train.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Reload the `EC_estimator.py` code to consider any changes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n"
     ]
    }
   ],
   "source": [
    "#reload model magic\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "import EC_estimator as annec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "import matplotlib.pyplot as plt\n",
    "from tabulate import tabulate\n",
    "import EC_estimator as annec\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.layers import Dense, Input\n",
    "from tensorflow.keras.layers.experimental import preprocessing\n",
    "from tensorflow.keras import Model"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load the dataset, preprocessing, and apply ANN model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load data\n",
    "df = annec.load_data('Inputs/inputs.csv')\n",
    "df_var = annec.df_by_variable(df)\n",
    "\n",
    "\n",
    "\n",
    "# Split data\n",
    "train_rows = df_var.shape[0] - 6039\n",
    "df_train, df_test = annec.split_data(df_var, train_rows, 6039)\n",
    "\n",
    "# Build model inputs\n",
    "inputs = annec.build_model_inputs(df_var)\n",
    "\n",
    "# Preprocessing layers\n",
    "layers = annec.preprocessing_layers(df_var, inputs)\n",
    "\n",
    "# Build model\n",
    "neurons = [8, 2, 1]  # This is an example; adapt it to your needs\n",
    "activations = ['sigmoid', 'sigmoid']\n",
    "model, tensorboard_cb = annec.build_model(layers, inputs)\n",
    "\n",
    "\n",
    "X_train = [ df_train.loc[:, pd.IndexSlice[feature,annec.lags_feature[feature]]] for feature in annec.feature_names()]\n",
    "\n",
    "y_train = df_train[('EC','0d')].to_frame()  # situation where one column is produced and coerced to a series\n",
    "# Otherwise you would do something like:\n",
    "#y.droplevel(level='lag',axis=1).rename(mapper={\"EC\":\"emm_ec\"})\n",
    "y_train.columns = [\"emm_ec\"]\n",
    "\n",
    "\n",
    "\n",
    "X_test = [ df_test.loc[:, pd.IndexSlice[feature,annec.lags_feature[feature]]] for feature in annec.feature_names()]\n",
    "\n",
    "\n",
    "y_test = df_test[('EC','0d')].to_frame()  # Only works for univariate dataset where it is a series\n",
    "y_test.columns = [\"emm_ec\"] \n",
    "\n",
    "\n",
    "# Train model\n",
    "history, model = annec.train_model(model, tensorboard_cb, X_train, y_train, X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Illustration of by-variable input (df_var)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lag           0d           1d           2d           3d           4d  \\\n",
      "0    22290.01548  22333.70068  22358.59029  22373.94862  22383.92889   \n",
      "1    22203.28407  22290.01548  22333.70068  22358.59029  22373.94862   \n",
      "2    21992.03107  22203.28407  22290.01548  22333.70068  22358.59029   \n",
      "3    18938.58199  21992.03107  22203.28407  22290.01548  22333.70068   \n",
      "\n",
      "lag           5d           6d           7d         1ave         2ave  \\\n",
      "0    22390.61113  22395.12267  22398.10862  22391.45013  20455.31666   \n",
      "1    22383.92889  22390.61113  22395.12267  22394.40549  21187.05572   \n",
      "2    22373.94862  22383.92889  22390.61113  22396.15244  21864.78865   \n",
      "3    22358.59029  22373.94862  22383.92889  22396.79180  22157.44139   \n",
      "\n",
      "lag         3ave         4ave         5ave         6ave         7ave  \\\n",
      "0    13808.89333  13727.31347  15871.35388  18889.58980  19086.62041   \n",
      "1    13863.05279  13723.46477  15415.59194  18875.79859  19027.33307   \n",
      "2    13971.76887  13721.41293  14965.45140  18855.52854  18992.74727   \n",
      "3    14465.72487  13720.89280  14528.43113  18821.66279  18970.20303   \n",
      "\n",
      "lag         8ave         9ave        10ave  \n",
      "0    27995.04227  29834.38514  30086.23375  \n",
      "1    27067.75544  29827.11331  29963.14182  \n",
      "2    26117.13349  29818.32929  29918.20477  \n",
      "3    25156.58976  29807.29370  29895.15910  \n"
     ]
    }
   ],
   "source": [
    "df = annec.load_data('Inputs/inputs.csv')\n",
    "\n",
    "# This will break up the incoming DataFrame into a MultiIndex column\n",
    "# with the site/variable (e.g. Sac) as the 0-level index and\n",
    "# lag (e.g. '3d') as the fast index.\n",
    "df_var = annec.df_by_variable(df)\n",
    "\n",
    "# This is an illustration of how to extract a single location/variable\n",
    "# so that it (with its lags) can be used as a single input in the ANN. \n",
    "# In the traditional CalSIM model would be 18 values \n",
    "# The droplevel gets rid of the vestiges of the multiindex\n",
    "sac = df_var.loc[:, pd.IndexSlice['sac',:]].droplevel(level=\"var\",axis=1)\n",
    "print(sac.head(4))\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluation of the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "867/867 [==============================] - 2s 3ms/step\n",
      "189/189 [==============================] - 1s 3ms/step\n",
      "predictions\n",
      "[[3295.7021]\n",
      " [3430.7017]\n",
      " [3538.958 ]]\n",
      "[[2290.895 ]\n",
      " [2287.8262]\n",
      " [2455.1255]]\n",
      "input data\n",
      "var            sac                                                      \\\n",
      "lag             0d           1d           2d           3d           4d   \n",
      "33754  7785.670076  7799.537119  7808.753521  7815.169612  7819.799191   \n",
      "33755  7763.476647  7785.670076  7799.537119  7808.753521  7815.169612   \n",
      "33756  7724.776026  7763.476647  7785.670076  7799.537119  7808.753521   \n",
      "\n",
      "var                                                                     \\\n",
      "lag             5d           6d           7d         1ave         2ave   \n",
      "33754  7823.235474  7825.844312  7827.861244  7833.139374  7798.669268   \n",
      "33755  7819.799191  7823.235474  7825.844312  7832.484449  7807.586937   \n",
      "33756  7815.169612  7819.799191  7823.235474  7831.658644  7816.464872   \n",
      "\n",
      "var                                                                     \\\n",
      "lag           3ave         4ave         5ave         6ave         7ave   \n",
      "33754  7737.682238  7753.917794  8876.896143  9669.167134  9693.248487   \n",
      "33755  7737.442994  7749.747103  8708.026372  9665.952129  9688.680497   \n",
      "33756  7737.290866  7746.743006  8539.160977  9661.609439  9685.582948   \n",
      "\n",
      "var                                           \n",
      "lag           8ave         9ave        10ave  \n",
      "33754  10380.62483  10598.11380  10549.39818  \n",
      "33755  10302.02801  10597.64112  10577.22046  \n",
      "33756  10222.09538  10596.89333  10586.94983  \n",
      "var           sac                                                      \\\n",
      "lag            0d           1d           2d           3d           4d   \n",
      "6035  7183.190302  7178.489123  7169.338157  7147.221362  7067.043165   \n",
      "6036  7185.981787  7183.190302  7178.489123  7169.338157  7147.221362   \n",
      "6037  7187.839785  7185.981787  7183.190302  7178.489123  7169.338157   \n",
      "\n",
      "var                                                                    \\\n",
      "lag            5d           6d           7d         1ave         2ave   \n",
      "6035  6879.099909  6855.996998  6846.414762  6836.786803  6852.697205   \n",
      "6036  7067.043165  6879.099909  6855.996998  6837.654798  6848.232893   \n",
      "6037  7147.221362  7067.043165  6879.099909  6839.453782  6844.936756   \n",
      "\n",
      "var                                                                     \\\n",
      "lag          3ave         4ave         5ave          6ave         7ave   \n",
      "6035  8228.374296  9576.938384  9604.657542  10053.670210  10211.79019   \n",
      "6036  7988.296654  9571.021454  9601.022381  10002.754410  10207.59661   \n",
      "6037  7749.724416  9562.502747  9598.449048   9950.937071  10203.87662   \n",
      "\n",
      "var                                          \n",
      "lag          8ave         9ave        10ave  \n",
      "6035  10620.48233  19579.94320  20417.26775  \n",
      "6036  10453.77877  18823.16760  20417.52532  \n",
      "6037  10370.22916  17983.96183  20415.15283  \n",
      "end predictions\n"
     ]
    }
   ],
   "source": [
    "# Calculate metrics for ANN\n",
    "#print(model.summary())\n",
    "metrics = []\n",
    "y_train_pred = model.predict(X_train)\n",
    "y_test_pred = model.predict(X_test)\n",
    "\n",
    "print(\"predictions\")\n",
    "print(y_train_pred[-4:-1])\n",
    "print(y_test_pred[-4:-1])\n",
    "print(\"input data\")\n",
    "print(X_train[0].iloc[-4:-1])\n",
    "print(X_test[0].iloc[-4:-1])\n",
    "print(\"end predictions\")\n",
    "metrics.append(annec.calculate_metrics('ANN', y_train, y_train_pred, y_test, y_test_pred))\n",
    "\n",
    "df3 = pd.DataFrame(metrics)\n",
