experiment_point_addition_and_removal.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "ca35b758",
   "metadata": {},
   "source": [
    "# Point Addition and Removal\n",
    "\n",
    "We perform point addition and removal experiments which are commonly used in evaluation of data valuation techniques. Particularly, we measure the **expected** model performance after data deletion to adapt to our problem setting. A sample run is given below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "229af6d7",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2023-05-16 14:03:43,602 - 29 - INFO - Program has started.\n",
      "2023-05-16 14:03:43,602 - 29 - INFO - Program has started.\n",
      "2023-05-16 14:03:43,602 - 29 - INFO - Program has started.\n",
      "2023-05-16 14:03:43,602 - 29 - INFO - Program has started.\n",
      "2023-05-16 14:03:43,602 - 29 - INFO - Program has started.\n",
      "2023-05-16 14:03:43,625 - 53 - INFO - ====> Iteration 0 =====\n",
      "2023-05-16 14:03:43,625 - 53 - INFO - ====> Iteration 0 =====\n",
      "2023-05-16 14:03:43,625 - 53 - INFO - ====> Iteration 0 =====\n",
      "2023-05-16 14:03:43,625 - 53 - INFO - ====> Iteration 0 =====\n",
      "2023-05-16 14:03:43,625 - 53 - INFO - ====> Iteration 0 =====\n",
      "2023-05-16 14:03:43,747 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:03:43,747 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:03:43,747 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:03:43,747 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:03:43,747 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:03:43,965 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:03:43,965 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:03:43,965 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:03:43,965 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:03:43,965 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9982031842987031\n",
      "---------> Number of values that have not converged: 50\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9991201504813261\n",
      "---------> Number of values that have not converged: 24\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.99954421739141\n",
      "---------> Number of values that have not converged: 8\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 1.0000599142695699\n",
      "---------> Number of values that have not converged: 5\n",
      "====> Monte-Carlo Round 2500 - Average convergence rate = 0.9999936845398261\n",
      "---------> Number of values that have not converged: 2\n",
      "====> Monte-Carlo Round 3000 - Average convergence rate = 1.000365374327927\n",
      "---------> Number of values that have not converged: 2\n",
      "2023-05-16 14:13:44,067 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:13:44,067 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:13:44,067 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:13:44,067 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:13:44,067 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9993411812113099\n",
      "---------> Number of values that have not converged: 25\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9988237319405913\n",
      "---------> Number of values that have not converged: 6\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9998444586377943\n",
      "---------> Number of values that have not converged: 2\n",
      "2023-05-16 14:20:24,883 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:20:24,883 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:20:24,883 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:20:24,883 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:20:24,883 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:20:24,889 - 53 - INFO - ====> Iteration 1 =====\n",
      "2023-05-16 14:20:24,889 - 53 - INFO - ====> Iteration 1 =====\n",
      "2023-05-16 14:20:24,889 - 53 - INFO - ====> Iteration 1 =====\n",
      "2023-05-16 14:20:24,889 - 53 - INFO - ====> Iteration 1 =====\n",
      "2023-05-16 14:20:24,889 - 53 - INFO - ====> Iteration 1 =====\n",
      "2023-05-16 14:20:25,005 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:20:25,005 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:20:25,005 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:20:25,005 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:20:25,005 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:20:25,214 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:20:25,214 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:20:25,214 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:20:25,214 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:20:25,214 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9999469829256139\n",
      "---------> Number of values that have not converged: 50\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9993283981684996\n",
      "---------> Number of values that have not converged: 21\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9997464959130337\n",
      "---------> Number of values that have not converged: 5\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9999430126877109\n",
      "---------> Number of values that have not converged: 2\n",
      "====> Monte-Carlo Round 2500 - Average convergence rate = 0.9998656689677994\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 14:28:57,726 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:28:57,726 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:28:57,726 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:28:57,726 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:28:57,726 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0017679565890054\n",
      "---------> Number of values that have not converged: 15\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0006432148615476\n",
      "---------> Number of values that have not converged: 7\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9997405489269964\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 14:34:52,745 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:34:52,745 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:34:52,745 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:34:52,745 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:34:52,745 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:34:52,750 - 53 - INFO - ====> Iteration 2 =====\n",
      "2023-05-16 14:34:52,750 - 53 - INFO - ====> Iteration 2 =====\n",
      "2023-05-16 14:34:52,750 - 53 - INFO - ====> Iteration 2 =====\n",
      "2023-05-16 14:34:52,750 - 53 - INFO - ====> Iteration 2 =====\n",
      "2023-05-16 14:34:52,750 - 53 - INFO - ====> Iteration 2 =====\n",
      "2023-05-16 14:34:52,865 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:34:52,865 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:34:52,865 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:34:52,865 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:34:52,865 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:34:53,093 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:34:53,093 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:34:53,093 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:34:53,093 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:34:53,093 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.000150027676932\n",
