDILS7eb97cbFdjfy8GNLODIMt2BDv4oaJiIicORQA1UVFvUC7SuYBXdIjHID5fygAEhERqSoFQHVRu/PMz92/gGUV23VhtzAcdhub41PZfSTdDY0TERGp/xQA1UWRQ8DhDakHS6wL1tjfi6HtQwANg4mIiFSVAqC6yNMXIgeb96XMBht9YhjsWw2DiYiIVIkCoLqqaBislDygkd1C8bDb2JqQxs5EDYOJiIhUlgKguqqoIOLe5ZCfW2xXIz8vzuqgYTAREZGqUgBUV4V2B/9mkJcBB34vsVuzwURERKpOAVBdZbefMh2+ZB7QyK5heDpsbDucxo7Daa5tm4iISD2nAKguc+YBlQyAgv08GdahGQDzNQwmIiJSKQqA6rKiPKBDsZBZcgFUDYOJiIhUjQKguiwoHJp1AawSy2IAnN81FC+HnR2J6WzXMJiIiEiFKQCq606tCv0nwb6eDO9oZoOpJpCIiEjFKQCq69qdGAbbVXJZDIBLehYNgx3CKmW/iIiIlKQAqK6LHAIOL0jZD0d3lth9fpdQvDzs7DqSwTYNg4mIiFSIAqC6zssfWg8y70upCh3o48nZHU/MBtMwmIiISIUoAKoPimaDlTIdHmB0z5OzwTQMJiIicnoKgOqDokTovcugIK/E7hEnhsF2J2WwJV7DYCIiIqejAKg+COsJfk0hNx0OrC6xO8Dbg3M7FRVFPOTq1omIiNQ7CoDqg9MsiwFwSc8WgIbBREREKkIBUH3hXBajZCI0wIjOzfH2sLP3aCab41Nd2DAREZH6RwFQfeFcFmMdZB0vsdvf24PzOjcHNBtMRETkdBQA1RfBLSGkE1iFsGdpqYcUFUX8VsNgIiIi5VIAVJ+0K386/HmdmxPg7UHcsUwWbz/iwoaJiIjULwqA6hNnHlDpAZCflwfX9o8A4L1lu13VKhERkXpHAVB9EjkU7B6QHAfH95V6yC1nReGw2/h151E2HkxxcQNFRETqBwVA9Yl3AIT3Nu/3/VrqIS0b+TorQ6sXSEREpHQKgOqbNkPNz72lB0AAfxnWFoBv/ojnUHKWK1olIiJSrygAqm8izzI/9y0v85DuLYMZ3LYpBYUWM1fsdU27RERE6hEFQPVN60Fgs8PxvZBysMzD7hhueoE+/S2O1OyS64eJiIg0ZAqA6hufILM2GJSZBwRwdsdmdGgeQHpOPrN/3++ixomIiNQPCoDqozYnhsH2lj0MZrfbnLlAH/y6h7yCQle0TEREpF5QAFQfFQVA5fQAAVzepwUhAd7Ep2RreQwREZFTKACqj1oPBmxwdCekJZR5mLeHg5uHRALw7rLdWh5DRETkBAVA9ZFvIwjrbt6fphdo/MBIfD0dbDqUyspdR2u/bSIiIvWAAqD6KvL0eUAAjf29GNuvFQDvqDCiiIgIoACo/qpAQcQit54Vhd0Gi7cdYVtCWi03TEREpO6rdACUl5fHueeey/bt22ujPVJRrYeYn0nbIL38ld8jm/pzUfcwQMtjiIiIQBUCIE9PTzZu3IjNZquN9khF+TeF5l3N+9PkAcHJ5TH+F3uIxNTs2myZiIhInVelIbAbb7yR999/v6bbIpVVwenwAH1aN6ZfZGNyCwr5cOXe2m2XiIhIHedRlZNyc3N57733WLRoEf369cPf37/Y/pdffrlGGienETkUfn+nQnlAAH8Z3pY1H6/lk1VxTDy7HYE+nrXcQBERkbqpSgHQxo0b6du3L0CJXCANjblQ5IlE6MRNkHkM/JqUe/j5XUJp18yfXUcymPHrXu4b0cEFjRQREal7qhQA/fLLLzXdDqmKgGYQ0skkQu9bAV1Gl3u4w25j8vkdufezGN5dtpubBrch2E+9QCIi0vBUexr8gQMHOHiw7FXJpZYVTYevQB4QwCU9wukcFkhadj7vakaYiIg0UFUKgAoLC3nmmWcIDg4mMjKS1q1b06hRI5599lkKC7XopksVDYOdpiBiEbvdxgMXdARgxq97OJqeU1stExERqbOqFAA9/vjjvP766zz//PPExMSwbt06/vnPf/Kf//yHJ554oqbbKOUpmgmWsAGyjlfolJFdQ+nRMpiM3ALeXqpeIBERaXiqFAB9+OGHvPfee9x111307NmTXr16cffdd/Puu+8yc+bMSl1r+vTpREVF4ePjQ3R0NMuWLSvz2MWLF2Oz2Uq8tm7d6jxm5syZpR6TnX2G1r4JDIMm7QAL4lZV6BSbzcaDJ3qBPlq5V3WBRESkwalSAHTs2DE6d+5cYnvnzp05duxYha8ze/ZsJk+ezOOPP05MTAzDhg1j1KhRxMXFlXvetm3biI+Pd746dCg+mykoKKjY/vj4eHx8fCrcrnqnTeWGwQDO6dSMvq0bkZ1XyPTFu2qpYSIiInVTlQKgXr168frrr5fY/vrrr9OrV68KX+fll1/mtttu4/bbb6dLly5MmzaNiIgI3nzzzXLPa968OWFhYc6Xw+Eott9msxXbHxYWVuE21UtthpmfFUyEBvOMHhrZCYBPf4vjUHJWbbRMRESkTqpSAPTiiy/ywQcf0LVrV2cA07VrV2bOnMm//vWvCl0jNzeXtWvXMnLkyGLbR44cyYoVK8o9t0+fPoSHhzNixIhSp+Snp6cTGRlJq1atGD16NDExMeVeLycnh9TU1GKveqUoETp+PWRXvO1D2jVlYFQTcgsKef2XnbXUOBERkbqnSgHQ2Wefzfbt27niiitITk7m2LFjXHnllWzbto1hw4ZV6BpJSUkUFBQQGhpabHtoaCgJCQmlnhMeHs4777zDnDlzmDt3Lp06dWLEiBEsXbrUeUznzp2ZOXMmX3/9NZ999hk+Pj4MHTqUHTt2lNmWqVOnEhwc7HxFRERU6DPUGcEtoXEbsAph/28VPu3UXqD/rt5P3NHMWmqgiIhI3WKzLMuqzAl5eXmMHDmSt99+m44dO1b5xocOHaJly5asWLGCwYMHO7f/4x//4OOPPy6W2FyeSy+9FJvNxtdff13q/sLCQvr27cvw4cN57bXXSj0mJyeHnJyT08FTU1OJiIggJSWFoKCgSnwqN/rqHoj9BIZOhguertSpE97/jWU7krg6uhX/vqbiQ5giIiJ1SWpqKsHBwRX6/nbbavAhISE4HI4SvT2JiYkleoXKM2jQoHJ7d+x2O/379y/3GG9vb4KCgoq96p1KFkQ8VVEv0Nx1B9h1JL0mWyUiIlInuW01eC8vL6Kjo1m0aFGx7YsWLWLIkCEVvk5MTAzh4eFl7rcsi9jY2HKPOSMU5QEdioGcygUxvSMacX6X5hRa8OqPZQeKIiIiZwq3rgb/4IMPMmHCBPr168fgwYN55513iIuLY+LEiQBMmTKFgwcP8tFHHwEwbdo02rRpQ7du3cjNzeWTTz5hzpw5zJkzx3nNp59+mkGDBtGhQwdSU1N57bXXiI2N5Y033qjKR60/GkdCcASk7Dd5QO1HVOr0By7oyI9bEvnmj0Pcc257OoUF1lJDRURE3M+tq8GPGzeOo0eP8swzzxAfH0/37t1ZsGABkZGRAMTHxxerCZSbm8vDDz/MwYMH8fX1pVu3bsyfP5+LL77YeUxycjJ33HEHCQkJBAcH06dPH5YuXcqAAQOq8lHrl8ih8MfnZhiskgFQtxbBXNwjjAUbEnhl0XbemhBdS40UERFxv0onQRcUFLB8+XJ69OhBkyZNaqtdblWZJKo6Zd3H8PUkiBgEt/1Q6dO3H07jwmlLsSz4YuJg+rc5M/98RUTkzFSrSdAOh4MLL7yQlJSUKjdQaklRIvTBtZCVXOnTO4YGMq6fKQHw93kbySvQwrYiInJmqlISdI8ePdi9W4to1jmNo6B5VyjMg41zTn98Kf52UWca+3my7XAaH67YW7PtExERqSOqFAD94x//4OGHH+bbb78lPj6+fldRPpPYbNB7vHkf+2mVLtHY34tHR5l13l5ZtJ34FC2RISIiZ55K5wCBqa3jvMApSc+WZWGz2SgoKKiZ1rlJvc0BAkhPhJc6g1UA9/wOzTpV+hKFhRZXv7WCdXHJXNIjnDfG962FhoqIiNSsynx/V2kWWGnrb0kdEdAcOoyE7d9B7Cy44JlKX8Jut/HcmB6M/s8y5m+IZ+z2I5zdsVktNFZERMQ9qrwWmN1u59133+XRRx+lffv2nH322cTFxZVYmV3coM+JYbD1s6Egv0qX6NoiiJuHRAHw5P82kp1Xv3v1RERETlWlAGjOnDlceOGF+Pr6EhMT41xHKy0tjX/+85812kCpgg4Xgl9TSE+A3VXvrXvggg40D/Rm79FM3lqyqwYbKCIi4l5VCoCee+453nrrLd599108PT2d24cMGcK6detqrHFSRR5e0GOseR/zSZUvE+jjyROjuwIwffEu9iZl1ETrRERE3K5KAdC2bdsYPnx4ie1BQUEkJydXt01SE3pfb35uWwCZx6p8mdE9wzmrfQi5+YX839ebqELOvIiISJ1TpQAoPDycnTt3lti+fPly2rZtW+1GSQ0I7wmhPaAgt8o1gcDM8nvm8m54Oews3X6E7zcm1GAjRURE3KNKAdCdd97J/fffz2+//YbNZuPQoUPMmjWLhx9+mLvvvrum2yhVVZQMHTurWpdp2yyAO882ge3T32wmPadqidUiIiJ1RZUCoL/+9a+MGTOGc889l/T0dIYPH87tt9/OnXfeyaRJk2q6jVJVPa4BuwccioHDm6t1qXvObU9EE18SUrN57acdNdRAERER96hSAASmGnRSUhK///47q1at4siRIzz77LM12TapLv8Q6HiReb++apWhi/h4Onjq0m4AvL98DxsPai04ERGpv6ocAAH4+fnRr18/BgwYQEBAQE21SWpS71NrAuVV61IjuoRycY8wCgotHvxvrGoDiYhIvVWtAEjqgQ4XgH8zyEiEnT9V+3LPXt6dkAAvth9O55VF22uggSIiIq6nAOhM5/CEnuPM+9iq1wQq0jTAm39e0QOAd5btZvXeqk+xFxERcRcFQA2BsybQ95BxtNqXG9ktjKv6tsKy4KH/ridDs8JERKSeUQDUEIR2g/DeUJgHG7+skUs+eVlXWgT7EHcsk6nfbamRa4qIiLiKAqCGoigZuhpLY5wqyMeTf13TC4BPVsWxdPuRGrmuiIiIKygAaih6XA0OL0j4AxI21Mglh7YP4abBkQD89cs/SMms3iwzERERV1EA1FD4NYFOo8z72M9q7LKPjupCVIg/CanZPPXNphq7roiISG1SANSQFA2D/TEb8nNr5JK+Xg5eGtsLuw3mxRzk+43xNXJdERGR2qQAqCFpNwICwiAzCXYsrLHL9m3dmIlntwPgsXkbOZKWU2PXFhERqQ0KgBoShwf0KqoJVL0FUv/s/vM70DkskGMZuTw2bwOWZdXo9UVERGqSAqCGpvcN5uf2HyDtcI1d1tvDwSvjeuPpsLFo82HmrjtYY9cWERGpaQqAGppmHaHVALAKTC5QDeoSHsTk8zsC8NQ3m4hPyarR64uIiNQUBUANUZ8TydCxs6CGh6ruHN6WXhGNSMvO529zNBQmIiJ1kwKghqjbleDhC0e2wsG1NXppD4edl67piZeHnaXbj/D56v01en0REZGaoACoIfIJgq6Xmfc1VBn6VO2bB/LXCzsB8Ny3m9l/LLPG7yEiIlIdCoAaqj4nkqE3zoHcmg9QbhkaRf82jcnILeCvX/5BYaGGwkREpO5QANRQRZ4FjVpDTips/bbGL++w2/j3Nb3w9XSwcvdRPlq5t8bvISIiUlUKgBoqu73GF0j9s8im/ky5uDMAz3+/lT1JGbVyHxERkcpSANSQ9brO/NyzFI7vq5Vb3DAwkiHtmpKdV8jDX6ynQENhIiJSBygAasgaR0LU2YAF62tugdRT2e02Xry6JwHeHqzdd5z3l++ulfuIiIhUhgKghq4oGTp2FhQW1sotWjX244nRXQD498Lt7DicViv3ERERqSgFQA1d59HgHQTJcbBvea3dZmy/CM7t1Izc/EIe+mI9ufm1E2yJiIhUhAKghs7LD7pfZd7XUjI0gM1m4/mrehLk48EfB1K4deZq0rLzau1+IiIi5VEAJCeHwTZ/DdkptXab0CAfpo+Pxs/LwfKdSYx7exWJqdm1dj8REZGyKAASaBkNIZ0gPws2zavVW53VIYTZdwwmJMCLzfGpXDF9BTsT02v1niIiIn+mAEjAZju5QGotDoMV6dEqmLl3DSUqxJ+DyVlc/dYK1u47Vuv3FRERKaIASIye14LNAQdWw5FttX671k39+HLiYHpHNCI5M4/r3/2NHzYl1Pp9RUREQAGQFAkMhQ4jzfvYWS65ZdMAbz77yyBGdG5OTn4hd32ylo9X1U5BRhERkVMpAJKTiobB1n8O+TkuuaWvl4O3J0Rz3YAICi144quN/OuHrViWKkaLiEjtUQAkJ3W4EAJCIf0w/PIPl93Ww2Hnn1f04IHzOwLwxi+7eHupKkaLiEjtUQAkJ3l4wSUvm/e/vgZ7a68w4p/ZbDbuP78DT17aFYAXv9/KrzuTXHZ/ERFpWBQASXFdRp+oC2TBvIm1WheoNDcPacM10a0otODez2I4mJzl0vuLiEjDoABISrroeWjcBlL2w3d/c+mtbTYbz47pTveWQRzLyOXuT9aSnVfg0jaIiMiZTwGQlOQdCFe8DTa7WSV+01cuvb2Pp4M3x0fTyM+T9QdSePqbTS69v4iInPkUAEnpWg+Csx4w77+dDKnxLr19RBM/Xru2DzYbfPb7fmavjnPp/UVE5MymAEjKdvajEN4Lso7D/+4BF09NH96xGQ9dYGaGPfG/TfxxINml9xcRkTOXAiApm4cXXPkuePjArp/g93dd3oS7z2nP+V1Cyc0v5K5P1nEsI9flbRARkTOPAiApX7NOcMEz5v2iJ1yyTMap7HYbL4/r5Vw37L7PYigoVJFEERGpHgVAcnr9/wLtzoP8bJj7F8h3bS9MkI8nb90Qja+ng+U7k3hpoWuDMBEROfMoAJLTs9vh8ung0wji18OSF1zehE5hgbxwdU8Api/exae/KSlaRESqTgGQVExQOFw6zbxf/jIk7XR5Ey7r1YI7h7cF4LF5G5i+eKfWDBMRkSpRACQV1+0KaH8+WIUQ87FbmvDoqM7cc247AF78fhv/XLBFQZCIiFSaAiCpnL43mZ/rP4OCfJff3maz8ciFnfn7JV0AeHfZHh758g/yCwpd3hYREam/FABJ5XS8CPyamhXjd/7otmbcPqwt/76mFw67jS/XHuCuWeu0ZIaIiFSYAiCpHA8v6Hmtee+mYbAiV0e34q0bovHysLNo82FunvE7adl5bm2TiIjUDwqApPL63GB+bv8e0o+4tSkXdA3lo1sHEODtwardx7ju3VUkpee4tU0iIlL3KQCSygvtCi36QmE+/DHb3a1hUNumfH7HIJr6e7HxYCpj31pJ3NFMdzdLRETqMAVAUjV9J5ifMR+7fI2w0nRvGcwXEwfTspEvu5MyuPT15SzZ7t7eKRERqbsUAEnVdL/KrBF