    "\n",
    "#print(tabulate(df3, headers='keys', tablefmt='psql'))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1sAAAHUCAYAAADMRTIhAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAACc3klEQVR4nOzdd3gU1dvG8e+m9wqkQOi9916ldwQFQUEUK4oGGzYU9adYwYL9VcCCoCKI9I703nsJSSihphDSk3n/GFgMoQTYsEm4P9c1V3Znzsw8QxbInXPmjMUwDAMRERERERGxKQd7FyAiIiIiIlIYKWyJiIiIiIjkAYUtERERERGRPKCwJSIiIiIikgcUtkRERERERPKAwpaIiIiIiEgeUNgSERERERHJAwpbIiIiIiIieUBhS0REREREJA8obIlIvmCxWHK1LF269JbOM2rUKCwWy03tu3TpUpvUkN8NHjyY0qVLX3X7qVOncHFx4b777rtqm4SEBDw8POjRo0euzzthwgQsFguHDx/OdS3/ZbFYGDVqVK7Pd9GxY8cYNWoUW7ZsybHtVj4vt6p06dJ069bNLue+UWfOnOGVV16hatWqeHh44OPjQ+PGjfnyyy9JT0+3d3k5tG7d+qr/xuT285aXLn7uTp8+be9SROQWOdm7ABERgNWrV2d7/84777BkyRIWL16cbX3VqlVv6TyPPPIInTp1uql969aty+rVq2+5hoKuaNGi9OjRg+nTpxMbG4u/v3+ONpMnTyY5OZkhQ4bc0rlGjhzJs88+e0vHuJ5jx47x1ltvUbp0aWrXrp1t2618Xu4Ue/bsoUOHDiQmJvL888/TtGlTkpOTmTlzJs8++yx//PEHs2fPxsPDw96lZlO2bFl+/fXXHOtdXV3tUI2IFFYKWyKSLzRu3Djb+6JFi+Lg4JBj/eWSkpJu6Ie4EiVKUKJEiZuq8eJv6wWGDBnC1KlT+fXXX3n66adzbP/xxx8JCgqia9eut3SecuXK3dL+t+pWPi93gszMTPr06UNCQgLr1q2jYsWK1m1dunShVatW3HfffTz33HN88803t60uwzBISUnB3d39qm3c3d3191lE8pyGEYpIgdG6dWuqV6/Ov//+S9OmTfHw8ODhhx8GYMqUKXTo0IGQkBDc3d2pUqUKL7/8MufPn892jCsNC7s4XGvu3LnUrVsXd3d3KleuzI8//pit3ZWGEQ4ePBgvLy8OHDhAly5d8PLyIiwsjOeff57U1NRs+x85coR77rkHb29v/Pz8uP/++1m/fj0Wi4UJEyZc89pPnTrF0KFDqVq1Kl5eXhQrVoy77rqL5cuXZ2t3+PBhLBYLH3/8MWPGjKFMmTJ4eXnRpEkT1qxZk+O4EyZMoFKlSri6ulKlShV++umna9ZxUceOHSlRogTjx4/PsW337t2sXbuWQYMG4eTkxIIFC+jZsyclSpTAzc2N8uXL8/jjj+dqiNSVhhEmJCTw6KOPEhgYiJeXF506dWLfvn059j1w4AAPPfQQFSpUwMPDg+LFi9O9e3e2b99ubbN06VIaNGgAwEMPPWQdSnZxOOKVPi9ZWVl8+OGHVK5cGVdXV4oVK8agQYM4cuRItnYXP6/r16+nRYsWeHh4ULZsWd5//32ysrKue+25kZKSwiuvvEKZMmVwcXGhePHiPPXUU8TFxWVrt3jxYlq3bk1gYCDu7u6ULFmSPn36kJSUZG3z9ddfU6tWLby8vPD29qZy5cq8+uqr1zz/tGnT2LVrFy+//HK2oHVRv3796NChAz/88AMxMTGkp6dTrFgxBg4cmKNtXFwc7u7uPPfcc9Z1CQkJvPDCC9muLzw8PMffa4vFwtNPP80333xDlSpVcHV1ZeLEibn5I7ymi0NbFyxYwEMPPURAQACenp50796dQ4cO5Wj/448/UqtWLdzc3AgICODuu+9m9+7dOdqtXbuW7t27ExgYiJubG+XKlSM8PDxHuxMnTtC/f398fX0JCgri4YcfJj4+PlubP/74g0aNGuHr62v9jF38d1FE7E9hS0QKlOPHj/PAAw8wYMAAZs+ezdChQwHYv38/Xbp04YcffmDu3LmEh4fz+++/071791wdd+vWrTz//PMMHz6cv//+m5o1azJkyBD+/fff6+6bnp5Ojx49aNu2LX///TcPP/wwY8eO5YMPPrC2OX/+PG3atGHJkiV88MEH/P777wQFBdGvX79c1Xf27FkA3nzzTWbNmsX48eMpW7YsrVu3vuI9ZF9++SULFizg008/5ddff+X8+fN06dIl2w9qEyZM4KGHHqJKlSpMnTqV119/nXfeeSfH0M0rcXBwYPDgwWzatImtW7dm23YxgF38ge/gwYM0adKEr7/+mvnz5/PGG2+wdu1amjdvfsP38xiGQa9evfj55595/vnnmTZtGo0bN6Zz58452h47dozAwEDef/995s6dy5dffomTkxONGjVi7969gDk09GK9r7/+OqtXr2b16tU88sgjV63hySefZMSIEbRv354ZM2bwzjvvMHfuXJo2bZojQMbExHD//ffzwAMPMGPGDDp37swrr7zCL7/8ckPXfa0/i48//piBAwcya9YsnnvuOSZOnMhdd91lDfuHDx+ma9euuLi48OOPPzJ37lzef/99PD09SUtLA8xhn0OHDqVVq1ZMmzaN6dOnM3z48Byh5nILFiwAoFevXldt06tXLzIyMli6dCnOzs488MADTJ06lYSEhGztfvvtN1JSUnjooYcAs9e6VatWTJw4kWeeeYY5c+YwYsQIJkyYQI8ePTAMI9v+06dP5+uvv+aNN95g3rx5tGjR4rp/hhkZGTmWKwXhIUOG4ODgwKRJk/j0009Zt24drVu3zhZqR48ezZAhQ6hWrRp//fUXn332Gdu2baNJkybs37/f2u5ibVFRUYwZM4Y5c+bw+uuvc+LEiRzn7dOnDxUrVmTq1Km8/PLLTJo0ieHDh1u3r169mn79+lG2bFkmT57MrFmzeOONN8jIyLjutYvIbWKIiORDDz74oOHp6ZltXatWrQzAWLRo0TX3zcrKMtLT041ly5YZgLF161brtjfffNO4/J++UqVKGW5ubkZkZKR1XXJyshEQEGA8/vjj1nVLliwxAGPJkiXZ6gSM33//Pdsxu3TpYlSqVMn6/ssvvzQAY86cOdnaPf744wZgjB8//prXdLmMjAwjPT3daNu2rXH33Xdb10dERBiAUaNGDSMjI8O6ft26dQZg/Pbbb4ZhGEZmZqYRGhpq1K1b18jKyrK2O3z4sOHs7GyUKlXqujUcOnTIsFgsxjPPPGNdl56ebgQHBxvNmjW74j4XvzeRkZEGYPz999/WbePHjzcAIyIiwrruwQcfzFbLnDlzDMD47LPPsh333XffNQDjzTffvGq9GRkZRlpamlGhQgVj+PDh1vXr16+/6vfg8s/L7t27DcAYOnRotnZr1641AOPVV1+1rrv4eV27dm22tlWrVjU6dux41TovKlWqlNG1a9erbp87d64BGB9++GG29VOmTDEA47vvvjMMwzD+/PNPAzC2bNly1WM9/fTThp+f33VrulynTp0MwEhJSblqm4vfsw8++MAwDMPYtm1btvouatiwoVGvXj3r+9GjRxsODg7G+vXrs7W7eD2zZ8+2rgMMX19f4+zZs7mq++L35krLkCFDrO0ufib/+3fMMAxj5cqVBmD873//MwzDMGJjYw13d3ejS5cu2dpFRUUZrq6uxoABA6zrypUrZ5QrV85ITk6+an0XP3eXf2+HDh1quLm5Wf/OfvzxxwZgxMXF5eq6ReT2U8+WiBQo/v7+3HXXXTnWHzp0iAEDBhAcHIyjoyPOzs60atUK4IrDeC5Xu3ZtSpYsaX3v5uZGxYoViYyMvO6+FoslRw9azZo1s+27bNkyvL29c0y20L9//+se/6JvvvmGunXr4ubmhpOTE87OzixatOiK19e1a1ccHR2z1QNYa9q7dy/Hjh1jwIAB2YbJlSpViqZNm+aqnjJlytCmTRt+/fVXaw/JnDlziImJyTaM6eTJkzzxxBOEhYVZ6y5VqhSQu+/Nfy1ZsgSA+++/P9v6AQMG5GibkZHBe++9R9WqVXFxccHJyQkXFxf2799/w+e9/PyDBw/Otr5hw4ZUqVKFRYsWZVsfHBxMw4YNs627/LNxsy72QF5ey7333ounp6e1ltq1a+Pi4sJjjz3GxIkTrzj8rWHDhsTFxdG/f3/+/vtvm86CZ1zogbr4OatRowb16tXLNgR19+7drFu3LtvnZubMmVSvXp3atWtn63nq2LHjFWcFveuuu644WcvVlCtXjvXr1+dYRo4cmaPt5Z+3pk2bUqpUKevnYfXq1SQnJ+f4XoSFhXHXXXdZvxf79u3j4MGDDBkyBDc3t+vWePlsnjVr1iQlJYWTJ08CWIfA9u3bl99//52jR4/m7uJF5LZR2BKRAiUkJCTHusTERFq0aMHatWv53//+x9KlS1m/fj1//fUXAMnJydc9bmBgYI51rq6uudrXw8Mjxw9Orq6upKSkWN+fOXOGoKCgHPtead2VjBkzhieffJJGjRoxdepU1qxZw/r16+nUqdMVa7z8ei7OsHax7ZkzZwAzDFzuSuuuZsiQIZw5c4YZM2YA5hBCLy8v+vbtC5j3N3Xo0IG//vqLl156iUWLFrFu3Trr/WO5+fP9rzNnzuDk5JTj+q5U83PPPcfIkSPp1asX//zzD2vXrmX9+vXUqlXrhs/73/PDlT+HoaGh1u0X3crnKje1ODk5UbRo0WzrLRYLwcHB1lrKlSvHwoULKVasGE899RTlypWjXLlyfPbZZ9Z9Bg4cyI8//khkZCR9+vShWLFiNGrUyDpM8Gou/oIiIiLiqm0uTuUfFhZmXffwww+zevVq9uzZA5ifG1dX12y/fDhx4gTbtm3D2dk52+Lt7Y1hGDkC4ZW+J9fi5uZG/fr1cywXfxHwX1f7e3Lxzzi3n4tTp04B5HrSlev9PW7ZsiXTp08nIyODQYMGUaJECapXr85vv/2Wq+OLSN7TbIQiUqBc6ZlHixcv5tixYyxdutTamwXkmCTAngIDA1m3bl2O9TExMbna/5dffqF169Z8/fXX2dafO3fupuu52vlzWxNA79698ff358cff6RVq1bMnDmTQYMG4eXlBcCOHTvYunUrEyZM4MEHH7Tud+DAgZuuOyMjgzNnzmT7QfRKNf/yyy8MGjSI9957L9v606dP4+fnd9PnB/Pewct/YD527BhFihS5qePebC0ZGRmcOnUqW+AyDIOYmBhrrwdAixYtaNGiBZmZmWzYsIEvvviC8PBwgoKCrM9Le+ihh3jooYc4f/48//77L2+++SbdunVj3759VwwgAO3bt+e7775j+vTpvPzyy1dsM336dJycnGjdurV1Xf/+/XnuueeYMGEC7777Lj///DO9evXK1jNVpEgR3N3dc0xU89/t/5WXz0O72t+T8uXLA9k/F5f77+fi4vfp8slUbkXPnj3p2bMnqamprFmzhtGjRzNgwABKly5NkyZNbHYeEbk56tkSkQLv4g9Zlz8f59tvv7VHOVfUqlUrzp07x5w5c7Ktnzx5cq72t1gsOa5v27ZtOZ5PlluVKlUiJCSE3377LdtEA5GRkaxatSrXx3Fzc2PAgAHMnz+fDz74gPT09GxDwWz9vWnTpg1AjucjTZo0KUfbK/2ZzZo1K8dQq8t7C67l4hDWyye4WL9+Pbt376Zt27bXPYatXDzX5bVMnTqV8+fPX7EWR0dHGjVqxJdffgnApk2bcrTx9PSkc+fOvPbaa6SlpbFz586r1nD33XdTtWpV3n///SvOCDllyhTmz5/PI488kq13yN/fn169evHTTz8xc+bMHENPAbp168bBgwcJDAy8Yg/U7Xz48OWft1WrVhEZGWkNkE2aNMHd3T3H9+LIkSMsXrzY+r2oWLEi5cqV48cff8wxW+mtcnV1pVWrVtaJeTZv3mzT44vIzVHPlogUeE2bNsXf358nnniCN998E2dnZ3799dccs+TZ04MPPsjYsWN54IEH+N///kf58uWZM2cO8+bNA8zZ/a6lW7duvPPOO7z55pu0atWKvXv38vbbb1OmTJmbmnnMwcGBd955h0ceeYS7776bRx99lLi4OEaNGnVDwwjBHEr45ZdfMmbMGCpXrpztnq/KlStTrlw5Xn75ZQzDICAggH/++ee6w9OupkOHDrRs2ZKXXnqJ8+fPU79+fVauXMnPP/+co223bt2YMGEClStXpmbNmmzcuJGPPvooR49UuXLlcHd359dff6VKlSp4eXkRGhpKaGhojmNWqlSJxx57jC+++AIHBwc6d+7M4cOHGTlyJGFhYdlmirOFmJgY/vzzzxzrS5cuTfv27enYsSMjRowgISGBZs2asW3bNt58803q1KljnV79m2++YfHixXTt2pWSJUuSkpJi7S1q164dAI8++iju7u40a9aMkJAQYmJiGD16NL6+vtl6yC7n6OjI1KlTad++PU2aNOH555+nSZMmpKam8s8///Ddd9/RqlUrPvnkkxz7Pvzww0yZMoWnn36aEiVKWGu5KDw8nKlTp9KyZUuGDx9OzZo1ycrKIioqivnz5/P888/TqFGjm/6zTU5OvuLjECDnc/82bNjAI488wr333kt0dDSvvfYaxYsXt86G6ufnx8iRI3n11VcZNGgQ/fv358yZM7z11lu4ubnx5ptvWo/15Zdf0r17dxo3bszw4cMpWbIkUVFRzJs374oPWb6WN954gyNHjtC2bVtKlChBXFwcn332WbZ7VkXEzuw6PYeIyFVcbTbCatWqXbH9qlWrjCZNmhgeHh5G0aJFjUceecTYtGlTjlnmrjYb4ZVmfWvVqpXRqlUr6/urzUZ4eZ1XO09UVJTRu3dvw8vLy/D29jb69OljzJ49O8esfFeSmppqvPDCC0bx4sUNNzc3o27dusb06dNzzNZ3cTbCjz76KMcxuMJsff/3f/9nVKhQwXBxcTEqVqxo/PjjjzmOmRt16tS54uxphmEYu3btMtq3b294e3sb/v7+xr333mtERUXlqCc3sxEahmHExcUZDz/8sOHn52d4eHgY7du3N/bs2ZPjeLGxscaQIUOMYsWKGR4eHkbz5s2N5cuX5/i+GoZh/Pbbb0blypUNZ2fnbMe50vcxMzPT+OCDD4yKFSsazs7ORpEiRYwHHnjAiI6Oztbuap/X3P75lipV6qoz5j344IOGYZizZo4YMcIoVaqU4ezsbISEhBhPPvmkERsbaz3O6tWrjbvvvtsoVaqU4erqagQGBhqtWrUyZsyYYW0zceJEo02bNkZQUJDh4uJihIaGGn379jW2bdt23ToNwzBOnz5tvPzyy0blypUNNzc3w8vLy2jYsKExbtw4Iy0t7Yr7ZGZmGmFhYQZgvPbaa1dsk5iYaLz++utGpUqVDBcXF8PX19eoUaOGMXz4cCMmJsbaDjCeeuqpXNVqGNeejRAw0tPTDcO49JmcP3++MXDgQMPPz8866+D+/ftzHPf//u//jJo1a1pr7dmzp7Fz584c7VavXm107tzZ8PX1NVxdXY1y5cplmyHz4ufu1KlT2fa7/O/IzJkzjc6dOxvFixc3XFxcjGLFihldunQxli9fnus/CxHJWxbDuOxBFSIictu89957vP7660RFReX6pnkRuT0uPotu/fr11K9f397liEgBpGGEIiK3ybhx4wBzaF16ejqLFy/m888/54EHHlDQEhERKYQUtkREbhMPDw/Gjh3L4cOHSU1NpWTJkowYMYLXX3/d3qWJiIhIHtAwQhERERERkTygqd9FRERERETygMKWiIiIiIhIHlDYEhERERERyQOaICOXsrKyOHbsGN7e3lgsFnuXIyIiIiIidmIYBufOnSM0NBQHh6v3Xyls5dKxY8cICwuzdxkiIiIiIpJPREdHX/PxLQpbueTt7Q2Yf6A+Pj52rkZEREREROwlISGBsLAwa0a4GoWtXLo4dNDHx0dhS0RERERErnt7kSbIEBERERERyQMKWyIiIiIiInlAYUtERERERCQP6J4tERERESmQDMMgIyODzMxMe5cihYyjoyNOTk63/MgnhS0RERERKXDS0tI4fvw4SUlJ9i5FCikPDw9CQkJwcXG56WMobImIiIhIgZKVlUVERASOjo6Ehobi4uJyyz0QIhcZhkFaWhqnTp0iIiKCChUqXPPBxdeisCUiIiIiBUpaWhpZWVmEhYXh4eFh73KkEHJ3d8fZ2ZnIyEjS0tJwc3O7qeNoggwRERERKZButrdBJDds8fnSJ1RERERERCQPKGyJiIiIiIjkAYUtEREREZECrHXr1oSHh+e6/eHDh7FYLGzZsiXPahKTwpaIiIiIyG1gsViuuQwePPimjvvXX3/xzjvv5Lp9WFgYx48fp3r16jd1vtxSqNNshCIiIiIit8Xx48etr6dMmcIbb7zB3r17revc3d2ztU9PT8fZ2fm6xw0ICLihOhwdHQkODr6hfeTmqGeroDEM+LwufH8X/NoXpg+F+SNh5Wew+VfYNw+OboTYSEjTQ/5ERETkzmAYBklpGXZZDMPIVY3BwcHWxdfXF4vFYn2fkpKCn58fv//+O61bt8bNzY1ffvmFM2fO0L9/f0qUKIGHhwc1atTgt99+y3bcy4cRli5dmvfee4+HH34Yb29vSpYsyXfffWfdfnmP09KlS7FYLCxatIj69evj4eFB06ZNswVBgP/9738UK1YMb29vHnnkEV5++WVq1659U98vgNTUVJ555hmKFSuGm5sbzZs3Z/369dbtsbGx3H///RQtWhR3d3cqVKjA+PHjAXP6/6effpqQkBDc3NwoXbo0o0ePvula8op6tgqalDg4ezD37Z09wTMQPIuCRxHzq1dR8Aq68LrYpdfu/qAHAoqIiEgBlJyeSdU35tnl3Lve7oiHi21+rB4xYgSffPIJ48ePx9XVlZSUFOrVq8eIESPw8fFh1qxZDBw4kLJly9KoUaOrHueTTz7hnXfe4dVXX+XPP//kySefpGXLllSuXPmq+7z22mt88sknFC1alCeeeIKHH36YlStXAvDrr7/y7rvv8tVXX9GsWTMmT57MJ598QpkyZW76Wl966