      "---------> Number of values that have not converged: 55\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0009043046259338\n",
      "---------> Number of values that have not converged: 30\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.0003033506761796\n",
      "---------> Number of values that have not converged: 11\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 1.000090368544871\n",
      "---------> Number of values that have not converged: 5\n",
      "2023-05-16 14:41:57,021 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:41:57,021 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:41:57,021 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:41:57,021 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:41:57,021 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9995644300965083\n",
      "---------> Number of values that have not converged: 14\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9997891518852\n",
      "---------> Number of values that have not converged: 2\n",
      "2023-05-16 14:46:19,052 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:46:19,052 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:46:19,052 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:46:19,052 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:46:19,052 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:46:19,057 - 53 - INFO - ====> Iteration 3 =====\n",
      "2023-05-16 14:46:19,057 - 53 - INFO - ====> Iteration 3 =====\n",
      "2023-05-16 14:46:19,057 - 53 - INFO - ====> Iteration 3 =====\n",
      "2023-05-16 14:46:19,057 - 53 - INFO - ====> Iteration 3 =====\n",
      "2023-05-16 14:46:19,057 - 53 - INFO - ====> Iteration 3 =====\n",
      "2023-05-16 14:46:19,147 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:46:19,147 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:46:19,147 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:46:19,147 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:46:19,147 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:46:19,357 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:46:19,357 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:46:19,357 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:46:19,357 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:46:19,357 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9996137788697331\n",
      "---------> Number of values that have not converged: 56\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0000541326987549\n",
      "---------> Number of values that have not converged: 18\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.0002053312426462\n",
      "---------> Number of values that have not converged: 8\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9997867380951512\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 14:53:11,128 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:53:11,128 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:53:11,128 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:53:11,128 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 14:53:11,128 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9996341462700888\n",
      "---------> Number of values that have not converged: 42\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9994879216852746\n",
      "---------> Number of values that have not converged: 6\n",
      "2023-05-16 14:57:22,459 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:57:22,459 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:57:22,459 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:57:22,459 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:57:22,459 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 14:57:22,463 - 53 - INFO - ====> Iteration 4 =====\n",
      "2023-05-16 14:57:22,463 - 53 - INFO - ====> Iteration 4 =====\n",
      "2023-05-16 14:57:22,463 - 53 - INFO - ====> Iteration 4 =====\n",
      "2023-05-16 14:57:22,463 - 53 - INFO - ====> Iteration 4 =====\n",
      "2023-05-16 14:57:22,463 - 53 - INFO - ====> Iteration 4 =====\n",
      "2023-05-16 14:57:22,560 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:57:22,560 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:57:22,560 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:57:22,560 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:57:22,560 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 14:57:22,756 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:57:22,756 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:57:22,756 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:57:22,756 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 14:57:22,756 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9993036642565316\n",
      "---------> Number of values that have not converged: 75\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9998027313819956\n",
      "---------> Number of values that have not converged: 22\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.000053813514052\n",
      "---------> Number of values that have not converged: 6\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9997870386316581\n",
      "---------> Number of values that have not converged: 4\n",
      "====> Monte-Carlo Round 2500 - Average convergence rate = 0.9995738150591356\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 15:04:50,735 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:04:50,735 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:04:50,735 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:04:50,735 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:04:50,735 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0010702145150754\n",
      "---------> Number of values that have not converged: 42\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.000086026692853\n",
      "---------> Number of values that have not converged: 2\n",
      "2023-05-16 15:08:44,692 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:08:44,692 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:08:44,692 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:08:44,692 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:08:44,692 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:08:44,698 - 53 - INFO - ====> Iteration 5 =====\n",
      "2023-05-16 15:08:44,698 - 53 - INFO - ====> Iteration 5 =====\n",
      "2023-05-16 15:08:44,698 - 53 - INFO - ====> Iteration 5 =====\n",
      "2023-05-16 15:08:44,698 - 53 - INFO - ====> Iteration 5 =====\n",
      "2023-05-16 15:08:44,698 - 53 - INFO - ====> Iteration 5 =====\n",
      "2023-05-16 15:08:44,800 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:08:44,800 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:08:44,800 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:08:44,800 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:08:44,800 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:08:45,003 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:08:45,003 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:08:45,003 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:08:45,003 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:08:45,003 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9991798757495917\n",
      "---------> Number of values that have not converged: 59\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9999301613278866\n",