2ZCscXOvu1gDQtlkAc+4aQq+IRqRk5XHzjN954xfVChIRkZLcHgBNnz6dqKgofHx8iI6OZtmyZWUeu3jxYmw2W4nX1q1bix03Z84cunbtire3N127dmXevHm1/TEaHp9g6Hq5ee/mZOhThQX78N87B3Ft/wgsC/71wzbunrWO9BzXT9kXEZG6y60B0OzZs5k8eTKPP/44MTExDBs2jFGjRhEXV/4yB9u2bSM+Pt756tChg3PfypUrGTduHBMmTGD9+vVMmDCBsWPH8ttvv9X2x2l4ipKhN8yB3LqTc+Pt4eD5q3ryzyt64Omw8d3GBK5441d2H0l3d9NERKSOsFluHB8YOHAgffv25c0333Ru69KlC2PGjGHq1Kkljl+8eDHnnnsux48fp1GjRqVec9y4caSmpvLdd985t1100UU0btyYzz77rNRzcnJyyMk5OXMoNTWViIgIUlJSCAoKquKnawAKC+G13pC8D654G3pd6+4WlbB233HunrWWw6k5BHp7MO3a3ozoEuruZomISC1ITU0lODi4Qt/fbusBys3NZe3atYwcObLY9pEjR7JixYpyz+3Tpw/h4eGMGDGCX375pdi+lStXlrjmhRdeWO41p06dSnBwsPMVERFRyU/TQNntJ3uBYj5xb1vKEB3ZmG/uPYv+bRqTlpPPbR+u4ZVF2yksVF6QiEhD5rYAKCkpiYKCAkJDi/9rPDQ0lISEhFLPCQ8P55133mHOnDnMnTuXTp06MWLECJYuXeo8JiEhoVLXBJgyZQopKSnO1/79+6vxyRqY3tcDNti7DI7tdndrStU80IdZtw/ixsGRALz60w7u+XQdWbmqHC0i0lB5uLsBNput2O+WZZXYVqRTp0506tTJ+fvgwYPZv38///73vxk+fHiVrgng7e2Nt7d3VZovwa2g3Xmw6yeImQUjnnB3i0rl5WHnmcu706NlMI/P28h3GxM4lLySd2/qR/NAH3c3T0REXMxtPUAhISE4HI4SPTOJiYklenDKM2jQIHbs2OH8PSwsrNrXlEoqGgaL/RQK63avyjX9Ivjk9oE09vNk/YEUrnhjBVsTUt3dLBERcTG3BUBeXl5ER0ezaNGiYtsXLVrEkCFDKnydmJgYwsPDnb8PHjy4xDUXLlxYqWtKJXW+BHwbQ9oh2PXL6Y93swFRTZh391DahvhzMDmLq99cyeJtie5uloiIuJBbp8E/+OCDvPfee3zwwQds2bKFBx54gLi4OCZOnAiY3Jwbb7zRefy0adP46quv2LFjB5s2bWLKlCnMmTOHSZMmOY+5//77WbhwIS+88AJbt27lhRde4Mcff2Ty5Mmu/ngNh4c39Bhr3tehmkDlaRPiz9y7hzCobRPSc/K5deZqPl65193NEhERF3FrDtC4ceM4evQozzzzDPHx8XTv3p0FCxYQGWmSVePj44vVBMrNzeXhhx/m4MGD+Pr60q1bN+bPn8/FF1/sPGbIkCF8/vnn/P3vf+eJJ56gXbt2zJ49m4EDB7r88zUofSfA72/D1vmQcRT8m7q7RafVyM+Lj24dyGPzNvDl2gM88b9N7EnK5PFLuuCwl50zJiIi9Z9b6wDVVZWpIyCneHs4xK+Hi16AQRPd3ZoKsyyL6Yt38a8ftgEwonNzXhrbi0Z+Xm5umYiIVEZlvr/dPgtMziB9JpgAaO0MCGoBeVmQn2V+Fr3ys0y+0OB7waNuBBg2m417zm1Pm6b+PPjfWH7amsiQ53/m+gGtuW1YFOHBvu5uooiI1DD1AJVCPUBVlHkMXuoMBTmnP3bQ3XBRyWrf7ha7P5nH5m5gc7yZGebpsHFFn5bcMbwd7ZsHuLl1IiJSnsp8fysAKoUCoGpY9Rb88Tk4vMHT9+TL48TPwnzTQwRw7admBlkdY1kWS3ck8ebinazafQwAmw0u7BrGfX3sdI39B5zzKLTq5+aWiojIqRQAVZMCoFr2/WOw6g2zovzE5dCotbtbVKZ1ccd5a/EuFm4+DMDTHjO4yWMRGc2j8b/7Zze3TkRETlUv1gKTBuz8p6BlNGSnwJe3QkGeu1tUpr6tG/POjf1Y9MBwro5uRV/7TgD8E9fyxuxvSM/Jd3MLRUSkKhQAiet5eMHVH4B3MBxYDT894+4WnVaH0ED+fXkHunucXCfOe8MsLnh5CT+e6B0SEZH6QwGQuEfjNjDmDfN+xWuw/Qe3NqdC4tdjK8wHTI2gqz2WczQljds/WsPds9aSmJrt3vaJiEiFKQAS9+lyKQy407yfNxFSDrq3PadzcI352fEiCGpJI9J4sft+HHYbCzYkMOKlJXyyah+FhUqrExGp6xQAiXuNfBbCe0PWMZhzGxTU4ZyaA6vNz9YDnQvAjin8kW8mnUWvVsGk5eTz9682cs3bK/l1ZxIFCoREROosBUDiXh7ecM0M8A6CuJXwyz/c3aKyHTjRA9SqP/QeD9hg92K6+hxl7t1DefLSrvh7OVi77zjj3/uNoc//zNQFW7TavIhIHaQASNyvSVu47DXzfvnLsPNH97anNCkHIfUg2OzQog80joR255p9MZ/gsNu4ZWgUix48m+sHtibIx4OE1GzeXrqbi6YtY9Sry3h36W4OK09IRKROUB2gUqgOkJt8+yCseR/8QuAvP5sgo67Y/D/4740Q1sPULgLY9BV8cRMEhsPkjeA4ubJMTn4Bv2xNZO66g/yyLZG8AvOfmd0GQ9uHMH5gay7oGqZFV0VEapDWApP66cJ/woHfIWEDzLwEbvwfNG3n7lYZRfk/rfqf3NbpYvBrCmnxsHMRdBrl3OXt4eCi7uFc1D2c4xm5zN8Qz7yYg6zdd5xlO5JYtiOJqBB/bh8WxVV9W+Hj6XDxBxIRadg0BCZ1h6cPXP8FNO0AKftNEHRku7tbZZya/1PEwwt6XWfer/uozFMb+3txw6BI5tw1hCWPnMPd57QjyMeDPUkZPD5vI2e98DOv/7yD5MzcWvwAIiJyKg2BlUJDYG6WnggfXgZHtoB/c9MTFNrVfe0pyIOprSA/GyatgZAOJ/cd2QZvDACbAx7YBEHhFbpkRk4+s1fv5/3leziYnAWAn5eDcf0juO2sKFo19quNTyIickbTUhhSvwU0h5u/hdAekJEIH442w2LucnijCX58GkGTPw3JNesErQeDVQCxsyp8SX9vD249K4rFj5zDtHG96RIeRGZuATN+3cvZ/1rMXZ+sZfG2RE2lFxGpJQqApG7yD4GbvjYzrjKPwszRcHCde9pSNPzVMhrspfwn0/dG8zPmYygsrNSlPR12xvRpyYL7zuKjWwdwVvsQCgotvtuYwM0zVjPshZ95ZdF2DhzPrOaHEBGRUykAkrrLr4kZ/mrVH7KT4aPLYf9q17ejtAToU3W93NQxOr4X9i6t0i1sNhvDOzbjk9sH8v3kYdw8pA3Bvp4cSsnm1Z92MOzFX5jw/m/M/yOenPyCqn0OERFxUgAkdZtPMEyYB62HQE4qfDwG9q1wbRtKS4A+lZc/9LjGvC8nGbqiOocF8dRl3fjtsRG8em1vhrZvimXBsh1J3PPpOgZP/ZlnvtnMpkMp1b6XiEhDpSToUigJug7KzYDProU9S8HTD278GiLKCEhqUuYxeDHKvP/rHtMrVZpDsfDO2eDwgge3gn/TGm1G3NFM/rtmP1+s3c/h1Bzn9s5hgVwd3YrLe7ekWaB3jd5TRKS+qcz3twKgUigAqqPysuDz8bDrJwgIgzsWV3jWVZVtXwifXmOm5t+7pvxj3xoGCX