SWmTp3KxIkTKVWqFB9++CEdO3bkwIEDBAQEMHLkSHbt2sWcOXMoUqQIBw4cIDk5GYDPP/+cGTNm8Pvvv1OyZEmio6OJjo6+6VryisJWQePsCQ/Ph6TTcP70ha9nLnt/YclMhfTzEHce4qKuf2xHF/AKBu+LS0j2rz7FwScUXL3y/jpFRERE7kDh4eH07t0727oXXnjB+nrYsGHMnTuXP/7445phq0uXLgwdOhQwA9zYsWNZunTpNcPWu+++S6tWrQB4+eWX6dq1KykpKbi5ufHFF18wZMgQHnroIQDeeOMN5s+fT2Ji4k1d5/nz5/n666+ZMGECnTt3BuD7779nwYIF/PDDD7z44otERUVRp04d6tevD5g9dhdFRUVRoUIFmjdvjsVioVSpUjdVR15T2CponFyg5NX/YlkZBqQlwvlTF8LXxa8nIfEUJJ4w1yWeNNelxENmGsRHmcu1uPleCl4+xS8sIRdC2YXFI0C9ZCIiInLbuDs7suvtjnY7t61cDBYXZWZm8v777zNlyhSOHj1KamoqqampeHp6XvM4NWvWtL6+OFzx5MmTud4nJCQEgJMnT1KyZEn27t1rDW8XNWzYkMWLF+fqui538OBB0tPTadasmXWds7MzDRs2ZPfu3QA8+eST9OnTh02bNtGhQwd69epF06ZNARg8eDDt27enUqVKdOrUiW7dutGhQ4ebqiUvKWwVVhYLuHqbS0DZ67dPTzFD17kTcO74f5YY82vCcUg4aga4lHhzObnr6sdzdL3UK+ZfGgLLQ2A5cwkop94xERERsSmLxWKzoXz2dHmI+uSTTxg7diyffvopNWrUwNPTk/DwcNLS0q55nMsn1rBYLGRlZeV6H8uFX5r/dx/LZb9Iz+29aldycd8rHfPius6dOxMZGcmsWbNYuHAhbdu25amnnuLjjz+mbt26REREMGfOHBYuXEjfvn1p164df/75503XlBcK/idSbMPZDfxKmsu1pCRAwjFIOHLh6zGIP3IplJ07DklnzCGMcZHmEr0m53G8Q8wAFlD2Qggrb4awgDLg5Jo31ygiIiJSwCxfvpyePXvywAMPAGb42b9/P1WqVLmtdVSqVIl169YxcOBA67oNGzbc9PHKly+Pi4sLK1asYMCAAYA5++KGDRuyTfZRtGhRBg8ezODBg2nRogUvvvgiH3/8MQA+Pj7069ePfv36cc8999CpUyfOnj17w7Mz5iWFLbkxbj7mUuzq433JSL3UK5ZwFM4egjOH4MwBc3KPpDOXgtnh5dn3tTiAbwkzeAWWM8PYxcWvlBkKRURERO4Q5cuXZ+rUqaxatQp/f3/GjBlDTEzMbQ9bw4YN49FHH6V+/fo0bdqUKVOmsG3bNsqWvf4IqstnNQSoWrUqTz75JC+++CIBAQGULFmSDz/8kKSkJIYMGQKY94XVq1ePatWqkZqaysyZM63XPXbsWEJCQqhduzYODg788ccfBAcH4+fnZ9PrvlUKW2J7Tq7m0EH/0lfennT2QgA7AGcOmgHszAEzkKWdMyfziIuCQ0su29Fi3h8WUOZSb1hgBShSwQxijvo4i4iISOEycuRIIiIi6NixIx4eHjz22GP06tWL+Pj421rH/fffz6FDh3jhhRdISUmhb9++DB48mHXr1l133/vuuy/HuoiICN5//32ysrIYOHAg586do379+sybNw9/f38AXFxceOWVVzh8+DDu7u60aNGCyZMnA+Dl5cUHH3zA/v37cXR0pEGDBsyePRsHh/z1ZCuLcSuDLe8gCQkJ+Pr6Eh8fj4+Pj73LKZwMw5y0wxrCDv1niTCD2NU4OF8Yklje7HUrXg+K1wfvoNtXv4iIiNwWKSkpREREUKZMGdzcNOrFXtq3b09wcDA///yzvUvJE9f6nOU2G6grQPIPi+XCc7+KQamm2bcZhjmbYmzEpd6w0/svBLMDkJECp/eay95Zl/bzDTODV4n6ZvgKqQUuHrf3ukREREQKuKSkJL755hs6duyIo6Mjv/32GwsXLmTBggX2Li1fU9iSgsFiufAw5qIQ1jD7tqwsc8KOMwfg9AE4sR2ObICTuyE+2lx2Tb9wHEcIrm4GrxINzCWwnKapFxEREbkGi8XC7Nmz+d///kdqaiqVKlVi6tSptGvXzt6l5WsKW1LwOThcmkmx3F2X1qeeg2ObzeB1dCMcWW8+X+z4VnPZ8IPZzs3vPz1fNSG4htkjpgAmIiIiAoC7uzsLFy60dxkFjsKWFF6u3lCmpbmAORQx/ggc3WAGsCMb4PgWSImDAwvN5SI3XwiqYfaCBdcwl6JVzIdKi4iIiIjkgsKW3DksFvALM5dqd5vrMtPhxI4LvV+bzCGIJ/eYD22OXGEuFzm6muErtA6E1jW/Fq0EDrZ7aryIiIiIFB4KW3Jnc3S+EJ7qXFqXkQan90HMdnM5sR2ObzN7wI5uNJeLnD3NSTdK1DfvJSvRUDMgioiIiAigsCWSk5PLheGD1YH+5jrDMGdCPLbZ7AE7ttm87ystEaJWmctF/qXN0BXWEMIaQVA19X6JiIiI3IEUtkRyw2Ixn+MVUBaq9zHXZWWa089fnHwjeh2c3AWxh81l++9mO/cAKN8OKnSA8m3BI8BeVyEiIiIit5HClsjNcnA0H6BcrDLUud9clxJvhq/odeZyZD0knzWD1/bfweJgTjdfoYO5BNfQrIciIiIihZTClogtufma089fnII+M8MMXPvnwf4F5mQc0WvNZfE74OIFRSpAkYqXlqKVwL+MZj4UERGRK2rdujW1a9fm008/BaB06dKEh4cTHh5+1X0sFgvTpk2jV69et3RuWx3nTqGwJZKXHJ2gVBNzaTfKnHp+/wJzObTUvOfr2GZz+S+LI4TWhlr9zWGLGnooIiJS4HXv3p3k5OQrPq9q9erVNG3alI0bN1K3bt0bOu769evx9PS0VZkAjBo1iunTp7Nly5Zs648fP46/v79Nz3W5CRMmEB4eTlxcXJ6e53ZwsOfJ//33X7p3705oaCgWi4Xp06dbt6WnpzNixAhq1KiBp6cnoaGhDBo0iGPHjmU7RmpqKsOGDaNIkSJ4enrSo0cPjhw5kq1NbGwsAwcOxNfXF19fXwYOHFgovnlSAPmWgPoPQf9J8HIkPLUO+v0Cd42EmveZsyK6eIGRaQ5HnP0CfFIJ/hhsBrSsTHtfgYiIiNykIUOGsHjxYiIjI3Ns+/HHH6ldu/YNBy2AokWL4uHhYYsSrys4OBhXV9fbcq7CwK5h6/z589SqVYtx48bl2JaUlMSmTZsYOXIkmzZt4q+//mLfvn306NEjW7vw8HCmTZvG5MmTWbFiBYmJiXTr1o3MzEs/lA4YMIAtW7Ywd+5c5s6dy5YtWxg4cGCeX5/INTk6m0MGq3SHli9A72/hsaXwyhEI3w4dR0NQdchMg53T4Nd7YGw1WPAmnNhpzpAoIiIiJsOAtPP2WXL5f3K3bt0oVqwYEyZMyLY+KSmJKVOmMGTIEM6cOUP//v0pUaIEHh4e1KhRg99+++2axy1durR1SCHA/v37admyJW5ublStWpUFCxbk2GfEiBFUrFgRDw8PypYty8iRI0lPTwfMnqW33nqLrVu3YrFYsFgs1pov7yDZvn07d911F+7u7gQGBvLYY4+RmJho3T548GB69erFxx9/TEhICIGBgTz11FPWc92MqKgoevbsiZeXFz4+PvTt25cTJ05Yt2/dupU2bdrg7e2Nj48P9erVY8OGDQBERkbSvXt3/P398fT0pFq1asyePfuma7keuw4j7Ny5M507d77iNl9f3xwfjC+++IKGDRsSFRVFyZIliY+P54cffuDnn3+mXbt2APzyyy+EhYWxcOFCOnbsyO7du5k7dy5r1qyhUaNGAHz//fc0adKEvXv3UqlSpby9SJEbZbGAX0loMtRcjm+FLZNg2+9w7jis/NRcPAKhVFMo1cz8GlRdU8yLiMidKz0J3gu1z7lfPQYu1x/G5+TkxKBBg5gwYQJvvPEGlguTZP3xxx+kpaVx//33k5SURL169RgxYgQ+Pj7MmjWLgQMHUrZsWevPsteSlZVF7969KVKkCGvWrCEhIeGK93J5e3szYcIEQkND2b59O48++ije3t689NJL9OvXjx07djB37lzrkEdfX98cx0hKSqJTp040btyY9evXc/LkSR555BGefvrpbIFyyZIlhISEsGTJEg4cOEC/fv2oXbs2jz766HWv53KGYdCrVy88PT1ZtmwZGRkZDB06lH79+rF06VIA7r//furUqcPXX3+No6MjW7ZswdnZGYCnnnqKtLQ0/v33Xzw9Pdm1axdeXl43XEduFah7tuLj47FYLPj5+QGwceNG0tPT6dChg7VNaGgo1atXZ9WqVXTs2JHVq1fj6+ub7cPZuHFjfH19WbVq1VXDVmpqKqmpqdb3CQkJeXNRItcTUstc2r9jTrSxZRIcXAJJZ2D3P+YC4OoLJRtDmZZQ7W7wLW7fukVERCSHhx9+mI8++oilS5fSpk0bwBxC2Lt3b/z9/fH39+eFF16wth82bBhz587ljz/+yFXYWrhwIbt37+bw4cOUKFECgPfeey9HB8frr79ufV26dGmef/55pkyZwksvvYS7uzteXl44OTkRHBx81XP9+uuvJCcn89NPP1nvGRs3bhzdu3fngw8+ICgoCAB/f3/GjRuHo6MjlStXpmvXrixatOimwtbChQvZtm0bERERhIWFAfDzzz9TrVo11q9fT4MGDYiKiuLFF1+kcuXKAFSoUMG6f1RUFH369KFGjRoAlC1b9oZruBEFJmylpKTw8ssvM2DAAHx8fACIiYnBxcUlx016QUFBxMTEWNsUK1Ysx/GKFStmbXMlo0eP5q233rLhFYjcIicXc8hhle6QkQbHt0DkSji8EqLWQGr8hVkP58H8183QVes+s72rt72rFxERyVvOHmYPk73OnUuVK1emadOm/Pjjj7Rp04aDBw+yfPly5s+fD0BmZibvv/8+U6ZM4ejRo9YOgNxOgLF7925KlixpDVoATZo0ydHuzz//5NNPP+XAgQMkJiaSkZFh/Rk7t3bv3k2tWrWy1dasWTOysrLYu3evNWxVq1YNR8dLo29CQkLYvn37DZ3rv+cMCwuzBi2AqlWr4ufnx+7du2nQoAHPPfccjzzyiHX027333ku5cuUAeOaZZ3jyySeZP38+7dq1o0+fPtSsWfOmaskNu96zlVvp6encd999ZGVl8dVXX123vWEY1m5ZINvrq7W53CuvvEJ8fLx1iY6OvrniRfKCkwuENYTmw+GBP83JNh5bBh3fM4cVYkDEMpj+JHxUAaY+AvsXmlPRi4iIFEYWizmUzx7LDT4zc8iQIUydOpWEhATGjx9PqVKlaNu2LQCffPIJY8eO5aWXXmLx4sVs2bKFjh07kpaWlqtjG1e4f+zyn3nXrFnDfffdR+fOnZk5cyabN2/mtddey/U5/nuuq/08/d/1F4fw/XdbVlbWDZ3reuf87/pRo0axc+dOunbtyuLFi6latSrTpk0D4JFHHuHQoUMMHDiQ7du3U79+fb744oubqiU38n3PVnp6On379iUiIoLFixdnS9zBwcGkpaURGxubrXfr5MmTNG3a1NrmvzfMXXTq1Clr2r4SV1dXzbQiBYfDhaniQ2tDk6cg9jBs+wO2TYYzB2D7H+bi6nPhPwUHc3p5i8V87eAIzu5QuRvUfRB8Qux8QSIiIoVX3759efbZZ5k0aRITJ07k0UcftQaF5cuX07NnTx544AHAvAdr//79VKlSJVfHrlq1KlFRURw7dozQUPMettWrV2drs3LlSkqVKsVrr71mXXf5DIkuLi7ZJpy72rkmTpzI+fPnrb1bK1euxMHBgYoVK+aq3ht18fqio6OtvVu7du0iPj4+259RxYoVqVixIsOHD6d///6MHz+eu+++G4CwsDCeeOIJnnjiCV555RW+//57hg0blif15uuerYtBa//+/SxcuJDAwMBs2+vVq4ezs3O2iTSOHz/Ojh07rGGrSZMmxMfHs27dOmubtWvXEh8fb20jUuj4l4ZWL8LTG+CRRdDgUXAPgNQEc5KNhKMQHwVxkRAbYQaymO2wdLQ54+GUB8z7wm7yt04iIiJydV5eXvTr149XX32VY8eOMXjwYOu28uXLs2DBAlatWsXu3bt5/PHHr3nry+XatWtHpUqVGDRoEFu3bmX58uXZQtXFc0RFRTF58mQOHjzI559/bu35uah06dJERESwZcsWTp8+nW0ug4vuv/9+3NzcePDBB9mxYwdLlixh2LBhDBw48JqdGrmRmZnJli1bsi27du2iXbt21KxZk/vvv59Nmzaxbt06Bg0aRKtWrahfvz7Jyck8/fTTLF26lMjISFauXMn69eutQSw8PJx58+YRERHBpk2bWLx4ca6D7M2wa89WYmIiBw4csL6/+A0NCAggNDSUe+65h02bNjFz5kwyMzOtH7SAgABcXFzw9fVlyJAhPP/88wQGBhIQEMALL7xAjRo1rLMTVqlShU6dOvHoo4/y7bffAvDYY4/RrVs3zUQohZ/FAiXqm0vH98xQlZUORpYZpIyLSybERcHGiRC16tLEGwHlzOeC1b5fD1YWERGxoSFDhvDDDz/QoUMHSpYsaV0/cuRIIiIi6NixIx4eHjz22GP06tWL+Pj4XB3XwcGBadOmMWTIEBo2bEjp0qX5/PPP6dSpk7VNz549GT58OE8//TSpqal07dqVkSNHMmrUKGubPn368Ndff9GmTRvi4uIYP358tlAI4OHhwbx583j22Wdp0KABHh4e9OnThzFjxtzSnw2YOaFOnTrZ1pUqVYrDhw8zffp0hg0bRsuWLXFwcKBTp07WoYCOjo6cOXOGQYMGceLECYoUKULv3r2tczFkZmby1FNPceTIEXx8fOjUqRNjx4695XqvxmJcaWDnbfLfWVj+68EHH2TUqFGUKVPmivstWbKE1q1bA+bEGS+++CKTJk0iOTmZtm3b8tVXX2W7ae7s2bM888wzzJgxA4AePXowbtw466yGuZGQkICvry/x8fE3fPOgSIFyYhds+BG2Toa0c+Y6R1eo3AVq9IXy7cx7xkREROwkJSWFiIgIypQpg5ubm73LkULqWp+z3GYDu4atgkRhS+44qYmw409Y/wPEbLu03t3fnFq+Rl8IawQO+Xo0soiIFEIKW3I72CJs5fsJMkTETly9oN5gc8KM41vMCTd2/AmJJ8yerw0/gm9JqHEP1HvQvE9MRERERKz0K2kRuTaLBULrQKf34LndMHA61BoALt7mJBsrxsDndeGvx+HkHntXKyIiIpJvKGyJSO45OEK5NnD31/DifrhnPJRtY06wsW0yfNXYnMnw2GZ7VyoiIiJidwpbInJznN2hem8YNB0eXWI+owvDnMXwu9bwc2+IXGXnIkVEpDDT1AOSl2zx+VLYEpFbV7wu3PcrDF0DNfuZD0w+uAjGd4bv28KWSZCebO8qRUSkkHB2dgYgKSnJzpVIYXbx83Xx83YzNBthLmk2QpEbcDYCVn5qhqzMNHOdm5/5vK76D0OR8vasTkRECoHjx48TFxdHsWLF8PDwwGKx2LskKSQMwyApKYmTJ0/i5+dHSEhIjjaa+t3GFLZEbkLiSdj0k/mw5PioS+vLtIT6Q6ByV3C8+d8WiYjIncswDGJiYoiLi7N3KVJI+fn5ERwcfMUgr7BlYwpbIrcgKxMOLDSni983D7jwz46rD5RuDmVbQ5lWULSSOfuhiIhILmVmZpKenm7vMqSQcXZ2xtHR8arbFbZsTGFLxEbiosyerk0/wfmT2bd5BUPZVmb4KtcWvIPsUqKIiIjItShs2ZjCloiNZWVCzDY4tBQOLYOo1ZCRcmm7gxNU6w1NnoLQ2vaqUkRERCQHhS0bU9gSyWPpKXBknRm+DiyC41subSvdwgxdFTqCgyZRFREREftS2LKx/BK2UtIzeeznjRiGgWFAlmFcWLCuc7BYcHV2wNXJEVdnB9ycHHG78N7bzYkS/u6EBXgQFuBBsI8bjg66R0byoWObYfWXsHMaZGWY6wLLQ+OhUKs/uHjYtz4RERG5Yyls2Vh+CluVR8612fGcHS2E+rlTMsCDEv4ehPq6EeLnTqifG6G+7gT7uuHmfPWbA0XyXPwRWPcdbJgAqfHmOldfqNINqvYy7+9ycrFjgSIiInKnUdiysfwStjKzDKZvPoqDg9mDZbFYcLBceA1YLBayDIPUjExS0rNISTe/Xnwfn5xG9NlkomOTOBqbTEbW9b/9gZ4uBPu64enqhIeLIx4ujrg7X3rt4eKEv6czAZ4uBHi4EODlQoCnC/4eLjg7asiX2EhqImz5FdZ8BbGHL61384XK3aFaL3NGQwUvERERyWMKWzaWX8KWLWVmGcQkpBB9Nonos0kciU3meHwyx+NTOBqXzPG4FJLTM2/pHL7uzpTwd6d0EU/KFvGkdKAnZYqar/089EOx3ISsLIheAzunw67pkHji0jY3P6jaA5qFQ2A5+9QnIiIihZ7Clo0VxrB1PYZhEJ+cztG4ZE4mpHI+LYOktEyS0zIvfDXfn0/LJC4pjTPn0zh7YYlNSuN6nyw/D2fC/D0I9nUjxNft0lcfd0J83Qjxc8PVSUMY5RqyMiFqjXlf166/L00lb3GEugOh1QjwCbVvjSIiIlLoKGzZ2J0Ytm5FZpZBXFIapxPTiDqbRMTpRCJOJ3H49HkiTp8nJiHlusdwsEBxf3ezN6yIuZQu4kmZQE9K+LvjpCGK8l9ZmRC5ElZ+DgcWmOuc3KDho9D8OfAIsG99IiIiUmgobNmYwpZtJaVlcPh0knXY4sWvMReWY/HJpKRnXXV/F0cHyhb1pHwxLyoU86ZikBcVgrwoFeip+8QEIlfBwrfM4YYArj7QdJg5k6Grl31rExERkQJPYcvGFLZuL8MwOJWYyuHTOXvFDp85T2rGlYOYk4OFkoHmlPZBF5ZgH1fz9YVhikHebjhouvvCzzBg/wJY9Dac2G6u8yhiPq+rwRBzYg0RERGRm6CwZWMKW/lHVpbB0bhk9p88x/4Tiew/mcj+E+fYfzKRpLTrT+jh5uxA6UBPyhb1tA5RLFvUk7JFvPD31KQdhU5WFuyaBovfhbMHzXWuvtDwEWj0JHgVtW99IiIiUuAobNmYwlb+l5VlcDwhhcjT5zlxLoWY+FROJKRwIiGFmIQUTiakEpOQQuY1prsvV9STxmUDaVw2kEZlAyjm7XYbr0DyVGY67JgKK8bCqT3mOid3qDvIHGLoF2bf+kRERKTAUNiyMYWtwiE9M4sjsck5hiZGnD7PsfjkHDMolr0QvhqVCaBSsDelAz31kOeCLisL9s6GFWPg6EZznYMT1OgLLZ6DIhXsW5+IiIjkewpbNqawVfjFJ6Wz7vBZ1hw6w5pDZ9h1POGK09eH+rpR5rIhiDVL+FHEy/X2Fy03zzAgYhksH2N+BcAC1e6Gli9CUFW7liciIiL5l8KWjSls3Xn+G742RMZy6FQi51Iyrtq+XFFPGpYJuLAEUtzP/TZWK7fkyEZY/gnsnXVpXeVuZugKrW23skRERCR/UtiyMYUtMQyDs+fTOHzmPIdOmbMiRpw+z4GTiew7kZijfXE/dxqVCaB2ST+qhfpSJcQbDxcnO1QuuRazA/79yHxAMhf+aazQAVq+BGEN7FqaiIiI5B8KWzamsCXXEpeUxobDsaw7fJa1EWfZcTQ+x0QcDhYoU8ST6sV9qRbqQ/VQX6oV98XX3dlOVctVndpr9nRt/wOMC48ZCCwPZVpB2VZQuoUekiwiInIHU9iyMYUtuRHnUzPYFBXL+oizbD8az85jCZw8l3rFtmWLeFKzhC+1wvyoFeZH1RAfTcKRX5w5aM5euPU3yPrvEFILhNS8EL5aQ+nm4KR79kRERO4UCls2prAlt+rkuRR2Hktg54Xwtf1oPEdik3O0c3KwUDnEmzph/tQv7U/DMgGE+Or+L7tKjoPIlXBomTmZxsWp4y/yKAL1H4L6Q8AnxC4lioiIyO2jsGVjCluSF86eT2PbkTi2Rsez9UgcW6PjOHM+LUe74n7uNCwTQIPSATQo7U/5Yl5YLBY7VCwAnIuBiH/N8HVwEZw7bq53cIKqvaDRE7rHS0REpBBT2LIxhS25HQzD4GhcMlui49gUGcf6w2fZeSyey5/D7OfhTP1S/tQrZYav6sV9NfTQXjIzzFkM13wDUasurS9ezwxdVXtqiKGIiEgho7BlYwpbYi+JqRlsvnD/1/rDsWyOjiUlPStbGxdHB2qU8KV+aX/qlwqgbkk/AvXcr9vv+FZY+505sUbmhXv03PygSneo3htKtwRHzUgpIiJS0Cls2ZjCluQXaRlZ7DqewIbDZ9lwOJYNkWc5nZhz6GGZIp7ULelPvVLmUqGYFw4OGnp4W5w/DRvHw/of4dyxS+s9ipg9XdV7Q8mm4OBgvxpFRETkpils2ZjCluRXhmEQeSaJDZGxbDh8lo2Rsew/mfO5X95uTtQp6U+zcoE0K1+EqiE+Cl95LSsTIlfBjqmwewYknbm0zTsE6g2GFi+ot0tERKSAUdiyMYUtKUjik9LZFB3LpshYNkbGsiU6jqS0zGxt/D2caVq+CC3KF6FZ+SKEBXjYqdo7RGa6OanGjr9gzz+QEm+uL9Uc7vkRvIPsW5+IiIjkmsKWjSlsSUGWkZnFnphzrIs4y8oDp1lz6AznLwtfpQI9aFcliM7Vg6lb0l+9XnkpIxV2ToNZL0DaOfAKhr4ToWRje1cmIiIiuaCwZWMKW1KYpGdmsTU6jhUHTrNi/2k2R8eR+Z8pD4t6u9KhahCdq4fQqGwAzo66tyhPnN4PUx4wn9vl4AQd34OGj4Gm9RcREcnXFLZsTGFLCrNzKemsPHCaeTtPsHD3Cc6lZFi3+bo7065KEC0rFqFJ2UCK+bjZsdJCKDURZgyDnX+Z72vcC90/AxdP+9YlIiIiV6WwZWMKW3KnSMvIYtXB08zbGcP8nSdyPGS5bFFPGpcNvLAEUMxb4euWGQas+Rrmvw5GJhSrCn1/hiLl7V2ZiIiIXIHClo0pbMmdKDPLYMPhsyzYdYI1EWfYeSyBy//FKFfUkyblAmlWrgiNywbi7+lin2ILg8hV8MdgSDwBWCCoGpRqCiWbmF+9g+1doYiIiKCwZXMKWyLmLIfrDp9lzaEzrD54ht0x2cOXxQJVQ3xoWi6QpuWL0LB0AJ6umtb8hpyLgb8eg4hlObcFlL0UvorXhyIV9awuERERO1DYsjGFLZGc4pLSWBtxltUHz7Dq4Gn2ncj+fC8nBwuNygbQvWYonaoH4+ehXq9cO3cColZB5Gqzx+vEDuCyf65dvKF4HShezwxfxeuBT4hdyhUREbmTKGzZmMKWyPWdPJdiBq8DZ1h16DTRZ5Ot25wdLbSsUJTutUJpVzUIL/V43ZjkOIheB5Er4ch6OLYZ0pNytgssD33+D0Lr3PYSRURE7hQKWzamsCVy4w6fPs+s7cf5Z+sx9sScs653dXKgbZVidKkRQutKxRS8bkZmhjll/NENcHQjHNkIp3aDkQUeRWDIfAgsZ+8qRURECiWFLRtT2BK5NftPnOOfbWbwijh93rrexcmB5uWL0KFqEO2qBlHEy9WOVRZwSWfhp54Qsw38S8OQBeBVzN5ViYiIFDoKWzamsCViG4ZhsPNYAv9sO8a8HTEcPnNpKJyDBeqXCqBDtSA6VgsmLMDDjpUWUOdOwA/tIS4SQmrB4Fng6m3vqkRERAoVhS0bU9gSsT3DMNh/MpF5O2KYv+sE24/GZ9tes4QvXWqE0LVGiILXjThzEH7oAEmnoWxrGPAHOGlyEhEREVtR2LIxhS2RvHc0LpkFO2OYt/MEayPOkPWff51qlvCla40Quih45c7RTTChG6Sfh+r3QO/vNU28iIiIjShs2ZjClsjtdToxlbk7Ypi9/ThrDuUMXu2rmPd4VQ72xmKx2K/Q/OzAIpjUF7IyoPFQ6Pie+TA0ERERuSUKWzamsCViP6fOpTJvZwyzth3P0eNVwt+ddlWCaF81iIZlAnB2VO9NNtt+h78eNV+3GwVNn7mw4ULoslgUwERERG6QwpaNKWyJ5A+nzqWyaPcJFu4+wfL9p0nNyLJu83Zz4q7KxehTtwTNyhfB0UEhAoBV42D+a9du4+QOrV+G5uG3pSQREZGCTGHLxhS2RPKfpLQMVuw/zcLdJ1i0+yRnzqdZt4X6utGnXgnurRdGyUDd48Xid2H5J2BkXrtdq5fN0KXeLhERkavKbTaw63ibf//9l+7duxMaGorFYmH69OnZthuGwahRowgNDcXd3Z3WrVuzc+fObG1SU1MZNmwYRYoUwdPTkx49enDkyJFsbWJjYxk4cCC+vr74+voycOBA4uLi8vjqRCSvebg40aFaMB/eU4t1r7Xjzyea8GCTUvi6O3MsPoUvFh+g5UdLuO+71fy16QjJadcJGoXZXa/By1HwUoS5vHjownLQXNqNMtstex+WvAv6PZyIiMgts2vYOn/+PLVq1WLcuHFX3P7hhx8yZswYxo0bx/r16wkODqZ9+/acO3fO2iY8PJxp06YxefJkVqxYQWJiIt26dSMz89IPVQMGDGDLli3MnTuXuXPnsmXLFgYOHJjn1ycit4+jg4X6pQN4q2d11r7ali/616FFhSJYLLDm0Fme+30rDd5dyIt/bGXlgdNkZt2BYcLVCzwCzMUz8MJSxFyaD4cO75rt/v0IFr2lwCUiInKL8s0wQovFwrRp0+jVqxdg9mqFhoYSHh7OiBEjALMXKygoiA8++IDHH3+c+Ph4ihYtys8//0y/fv0AOHbsGGFhYcyePZuOHTuye/duqlatypo1a2jUqBEAa9asoUmTJuzZs4dKlSrlqj4NIxQpmI7GJTN14xH+2BhN9Nlk6/ogH1d61AqlZ+3iVAv10YyGF635Gua+bL5u+gy0f1tDCkVERC5TIIYRXktERAQxMTF06NDBus7V1ZVWrVqxatUqADZu3Eh6enq2NqGhoVSvXt3aZvXq1fj6+lqDFkDjxo3x9fW1trmS1NRUEhISsi0iUvAU93PnmbYVWPZCG35/vAkDGpXE192ZEwmpfL88gm5frKDD2H/5cskBTiSk2Ltc+2v8JHT52Hy96nOY95p6uERERG5Svg1bMTExAAQFBWVbHxQUZN0WExODi4sL/v7+12xTrFixHMcvVqyYtc2VjB492nqPl6+vL2FhYbd0PSJiXw4OFhqWCeC9u2uw7rW2fDewHl1qBOPi5MD+k4l8NG8vTd9fzBM/b2TlgdPkk05/+2j4KHQdY75e8yXMGaHAJSIichOc7F3A9Vw+tMcwjOsO97m8zZXaX+84r7zyCs8995z1fUJCggKXSCHh6uRIh2rBdKgWTEJKOnO3x/DnxiOsO3yWuTtjmLszhrJFPBnQqCT31gvD18PZ3iXffg2GgIMT/PMsrPsWTuyE+g9B5W7g7Gbv6kRERAqEfBu2goODAbNnKiQkxLr+5MmT1t6u4OBg0tLSiI2Nzda7dfLkSZo2bWptc+LEiRzHP3XqVI5es/9ydXXF1dXVJtciIvmXj5szfRuE0bdBGHtjzvHLmkimbT7KodPn+d+s3Xw0by/da4XSu05xGpYJwOlOemhyvQfBwdEMXJErzMXdH2r1h7qDoFgVe1coIiKSr+XbnxrKlClDcHAwCxYssK5LS0tj2bJl1iBVr149nJ2ds7U5fvw4O3bssLZp0qQJ8fHxrFu3ztpm7dq1xMfHW9uIiABUCvbmnV7VWfNqW969uzqVg71Jzcjiz41HGPB/a2nw7kJe+nMrS/acJDXjDplGvs4D8Mxm8/lbPiUgORbWfAVfNYb/awebfoa08/auUkREJF+y62yEiYmJHDhwAIA6deowZswY2rRpQ0BAACVLluSDDz5g9OjRjB8/ngoVKvDee++xdOlS9u7di7e3NwBPPvkkM2fOZMKECQQEBPDCCy9w5swZNm7ciKOjIwCdO3fm2LFjfPvttwA89thjlCpVin/++SfXtWo2QpE7j2EYbIqKZcr6aBbsOkFsUrp1m5erE3dVLkbn6sG0qVwMN2dHO1Z6m2RlwsHFsGki7J0DWRnmer+S0OsbKN3MvvWJiIjcJrnNBnYNW0uXLqVNmzY51j/44INMmDABwzB46623+Pbbb4mNjaVRo0Z8+eWXVK9e3do2JSWFF198kUmTJpGcnEzbtm356quvst1fdfbsWZ555hlmzJgBQI8ePRg3bhx+fn65rlVhS+TOlpGZxbqIC/d07Yjh5LlU6zY/D2f6NyzJwMalCPVzt2OVt9G5E7D1N1j3PSQcASzQdBjc9To4aQi2iIgUbgUibBUkClsiclFWlsHm6Djm7Yxh1rbjHI0zn9/l6GChU7VgHmpWmnql/O+MZ3elnoO5r8Dmn833xapB7+8guPq19xMRESnAFLZsTGFLRK4kM8tg4e4TjF8ZwZpDZ63raxT3ZXDT0nSrFYKr0x0wxHDPLJjxDCSdBkcXaPOa2dPlcAdcu4iI3HEUtmxMYUtErmfXsQQmrjrMtC1HScvIAsDfw5nedUvQr0EYFYO87VxhHks8Bf88A3tnm+9LNoWe4yCwnH3rEhERsTGFLRtT2BKR3Dp7Po3f1kXx8+pIYhJSrOvrlPSjX/0wutUKxcs13z5549YYBmz+Bea+DGmJgAXK3QV1B0KlLrqfS0RECgWFLRtT2BKRG5WRmcWyfaeYsj6axXtOkpFl/nPr4eJI95qh9G0QRt2SfoXz3q6zETDreTi46NI69wCodR/UGQhBVe1Xm4iIyC1S2LIxhS0RuRUnz6Xw16aj/L4+mkOnLz2XqlKQN/c1DOPuOsXx83CxY4V55Owhs6dryyQ4d/zS+uL1zAcjV+8DroV8eKWIiBQ6Cls2prAlIrZgGAbrD8cyeX0Us7cfJyXdvLfLxcmBrjVCuK9BGA3LBBS+3q7MDLOXa9NPsG/upWd0uXhBjXuh/kMQUsu+NYqIiOSSwpaNKWyJiK3FJ6fz95aj/LYumt3HE6zryxb1pF/9MO6uW5xi3m52rDCPJJ6ErZPNhyOfOXBpfWhdM3RV7wMunvarT0RE5DoUtmxMYUtE8ophGGw7Es9v66KYsfUYSWmZgPncrlYVi9KnbgnaVimGm3Mhm0bdMODwctgwHnb/A1np5npXH6h9P7R5Fdz0762IiOQ/Cls2prAlIrdDYmoGM7Yc48+N0WyKirOu93V3pketUO6pV4KaJXwL3zDD86dhy6+wcYJ5nxeAb0m4+2so3dyupYmIiFxOYcvGFLZE5HY7eCqRqRuP8Nemo9mmkK8U5M2wtuXpUj0EB4dCFrqysuDgYpj9PMQeBizQ5Cm4ayQ4F8IhlSIiUiApbNmYwpaI2EtmlsGqg6f5c+MR5u6IIfXCA5OrhvjwYsdKtK5UtPD1dKWeg3mvmfd1ARStAr2/1SQaIiKSLyhs2ZjClojkB/HJ6YxfGcH/LY8gMdWc0a9+KX9e7FiJRmUD7VxdHtg7F2YMg/MnwcEZWr8MzcLBsZA+FFpERAoEhS0bU9gSkfzk7Pk0vll2kImrDlt7ulpWLMoLHSpSs4SffYuztfOn4Z9nYc9M832JhtD7OwgoY9+6RETkjqWwZWMKWyKSH8XEp/DF4v1MWR9NRpb5z3nz8kUY0rwMrSoWLTz3dBmGOV38nJcgNQFcvKHrx1CzHxS2IZQiIpLvKWzZmMKWiORnkWfO8+nC/fy95SgXMhdli3rycLMy9KlbAneXQjJtfFwU/PUYRK0231e/B7qNATdf+9YlIiJ3FIUtG1PYEpGCIPpsEhNXHWbK+mjOXbiny8/DmQENSzKoSWmCfQvBjH5ZmbB8DCwdDUamOUV8n++hZGN7VyYiIncIhS0bU9gSkYIkMTWD39dHM35VBNFnkwFwcrDQtWYIDzUrQ+0wP/sWaAvR62HqEIiLBIsDtHwRWr6kyTNERCTPKWzZmMKWiBREmVkGC3ad4MeVEayLOGtdX6ekHw81K0Pn6sE4OzrYscJblJIAs1+EbZPN9yUamlPEB5S1b10iIlKoKWzZmMKWiBR0O47G8+PKCGZuPU5apjmDYbCPGwOblKJ/w5IEeLrYucJbsP1PmDncnDzD2RM6vgv1BmvyDBERyRMKWzamsCUihcWpc6n8ujaSX9ZEcToxFQBXJwd61g7lwaalqRZaQCebiI2E6UMhcoX5vkIH6PEFeAfbty4RESl0FLZsTGFLRAqb1IxMZm07zviVh9l+NN66vkFpfx5sWpqO1QrgEMOsLFjzFSx6GzJTwd0fuo2FanfbuzIRESlEFLZsTGFLRAorwzDYFBXLhFWRzNl+3Pq8riAfV+5vZA4xLOrtaucqb9DJ3eYU8THbzPc17oUuH5nhS0RE5BYpbNmYwpaI3AlOJKTw69ooJq2N5HRiGgAujg7cXac4Q9uUo1Sgp50rvAEZafDvh7D8EzCywDsUOv4Pqt4NDgWsx05ERPIVhS0bU9gSkTtJakYms7cfZ8KqSLZGxwHg6GChZ+1QnmpTnnJFvexb4I2IXg/THoezB833wTWh3ZtQrq0m0BARkZuisGVjClsicqfacPgsXyw+wLJ9pwAzn3SrGcqwu8pTMcjbztXlUtp5WP0lrPwc0s6Z60q3gLZvQlgD+9YmIiIFjsKWjSlsicidbmt0HF8sPsDC3Ses6zpVC+bZdhWoElJA/l08fwZWjIF130GmOUySyt3grpFQrLJ9axMRkQJDYcvGFLZEREw7j8UzbvEB5uyIAcyert51SvBch4oU93O3c3W5FBcNS9+HrZPM+7ksDtDoCbOny9nN3tWJiEg+p7BlYwpbIiLZ7Ttxjs8W7mfW9uMAuDg58FCz0gxtVR5fD2c7V5dLJ/fA4ndgz0zzfdEq0Od7CK5h37pERCRfU9iyMYUtEZEr2xIdx+jZu1kbcRYAX3dnnm5TnoFNSuHm7Gjn6nJp3zz4+yk4fwocXcxhhU2e1qyFIiJyRQpbNqawJSJydYZhsGTvSd6fs4d9JxIBKO7nTni7CvSsXRwXpwIQWs6fhhnPwN5Z5vvSLeDub8C3hH3rEhGRfEdhy8YUtkREri8zy2DqxiOMWbCPmIQUAIJ93BjSvAz3NQzD2y2fDy80DNj0E8x9GdKTwM0Xuo6BGvfYuzIREclHFLZsTGFLRCT3ktMymbj6MD+siODUuVQAvN2ceKBxKR5qWppiPvl8EoozB+Gvx+DoBvN9cA0o1RxKNTUXzyL2rU9EROxKYcvGFLZERG5cakYmf28+xrf/HuTgqfMAuDg6cHed4jzasizli+XjhyNnZsDyj2HZh2BkZt9WpNKl4FWpM7gWkOeNiYiITShs2ZjClojIzcvKMli05yTfLjvIhshYABws0LN2cZ5pW4EyRTztXOE1nDsBkSsgchVEroaTO7NvL1oFHluqKeNFRO4gCls2prAlImIbGyPP8vXSQ9aHIzs6WOhdxwxdYQEedq4uF5LOQtQaiFwJWyZB8llo9TK0ecXelYmIyG2isGVjClsiIra142g8YxfsY9GekwA4OVi4t34YT99VvuA8HHnHVPjzYXO6+CdXQZEK9q5IRERuA4UtG1PYEhHJG5ujYhm7cD//7jsFmPd03dcwjKfblM//E2kYBvzSBw4ugjItYdAMsFjsXZWIiOQxhS0bU9gSEclb6w+fZeyCfaw6eAYAN2cHBjctwxOtyuLn4WLn6q7hbAR81RgyUuDub6HWffauSERE8pjClo0pbImI3B6rD57ho3l72BQVB5hTxj/esiwPNSuDp6uTfYu7mn8/hsXvgEcReHo9eATYuyIREclDCls2prAlInL7GIbB4j0n+WjeXvbEnAOgiJcLT7Upz4BGJXF1crRzhZfJSINvW8CpPVB3EPT4wt4ViYhIHlLYsjGFLRGR2y8ry+CfbccYs2AfkWeSACju586zbSvQu25xnBwd7Fzhf0SuhvGdzNcPzYVSTexbj4iI5BmFLRtT2BIRsZ/0zCz+2HCEzxbt40RCKgBli3gS3r4i3WqE4OCQTyal+Ptp2Pyz+eytJ5aDo7O9KxIRkTygsGVjClsiIvaXkp7JL2si+XLJAWKT0gGoHOzNCx0q0bZKMSz2ngkw6SyMqw9JZ6DdKGg+3L71iIhInlDYsjGFLRGR/CMxNYMfV0Tw/b+HOJeaAUCdkn680KESzcoXsW9xW36D6U+Akzs8tQb8S0NWJiTHwvnTkHTa/BpQBkJq2bdWERG5KQpbNqawJSKS/8QlpfHNskNMWBVBSnoWAI3KBDC8fUUalw20T1GGARO7w+Hl4H5hVsLkWOAK/922ehlajQCHfHTvmYiIXJfClo0pbImI5F8nE1L4cskBflsXTVqmGboalw1geLuKNLJH6Dq9H75pbj5767/c/MCzCLh4wfEt5rrK3cznc7l63e4qRUTkJils2ZjClohI/ncsLpmvlx5kyvpLoatJ2UCGt69IwzK3+dlXZw5CwlHz2VueRcDdP/uEGZt/hZnhkJkGxapB/0nmkEMREcn3FLZsTGFLRKTgOBaXzFdLDzBlfTTpmeZ/c03LBTKiU2VqhfnZt7j/il4HUx6AxBPmkMO+E6FMS3tXJSIi16GwZWMKWyIiBc/RuGS+WnKA3zdcCl296xZnRKfKBPm42bm6CxKOweQBcGwzWByh8wfQ4BGw98yKIiJyVQpbNqawJSJScB2JTWLMgn38tekoAO7OjgxtXY5HW5bFzdnRztUB6ckw4xnY/rv5vlZ/aPAohNbR5BkiIvmQwpaNKWyJiBR8W6LjePufnWyKigOguJ87L3euTLeaIfZ/RpdhwMrPYOEorDMXehaF8u2hYgcodxe4+dqzQhERuSC32SBf/7osIyOD119/nTJlyuDu7k7ZsmV5++23ycrKsrYxDINRo0YRGhqKu7s7rVu3ZufOndmOk5qayrBhwyhSpAienp706NGDI0eO3O7LERERO6sd5sfUJ5vy2X21CfF142hcMsN+28y936xmS3ScfYuzWKB5ODw4A6r2BFcfOH8Ktk6CPwbDh2VhQjdY/wP85/9BERHJv/J12Prggw/45ptvGDduHLt37+bDDz/ko48+4osvvrC2+fDDDxkzZgzjxo1j/fr1BAcH0759e86dO2dtEx4ezrRp05g8eTIrVqwgMTGRbt26kZmZaY/LEhERO7JYLPSsXZzFz7dmeLuKuDs7siEyll5fruSJnzdy4OS56x8kL5VpCX1/ghcPwoP/QJOnIbACZGWYz+6a9Rys/uL6xxEREbvL18MIu3XrRlBQED/88IN1XZ8+ffDw8ODnn3/GMAxCQ0MJDw9nxIgRgNmLFRQUxAcffMDjjz9OfHw8RYsW5eeff6Zfv34AHDt2jLCwMGbPnk3Hjh1zVYuGEYqIFE7H45P5ZP4+/tp0hCwDHCxwT70ShLerSKifu73Lu+TMQdg4AVZ9Di7eMGwjeAfZuyoRkTtSoRhG2Lx5cxYtWsS+ffsA2Lp1KytWrKBLly4AREREEBMTQ4cOHaz7uLq60qpVK1atWgXAxo0bSU9Pz9YmNDSU6tWrW9tcSWpqKgkJCdkWEREpfEJ83fn43lrMDW9J+6pBZBnw+4YjtP54Ke/O2kXs+TR7l2gKLAft3oLQupB2Dha/be+KRETkOvJ12BoxYgT9+/encuXKODs7U6dOHcLDw+nfvz8AMTExAAQFZf/NXlBQkHVbTEwMLi4u+Pv7X7XNlYwePRpfX1/rEhYWZstLExGRfKZikDffD6rP1Ceb0rBMAGkZWXy/PIKWHy7hq6UHSM3IB0PPHRyg84fm682/wtFN9q1HRESuKV+HrSlTpvDLL78wadIkNm3axMSJE/n444+ZOHFitnaXzyBlGMZ1Z5W6XptXXnmF+Ph46xIdHX3zFyIiIgVGvVL+THmsMeMfakCVEB/OpWbw4dy9dBj7Lwt2ncDuo+/DGkDN+wAD5r5szmIoIiL5Ur4OWy+++CIvv/wy9913HzVq1GDgwIEMHz6c0aNHAxAcHAyQo4fq5MmT1t6u4OBg0tLSiI2NvWqbK3F1dcXHxyfbIiIidwaLxUKbSsWYNaw5n9xbi2LerkSeSeLRnzYw6Md19p9Eo92b4OwJ0Wth+5/2rUVERK4qX4etpKQkHC57mKOjo6N16vcyZcoQHBzMggULrNvT0tJYtmwZTZs2BaBevXo4Oztna3P8+HF27NhhbSMiInIlDg4W+tQrweIXWvNk63K4ODqwfP9pOn66nLf+2Ul8crp9CvMJhRbPma8XvAFp5+1Th4iIXFO+Dlvdu3fn3XffZdasWRw+fJhp06YxZswY7r77bsD8zWN4eDjvvfce06ZNY8eOHQwePBgPDw8GDBgAgK+vL0OGDOH5559n0aJFbN68mQceeIAaNWrQrl07e16eiIgUEF6uTozoVJn5w1vSrkoQmVkG41ceps3HS/ltXRRZWXYYytfkafArBeeOwYqxt//8IiJyXfl66vdz584xcuRIpk2bxsmTJwkNDaV///688cYbuLi4AOa9V2+99RbffvstsbGxNGrUiC+//JLq1atbj5OSksKLL77IpEmTSE5Opm3btnz11Vc3NOmFpn4XEZGL/t13irdn7uLAyUQAapXw5e2e1akV5nd7C9n9D0x5ABxd4en14F/q9p5fROQOldtskK/DVn6isCUiIv+VnpnFxFWH+XThfhJTM7BY4L4GYbzYsTIBni63pwjDgJ96QMS/ULWn+TBkERHJc4XiOVsiIiL5lbOjA4+0KMviF1rRu05xDAN+WxfNXZ8s5Zc1kWTejqGFFgt0+gAsDrDrbzN0iYhIvqGwJSIicguKebsxpl9t/niiCZWDvYlLSuf16Tvo9eVKNkXFXv8AtyqoKtQfYr6e+wpkZuT9OUVEJFcUtkRERGygQekAZg5rzqjuVfF2c2L70Xh6f7WKZ37bzJHYpLw9eZtXwc0PTuyAX/tAzI68PZ+IiOSK7tnKJd2zJSIiuXXqXCofzt3Dn5uOYBjg4uTAkOZlGNq6HN5uznlz0u1/wrQnICsdsECd+6HN6+ATkjfnExG5g2mCDBtT2BIRkRu181g8787azaqDZwAI9HRhePuK3NcgDCfHPBhccvYQLHwLdk033zt7QNNnoOkwcPWy/flERO5QCls2prAlIiI3wzAMFu0+yXtzdnPolPnw4QrFvHitaxVaVyqWNyeNXgfzXoMj68z3XkHQ5jWo8wA4OObNOUVE7iAKWzamsCUiIrciPTOLSWuj+HThPmKT0gHoVC2YN7pXJdTP3fYnNAxzhsKFb0LsYXNdmZbQ5wfwyqOQJyJyh1DYsjGFLRERsYX45HS+WLSf8asOk5ll4OHiyPB2FRncrDTOeTG0MCMV1n0PS96D9PPgFQz3/Ailm9n+XCIidwiFLRtT2BIREVvaE5PA69N2sCHSnB6+crA3795dnXqlAvLmhKf2wu+D4NQesDhC2zeg2bPms7pEROSGKGzZmMKWiIjYWlaWwR8boxk9Zw9xF4YW3tcgjBGdKuPv6WL7E6adh5nDYdsU832lLtDrK3D3t/25REQKsdxmg5sarxAdHc2RI0es79etW0d4eDjffffdzRxORETkjuTgYKFfg5Isfr41feuXAGDy+mjajlnG7+ujycqy8e9DXTzh7m+h26fg6AJ7Z8O3reDYZtueR0REgJsMWwMGDGDJkiUAxMTE0L59e9atW8err77K22+/bdMCRURECrsATxc+vKcWfzzRhIpBXpw9n8ZLU7dx77er2Xks3rYns1ig/kMwZAH4lYK4SPihg/mcLhERsambCls7duygYcOGAPz+++9Ur16dVatWMWnSJCZMmGDL+kRERO4YDUoHMOuZFrzapTIeLo5sjIyl+xcrGDVjJwkp6bY9WWhteHyZOZQwMw2mPwlRa2x7DhGRO9xNha309HRcXV0BWLhwIT169ACgcuXKHD9+3HbViYiI3GGcHR14rGU5Fj3fiq41Q8gyYMKqw9z18TKmbT6CTW+1dveHfr9C1Z5m4Jp8P8RF2e74IiJ3uJsKW9WqVeObb75h+fLlLFiwgE6dOgFw7NgxAgMDbVqgiIjInSjE150vB9Tl5yENKVvEk9OJqQyfspX7vlvDnpgE253IwQF6fQ3BNSHpNPzWH1ITbXd8EZE72E2FrQ8++IBvv/2W1q1b079/f2rVqgXAjBkzrMMLRURE5Na1qFCUOeEteLFjJdycHVgbcZYuny3njb93EJeUZpuTuHhC/9/Asxic2AF/PQZZWbY5tojIHeymp37PzMwkISEBf/9L08UePnwYDw8PihUrfE+m19TvIiJib0dik3hv9m5mb48BwM/DmefbV6R/w5I42eKByNHrYUJXyEyF5s9Buzdv/ZgiIoVQnk79npycTGpqqjVoRUZG8umnn7J3795CGbRERETygxL+Hnx1fz0mPdqISkHexCWlM/LvnXT7YgWrD5659ROENYAeX5ivV4yBbb/f+jFFRO5gNxW2evbsyU8//QRAXFwcjRo14pNPPqFXr158/fXXNi1QREREsmtargiznmnO2z2r4evuzJ6Yc/T/fg1P/bqJY3HJt3bwWv2g+XDz9d9Pw5ENt16wiMgd6qbC1qZNm2jRogUAf/75J0FBQURGRvLTTz/x+eef27RAERERycnJ0YFBTUqz9IXWDGxcCgcLzNp+nHZjlvHtsoOkZ97CPVd3vQGVLgwn/K0/xB+xXeEiIneQmwpbSUlJeHt7AzB//nx69+6Ng4MDjRs3JjIy0qYFioiIyNX5e7rwTq/qzHqmBQ1K+5OUlsnoOXvo+vly1kWcvbmDOjhA72+hWDU4fxLGdzHv5xIRkRtyU2GrfPnyTJ8+nejoaObNm0eHDh0AOHnypCaPEBERsYMqIT5MeawJH91TkwBPF/adSKTvt6t5/vetnElMvfEDunqbMxT6lYS4SPixI/z7MWRl2r54EZFC6qbC1htvvMELL7xA6dKladiwIU2aNAHMXq46derYtEARERHJHQcHC/fWD2PRc63o37AkAFM3HeGuT5bx69pIsrJucAJi/1Lw+HKo1huMTFj8DvzUExKO5UH1IiKFz01P/R4TE8Px48epVasWDg5mZlu3bh0+Pj5UrlzZpkXmB5r6XURECppNUbG8Pm0Hu46bD0GuW9KP9/vUpGKQ940dyDBgy68w+yVIPw/u/tDzK6jcJQ+qFhHJ/3KbDW46bF105MgRLBYLxYsXv5XD5HsKWyIiUhBlZGbx0+pIxizYR2JqBs6OFp5sVY6hbcrj5ux4Ywc7fQCmPgzHt5rvGzwKHd4BZ3fbFy4iko/l6XO2srKyePvtt/H19aVUqVKULFkSPz8/3nnnHbL0xHkREZF8w8nRgYebl2HBcy1pVyWI9EyDzxcfoMvny1l76AafzVWkPAxZAE2eNt+v/x4+rQkzn4NDSyEzw+b1i4gUZDfVs/XKK6/www8/8NZbb9GsWTMMw2DlypWMGjWKRx99lHfffTcvarUr9WyJiEhBZxgGc3bE8OaMnZw6Z06a0b9hGC93roKvu/ONHWz/Qvh7KCSeuLTOPcAcWlilJ5RtDU4uVyoCMlLB0RkcbrBnTUQkn8jTYYShoaF888039OjRI9v6v//+m6FDh3L06NEbrzifU9gSEZHCIj4pnffn7ua3ddEAFPV25Z2e1ehUPeTGDpSRBhHLYNffsGcWJP9nqnlXHyhaGdKTzfu80pIgPQnSzpuTbXiHwCMLwbeEDa9MROT2yNOw5ebmxrZt26hYsWK29Xv37qV27dokJ9/i0+vzIYUtEREpbNYeOsMrf23n0OnzAHSvFcpbPaoR4HmFHqnrycyAyBWwawbsmZm9x+tqyraGB6aZz/USESlA8jRsNWrUiEaNGvH5559nWz9s2DDWrVvH2rVrb7zifE5hS0RECqOU9Ew+X7Sfb5YdJMuAIl4u/K9X9Rvv5fqvrEw4ssEMXC4e4OwJLhcWZw84fwp+6AAZydDlY2j4qO0uSETkNsjTsLVs2TK6du1KyZIladKkCRaLhVWrVhEdHc3s2bNp0aLFLRWfHylsiYhIYbY1Oo4X/tjK/pOJwC32cuXG2m9hzkvg5A5PrDAn3xARKSDydDbCVq1asW/fPu6++27i4uI4e/YsvXv3ZufOnYwfP/6mixYRERH7qBXmx8xnmjO0dTkcLPDP1mN0GLuMuTti8uaEDR6FMq3M3q3pT2gmQxEplG75OVv/tXXrVurWrUtmZqatDplvqGdLRETuFJf3cnWrGcJbPaoR6OVq2xPFH4GvmkJqPNw1Elq+YNvji4jkkTzt2RIREZHCq1aYH/8Mu9TLNXPbcdqNWcbfW45iw9/RmjMRdv7AfL30fTi+zXbHFhHJBxS2REREJAc3Z0de6lSZ6U81o3KwN7FJ6Tw7eQtDJm7gWJwNZx2udR9U7gZZ6TDtcfMZXCIihYTCloiIiFxVzRJmL9fz7Svi4ujA4j0n6TD2X35dG0lWlg16uSwW6P4ZeBSBk7tgybu3fkwRkXzihu7Z6t279zW3x8XFsWzZMt2zJSIiUgjtP3GOl6ZuY3NUHACNygTwfp+alCnieesH3zMLJg8ALPDQHCjV5NaPKSKSR/Lkni1fX99rLqVKlWLQoEG3XLyIiIjkPxWCvPnziaa82b0q7s6OrI04S6dP/+XbZQfJyMy6tYNX7gq17wcMczjhkY02qVlExJ5sOhthYaaeLRERkUuizybxyl/bWXHgNAA1ivvyQZ+aVA29hf8jU+Lh62YQH22+L9kUmg6Dip3AQXc+iEj+kacPNb4TKWyJiIhkZxgGf2w8wv9m7iIhJQMnBwtPti7H03eVx9XJ8eYOGnsYln0I2343J80ACCwPTZ42J9NwdrdZ/SIiN0thy8YUtkRERK7sZEIKb/y9k7k7zQcglyvqyYf31KReqYCbP2jCcVj3Laz/0XwOF5iTaDR6ApqHg6PzrRcuInKTFLZsTGFLRETk2uZsP87Iv3dyOjEViwUebFKaEZ0q4+5yk71cAKnnYPMvsPoriI8y1zV9Bjq8Y5uiRURugsKWjSlsiYiIXF9cUhrvztrNHxuPAFC2iCef9K1FnZL+t3bgzAzYNAFmPQ8WB3hkERSve+sFi4jchDyZjVBERETkWvw8XPjo3lpMfLghQT6uHDp9nj5fr+KT+XtJy7iFGQsdnaDBI1C9DxhZMGMYZKbbrnARkTygsCUiIiI216piUeaHt6Jn7VCyDPhi8QF6f72SfSfO3dqBO30A7gFwYges/Mw2xYqI5BGFLREREckTvh7OfHZfHcYNqIOfhzM7jibQ7YsV/N/yQ2Rl3eRdDF5FofMH5utlH8CpfdffJ+ksrPkGEo7d3DlFRG6SwpaIiIjkqW41Q5kf3pI2lYqSlpHF/2bt5r7v13D49PmbO2CNe6F8e8hMM4cTZl1jeOKZg/B/7WDuCJgz4ubOJyJykxS2REREJM8V83Hjx8ENGN27Bh4ujqyLOEunz/7l/5YfIvNGe7ksFug2Fly8IHoNbPjhyu0Or4D/awtnD5rv982DlIRbuxARkRugsCUiIiK3hcVioX/DkswLb0nTcoGkpJu9XPd8s4r9N3ovl18YtBtlvl44CuKism/fMgl+6gXJsVC8HviXgcxU2DvHBlciIpI7ClsiIiJyW4UFePDrI40Y3bsG3q5ObI6Ko+vnK/hyyQHSM29gxsL6QyCsMaQlwszhYBjmkMJF78D0JyErHar2gsGzoNZ95j47pubJNYmIXInCloiIiNx2F3u55j/XkrsqFyMtM4uP5u2l57iV7DwWn7uDODhAz3Hg6AoHFsLGCfDnQ7D8Y3N7i+fhnvHg7A7VepvrDi4yJ8wQEbkN8n3YOnr0KA888ACBgYF4eHhQu3ZtNm7caN1uGAajRo0iNDQUd3d3Wrduzc6dO7MdIzU1lWHDhlGkSBE8PT3p0aMHR44cud2XIiIiIpcJ8XXnhwfr82m/2vh5OLPreAI9x63ks4X7ychNL1eRCtDqJfP1zHDYNR0cnKHX19D2DTOQARStCEE1ICsDdv+TV5cjIpJNvg5bsbGxNGvWDGdnZ+bMmcOuXbv45JNP8PPzs7b58MMPGTNmDOPGjWP9+vUEBwfTvn17zp27NPY7PDycadOmMXnyZFasWEFiYiLdunUjMzPTDlclIiIi/2WxWOhVpzgLhreiS41gMrIMxi7cxz3frObQqcTrH6DZs2aQAnD3h0HTofaAnO2qX+jd0lBCEblNLIZh3OSDLvLeyy+/zMqVK1m+fPkVtxuGQWhoKOHh4YwYYU7nmpqaSlBQEB988AGPP/448fHxFC1alJ9//pl+/foBcOzYMcLCwpg9ezYdO3bMVS0JCQn4+voSHx+Pj4+PbS5QREREsjEMg7+3HGPk3zs4l5KBm7MDr3WpwgONS2GxWK6+Y1wUbJxohqzAclduE3sYPqsFFgd4fi94FcuTaxCRwi+32SBf92zNmDGD+vXrc++991KsWDHq1KnD999/b90eERFBTEwMHTp0sK5zdXWlVatWrFq1CoCNGzeSnp6erU1oaCjVq1e3trmS1NRUEhISsi0iIiKSty72cs0Lb0mz8uaMhSP/3smgH9cRE59y9R39SkLbkVcPWgD+paF4fTCyYNffNq9dRORy+TpsHTp0iK+//poKFSowb948nnjiCZ555hl++uknAGJiYgAICgrKtl9QUJB1W0xMDC4uLvj7+1+1zZWMHj0aX19f6xIWFmbLSxMREZFrCPVz5+eHG/Fm96q4OjmwfP9pOn76LzO2Hru1A2sooYjcRvk6bGVlZVG3bl3ee+896tSpw+OPP86jjz7K119/na3d5cMKDMO49lCDXLR55ZVXiI+Pty7R0dE3fyEiIiJywxwcLDzUrAyznmlOjeK+xCen88xvm3nmt83EJ6Xf3EGr3Q1YIGo1xGuyLBHJW/k6bIWEhFC1atVs66pUqUJUlPngwuDgYIAcPVQnT5609nYFBweTlpZGbGzsVdtciaurKz4+PtkWERERuf3KF/Pmr6FNeaZtBRwdLMzYeoyOn/7L8v2nbvxgPqFQqqn5euc02xYqInKZfB22mjVrxt69e7Ot27dvH6VKlQKgTJkyBAcHs2DBAuv2tLQ0li1bRtOm5j+k9erVw9nZOVub48ePs2PHDmsbERERyd+cHR14rn1F/nyiCWWKeBKTkMLAH9bx5t87SE67wdmFNZRQRG6TfB22hg8fzpo1a3jvvfc4cOAAkyZN4rvvvuOpp54CzOGD4eHhvPfee0ybNo0dO3YwePBgPDw8GDDAnPLV19eXIUOG8Pzzz7No0SI2b97MAw88QI0aNWjXrp09L09ERERuUJ2S/sx6pjkDG5u/eJ24OpKuXyxna3Rc7g9SpSdYHOHYZjhzMG8KFREhn4etBg0aMG3aNH777TeqV6/OO++8w6effsr9999vbfPSSy8RHh7O0KFDqV+/PkePHmX+/Pl4e3tb24wdO5ZevXrRt29fmjVrhoeHB//88w+Ojo72uCwRERG5BR4uTrzTqzoTHmpAMW9XDp06T++vVzF2wT7Sc/MgZK+iUKal+XrnX3lbrIjc0fL1c7byEz1nS0REJP+JPZ/G63/vYNa24wDUKuHLmH61KVfU69o7bvoZZjwNxarB0Ks/CkZE5EoKxXO2RERERK7F39OFcf3r8Nl9tfFxc2LrkXi6fr6ciasOc83fJ1fpBg7OcHInnNx95TYZqbB7JkSuBv1uWkRugsKWiIiIFGgWi4WetYszb3hLmpcvQkp6Fm/OuM6DkN39ofyFe7d3XDaUMDPD7Pn6oh5MuR/Gd4LvWsG2PyDzJqecF5E7ksKWiIiIFAohvu789HBDRuX2Qcj/nZXQMCArywxUXzY0hxjGR4NnMXByh+Nb4a9H4LNasPIzSIm/fRcmIgWW7tnKJd2zJSIiUnAcOJnIc79vYdsRMxR1rxXKOz2r4efhcqlR6jn4qDxkpEC7UbDtdzi5y9zmUQRaPAf1H4a0JNjwI6z7Ds6fNLe7eEHdQdDkKfAtcXsvTkTsLrfZQGErlxS2RERECpb0zCy+XHKALxYfIDPLIMjHlQ/61KR1pWKXGv0+CHb9fem9my80fQYaPQGul02ykZ4C2/+A1V/CqQv3ebkHwCMLIbBc3l+QiOQbCls2prAlIiJSMG2NjmP471s4dOo8AAMaleS1LlXwdHWC/Qvh1z5mT1XjJ6HJ0+Dud+0DGgYcXAQL3oQTOyCwPAxZAB4BeX8xIpIvKGzZmMKWiIhIwZWSnskHc/cwfuVhAEoGePBJ31o0KB0AMTvAJ/TGw9K5E/B/bc17u0q3gAf+AieX6+8nIgWepn4XERERucDN2ZE3u1dj0iONKO7nTtTZJPp+u5rRs3eTEljl5nqlvINgwBRw8YbDy2HmcE0RLyLZKGyJiIjIHaNp+SLMCW/BPfVKYBjw7b+H6DFuBTuO3uTsgkHV4N4JYHGALb/AirE2rVdECjaFLREREbmj+Lg58/G9tfhuYD2KeLmw70Qid3+1ki+XHCAjM+vGD1ihHXT+0Hy96C3YOd2m9YpIwaWwJSIiInekDtWCmRfeko7VgkjPNPho3l76fruaw6fP3/jBGj4KjZ40X097HI5stG2xIlIgKWyJiIjIHSvQy5VvHqjHJ/fWwtvViU1RcXT+bDm/rInkhucQ6/guVOhoPrfrt/sgLipvihaRAkNhS0RERO5oFouFPvVKMHd4S5qUDSQ5PZPXp+9g8Pj1nEhIyf2BHBzhnh8gqIb58ONf+0LiqbwrXETyPYUtEREREaC4nzu/PtKIkd2q4uLkwLJ9p+gw9l/+2Xos9wdx9TZnKPQOMR98PKELJNzA/iJSqChsiYiIiFzg4GBhSPMyzBrWnBrFfYlPTmfYb5sZ9ttm4pLScncQ3+IweBb4lIDT++DHTnA2Im8LF5F8SWFLRERE5DIVgrz5a2hTnrmrPI4OFv7ZeowOY/9l6d6TuTtAYDl4eA4ElIW4SBjfGU7tzduiRSTfUdgSERERuQJnRwee61CJqU82pWxRT06eS2Xw+PW8Nm0751Mzrn8Av5Lw0BwoWgXOHTcD1/FteV+4iOQbClsiIiIi11A7zI9Zw1rwULPSAPy6Noouny9nw+Gz19/ZO9gcUhhSC5LOwMRuEL0+bwsWkXxDYUtERETkOtxdHHmzezUmPdKIUF83Is8k0ffb1Xw4dw/p13sQsmcgPPgPhDWGlHj4qSccWnZ7ChcRu1LYEhEREcmlpuWLMHd4S/rULUGWAV8tPciA79cQE3+dKeLdfGHgX1C2DaSfh0l94djm21O0iNiNwpaIiIjIDfBxc+aTvrX4ckBdvFydWH84li6fL+fffdd5ppaLJ/SfDOXbmQ8+/mOw2dMlIoWWwpaIiIjITehaM4SZw5pTNcSHs+fTeHD8OsYs2EdmlnH1nZzdoM//mZNnxB6Gv58G4xrtRaRAU9gSERERuUmli3jy19Cm9G9YEsOAzxftZ+APazl1LvXqO7n7wz0TwMEZds+Add/dtnpF5PZS2BIRERG5BW7OjozuXYNP+9XGw8WRVQfP0OXz5aw+eObqO5WoBx3eMV/Pew2Obro9xYrIbaWwJSIiImIDveoUZ8bTzakY5MWpc6nc/39r+HzR/qsPK2z0BFTpDlnp5v1byXG3s1wRuQ0UtkRERERspHwxL6Y/1Yx76pmzFY5ZsI9BP15lWKHFAj3GgV8piIuEv5/S/VsihYzCloiIiIgNebg48fG9tfj43lq4Ozuy8sAZOn+2nFUHTuds7O4H904ARxfYMxPWfnO7yxWRPKSwJSIiIpIH7qlXghlPN6NikBenE1O5/4e1V56tsHhd6PCu+Xr+SDiy8fYXKyJ5wmIY6q/OjYSEBHx9fYmPj8fHx8fe5YiIiEgBkZyWyagZO5myIRqAxmUD+Py+OhTzcbvUyDDgjwdh19/gFQTF64GRlXNx84XWr0CxKna6GhGB3GcDha1cUtgSERGRWzF981FenbadpLRMAj1deK93DTpWC77UICUevm0FsRHXPpCbr/lw5FJN87ZgEbkqhS0bU9gSERGRW3XwVCJP/bqJPTHnAOhdtzhvdq+Gr7uz2SDpLOyZBUYmWBxyLut/gOg14OgK9/wIVbrZ8WpE7lwKWzamsCUiIiK2kJqRyZgF+/ju30MYBoT4uvHRPbVoXqHI9XdOT4Y/H4a9s83w1fUTqP9w3hctItnkNhtoggwRERGR28jVyZFXOlfhj8ebUCrQg+PxKTzww1re+HsHSWkZ197Z2R36/gx1B5n3cM0cDktGa8p4kXxKYUtERETEDuqXDmDOsy0Y2LgUAD+tjqTLZ8vZGBl77R0dnaD759DyJfP9svdhZjhkZeZtwSJywxS2REREROzEw8WJd3pV5+chDQnxdePwmSTu/WYVYxfsIyMz6+o7Wixw12vmMEIssHEC/D7IHGYoIvmGwpaIiIiInbWoUJS54S3pXac4WQZ8tmg/A75fy7G464SnBo9A35/MCTP2zDTv51IPl0i+obAlIiIikg/4ujszpl9tPruvNl6uTqw7fJYuny9n/s6Ya+9YtQc8MNUMXHtnw4I3bk/BInJdClsiIiIi+UjP2sWZ9UxzapbwJS4pncd+3sgbf+8gJf0aPVZlWkCvr8zXq8eZU8SLiN0pbImIiIjkM6UCPfnziaY82qIMYE6e0evLlRw4ee7qO9W4B9q8br6e/SIcWHgbKhWRa1HYEhEREcmHXJwceK1rVcY/1IBATxf2xJyj+xcr+WvTkavv1PIFqNXffCjyHw/BiV23r2ARyUFhS0RERCQfa1OpGHOebUHz8kVITs/kud+3MnL6DtIyrjBbocUC3T+DUs0gNQEm9YPEk7e/aBEBFLZERERE8r1iPm5MfLghz7StAMDPayLp991qjsdfYbZCJ1fo9wsElIP4KPitv6aEF7EThS0RERGRAsDRwcJz7Svyw4P18XFzYnNUHN2/WMGqg6dzNvYIgPv/AHd/OLoBpj0BWdd4bpeI5AmFLREREZECpG2VIP4Z1pwqIT6cTkzjgf9by7fLDmIYRvaGgeWg36/g4Ay7psOcF/UMLpHbTGFLREREpIApFejJX082pXdd8yHIo+fs4clfNnEuJT17w9LNoOc48/X6/4MpAyHt/O0vWOQOpbAlIiIiUgC5uzjyyb21+F+v6jg7Wpi7M4bOny1nw+Gz2RvWug/unXDhocezYEI3TZohcpsobImIiIgUUBaLhQcal+L3x5tQwt+dI7HJ9P12NZ/M30t65n/u0ap2Nzz4D7gHwLFN8H9t4dRe+xUucodQ2BIREREp4OqU9Gf2sy2swwq/WHyAe75exaFTiZcalWwEjyyEgLIQFwU/tIeIf+1XtMgdQGFLREREpBDwcXNmTN/ajBtQBx83J7Yeiafr5yuYtDbq0uQZgeVgyEIIawwp8fBzb9g62b6FixRiClsiIiIihUi3mqHMG96SpuUCSU7P5NVp23n0p42cSUw1G3gGwqC/zaGFWekw7XGY+yqkJNi3cJFCSGFLREREpJAJ8XXnlyGNeK1LFVwcHVi4+wRdP1/B+ouTZzi7QZ8fodmz5vs1X8IXdWHjRE0PL2JDBSpsjR49GovFQnh4uHWdYRiMGjWK0NBQ3N3dad26NTt37sy2X2pqKsOGDaNIkSJ4enrSo0cPjhw5cpurFxEREbl9HBwsPNqyLNOfaka5op7EJKRw33dr+GbZQbKyDHBwgPZvw4DfIaAcnD8F/zwD37WCwyvsXb5IoVBgwtb69ev57rvvqFmzZrb1H374IWPGjGHcuHGsX7+e4OBg2rdvz7lz56xtwsPDmTZtGpMnT2bFihUkJibSrVs3MjP1mxsREREp3KqG+jDj6eb0qh1KZpbB+3P28MhPG4g9n2Y2qNgRhq6Bju+Bqy/EbIcJXc1ncsUetmvtIgVdgQhbiYmJ3H///Xz//ff4+/tb1xuGwaeffsprr71G7969qV69OhMnTiQpKYlJkyYBEB8fzw8//MAnn3xCu3btqFOnDr/88gvbt29n4cKF9rokERERkdvG09WJsf1qM7p3DVycHFi85yTdvljB5qhYs4GTCzR5Cp7ZDPWHgMUBds+AcQ1gzdf2LV6kACsQYeupp56ia9eutGvXLtv6iIgIYmJi6NChg3Wdq6srrVq1YtWqVQBs3LiR9PT0bG1CQ0OpXr26tc2VpKamkpCQkG0RERERKagsFgv9G5Zk2tCmlA704Gic+UyuH1ZEXJqt0DMQuo2BJ1ZC2daQmQbzXtUzuURuUr4PW5MnT2bTpk2MHj06x7aYmBgAgoKCsq0PCgqybouJicHFxSVbj9jlba5k9OjR+Pr6WpewsLBbvRQRERERu6sW6ss/w5rTtUYI6ZkG78zcxWM/b7w0rBAgqCoMnA6VuoKRBYvfsVu9IgVZvg5b0dHRPPvss/zyyy+4ubldtZ3FYsn23jCMHOsud702r7zyCvHx8dYlOjr6xooXERERyae83ZwZN6AOb/WohrOjhQW7TtD5s+WsPnjmUiOLBdqOvDCk8B84stF+BYsUUPk6bG3cuJGTJ09Sr149nJyccHJyYtmyZXz++ec4OTlZe7Qu76E6efKkdVtwcDBpaWnExsZetc2VuLq64uPjk20RERERKSwsFgsPNi3NtKHNKFvEnK1wwP+t4aN5e0jPzDIbFasCtfqbrxe+CReHG4pIruTrsNW2bVu2b9/Oli1brEv9+vW5//772bJlC2XLliU4OJgFCxZY90lLS2PZsmU0bdoUgHr16uHs7JytzfHjx9mxY4e1jYiIiMidqnpxc1hh3/olMAz4cslB+n67muizSWaD1i+DowscXg6Hlti3WJECxsneBVyLt7c31atXz7bO09OTwMBA6/rw8HDee+89KlSoQIUKFXjvvffw8PBgwIABAPj6+jJkyBCef/55AgMDCQgI4IUXXqBGjRo5JtwQERERuRN5ujrx4T21aFmxKK/8tZ3NUXF0+Ww5/7u7Oj1rl4QGj8Car2DhW1CmtfmMLhG5rnwdtnLjpZdeIjk5maFDhxIbG0ujRo2YP38+3t7e1jZjx47FycmJvn37kpycTNu2bZkwYQKOjo52rFxEREQkf+lWM5RaJfwIn7KFjZGxPDt5C8v2neKtds/gveknOL4Fdv8N1e62d6kiBYLFMDT4NjcSEhLw9fUlPj5e92+JiIhIoZaRmcXniw8wbvF+sgwo7ufO5EpLCdv6GQSWh6FrwbHA/85e5KblNhuoD1hEREREsnFydOC59hWZ8ngTwgLcORqXTOd1NUly8oMzB2DLr/YuUaRAUNgSERERkStqUDqAOc+2pF/9MBINdz5K7g5A+uL3ID3ZztWJ5H8KWyIiIiJyVV6uTnxwT02+H1SfOa5dOGIUwfl8DKsnv09mlu5GEbkWhS0RERERua72VYOY+Vxb5hd9GIAqB75nyDcLiTqTZOfKRPIvhS0RERERyZUiXq489OQI4r3K4Wc5T71jv9Dps3+ZtDYKzbkmkpPCloiIiIjkmsXRCd+ubwPwpNM/3Jc5k1enbeOhCes5kZBi5+pE8heFLRERERG5MZW7Qs37cCKTN5x/5huXz9i4N5IOY//l7y1H1cslcoHCloiIiIjcGIsF7v4GOn8IDs50cljHPI83CE05wLOTt/DUpE2cPZ9m7ypF7E5hS0RERERunMUCjR6Hh+eCbxihWcf4x30U9zktZfb2GDqM/Zcle07au0oRu1LYEhEREZGbV6I+PP4vlG+PU1Yq7zt9x7feP5CYmMBDE9YzcvoOktMy7V2liF0obImIiIjIrfEIgAG/w10jweJAx/RFLPF/Dw9S+HlNJF2/WM72I/H2rlLktlPYEhEREZFb5+AALV+AQX+DRxFCkvczr85qgnxcOXTqPHd/tZIvlxzQg5DljqKwJSIiIiK2U6Yl9PoKgLA9PzJ/QBG61AgmI8vgo3l76fftaqLPXuVByCkJcGgpZGnYoRQOClsiIiIiYlsVO0KV7mBk4rtoBF/2r80n99bCy9WJDZH/3969R1VV530cfx84cLgIKCIgggrj/Voq5V3TIi1rzK5Omk7PTKMpSXaxyRpt8pLNZPOUqdV4mSc1zUnNyiwtxdRMQ1Hy2uQlvCBeEJA7nN/zx6lTBBY6HA7g57XWXu6z9+8cvrv1XSw+/fb+nQwG/u/nrNx1vPR7CnNg/s3wf7+FVQ8rcEmtoLAlIiIiIpVvwAvg5Q+p27AkL+bOzpF8NK4XsU3rcbGgmEeX7SZh6S6y84vAGFgdD+n7HO/ds9Tx2m537zWI/JcUtkRERESk8gVFwg1PO/bX/QVyzhIV7Mfbf+zK+Jta4OlhYVXySW555XOOf/R3+Ppd8LBCr8fA4gnJi+GDcQpcUqMpbImIiIiIa1w/CsLaQ16GI3ABVk8PHunfnHf+1JXIer5EZOwk/MtpANjjpkL/v8CQN8DiATv/Dz4cr8AlNZbCloiIiIi4hqcVBs107CcvhqNbnKc6Nwnmoweb8abfLKwWOytKejI0uQMnL+RB+7tg8FzAAkkL4KMnHLcaitQwClsiIiIi4jpR10HnkY79D8dDcaFjv7iAgPceJLAkgwuBLZlqeYgvjzoWz3g36Timwz3fr2pogR3/hLVPKXBJjaOwJSIiIiKu1X8S+IXAmQPwxSzHsbV/huM7wCeIuiOX8e4jN9IxMojMvCIeW76bB+ZvJ7XxYLj9Vcf4L+fCxxMVuKRGUdgSEREREdfyC4abpzr2E1+EjTPgq3mABe6cB8HRNA3x59+ju/PkgJZ4Wz34/JuzxL28iXm5PbEP+l/He7e9Bu8Mh4tn3HYpIpfDYoz+90BFZGVlERQURGZmJoGBge4uR0RERKRmMQb+dRsc/fzHYzdMhD5Plhl6+MxFnlqRwvYj5wHoGFWXua330HDzM2AvdsySDZoJbX5bVdWLlFLRbKCZLRERERFxPYsFbn0JPLwcr1sMhF6Plzs0pkEdlv6xK9PuaE+Azcru1Av0Wt+Et9rNx96gNeSehXcegH//D+Ser8KLELk8mtmqIM1siYiIiFSC5Lfh6GbHbYW+dX91eFpmPs++9zXr9p0GoGmQldcbr6fFf/6JxdihThjc9gq0HODiwkV+VNFsoLBVQQpbIiIiIu5hjOGjr9OY8sE+TmbmAzAkNI2pltfwzfzWMeia+2HgDLAFuLFSuVroNkIRERERqRUsFgu3tG/IZ4/35YmbW1LHZmVFejjXnH6Wj4PuxmBxfI/XqofdXapIKQpbIiIiIlIj+Hh5MuaGZmx8oi/Dujam2MPGn07fwbCiiZTgCftXw3/Wu7tMESeFLRERERGpUULq2JgyuD0fJ/TixtahbClpw8LiOAByVo2H4gI3VyjioLAlIiIiIjVSs9AA/jkiliV/uJ7V9UaQburif/EYq+c8xZlsBS5xP4UtEREREanRujcL4Z1xcexoMR6Am84uYsTM5azadQKtBSfupLAlIiIiIjWezerJrb97hIsNu+FrKSSheD4Jy5L5w7++Iu37FQxFqprCloiIiIjUDhYLde54GeNhJc4ziThrMp8eSOemmYm89cVRSuya5ZKqpbAlIiIiIrVHaGssXUcDMCt4GbGRfmQXFPPse3sZ9Opmvjx8zs0FytVEYUtEREREapc+EyCgId5Zx1jWdhvP3d6WIF8v9p/K4t43tjFmyU5OXMhzd5VyFVDYEhEREZHaxRYAN08FwGPzy4xoZdjwuOO7uTws8OGeU/R/aSP/u/4b8otK3Fys1GYKWyIiIiJS+7QdAtF9oKQA1j5FsL83Uwa354P4XlwXHUx+kZ2X1x+i/0uJfLjnlFYtFJdQ2BIRERGR2sdigVv+Dh5ecGgtfPkGFOXTJiKQZQ91ZdbvriUiyIcTF/IYs2Qnd8/9guTUC+6uWmoZi1GMr5CsrCyCgoLIzMwkMDDQ3eWIiIiISEWsmwRb/uHYtwVB299Ch3uhcXfyig1zE7/l9U3fkl9kB2DwNRE8MaAVjer6uq9mqfYqmg0UtipIYUtERESkBiougE1/h+TFkHXix+OBkdDhbuhwL2m2aP728UHe3XkcAJvVgz/2imFU399Qx2Z1U+FSnSlsVTKFLREREZEazG6HY1tgzzLY9x4UZP14rvVtcOc8UtLymfLhPr48ch6AkDo24vs1487OkQpdUorCViVT2BIRERGpJYryHc9xpSx3/GsvhuZxcO8ijKc3n+w7zfQ1+zl6LhcAf29PhnSKZHi3JrQIC3Bz8VIdKGxVMoUtERERkVro8EZYch8U5zkDF1YbhcV2lu74joVbj3L4TI5z+PXRwQzv1oS4NuF4W7XW3NVKYauSKWyJiIiI1FKXCFwAxhi2fnuOt744xrr9pymxO/50bhBg43fXNWZE96YE+3u7sXhxB4WtSqawJSIiIlKL/ULg+sGpzDze3p7K29u/40x2AQA+Xh7cF9uYP/aO0QqGVxGFrUqmsCUiIiJSy1UgcAEUldj5eG8aryceJuVEJgBWDwu/vaYRo/vG0CxUz3XVdgpblUxhS0REROQqUMHABY5bDLf85xyzN/6Hrd+ecx6PaxPGwzc045qoulVTs1