      "---------> Number of values that have not converged: 18\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.0003792085887853\n",
      "---------> Number of values that have not converged: 9\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9998476078841609\n",
      "---------> Number of values that have not converged: 2\n",
      "2023-05-16 15:15:32,739 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:15:32,739 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:15:32,739 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:15:32,739 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:15:32,739 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9994911860652403\n",
      "---------> Number of values that have not converged: 53\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9997055003481887\n",
      "---------> Number of values that have not converged: 18\n",
      "2023-05-16 15:19:27,378 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:19:27,378 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:19:27,378 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:19:27,378 - 85 - INFO - ---------> DeRDaVa done.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2023-05-16 15:19:27,378 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:19:27,383 - 53 - INFO - ====> Iteration 6 =====\n",
      "2023-05-16 15:19:27,383 - 53 - INFO - ====> Iteration 6 =====\n",
      "2023-05-16 15:19:27,383 - 53 - INFO - ====> Iteration 6 =====\n",
      "2023-05-16 15:19:27,383 - 53 - INFO - ====> Iteration 6 =====\n",
      "2023-05-16 15:19:27,383 - 53 - INFO - ====> Iteration 6 =====\n",
      "2023-05-16 15:19:27,468 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:19:27,468 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:19:27,468 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:19:27,468 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:19:27,468 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:19:27,677 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:19:27,677 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:19:27,677 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:19:27,677 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:19:27,677 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0011900263405686\n",
      "---------> Number of values that have not converged: 52\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9998305734478597\n",
      "---------> Number of values that have not converged: 20\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9999764147770679\n",
      "---------> Number of values that have not converged: 7\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 1.000017221860326\n",
      "---------> Number of values that have not converged: 3\n",
      "====> Monte-Carlo Round 2500 - Average convergence rate = 0.9999106860982603\n",
      "---------> Number of values that have not converged: 1\n",
      "====> Monte-Carlo Round 3000 - Average convergence rate = 1.0000471896067256\n",
      "---------> Number of values that have not converged: 2\n",
      "====> Monte-Carlo Round 3500 - Average convergence rate = 1.0000900626106002\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 15:30:56,666 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:30:56,666 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:30:56,666 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:30:56,666 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:30:56,666 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0006544107296884\n",
      "---------> Number of values that have not converged: 22\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9999209123974494\n",
      "---------> Number of values that have not converged: 5\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.000582787130873\n",
      "---------> Number of values that have not converged: 4\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9996538535922873\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 15:38:02,306 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:38:02,306 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:38:02,306 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:38:02,306 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:38:02,306 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:38:02,310 - 53 - INFO - ====> Iteration 7 =====\n",
      "2023-05-16 15:38:02,310 - 53 - INFO - ====> Iteration 7 =====\n",
      "2023-05-16 15:38:02,310 - 53 - INFO - ====> Iteration 7 =====\n",
      "2023-05-16 15:38:02,310 - 53 - INFO - ====> Iteration 7 =====\n",
      "2023-05-16 15:38:02,310 - 53 - INFO - ====> Iteration 7 =====\n",
      "2023-05-16 15:38:02,427 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:38:02,427 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:38:02,427 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:38:02,427 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:38:02,427 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:38:02,621 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:38:02,621 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:38:02,621 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:38:02,621 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:38:02,621 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9997392075479173\n",
      "---------> Number of values that have not converged: 57\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9995682083820857\n",
      "---------> Number of values that have not converged: 16\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9998910550413589\n",
      "---------> Number of values that have not converged: 7\n",
      "2023-05-16 15:43:51,971 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:43:51,971 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:43:51,971 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:43:51,971 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:43:51,971 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9991650187731789\n",
      "---------> Number of values that have not converged: 16\n",
      "2023-05-16 15:46:59,481 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:46:59,481 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:46:59,481 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:46:59,481 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:46:59,481 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 15:46:59,485 - 53 - INFO - ====> Iteration 8 =====\n",
      "2023-05-16 15:46:59,485 - 53 - INFO - ====> Iteration 8 =====\n",
      "2023-05-16 15:46:59,485 - 53 - INFO - ====> Iteration 8 =====\n",
      "2023-05-16 15:46:59,485 - 53 - INFO - ====> Iteration 8 =====\n",
      "2023-05-16 15:46:59,485 - 53 - INFO - ====> Iteration 8 =====\n",
      "2023-05-16 15:46:59,563 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:46:59,563 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:46:59,563 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:46:59,563 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:46:59,563 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 15:46:59,758 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:46:59,758 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:46:59,758 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:46:59,758 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 15:46:59,758 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0010914095365675\n",