/AhVNh8N210pz8gkKWbD/CnHUH+HFzIrkFJufIYbdxTsdmXBXdihFdmuPtobpCItLwqBCinJk8fWHsh/D+SEjcDLPHw80LTP2g2nK6/J9TRd8E8x8yw2CD7gJbzVd59nDYGdEllBFdQknOzOWbP+L5cu0B1u9P5qetify0NZFgX0+6twwiwNuDAG9PAn08zHsfD+f7bi2CaN88sMbbJyJSX6gHqBTqAarjju2Bd8+FrOPQ63oYM71Wgg0APhoDu3+BS16G/reVf2xWMrzUGfKz4LZFEDGgdtpUip2J6cxZd4B56w6SUMH1xjqHBXJprxZc1qsFEU1Ud0hE6j8NgVWTAqB6YNcv8MmVYBXCRc+bHpeaVlgIL0Sa5Os7l0F4z9OfM28irP8Met8AY96o+TadRkGhxZq9x0hIzSYtO5/0nHzSsvNIz84nLSef9Ox8kjPziNl/3Lk+GUCf1o24tGcLRvcMp3lQLfaoiYjUIgVA1aQAqJ5YOR1+mGKqMN8w5+Tq7DUlcStMH2iSrh/dX2yx0zLFrYIPLgQPX3hoK/g2qtk21ZDkzFx+2JTA1+sPsXLXUYrqLdptMKhtU6IjG+Pj6cDH04GvpwNfLzs+Hg58vBz4eDgIDfImsqm/FnMVkTpFOUDSMAy6y1SIXv8pfHkL/OUXaBJVc9cvyv9p0bdiwQ9AxEBo1sUs4/HHbBh4Z821pwY18vNiXP/WjOvfmsS0bOb/Ec/X6w8RE5fMil1HWbHr6Gmv4evpoHN4IF3Dg+jaIogu4UF0DgvEz0v/WxGRuk//p5L6y2aD0a/Aka1waB18fr3JvfEOqJnrOxOg+1WuTf1uhe8egTUzYMAdtZefVEOaB/pwy9Aobhkaxf5jmSzYEM/B5Cyy8wrIyiskO6/A+crKKyArt4CDyVlk5RUQE5dMTFyy81o2G0Q19adXRCOGdQhhWIdmmp4vInWShsBKoSGweib1ELxzDqQfhs6jYezHpS9ZUVnTh0DiJhg3C7qMrvh52SkmGTovE275HiIHV78tdUxBocWepAw2x6ey+VAqW+JT2RyfypG0nBLHdg0PYnjHZgzvGEK/yCZ4eaj+qojUDuUAVZMCoHpo/+8w8xIoyIVzHoNz/la96+WkwfOtTZL1Q9sgMKxy5//vHoj5BHqMhaverV5b6pEjaTlsjk9l1e6jLN1+hE2HUovt9/NyMLhtU4a2D6Ffm8Z0DQ/Cw6GASERqhgKgalIAVE+t+xi+nmTee/qZJSq8/MEr4JT3/hDYAs6aDEEtyr7WnqXw4aUQ3BoeqMJK9AfXwrvngcMbHtxS45Wh64uk9ByW70hi6fYjLN2RRFJ68R4iX08HvSMa0a9NY/pGNqZv68YE+3q6qbUiUt8pCVoapr4T4NguWD7NDD/lZULGkdKP3bsMblsI3mUUA6xK/s+pWvSF8F4Qv94kaQ+5t2rXqedCArwZ06clY/q0pLDQYktCKku3J/H7nqOs3Xec1Ox8Vu4+ysrdJunaZoOOzQOJCvF35htl5uWTmVNAZm4Bmbn5ZOUVEOjjyeW9W3Bt/9Z0ClNBRxGpPPUAlUI9QPVc1nGTh5ObceKVfvJ9Thos/ZfJF+o4Cq6dBfZSlo347DrYtqB6y1qsmQHfToYm7eDetXU+GdrVCgstdh5JZ83e46zZd4x1+46z92hmpa/Tp3Ujru0fweieLfD31r/pRBoyDYFVkwKgM9yBNTDjYijIgaH3wwXPFN9vWfDvDqb36LYfq77oak4avNQFctPM4q1tz65+289wR9JyWLvvOEfSsvH18sDPy4GvlwM/Twd+Xh7mvZeDbYfTmP37fn7ccpj8E0WM/L0cXNa7BeP6t6ZXq2BsCjhFGhwFQNWkAKgB2PAlzDmxtMWYN6H39Sf3Hd8Lr/YyK7tPOQAe1ZjG/e2DsOZ96DrGrGMmNSoxLZu56w4ye/V+9iRlOLe3auxLSIA3gaesfxbg7UmAjwdBPh408feiV0Qj2ob4K1ASOYMoB0jkdHpcbeoHLf0XfHM/NGkLrQeZfQdOrPoe1rN6wQ9Av1tMALT1W0g7DIGh1bueFNM80IeJZ7fjzuFt+X3PMT5fvZ8FG+I5cDyLA8ezTnt+Iz9P+rZuTN/Wjegb2ZherRppGE2kgdB/6dJwnfMYHNkGW76Gz8fDX36GxpHVT4A+VVgPs5L8gdUQ+wkMe6j615QSbDYbA9s2ZWDbpjx1WTc2H0o1a6DlFK2Hln9ibbQ80rLzOXg8iw0HU0jOzOPnrYn8vDURMEuBdAkPok/rRnRvEUzXFkF0DA3Ex7OUPDERqdcUAEnDZbfDFW+ZIa+EP0zi820/nBIAVTH358/63WquuXYmDJ1cetK11JhgX08Gtzt92YHc/EI2x6eybt9x1sUdZ92+4xxKyWbTodRi9YscdhvtmwXQrYVZ8qNo6Y9Gfl61+TFEpJYpB6gUygFqYFIOwDvnQkYidLgQdv0MhXlw/3po3Kb618/Lgpc6mZlp47+EDhdU/5pSKxJSslkXd5zY/clsPpTKpkMpHM/MK/XYYF9PIpv60bqJH5FN/Yhs4k/rpuZ9aKAPGbn5HM/I43hmLscyc0nOzOVYRh7Jmblk5BQQFeJH1xZBdAoLIkDDbiI1QknQ1aQAqAE6dWYYgH8zeHhHzU1d/+5R+O1N6HQJXPdpzVxTap1lWSSkZrPpoFnqY9OhFDbHp7L/2OnziyqjTVM/uoSb3qUu4UF0CA0gv9AiJSuP1Kw8UrPzSc3KM79n55GZU0CbEH96tAymW4sg5S2JnKAAqJoUADVQf3wBc2837ztdDNd9VnPXPrIN3hgANjtM3gjBLWvu2uJyGTn5xB3LZN/RTOKOZZz4aV4HjmdRcGJqvo+nnSZ+XjTy86KJvxeN/b1o7OeJt4ednYnpbIlPIyE1u1ptsdmgfbMAerQMpkerYHq2CqZreDC+XjU71JqSmcfP2w6zaPNhVuw6SmRTf+4Y1paLuofhsGsmndQNCoCqSQFQA7b4BVg89cTU+Otq9tozLoZ9v8LZj8K5U2r22lJn5BcUciwjl0AfzwoFIUfTc9gSn+ZcUHZLfCq7j2Tg7WknyMeTYF9Pgnw9CPLxJMjX/O7tYWdHYjobDqSUGkDZbODlsONht2G32/Cw23DY7Tjs4GG34+Gw0SLYlw6hAXQIDaRD8wA6NA+gaUDxWY/7j2WyaLMJen7fe8wZ2J0qsqkffxnWlqujW9VMsnhhIXz/N1OlfcT/Vf960qAoAKomBUANXE46eAfU/HWLag8FtoDJG8ChYQupvsTUbDYcTDGvAyn8cTCFI2k5pz+xFE38vWjfPIDIJn5sOJjC1oS0Yvs7hQZyQddQhndsxq87k/hw5V6ST+RIhQR4ccvQKG4YGEmwXzXWc9v1C3w8xry/eQG0GVr1a0mDowComhQASa3Iz4GXu0DmUbj2U+h8ibtbJGeoo+k5ZOUVUFgI+YWFFBRaFFgW+QUWBYUWuQWFxB3NZHtiGjsPp7MjMZ39xzP587eB3Qb92zThgq6hjOwaRuumfsX2Z+bmM3v1ft5btoeDySYvyt/LwXUDWnNe5+Z4ezrw8bTj6+nAx/my4+PhIDu/gKPpuSSl53A0PZejGTkkpecyfMMUehxbCEB800EkXP45ncOCanxIT85MCoCqSQGQ1JqFT8CK18zsspsXKBdI6oys3AJ2HUlnZ2I6u5MyiGzix3mdm9PY//TT/fMKCpn/RzxvLdlVoteoMoJIZ7X3PXjb8ii0bNhtFlfkPM16OtCuWQBdWwTRrUUQ3VqY5O+qliJIz8knL7+wQp9N6hcFQNWkAEhqTXoivHc+JO8z1advng9BLdzdKpEaYVkWi7cf4aMVezmYnEVWXgHZeYVk5xWQnVdAXkHxrxtvDzshAd6EBHjRNMCb0TkLuDL+ZY76d+CgX2d6HvmG5ba+3JD1cKn3a9nIl24tgujeMpjuLU1g1DzQu9jyJscyctl0KIWNB80svk2HUp3LprRu4kff1o3o07oxfVo3okt4EJ4Oe7mfMT0nn4SUbHLyC4ho4keQTzWG+6TGKQCqJgVAUquS42DmJeZn0/Zw07cQFO7uVonUuoJCi+y8ArLyCvD1NAvbFluL7Z1z4FAMXDgVOl4Ir/cDq5Cj1//A+sKoE7WZzCvuWGap9wgJ8KZ7yyA87HY2H0rhUErFZ9l5e9jp2SqYPq0b0yLYh8S0HBJSszmcmk1CSjaHU3NIz8kvdk5jP09aN/UnsokfbZr6mfdN/QgJ8CYzN5+MnAJnBfKi9+nZ+fgWZnJ2z3Z0baHvmJqkAKiaFABJrTu+zwRBKfuhaQe4+VsIDHN3q0TcJ2EjvDUU7J7w0FbwD4G5d8Afs0utn5WSlecsVrnpUCobD6aw60g6pUxUIyrEn64tguh+YuisW4sgPBx21u9PJiYumZj9x4mJSyYlq/Sil38W4O2Bl4edYxm5lf6Y3Wx7ed7zHbrY4ngi/xZimo1hTJ+WXN67BeHBvpW+XlVZlnVGLgSsAKiaFACJSxzfCzMugdQDENLR9ARpsVRpqL6fAqumQ5fLYNzHZtuR7aZ+FhZM/BXCupd7iazcArYkpLLpYAr5hZZz2ZLACgxTWZbF7qQMYuKSWRd3nKPpOYQG+RAa5ENYkA9hwSfeB/s4K3enZeeZ+k9HM9n3p7pQxzNy8fP2INDbA39vDxp7FXBd5ixGpnyJgwLnfafk3cZnBSOw2WBw26aM6dOSUd3DnG22LItjGbnsPXHtvUmm3lRSeo7JizrxGds3DyizDEFhocWOxHTnki/r4o6z92gm/SIbc2XflozqEX7GDOUpAKomBUDiMsf2mJ6g1IMQ0sn0BAU0d3erRFwrP9csF5N1DK7/AjqOPLnvi1tg01zoOgbGfui2JlbLnqXw9X1wfI/5vduV4NcUVr8LwNtB9zE1cZDzcG8POwOimnAsI5d9RzNLDLuVpmjNuqL16loHOdiUmE3M/mRi45JJK+ca3h52LugaylV9WzGsQwgep8mDqgh39TApAKomBUDiUsd2m56gtEPQrLPpCQpodnJ/eiIcXAcH15rX4Y1mPbFLXgEPzWKRM8Dm/8F/b4TAcFMp/dQaWYc3wZtDABvc8xs06+S2ZlZa1nEz8zPmRI9WYAsY/TJ0GgWWBT88DqveAODYeS/yWcEI5q47wK4jGcUuY7NBeJAPrZv60aapWXOuiZ8Xu46kn1iiJdVZjwngIY//codjPk/m38TnBecB4OvpoFdEMH1bN6Zv68a0burHos2HmRdzkJ2J6c5zQwK8uKxXS67o05KOYQF4e5y+/EBRz9vGE7WoNhxMYWdiOo38PGnXLIAOoQG0bxZA++aBtG8eQGiQd60FRwqAqkkBkLjc0V2mJygtHpp1gV7XwqF1JvBJ2V/6Oe3Og7EfmYq5IvXZrGtgx0I460E4/8mS+z8fD1u/hR5j4ap3Xd++qtj8P1jwCKQfNr/3vx1GPAk+p3yn/CkIYvQ0rOib2XgwldgDyYQH+dAmxI9Wjf3KrbJdtGbd5kOpONa8xzm7XgQgx+bDgrPm0KFzDzqHBZbas2NZFhsOpjB33UG+Xn+oRF5TgLcHTfy9aBrgRVN/s6RLE39vAn082H0kg40HU9h5JL3UKuFlCfT2oHdIAR0iW/N/l5U/rFlZCoCqSQGQuMWpQVAxNvOv3pbR0LIvePrD/IcgLwNa9DFDBqf2GInUJ6mH4JVuYBXCveugabuSxxyKhXfONmvpTVpT+jF1RU46fHMfbJxjfg/pCJe+BpGDSz++lCCIfrdU7d5bF8Ds8eZZ+jeHjESIOhtu/F+FFnbOKyhk6fYjzF13kB+3HCYnv7DCtw4J8KZHyyCzQG/LYLqEBZGclcvOxPRir33HMuls7eF9r3+xIuBCrnzk7ap91jIoAKomBUDiNkk7Yf6D4BN8IuCJhvBexf/VCHBgLXx6jakq3aQtTJhniiuK1DfLXoKfnoHWQ+DW78o+btZY2PED9B4PY6a7rn2VkbTD9FYlbQO7B5z1AAx7GDx9yj+vJoKgg2vNUHp+FvS9EYZOhjeHmt8v+4/ZVgmWZZGalc/RjByOZeSSlJ7LsYxcjmXkcDQjl5SsPCIa+zkX4f1z/aWy5G1ZgGPObdjzM0kPak/ApKXg5V+5z1oOBUDVpABI6oWknfDxFZASBwGhMP5LCO/p7laJVJxlwX/6mjy4y6dDn/FlH3tgDbw3AmwOuG9d3Qv4N38NX90NuWkml+maD6H1wIqfX50g6PheU2A14wi0GwHXzwaHJ6z4Dyz8O3gHm/wpd9cb++1t+P5R00PV9hzzjHwb1egtKvP9Xf1UbxFxj5D2cNtCCO1u8gxmXgJ7lrm7VSIVF7fSBD9eAdD18vKPbdXP5L1ZBbD8Fde0ryIK8mHR/8F/J5jgJ3Io3LGkcsEPmCGqC/8Bg+4xv3872dRBOra7/PMyj8EnV5vgJ6yHmSnnODGlfdDdphc5J8X0LLurv6OwAL77G3z3VxP89L3R/IOthoOfylIAJFKfBYWb5TQih0JOKnxyJWz6yt2tEqmYmE/Mz25XgHfA6Y8f/tcT582ClANVv29eFmSnVv38IulHzMr1v75qfh88yeTbVLWeV1EQNOQ+8/sfs+H1/vDNZEg5WPL4vGwz5HZ0BwS1MvmAp06KsDvgstdNccltC2DTvKq1qzpy0k0bf3vL/H7+0yYnyuH+ukMKgETqO99GcMNc6DwaCnLhi5vhq3tgxesmKTJxq/kfpUhdkpN28gu5z4SKnRM5GNoMg8K8k0FHZeVlm+GiV3tCWkLVrgGwfzW8PRz2LjMTE66ZaYKX6n6x22ww8lm4YzG0Px8K82HtDHitD3z/GGQkmeMKC+F/d0PcCvAOgvFflD7EFdoVhj1k3i94BDKOVq99lZF6CGaMgu3fgYePGfI6a3KFErJdQTlApVAOkNRLhQVmdtjaGaXstEFwK2gSZZKme11f+S76+i7ruPnSiltpvkAHT9LyI+607iP4+l6zFMyk1RX/Uty9BD66DBxecNdKMxRcGUv+Bb88Z96f9wQML32h1XL98V+T71OYZ9o/7hNo3rny16mIfSvgp2dNoAMm2Bp0l+nFWvWGSba+YY7JqSlLfq6ZRZe4GXqOgyvfqZ22niphg0lcTzsE/s3g2s8gon+t31ZJ0NWkAEjqLcsyXd0H15rcgWO74ehuk5twKoc33LkEmndxTztrm2WZxND9v0HcKvM6sqX4MQGh5l+kZU1Pltr1/kjz53P+06ZXoKIsCz65Cnb9BJFnwU3fgL2CgxnH98IbAyH/RI9oo0i4L7bi54MZ0nm5ixly7nIZXP5GyVmaNc2yzOf96VmIjy2+b8yb0Pv601/jwFp4/3yTg/Pnats17eBa+PAyyE03Fe7H/9dlSesKgKpJAZCcUSzLTJc/ussEROs+Mv+aDO0Of/kZPLxr7j5Hd8Gun8G3MfS42j1d3WtmwOLnIb2U4Y0m7aD1IFNg8sgW86/nkc/BwIl1plu+QTiyHd