Q5ha1KprAlIiIicpX4aeAKinJs/vXBLwT8Q37818MK+RcgP5PT6afZfySVrIwzBJKLAawhMVzT8VoCGraE4Gio2wS8fNx8cVIZFLYqmcKWiIiIyFXkcCK8PRSKcn59bIVZICgSGnWGW/4OdRpU4mdLVapoNtC3s4mIiIiI/FxMH0hIgfR9kHsWcr7ffrpvSsCnLvjWBZ+g7zfHfuq5LL5MSsL34jGaWk4T7ZGOH3mQmerY8jJg+Erw8HTzhYorKWyJiIiIiJTHvz5E97qit0YBjfoZ3vkqlWFrD5CRW0gIWfz+N1mMPj0ZjyOJsHE69HumcmuWakW3EVaQbiMUERERkStxIbeQlz45xOIvj2E3cLvHFl7xfg2AsR5Ps9OrCz5enti8PAn0sdI1pj43tAqlQ6MgPDwsbq5eyqNntiqZwpaIiIiI/De+PpHJ5NV7+epYBs9b5zPcup4MU4dBBVM5Qdnnt+r7e9OnZQP6tQqlV/MGBPl6uaFqKU9Fs4FHFdZ02aZPn05sbCwBAQGEhoYyePBgDh48WGqMMYbJkycTERGBr68vffv2Ze/evaXGFBQUEB8fT0hICP7+/tx+++0cP368Ki9FRERERK5y7RoF8e/R3dk9KY648fMpCO1IPctFPolawDv/cy3/evA6XhjSnoHtwqljs3Iup5AVO08wdskuOj2/jnte/4INB9LdfRlyGar1zNaAAQO47777iI2Npbi4mIkTJ5KSksK+ffvw9/cHYMaMGUydOpWFCxfSokULpkyZwqZNmzh48CABAY4vlBs9ejTvv/8+CxcupH79+jz22GOcP3+epKQkPD0r9lCiZrZEREREpFJlHIPXezuWj4/9I9z6d+epwmI7Sccy2HAwnQ0H0vkm/aLzXL9WofxlUBuahvi7oWiBWnob4ZkzZwgNDSUxMZHevXtjjCEiIoKEhAQmTJgAOGaxwsLCmDFjBn/605/IzMykQYMGvPXWW9x7770AnDx5kqioKNasWcPNN99c7s8qKCigoKDA+TorK4uoqCiFLRERERGpPIc+hiX3OPbvnAft7yp3WOr5XBZtO8b8LUcoKjF4e3rwh17RjLmhGf42rXlX1WrFbYQ/l5mZCUBwcDAAR44cIS0tjbi4OOcYm81Gnz592Lp1KwBJSUkUFRWVGhMREUG7du2cY8ozffp0goKCnFtUVJQrLklERERErmYtboZejzn2Vz8CZw6WOywq2I8/39KatQm96dOiAYUldmZv/Jb+LyXyXvIJLjV/kl9UwumsfIpL7K66AvkFNSYGG2MYP348PXv2pF27dgCkpaUBEBYWVmpsWFgYx44dc47x9vamXr16Zcb88P7y/PnPf2b8+PHO1z/MbImIiIiIVKq+T0Pqdjj6OSwbDg+8B4ENyx36mwZ1WPj7WNbvT+f5D/bx3flcxi1N5q0vjtEstA7ncgo5d7GAczmFnL9YSHZBMQBBvl70/X6xjb4tQgny02IbVaHGhK2xY8eyZ88eNm/eXOacxVJ6SUxjTJljP/drY2w2Gzab7cqKFRERERGpKE8r3DUf5vaCswfhH+0dtxN2fRgadigz3GKxcFObMHo1D+Gfnx9m1ob/8NWxDL46lnHJH5GZV8R7ySd5L/kknh4WujSpR//WofRvHUZMiP+v/u0sV6ZGhK34+HhWr17Npk2biIyMdB4PDw8HHLNXDRv+mP7T09Ods13h4eEUFhaSkZFRanYrPT2d7t27V9EViIiIiIj8gjqhcP9y+OhJ+O4L2P22Y4vuDd3GQrObwKP0E0A+Xp6M7decIZ0iWbYjFU8PC8H+3oTU8aZ+HZtj39+Gv82T5NQLrN+fzmcHTnPo9EW+PHKeL4+cZ9qaA7QMC+De2CiGdGpEXT9vN/0HqJ2q9QIZxhji4+NZuXIlGzdupHnz5mXOR0RE8Oijj/Lkk08CUFhYSGhoaJkFMhYtWsQ99zgePjx16hSRkZG/uEDGz2k1QhERERGpEseTYNtrsHcVmBLHsfrN4bqHoG5j8PAEi4dj8/AEiyd4WMHLF2x1wBYI3nXAaoNyZqxSz+fy2YF01u8/zZeHz1P4/fNc3lYPBrYL577YxnSNCf7F2S673REhrtYvXa4VqxE+/PDDLFmyhPfee4+WLVs6jwcFBeHr6ws4ln6fPn06CxYsoHnz5kybNo2NGzeWWfr9gw8+YOHChQQHB/P4449z7tw5Lf0uIiIiItXXhVTY/gYk/QsKMi///R5WR+jyCYToPtD9EWjQotSQzLwiVief4O3tqew7leU83rS+H/fGNqZtRCCnMvM4cSGfExl5nLyQx8nMPE5dyMfL00JsdDDdYurT7Tf1aRsRhOdVEr5qRdi6VJpesGABI0eOBByzW8899xyvv/46GRkZXH/99bz22mvORTQA8vPzeeKJJ1iyZAl5eXn079+f2bNnX9aCFwpbIiIiIuIWBdmwazHsXw1FeY7ZLmMHu92xby8BezEU5ULBRSjKucQHWaDVrdBjHERdV+qMMYaUE5m8vT2V1cknyCks+YWCDMM81xNpOcPSkhs4ahyP8wT4WLk+OpiuMfXp1KQeDerYqOvnRR2btdY9E1YrwlZ1orAlIiIiIjWCvQQKLzqCV+FFyDoBO+bBgQ9+HNO4uyN0NY8r8yxYTkExH+45xb+TjnMhr5BGdX2J+H5rFORD7NE5NNozCwCDhWS/rrySezMb8psDZUOVl6eFen7eBPt7U9fPi/r+NsICfWgY5ENYkOPf8EAfwgJ98LbWjG+mUtiqZApbIiIiIlKjnTkEW1+B3UvBXuQ41qA13PA0tLm9Yp+xYRokznDsR8bC8R3OU7n12/FF6H0syenEvtP5nM8ppKD48r7f64dAVtfXi7p+3gT5ejm3un5etGkYyPUx9S/rM11BYauSKWyJiIiISK2QdQq2zYavFkBhtuNYx6Ew8EXH812XsvEF2DjdsX/zNOg2xhHgts12rJxYnO84FxAB1w6DsLbkBzblvC2S80VeZOQWcj6nkLMXCzmdlU9a5vdblmMrrEAwG3pdFNOHlF0Ov6opbFUyhS0RERERqVXyM2HzP2DLPxzPgNVrCkPeLPM8FwAbZ8DGaY79uCnQPb70+ZxzkDQftr8JF0+XfX+dMAiOcWxh7aDzSPD2c542xpCRW0R6dj4Xcou4kFtEVl4RF/IKuZBbRGZeERfyiujdPIR7YxtX0n+AK6ewVckUtkRERESkVjr2Bax4CDK/cywj32cC9HrM8WXLAIl/gw1THPs3PQ89Hrn0ZxUXwNfvwuGNcP6wY8s9V3ZceAe4dxHUa1KxGvMz4dPvf3Zdha1aR2FLRERERGqt/Ez48HFIecfxOup6GPIGpPwbPnvecezG56BnwuV/dl4GnD/iCF7nvoXtrzsCmG8w3L0AYvr+8vsPrIEPx0P2KWh2Iwx79/JrqGQKW5VMYUtEREREar0978CHj0FBFlh9fnwOq/8k6DW+cn7GhVRYNgxOJTu+mPmmv0K3sWW/gPniGfjoSdi7wvE6OAZufxWa9qycOv4LFc0GNWNtRRERERERcb0O98CozRDV9SdB6y+VF7QA6kbBg2uh4+8cz4p98gy8+wcozHWcNwZ2L4PXYh1By+IJPRJg9NZqEbQuh2a2KkgzWyIiIiJy1Sgphl1vgS0A2t/lmp9hDOz4J6x9yvGlzGHtHCsibn4Z/rPOMSasPfz2VYi41jU1XCHdRljJFLZERERERFzg2FZ45wHIOfPjMU8b9HnS8cXLnl7uq+0SdBuhiIiIiIhUf026w0OJ0Kiz43VUV8etjL0fr5ZB63JY3V2AiIiIiIhc5YIawYMfw5kDENoWPGrHnJDCloiIiIiIuJ+nF4S3d3cVlap2REYREREREZFqRmFLRERERETEBRS2REREREREXEBhS0RERERExAUUtkRERERERFxAYUtERERERMQFFLZERERERERcQGFLRERERETEBRS2REREREREXEBhS0RERERExAUUtkRERERERFxAYUtERERERMQFFLZERERERERcQGFLRERERETEBazuLqCmMMYAkJWV5eZKRERERETEnX7IBD9khEtR2Kqg7OxsAKKiotxciYiIiIiIVAfZ2dkEBQVd8rzF/FocEwDsdjsnT54kICAAi8Xi1lqysrKIiooiNTWVwMBAt9YiNYf6Rq6UekeuhPpGroT6Rq5UVfeOMYbs7GwiIiLw8Lj0k1ma2aogDw8PIiMj3V1GKYGBgfpFJJdNfSNXSr0jV0J9I1dCfSNXqip755dmtH6gBTJERERERERcQGFLRERERETEBRS2aiCbzcakSZOw2WzuLkVqEPWNXCn1jlwJ9Y1cCfWNXKnq2jtaIENERERERMQFNLMlIiIiIiLiAgpbIiIiIiIiLqCwJSIiIiIi4gIKWyIiIiIiIi6gsFXDzJ49m+joaHx8fOjcuTOff/65u0uSamT69OnExsYSEBBAaGgogwcP5uDBg6XGGGOYPHkyERER+Pr60rdvX/bu3eumiqU6mj59OhaLhYSEBOcx9Y1cyokTJxg2bBj169fHz8+Pa665hqSkJOd59Y78XHFxMc888wzR0dH4+voSExPDX//6V+x2u3OM+kYANm3axG233UZERAQWi4VVq1aVOl+RPikoKCA+Pp6QkBD8/f25/fbbOX78eJVdg8JWDbJs2TISEhKYOHEiu3btolevXgwcOJDvvvvO3aVJNZGYmMiYMWPYtm0b69ato7i4mLi4OHJycpxjXnzxRWbOnMmsWbPYsWMH4eHh3HTTTWRnZ7uxcqkuduzYwRtvvEGHDh1KHVffSHkyMjLo0aMHXl5efPTRR+zbt4+XXnqJunXrOseod+TnZsyYwdy5c5k1axb79+/nxRdf5G9/+xuvvvqqc4z6RgBycnLo2LEjs2bNKvd8RfokISGBlStXsnTpUjZv3szFixcZNGgQJSUlVXMRRmqM6667zowaNarUsVatWpmnnnrKTRVJdZeenm4Ak5iYaIwxxm63m/DwcPPCCy84x+Tn55ugoCAzd+5cd5Up1UR2drZp3ry5WbdunenTp48ZN26cMUZ9I5c2YcIE07Nnz0ueV+9IeW699Vbz4IMPljo2ZMgQM2zYMGOM+kbKB5iVK1c6X1ekTy5cuGC8vLzM0qVLnWNOnDhhPDw8zNq1a6ukbs1s1RCFhYUkJSURFxdX6nhcXBxbt251U1VS3WVmZgIQHBwMwJEjR0hLSyvVRzabjT59+qiPhDFjxnDrrbdy4403ljquvpFLWb16NV26dOHuu+8mNDSUa6+9ljfffNN5Xr0j5enZsyeffvophw4dAmD37t1s3ryZW265BVDfSMVUpE+SkpIoKioqNSYiIoJ27dpVWS9Zq+SnyH/t7NmzlJSUEBYWVup4WFgYaWlpbqpKqjNjDOPHj6dnz560a9cOwNkr5fXRsWPHqrxGqT6WLl3Kzp072bFjR5lz6hu5lMOHDzNnzhzGjx/P008/zfbt23nkkUew2Ww88MAD6h0p14QJE8jMzKRVq1Z4enpSUlLC1KlTGTp0KKDfOVIxFemTtLQ0vL29qVevXpkxVfX3s8JWDWOxWEq9NsaUOSYCMHbsWPbs2cPmzZvLnFMfyU+lpqYybtw4PvnkE3x8fC45Tn0jP2e32+nSpQvTpk0D4Nprr2Xv3r3MmTOHBx54wDlOvSM/tWzZMhYtWsSSJUto27YtycnJJCQkEBERwYgRI5zj1DdSEVfSJ1XZS7qNsIYICQnB09OzTApPT08vk+hF4uPjWb16NRs2bCAyMtJ5PDw8HEB9JKUkJSWRnp5O586dsVqtWK1WEhMTeeWVV7Barc7eUN/IzzVs2JA2bdqUOta6dWvnwk36nSPleeKJJ3jqqae47777aN++PcOHD+fRRx9l+vTpgPpGKqYifRIeHk5hYSEZGRmXHONqCls1hLe3N507d2bdunWljq9bt47u3bu7qSqpbowxjB07lhUrVvDZZ58RHR1d6nx0dDTh4eGl+qiwsJDExET10VWsf//+pKSkkJyc7Ny6dOnC/fffT3JyMjExMeobKVePHj3KfL3EoUOHaNKkCaDfOVK+3NxcPDxK/wnq6enpXPpdfSMVUZE+6dy5M15eXqXGnDp1iq+//rrqeqlKluGQSrF06VLj5eVl5s2bZ/bt22cSEhKMv7+/OXr0qLtLk2pi9OjRJigoyGzcuNGcOnXKueXm5jrHvPDCCyYoKMisWLHCpKSkmKFDh5qGDRuarKwsN1Yu1c1PVyM0Rn0j5du+fbuxWq1m6tSp5ptvvjGLFy82fn5+ZtGiRc4x6h35uREjRphGjRqZDz74wBw5csSsWLHChISEmCeffNI5Rn0jxjhWyd21a5fZtWuXAczMmTPNrl27zLFjx4wxFeuTUaNGmcjISLN+/Xqzc+dO069fP9OxY0dTXFxcJdegsFXDvPbaa6ZJkybG29vbdOrUybmkt4gxjmVRy9sWLFjgHGO3282kSZNMeHi4sdlspnfv3iYlJcV9RUu19POwpb6RS3n//fdNu3btjM1mM61atTJvvPFGqfPqHfm5rKwsM27cONO4cWPj4+NjYmJizMSJE01BQYFzjPpGjDFmw4YN5f5dM2LECGNMxfokLy/PjB071gQHBxtfX18zaNAg891331XZNViMMaZq5tBERERERESuHnpmS0RERERExAUUtkRERERERFxAYUtERERERMQFFLZERERERERcQGFLRERERETEBRS2REREREREXEBhS0RERERExAUUtkRERERERFxAYUtERMTFLBYLq1atcncZIiJSxRS2RESkVhs5ciQWi6XMNmDAAHeXJiIitZzV3QWIiIi42oABA1iwYEGpYzabzU3ViIjI1UIzWyIiUuvZbDbCw8NLbfXq1QMct/jNmTOHgQMH4uvrS3R0NMuXLy/1/pSUFPr164evry/169fnoYce4uLFi6XGzJ8/n7Zt22Kz2WjYsCFjx44tdf7s2bPccccd+Pn50bx5c1avXu3aixYREbdT2BIRkaves88+y5133snu3bsZNmwYQ4cOZf/+/QDk5uYyYMAA6tWrx44dO1i+fDnr168vFabmzJnDmDFjeOihh0hJSWH16tU0a9as1M947rnnuOeee9izZw+33HIL999/P+fPn6/S6xQRkaplMcYYdxchIiLiKiNHjmTRokX4+PiUOj5hwgSeffZZLBYLo0aNYs6cOc5zXbt2pVOnTsyePZs333yTCRMmkJqair+/PwBr1qzhtttu4+TJk4SFhdGoUSN+//vfM2XKlHJrsFgsPPPMMzz//PMA5OTkEBAQwJo1a/TsmIhILaZntkREpNa74YYbSoUpgODgYOd+t27dSp3r1q0bycnJAOzfv5+OHTs6gxZAjx49sNvtHDx4EIvFwsmTJ+nfv/8v1tChQwfnvr+/PwEBAaSnp1/pJYmISA2gsCUiIrWev79/mdv6fo3FYgHAGOPcL2+Mr69vhT7Py8urzHvtdvtl1SQiIjWLntkSEZGr3rZt28q8btWqFQBt2rQhOTmZnJwc5/ktW7bg4eFBixYtCAgIoGnTpnz66adVWrOIiFR/mtkSEZFar6CggLS0tFLHrFYrISEhACxfvpwuXbrQs2dPFi9ezPbt25k3bx4A999/P5MmTWLEiBFMnjyZM2fOEB8fz/DhwwkLCwNg8uTJjBo1itDQUAYOHEh2djZbtmwhPj6+ai9URESqFYUtERGp9dauXUvDhg1LHWvZsiUHDhwAHCsFLl26lIcffpjw8HAWL15MmzZtAPDz8+Pjjz9m3LhxxMbG4ufnx5133snMmTOdnzVixAjy8/N5+eWXefzxxwkJCeGuu+6qugsUEZFqSasRiojIVc1isbBy5UoGDx7s7lJERKSW0TNbIiIiIiIiLqCwJSIiIiIi4gJ6ZktERK5qupteRERcRTNbIiIiIiIiLqCwJSIiIiIi4gIKWyIiIiIiIi6gsCUiIiIiIuICClsiIiIiIiIuoLAlIiIiIiLiAgpbIiIiIiIiLqCwJSIiIiIi4gL/D6x+wvUSZvm5AAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "annec.plot_history(history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot the ANN prediction versus the test target as a time series"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0    2757.110352\n",
      "Name: 6038, dtype: float32\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1a5536ebf10>"
      ]
     },
     "execution_count": 217,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "ax = y_test.plot()\n",
    "pred = pd.DataFrame(index=y_test.index,data= y_test_pred)\n",
    "print(pred.iloc[-1])\n",
    "pred.plot(ax=ax)\n",
    "plt.legend([\"target\",\"ann\"])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Save the final model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: ./Export/emmaton\\assets\n",
      "Model saved at location: ./Export/emmaton\n"
     ]
    }
   ],
   "source": [
    "############################################################################################\n",
    "############# Please change the model_save_path to the direction you saved the files #######\n",