      "---------> Number of values that have not converged: 55\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0001794918422708\n",
      "---------> Number of values that have not converged: 23\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.0001644136775223\n",
      "---------> Number of values that have not converged: 13\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9998645130543315\n",
      "---------> Number of values that have not converged: 3\n",
      "====> Monte-Carlo Round 2500 - Average convergence rate = 0.9998347009177276\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 15:54:24,583 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:54:24,583 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:54:24,583 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:54:24,583 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 15:54:24,583 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0012362421893675\n",
      "---------> Number of values that have not converged: 49\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0015101571622487\n",
      "---------> Number of values that have not converged: 23\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.0004670830877653\n",
      "---------> Number of values that have not converged: 11\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 1.0009514794522965\n",
      "---------> Number of values that have not converged: 8\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2023-05-16 16:02:10,095 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:02:10,095 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:02:10,095 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:02:10,095 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:02:10,095 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:02:10,100 - 53 - INFO - ====> Iteration 9 =====\n",
      "2023-05-16 16:02:10,100 - 53 - INFO - ====> Iteration 9 =====\n",
      "2023-05-16 16:02:10,100 - 53 - INFO - ====> Iteration 9 =====\n",
      "2023-05-16 16:02:10,100 - 53 - INFO - ====> Iteration 9 =====\n",
      "2023-05-16 16:02:10,100 - 53 - INFO - ====> Iteration 9 =====\n",
      "2023-05-16 16:02:10,207 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:02:10,207 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:02:10,207 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:02:10,207 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:02:10,207 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:02:10,405 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:02:10,405 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:02:10,405 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:02:10,405 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:02:10,405 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9998991814323606\n",
      "---------> Number of values that have not converged: 53\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9991834430613034\n",
      "---------> Number of values that have not converged: 18\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9995819276046299\n",
      "---------> Number of values that have not converged: 7\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9997067093645008\n",
      "---------> Number of values that have not converged: 1\n",
      "====> Monte-Carlo Round 2500 - Average convergence rate = 0.9995811383008596\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 16:09:29,671 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:09:29,671 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:09:29,671 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:09:29,671 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:09:29,671 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0001109763783562\n",
      "---------> Number of values that have not converged: 41\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9993886992124138\n",
      "---------> Number of values that have not converged: 11\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.0002075524995055\n",
      "---------> Number of values that have not converged: 5\n",
      "2023-05-16 16:14:52,582 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:14:52,582 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:14:52,582 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:14:52,582 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:14:52,582 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:14:52,587 - 53 - INFO - ====> Iteration 10 =====\n",
      "2023-05-16 16:14:52,587 - 53 - INFO - ====> Iteration 10 =====\n",
      "2023-05-16 16:14:52,587 - 53 - INFO - ====> Iteration 10 =====\n",
      "2023-05-16 16:14:52,587 - 53 - INFO - ====> Iteration 10 =====\n",
      "2023-05-16 16:14:52,587 - 53 - INFO - ====> Iteration 10 =====\n",
      "2023-05-16 16:14:52,696 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:14:52,696 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:14:52,696 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:14:52,696 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:14:52,696 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:14:52,891 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:14:52,891 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:14:52,891 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:14:52,891 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:14:52,891 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.999703135843713\n",
      "---------> Number of values that have not converged: 62\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9996418640393219\n",
      "---------> Number of values that have not converged: 21\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.0002299832965063\n",
      "---------> Number of values that have not converged: 11\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 1.000132950503223\n",
      "---------> Number of values that have not converged: 4\n",
      "====> Monte-Carlo Round 2500 - Average convergence rate = 0.9999901999378751\n",
      "---------> Number of values that have not converged: 3\n",
      "2023-05-16 16:22:35,747 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:22:35,747 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:22:35,747 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:22:35,747 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:22:35,747 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9987210735926635\n",
      "---------> Number of values that have not converged: 35\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0013441319136644\n",
      "---------> Number of values that have not converged: 15\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.000860957464856\n",
      "---------> Number of values that have not converged: 3\n",
      "2023-05-16 16:28:01,747 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:28:01,747 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:28:01,747 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:28:01,747 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:28:01,747 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:28:01,750 - 53 - INFO - ====> Iteration 11 =====\n",
      "2023-05-16 16:28:01,750 - 53 - INFO - ====> Iteration 11 =====\n",