7ob2Z0Pbi58j1xx/fB9EGQlwmXvgrRN1fsvM+uM/9AaD3EVJjOSTElJNqdV/F7r37P9LQ2aQuT1lYueKouyzIFIX/+h/n7e+7f4exHKn7+D4/DytchqCXcvap2Ares42ZoMDnODFeO+8Slyc6V+f72KHeviNR/NptZWds/xAx7tTsP3hxsltT4+VkTAFRVTppZ52jnT7DzR0jed3Lfrp9h9Cunr4FSkw7FnpjtUmgSP1v0hoiBJuiJGHhynbVTi9V9/6iZYn3ZazVaj0TKEXsi+bnDyKoNQzaONMNXP0yBhf8HHS48/RTvbd+b4MfuAZdOg9/fMcHM2g8rHgBZFvx2YvhowJ2uDX7A/Lfc5VLodLGZ9VXZZ3fu4yaAOr4XfnzKLMtRkywL/jfJBD+N28C1s0xNszpKSdAiDU1gqJkZAiZReveSyp1/dJeZhjxzNLwQBZ9fD2veN8GPwwsiBpmKves/PVHZuhqJppVRkG9ySqxCM6V6yn64/UeTmNrl0uKLzHoHwFXvw0XPmy/EjV+axNiju1zT1oasIA/Wf27e97mh6tcZeOfJKd4LTpPHk5dlpmADDL7HVFYv6jXaOt/M5qqI3YtNkUOvgIoNO9UWu6NqgaOXnymKCOa/2cXPm964mvLbWybAcniZpPA6HPyAAiCRhqnzxSe+ACyYN9HUEqmIjXPN0MOPT5nZL4V5Ziig/1/gutnw1z1w2w9mVppPIzi4Bt45x+QE1LZVb0DCH+a+F/8bPH3LP95mM8mkN31j8oESN5u2bp1f+21tyH5/19St8m8OHS+s+nXsDvNlbvcwX7qb/1f2scunmQA9qOXJqfRhPaBFX/N3eP2nFWz7id6fXtfVft5PbYkaDtEnCiwunmqGIl/vb6pxH1xX9VpBB9aahV8BLvynWaanjnN7ADR9+nSioqLw8fEhOjqaZcsqVsjt119/xcPDg969exfbPnPmTGw2W4lXdramAYsUc+E/TU5M2iH49oHT/49v5Rvw5S1mqn3rwSbIuHcd3BcDl/wbOl10spZLu3NNflGzzmZtsw9GwfrZtfdZju2BX6aa9yOfK97bczqRQ+DOpeYz5aSaHq0fnzbTjN0l9ZD7itbVpvQj5ksX4Ly/V3/KeGg3s9wEmCneWcdLHnNs98nCiRf+s3i9oaJeoLUfnv55H9sD204s1THgjmo12+0u/repxdNuhBkqTtpuliR591x4pTss+KspqlrR/wayjpvyG4V5pve1/+212vya4tYAaPbs2UyePJnHH3+cmJgYhg0bxqhRo4iLiyv3vJSUFG688UZGjBhR6v6goCDi4+OLvXx8XJiHIFIfePmblbXtHrD5q5PDEn9WWGjqj/zwmPl9wB2m+OKAv5S/KGXTdnDbIug4CgpyYN4dpix/YUHNfg7LMlVz87NM0mVVhlUCw0xP0KC7ze/LX4Y/yngetW3ldDPL6MdSVkWv7356ygSZ4b2rN/x1quGPmAVH0w+f7IEoYlnmy7wgB9qeW7LadPerzHDWsV2wd3n591n9HmCZfKFmHWum7e7i8IDom2DCXPjrLrjyPfNsPP0h9QD8/jZ8OBrevwDi/yj/WpZl6o6lnMj7uew/9WZCgVsDoJdffpnbbruN22+/nS5dujBt2jQiIiJ48803yz3vzjvv5Prrr2fw4NKnr9psNsLCwoq9ypOTk0Nqamqxl0iD0DIaznnUvF/wiEmOPFV+Dsy59eT6ROc/DaNeNMMPFeETBNd+ahaEBLM20adjISu5JlpvrP/c5GY4vM2MoKr+z9fhCRdNPTlEsqa0ekq17Nge+Olp8/7XVyufn1WXHVh7svLzxf+q+N+h0/HwNr0ZADEfF39m2xbAzkWml+Pif5X8u+EdYIIggHUfln2P3AxzbTDJz2cSn2DoeQ2M/cgEQ9d9Dr1vMIHhwTWmftD3U8yEh9KsehO2za83eT+nclsAlJuby9q1axk5sngtgpEjR7JixYoyz5sxYwa7du3iySfL/tdReno6kZGRtGrVitGjRxMTE1NuW6ZOnUpwcLDzFRERUbkPI1KfnfWgSVzOTYO5d5pkYjBBysdXmmq9dk/zr8SqVHG122HEE3D1DPDwNbPF/t0Rnm8NL7aDl7rAtB7wWl9To+XNs+C/N5mpzqeTkXSyZ+qcR8vvkaqo/reb6dkHfofELac/vqZYlpnBlp9tnhOYYnuVDRYTt5q1oYqGa+qCwsKTSci9roOIATV7/cjBJ4ddvrkfcjPN67sTwf3Q+yCkQ+nnFg2Dbf5f2blwf/wXslOgcZSZuXam8vSFTqNgzBswaQ10u9JMKlg13eQJbfqq+FDhgbVmHTSoN3k/p3JbAJSUlERBQQGhocXXTAkNDSUhofRZIzt27ODRRx9l1qxZeHiUPoO/c+fOzJw5k6+//prPPvsMHx8fhg4dyo4dO8psy5QpU0hJSXG+9u/fX/UPJlLf2B1w5dvgFQj7V5l8iZQD8MFFsG+52X7Dl+ZfidXR/UqTIN2otRmSyE6BzCSTg5QcZ4YhjmyFwxvMkNybQyH2s/JzM76fAlnHILQHDLm3eu0rEhhqvgQA1n1cM9esiA1fmtIBDm+zyG2TtmY4oihwqIjUeFMkcOci82zcmcd0qvWfmd4Er0A4/6nauceIJ02S8/E9sOR5k9OSEgfBESeXgihNiz4mIbogt/RhYMsyq5iDGfZ19dR3dwkKh2tmmAkNjaNMLt8XN8Gsq01eVT3N+zmV2+sA2f70r0nLskpsAygoKOD666/n6aefpmPHssdfBw0axKBBg5y/Dx06lL59+/Kf//yH1157rdRzvL298fauoWJwIvVR4zYmkXnenSZJdfV7ppBgYLhZYyisR83cJ7wX3BsDKfvNGkcFeeZ/oAX5J37mmSnLy14ywdhXE806QqOngV+T4tfa+SNs+K+Zcn/ZqzW7uGLfG83MovWfwflP1lyxyLJkHjP1iMDktIT3hCvegQ9GmgUxO40yC4aWJyfdDC+mnlgk9Pge2PcrRA2r3bafTnaKmTUIcPZfa2/5EZ8guORl+GycKe9gOxGoXDS1/PpONhv0vclMpV/3oZkZeOp30N5lpuigpz/0Hl87ba/L2o+Au1eafxgtf8X8dzd9MDRtXy/zfk7ltlA2JCQEh8NRorcnMTGxRK8QQFpaGmvWrGHSpEl4eHjg4eHBM888w/r16/Hw8ODnn38u9T52u53+/fuX2wMkIpg1grpdCVaBCX5COpkk5poKfoo4PMyaZCEdzEKN4b2gVbQpVhg1zJTov2WBKXRn9zBDE9MHm2KLRXIzzMw1MFWcW0bXbBvbjYDAFqZ3yRXT4n980vSGNesMQ+832yL6n+y5+PYB07tTloJ8M0Mv4Q/wC4GOF5nt6z6q3XZXxJIXISPRfGEOnFi79+p0kcnpsQpMQN3+fLNI8On0HGuGHY9sNctznKqo96fXtS6taFynePrCuY+ZtdeizjbDtIc31su8n1O5LQDy8vIiOjqaRYsWFdu+aNEihgwZUuL4oKAgNmzYQGxsrPM1ceJEOnXqRGxsLAMHlr6wo2VZxMbGEh5+miqhIg2dzWYqw7YeYr40bv0eGrkpH87uMItU3v7jiRk+CfDJlWZGT14W/PJPM2wW3NpUt61pDg/oc+Jf+7UdROxbcfIeo6eBh9fJfWf/zQSIWcfhf/eUPhxoWab3YsdC8yV+/WzT0wImeCxtarirHNlmiuMBXPRC8c9WWy56wSy+6elnEvYr0jPhE2yGaMFMiS+SHGcSqaH+T32vCSHt4cb/mXzAVgPMOmT1LO/nVG4dAnvwwQeZMGEC/fr1Y/DgwbzzzjvExcUxcaL5V8KUKVM4ePAgH330EXa7ne7duxc7v3nz5vj4+BTb/vTTTzNo0CA6dOhAamoqr732GrGxsbzxxhsu/Wwi9ZJvY7i1DiXPtugDdywxiZar3zXTc3f+aIZ3wARsp9Z1qUl9boCl/4Ldv5jZcbWxmGN+jknaBZOM++eFWR2eZijsnbPNYphr3i+Za/HrNFg7A7DBVe9Bq34mKArtbv6VvuFLk7viapZlhvUK883SDR3Od819A5rBxF9NTk9lAvi+N0HsLJP0f9FU09uz+j2TBNz2nNpb7b2+sdlMPmB1cwLrALdmc40bN45p06bxzDPP0Lt3b5YuXcqCBQuIjIwEID4+/rQ1gf4sOTmZO+64gy5dujBy5EgOHjzI0qVLGTCghmcdiIhrePmZ/KTxc0zF5mO7zJdS96uhwwW1d9/GbcwXH0DMrNq5x/Jppgidf/Oyk4ObdzblBwB++Dsk7Ty5b8OXJ/NrLnoeupwY7rHZTB4TlD+9uzZtW3AiqdvLLEfiSoGhle+9jBgAzbqYelIbvjCzyIp65s60qe8CaDX4Umk1eJE6KuOomfaecsDkHgQ0q937bZwDX95q8oEe2FhztWsAknbAm0NMT8VV70OPq8s+trAQPh4De5aYfKdbF5pclY/HmPMH3W16LU6VeQxe6mxm3N2xxCwM6yp52fDGALP8xLCHYMT/ue7e1bHqTdNrFdrdDHl9c5+ZtXhfbM3+2Uutqcz3dwOZzyciZwT/pmbK/i3zaz/4AZML5dvYTNU/NQm7PNsXmhICC/9u8ntKq3xtWSaxuSDXJOoWFeMri91u8i18gs26avMfMEt2FOSahV5HPlfyHL8mZh+4Phl6xX9M8BPYovwp6HVNz3GmDMHhjfDzs2bbgDsU/JyhFACJiJTFw9sU7oOKDSUd3mRqpcStNEHAjFHwr/Yw7y7Y/LWZqg4Q+6mZXu3hC5e8VLFE3eCWcPFLJ9ryEWQnQ6v+cOW7ZX9B951gfhYN6bjCoRhTxgBg5LPlT0Gva/yanFwuI+OISaSuqSU7pM5RACQiUp4+J4KI7d9D2uGyj8s8Znpl8jLNwqo9x5mV6bOOmdXG/zsBXmwLs66BhSdmrp07pXLJ1T2uNqUKwBSnu+7z8le9bzMcGkWa9be2fF3x+1TVoVj46HKTR9P23NP3bNVF0TedfN9znOkBlDOSAiARkfKEdjU9LYX5pjBiaQoLYM5tZrZYo9Zm/bMr34FHdsJN38Kge0ygU5BjpqtnHTd5JkWLr1aUzQaXv2Fet34P/iHlH2+3n+wFqu1hsKLgJzvFLK0y7uN6WRyPyKFmsVYPn9qvWyRupSToUigJWkSKWfcRfH0vNGkH964t+cW+6EkzHd3DF24vo3ikZZm6ONsWQMIGs3ZZs0613/bUQ/BKNzNzbtJaU8ulpsWvhw8vM8NyEQPhhjngHVjz93GV7BSz+GdwK3e3RCpJSdAiIjWp25Vmdexju0xi86k2zjXBD8Dlr5ddOdtmM1Pahz1o1lhyRfADENQC2p8oF1C0onlNiv/jRM9PsimON/7L+h38gEk2V/BzxlMAJCJyOt4BJysFnzqUlLDRVGgGsxhreVPZ3aloGCz2U7PeWnkyjpqFVJdPM0N65UnYAB9dZob0WvU3PT8+6jWX+kEBkIhIRfQ9kRy7+SvISi6e9Nz2HBjxlPvadjodLzLLQ2QkmhyksiRsgHfPgVXTzfpkr/aCd86FX18zy0L8+dgPTwQ/Lfsp+JF6RwGQiEhFtIyG5l3NQpB/zDZJz8n7zCyrq2eY9cPqKocn9L7evC8rGXrTV/D+SBPoNI4yi17a7HBoHSx6Aqb1gHfPM9P7d/50Ivg5Zp7LhLn1dkFMabiUBF0KJUGLSKmKKgU7vEwRwvKSnuuapB3wej8T1DywyeQGgakyvfifZt0zMNPXr/7A1MRJP2Kmz2+aB/t+NYnUp2rRFybMa7irpEudoyRoEZHa0HPcyeAHyk96rmtCOkDrISaIif3UbMtONcN4RcHP4Ekmidmvifk9oBn0vw1u/hYe3AoX/xsizwJsZthLwY/UY3W4z1ZEpI7xa2IWYV3/KQy5r+4mPZel7wSIW2Fmg3UdY4KfpG1m+YfLXoNe15Z9bmCoWVV+wF9M4OTpV7eH/UROQ0NgpdAQmIiUKTfDLHnRqn/9K/SXm2EWSM1JNYX+8rPNel3XfmJyeUTqOQ2BiYjUFi9/iBhQ/4IfMG0v6rXKzzZ1e+5YrOBHGiQFQCIiDcmge8zMtX4ncnsCQ93dIhG30ACuiEhDEtIeJv/h7laIuJ16gERERKTBUQAkIiIiDY4CIBEREWlwFACJiIhIg6MASERERBocBUAiIiLS4CgAEhERkQZHAZCIiIg0OAqAREREpMFRACQiIiINjgIgERERaXAUAImIiEiDowBIREREGhwFQCIiItLgeLi7AXWRZVkApKamurklIiIiUlFF39tF3+PlUQBUirS0NAAiIiLc3BIRERGprLS0NIKDg8s9xmZVJExqYAoLCzl06BCBgYHYbLYavXZqaioRERHs37+foKCgGr12Q6DnV316htWj51d9eobVo+dXNsuySEtLo0WLFtjt5Wf5qAeoFHa7nVatWtXqPYKCgvQXtxr0/KpPz7B69PyqT8+wevT8Sne6np8iSoIWERGRBkcBkIiIiDQ4CoBczNvbmyeffBJvb293N6Ve0vOrPj3D6tHzqz49w+rR86sZSoIWERGRBkc9QCIiItLgKAASERGRBkcBkIiIiDQ4CoBERESkwVEA5ELTp08nKioKHx8foqOjWbZsmbubVGctXbqUSy+9lBYtWmCz2fjqq6+K7bcsi6eeeooWLVrg6+vLOeecw6ZNm9zT2Dpo6tSp9O/fn8DAQJo3b86YMWPYtm1bsWP0DMv35ptv0rNnT2exucGDB/Pdd9859+v5Vc7UqVOx2WxMnjzZuU3PsHxPPfUUNput2CssLMy5X8+vehQAucjs2bOZPHkyjz/+ODExMQwbNoxRo0YRFxfn7qbVSRkZGfTq1YvXX3+91P0vvvgiL7/8Mq+//jqrV68mLCyMCy64wLmOW0O3ZMkS7rnnHlatWsWiRYvIz89n5MiRZGRkOI/RMyxfq1ateP7551mzZg1r1qzhvPPO4/LLL3d+wej5Vdzq1at555136NmzZ7Hteoan161bN+Lj452vDRs2OPfp+VWTJS4xYMAAa+LEicW2de7c2Xr00Ufd1KL6A7DmzZvn/L2wsNAKCwuznn/+eee27OxsKzg42Hrrrbfc0MK6LzEx0QKsJUuWWJalZ1hVjRs3tt577z09v0pIS0uzOnToYC1atMg6++yzrfvvv9+yLP0drIgnn3zS6tWrV6n79PyqTz1ALpCbm8vatWsZOXJkse0jR45kxYoVbmpV/bVnzx4SEhKKPU9vb2/OPvtsPc8ypKSkANCkSRNAz7CyCgoK+Pzzz8nIyGDw4MF6fpVwzz33cMkll3D++ecX265nWDE7duygRYsWREVFce2117J7925Az68maDFUF0hKSqKgoIDQ0NBi20NDQ0lISHBTq+qvomdW2vPct2+fO5pUp1mWxYMPPshZZ51F9+7dAT3DitqwYQODBw8mOzubgIAA5s2bR9euXZ1fMHp+5fv8889Zt24dq1evLrFPfwdPb+DAgXz00