    "############################################################################################\n",
    "\n",
    "model_save_path = r'./Export/emmaton'\n",
    "annec.save_model(model, model_save_path)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "dd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "867/867 [==============================] - 3s 4ms/step\n",
      "189/189 [==============================] - 1s 3ms/step\n",
      "+----+---------+------------+--------------+-------------------------+-----------+-------------+------------------------+\n",
      "|    | Model   |   Train_R2 |   Train_RMSE |   Train_Percentage_Bias |   Test_R2 |   Test_RMSE |   Test_Percentage_Bias |\n",
      "|----+---------+------------+--------------+-------------------------+-----------+-------------+------------------------|\n",
      "|  0 | ANN     |       0.98 |       134.25 |                    4.47 |      0.97 |      183.95 |                   2.05 |\n",
      "+----+---------+------------+--------------+-------------------------+-----------+-------------+------------------------+\n"
     ]
    }
   ],
   "source": [
    "def root_mean_squared_error(y_true, y_pred):\n",
    "    y_true = tf.cast(y_true, tf.float32)\n",
    "    return tf.keras.backend.sqrt(tf.keras.backend.mean(tf.keras.backend.square(y_pred - y_true)))\n",
    "\n",
    "loaded_model = tf.keras.models.load_model(model_save_path, custom_objects={'root_mean_squared_error': root_mean_squared_error})\n",
    "# Calculate metrics for ANN\n",
    "metrics = []\n",
    "y_train_pred = loaded_model.predict(X_train)\n",
    "y_test_pred = loaded_model.predict(X_test)\n",
    "metrics.append(annec.calculate_metrics('ANN', y_train, y_train_pred, y_test, y_test_pred))\n",
    "\n",
    "df3 = pd.DataFrame(metrics)\n",
    "print(tabulate(df3, headers='keys', tablefmt='psql'))\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load the saved model and test it with dataset and one sample\n",
    "\n",
    "Currently this tests Emmaton at a time when EC is 200-ish which is not a very compelling example. It also appears to use the same inputs as the test/training process, so the above example should cover it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "in\n",
      "var            sac                                                          \\\n",
      "lag             0d            1d            2d            3d            4d   \n",
      "0     22290.015480  22333.700680  22358.590290  22373.948620  22383.928890   \n",
      "1     22203.284070  22290.015480  22333.700680  22358.590290  22373.948620   \n",
      "2     21992.031070  22203.284070  22290.015480  22333.700680  22358.590290   \n",
      "3     18938.581990  21992.031070  22203.284070  22290.015480  22333.700680   \n",
      "4     18157.222480  18938.581990  21992.031070  22203.284070  22290.015480   \n",
      "...            ...           ...           ...           ...           ...   \n",
      "6034   7178.489123   7169.338157   7147.221362   7067.043165   6879.099909   \n",
      "6035   7183.190302   7178.489123   7169.338157   7147.221362   7067.043165   \n",
      "6036   7185.981787   7183.190302   7178.489123   7169.338157   7147.221362   \n",
      "6037   7187.839785   7185.981787   7183.190302   7178.489123   7169.338157   \n",
      "6038   7189.208433   7187.839785   7185.981787   7183.190302   7178.489123   \n",
      "\n",
      "var                                                                         \\\n",
      "lag             5d            6d            7d          1ave          2ave   \n",
      "0     22390.611130  22395.122670  22398.108620  22391.450130  20455.316660   \n",
      "1     22383.928890  22390.611130  22395.122670  22394.405490  21187.055720   \n",
      "2     22373.948620  22383.928890  22390.611130  22396.152440  21864.788650   \n",
      "3     22358.590290  22373.948620  22383.928890  22396.791800  22157.441390   \n",
      "4     22333.700680  22358.590290  22373.948620  22396.300940  22260.213840   \n",
      "...            ...           ...           ...           ...           ...   \n",
      "6034   6855.996998   6846.414762   6841.590561   6836.436623   6858.925169   \n",
      "6035   6879.099909   6855.996998   6846.414762   6836.786803   6852.697205   \n",
      "6036   7067.043165   6879.099909   6855.996998   6837.654798   6848.232893   \n",
      "6037   7147.221362   7067.043165   6879.099909   6839.453782   6844.936756   \n",
      "6038   7169.338157   7147.221362   7067.043165   6843.395764   6842.454345   \n",
      "\n",
      "var                                                                        \\\n",
      "lag           3ave          4ave          5ave          6ave         7ave   \n",
      "0     13808.893330  13727.313470  15871.353880  18889.589800  19086.62041   \n",
      "1     13863.052790  13723.464770  15415.591940  18875.798590  19027.33307   \n",
      "2     13971.768870  13721.412930  14965.451400  18855.528540  18992.74727   \n",
      "3     14465.724870  13720.892800  14528.431130  18821.662790  18970.20303   \n",
      "4     15149.350230  13721.815370  14125.991190  18753.022640  18954.27960   \n",
      "...            ...           ...           ...           ...          ...   \n",
      "6034   8468.318958   9581.265866   9610.524133  10102.485650  10216.69770   \n",
      "6035   8228.374296   9576.938384   9604.657542  10053.670210  10211.79019   \n",
      "6036   7988.296654   9571.021454   9601.022381  10002.754410  10207.59661   \n",
      "6037   7749.724416   9562.502747   9598.449048   9950.937071  10203.87662   \n",
      "6038   7515.059380   9549.369973   9596.417044   9898.797265  10200.40479   \n",
      "\n",
      "var                                          \n",
      "lag          8ave         9ave        10ave  \n",
      "0     27995.04227  29834.38514  30086.23375  \n",
      "1     27067.75544  29827.11331  29963.14182  \n",
      "2     26117.13349  29818.32929  29918.20477  \n",
      "3     25156.58976  29807.29370  29895.15910  \n",
      "4     24192.67620  29792.87911  29881.01984  \n",
      "...           ...          ...          ...  \n",
      "6034  11082.25968  20041.75615  20414.36335  \n",
      "6035  10620.48233  19579.94320  20417.26775  \n",
      "6036  10453.77877  18823.16760  20417.52532  \n",
      "6037  10370.22916  17983.96183  20415.15283  \n",
      "6038  10321.02636  17111.85977  20409.89260  \n",
      "\n",
      "[6039 rows x 18 columns]\n",
      "out\n",
      "Sample\n",
      "[array([[ 7717.142,  7716.127,  7715.153,  7714.168,  7713.113,  7711.911,\n",
      "         7710.454,  7708.58 ,  7413.865,  6964.834,  6990.893,  9085.646,\n",
      "        10543.928, 10538.482,  9573.437,  9400.652, 10259.991, 19202.16 ]]), array([[ 3505.366,  3505.366,  3505.366,  3505.366,  3505.366,  3505.366,\n",
      "         3505.366,  3505.366,  3228.763,  2744.708,  2744.708,  4691.516,\n",
      "         6313.855,  6313.855,  4611.389,  4233.064,  4233.064, 11767.55 ]]), array([[ 0.   ,  0.   ,  0.   ,  0.   ,  0.   ,  0.   ,  0.   ,  0.   ,\n",
      "         0.   ,  0.   ,  0.   , 16.364, 30.   , 30.   , 30.818, 31.   ,\n",
      "        31.   , 31.   ]]), array([[1632.914, 1632.914, 1632.914, 1632.914, 1632.914, 1632.914,\n",
      "        1632.914, 1632.914, 1364.777,  895.536,  895.536, 1093.966,\n",
      "        1259.325, 1259.325, 2029.012, 2200.054, 2200.054, 2565.653]]), array([[1277.638, 1277.781, 1277.942, 1278.132, 1278.363, 1278.656,\n",
      "        1279.04 , 1279.564, 1347.192, 1445.815, 1443.771, 1205.677,\n",
      "        1039.761, 1039.77 , 1108.564, 1121.629, 1104.088,  920.058]]), array([[6.56 , 6.184, 5.508, 5.083, 4.913, 5.024, 5.233, 5.991, 5.958,\n",
      "        5.254, 5.719, 4.515, 5.757, 5.255, 4.907, 6.271, 5.489, 6.276]]), array([[0.   , 0.   , 0.   , 0.   , 0.   , 0.   , 0.   , 0.   , 0.   ,\n",
      "        0.   , 0.   , 0.   , 0.   , 0.   , 0.818, 1.   , 1.   , 1.   ]])]\n",
      "1/1 [==============================] - 0s 174ms/step\n",
      "[[3275.4763]]\n"
     ]
    }
   ],
   "source": [
    "def root_mean_squared_error(y_true, y_pred):\n",
    "    y_true = tf.cast(y_true, tf.float32)\n",
    "    return tf.keras.backend.sqrt(tf.keras.backend.mean(tf.keras.backend.square(y_pred - y_true)))\n",
    "\n",
    "loaded_model = tf.keras.models.load_model(model_save_path, custom_objects={'root_mean_squared_error': root_mean_squared_error})\n",
    "# Calculate metrics for \n",
    "\n",
    "df_sample = pd.read_csv(\"F:/projects/ann_calsim_ec_estimator/Inputs/selected.csv\", header=None)\n",
    "input_sample = [frame.to_frame().to_numpy().transpose() for ndx,frame in df_sample.iterrows()] \n",
    "print(\"Sample\")\n",
    "print(input_sample)\n",
    "y_sample_pred = loaded_model.predict(input_sample)\n",
    "print(y_sample_pred)\n",
    "\n",
    "\n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "DWR_ML",
   "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": "3.10.12"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}