Ud07NiRw4cP89xzzzFkyBA2bdqk51cDFAC5kM1mK/a7ZVkltknF6XlWzKRJk/jjjz9Yvnx5iX16huXr1KkTsbGxJCcnM2fOHG666SaWLFni3K/nV7b9+/dz//33s3DhQnx8fMo8Ts+wbKNGjXK+79GjB4MHD6Zdu3Z8+OGHDBo0CNDzqw4NgblASEgIDoejRG9PYmJiiehdTq9oFoSe5+nde++9fP311/zyyy+0atXKuV3PsGK8vLxo3749/fr1Y+rUqfTq1YtXX31Vz68C1q5dS2JiItHR0Xh4eODh4cGSJUt47bXX8PDwcD4nPcOK8/f3p0ePHuzYsUN/B2uAAiAX8PLyIjo6mkWLFhXbvmjRIoYMGeKmVtVfUVFRhIWFFXueubm5LFmyRM/zBMuymDRpEnPnzuXnn38mKiqq2H49w6qxLIucnBw9vwoYMWIEGzZsIDY21vnq168f48ePJzY2lrZt2+oZVlJOTg5btmwhPDxcfwdrgtvSrxuYzz//3PL09LTef/99a/PmzdbkyZMtf39/a+/eve5uWp2UlpZmxcTEWDExMRZgvfzyy1ZMTIy1b98+y7Is6/nnn7eCg4OtuXPnWhs2bLCuu+46Kzw83EpNTXVzy+uGu+66ywoODrYWL15sxcfHO1+ZmZnOY/QMyzdlyhRr6dKl1p49e6w//vjDeuyxxyy73W4tXLjQsiw9v6o4dRaYZekZns5DDz1kLV682Nq9e7e1atUqa/To0VZgYKDze0PPr3oUALnQG2+8YUVGRlpeXl5W3759nVOSpaRffvnFAkq8brrpJsuyzBTQJ5980goLC7O8vb2t4cOHWxs2bHBvo+uQ0p4dYM2YMcN5jJ5h+W699Vbnf6/NmjWzRowY4Qx+LEvPryr+HADpGZZv3LhxVnh4uOXp6Wm1aNHCuvLKK61NmzY59+v5VY/NsizLPX1PIiIiIu6hHCARERFpcBQAiYiISIOjAEhEREQaHAVAIiIi0uAoABIREZEGRwGQiIiINDgKgERERKTBUQAkIiIiDY4CIBGRUixevBibzUZycrK7myIitUABkIiIiDQ4CoBERESkwVEAJCJ1kmVZvPjii7Rt2xZfX1969erFl19+CZwcnpo/fz69evXCx8eHgQMHsmHDhmLXmDNnDt26dcPb25s2bdrw0ksvFdufk5PDX//6VyIiIvD29qZDhw68//77xY5Zu3Yt/fr1w8/PjyFDhrBt2zbnvvXr13PuuecSGBhIUFAQ0dHRrFmzppaeiIjUJA93N0BEpDR///vfmTt3Lm+++SYdOnRg6dKl3HDDDTRr1sx5zCOPPMKrr75KWFgYjz32GJdddhnbt2/H09OTtWvXMnbsWJ566inGjRvHihUruPvuu2natCk333wzADfeeCMrV67ktddeo1evXuzZs4ekpKRi7Xj88cd56aWXaNasGRMnTuTWW2/l119/BWD8+PH06dOHN998E4fDQWxsLJ6eni57RiJSDW5ejV5EpIT09HTLx8fHWrFiRbHtt912m3XddddZv/zyiwVYn3/+uXPf0aNHLV9fX2v27NmWZVnW9ddfb11wwQXFzn/kkUesrl27WpZlWdu2bbMAa9GiRaW2oegeP/74o3Pb/PnzLcDKysqyLMuyAgMDrZkzZ1b/A4uIy2kITETqnM2bN5Odnc0FF1xAQECA8/XRRx+xa9cu53GDBw92vm/SpAmdOnViy5YtAGzZsoWhQ4cWu+7QoUPZsWMHBQUFxMbG4nA4OPvss8ttS8+ePZ3vw8PDAUhMTATgwQcf5Pbbb+f888/n+eefL9Y2EanbFACJSJ1TWFgIwPz584mNjXW+Nm/e7MwDKovNZgNMDlHR+yKWZTnf+/r6Vqgtpw5pFV2vqH1PPfUUmzZt4pJLLuHnn3+ma9euzJs3r0LXFRH3UgAkInVO165d8fb2Ji4ujvbt2xd7RUREOI9btWqV8/3x48fZvn07nTt3dl5j+fLlxa67YsUKOnbsiMPhoEePHhQWFrJkyZJqtbVjx4488MADLFy4kCuvvJIZM2ZU63oi4hpKghaROicwMJCHH36YBx54gMLCQs466yxSU1NZsWIFAQEBREZGAvDMM8/QtGlTQkNDefzxxwkJCWHMmDEAPPTQQ/Tv359nn32WcePGsXLlSl5//XWmT58OQJs2bbjpppu49dZbnUnQ+/btIzExkbFjx562jVlZWTzyyCNcffXVREVFceDAAVavXs1VV11Va89FRGqQu5OQRERKU1hYaL366qtWp06dLE9PT6tZs2bWhRdeaC1ZssSZoPzNN99Y3bp1s7y8vKz+/ftbsbGxxa7x5ZdfWl27drU8PT2t1q1bW//617+K7c/KyrIeeOABKzw83PLy8rLat29vffDBB5ZlnUyCPn78uPP4mJgYC7D27Nlj5eTkWNdee60VERFheXl5WS1atLAmTZrkTJAWkbrNZlmnDIqLiNQDixcv5txzz+X48eM0atTI3c0RkXpIOUAiIiLS4CgAEhERkQZHQ2AiIiLS4KgHSERERBocBUAiIiLS4CgAEhERkQZHAZCIiIg0OAqAREREpMFRACQiIiINjgIgERERaXAUAImIiEiD8/97Ts4yURSqeQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pd.DataFrame(model.history.history).plot()\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('error')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "d51a6ffb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11250/11250 [==============================] - 18s 2ms/step\n"
     ]
    }
   ],
   "source": [
    "prediction_proba=model.predict(X_test).reshape(len(y_test),)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "c45b4c5d",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ams_dat,max_ams=ams_curve(prediction_proba,y_test,weights,test_ratio=0.8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "853048e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.savetxt('Data/DNN_curve.csv',ams_dat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "8220671a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Final AMS = 2.160602095162433\n"
     ]
    }
   ],
   "source": [
    "print('Final AMS = {}'.format(max_ams))"
   ]
  }
 ],
 "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": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}