      "2023-05-16 16:28:01,750 - 53 - INFO - ====> Iteration 11 =====\n",
      "2023-05-16 16:28:01,750 - 53 - INFO - ====> Iteration 11 =====\n",
      "2023-05-16 16:28:01,750 - 53 - INFO - ====> Iteration 11 =====\n",
      "2023-05-16 16:28:01,837 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:28:01,837 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:28:01,837 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:28:01,837 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:28:01,837 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:28:02,030 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:28:02,030 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:28:02,030 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:28:02,030 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:28:02,030 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9996706156958125\n",
      "---------> Number of values that have not converged: 47\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9997931918219911\n",
      "---------> Number of values that have not converged: 25\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9995496743376528\n",
      "---------> Number of values that have not converged: 4\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9999153757155002\n",
      "---------> Number of values that have not converged: 2\n",
      "2023-05-16 16:34:02,077 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:34:02,077 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:34:02,077 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:34:02,077 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:34:02,077 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.000118927807521\n",
      "---------> Number of values that have not converged: 56\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9999641092594187\n",
      "---------> Number of values that have not converged: 23\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.000168284194373\n",
      "---------> Number of values that have not converged: 4\n",
      "2023-05-16 16:39:29,433 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:39:29,433 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:39:29,433 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:39:29,433 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:39:29,433 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:39:29,438 - 53 - INFO - ====> Iteration 12 =====\n",
      "2023-05-16 16:39:29,438 - 53 - INFO - ====> Iteration 12 =====\n",
      "2023-05-16 16:39:29,438 - 53 - INFO - ====> Iteration 12 =====\n",
      "2023-05-16 16:39:29,438 - 53 - INFO - ====> Iteration 12 =====\n",
      "2023-05-16 16:39:29,438 - 53 - INFO - ====> Iteration 12 =====\n",
      "2023-05-16 16:39:29,531 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:39:29,531 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:39:29,531 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:39:29,531 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:39:29,531 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:39:29,724 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:39:29,724 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:39:29,724 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:39:29,724 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:39:29,724 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0006441849348093\n",
      "---------> Number of values that have not converged: 54\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0000821827326443\n",
      "---------> Number of values that have not converged: 23\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.999780708340056\n",
      "---------> Number of values that have not converged: 7\n",
      "2023-05-16 16:45:12,138 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:45:12,138 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:45:12,138 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:45:12,138 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:45:12,138 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9994574598903153\n",
      "---------> Number of values that have not converged: 9\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0004507906669897\n",
      "---------> Number of values that have not converged: 5\n",
      "2023-05-16 16:49:48,196 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:49:48,196 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:49:48,196 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:49:48,196 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:49:48,196 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 16:49:48,200 - 53 - INFO - ====> Iteration 13 =====\n",
      "2023-05-16 16:49:48,200 - 53 - INFO - ====> Iteration 13 =====\n",
      "2023-05-16 16:49:48,200 - 53 - INFO - ====> Iteration 13 =====\n",
      "2023-05-16 16:49:48,200 - 53 - INFO - ====> Iteration 13 =====\n",
      "2023-05-16 16:49:48,200 - 53 - INFO - ====> Iteration 13 =====\n",
      "2023-05-16 16:49:48,280 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:49:48,280 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:49:48,280 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:49:48,280 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:49:48,280 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 16:49:48,476 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:49:48,476 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:49:48,476 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:49:48,476 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 16:49:48,476 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0003390957296674\n",
      "---------> Number of values that have not converged: 57\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0001683236056698\n",
      "---------> Number of values that have not converged: 28\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9999911418102942\n",
      "---------> Number of values that have not converged: 6\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 1.0000267311406117\n",
      "---------> Number of values that have not converged: 3\n",
      "====> Monte-Carlo Round 2500 - Average convergence rate = 1.00006798016108\n",
      "---------> Number of values that have not converged: 3\n",
      "====> Monte-Carlo Round 3000 - Average convergence rate = 1.000082482096539\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 16:58:59,841 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:58:59,841 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:58:59,841 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:58:59,841 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 16:58:59,841 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.999142173848411\n",
      "---------> Number of values that have not converged: 21\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9994929180904254\n",
      "---------> Number of values that have not converged: 7\n",
      "2023-05-16 17:03:44,848 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:03:44,848 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:03:44,848 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:03:44,848 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:03:44,848 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:03:44,853 - 53 - INFO - ====> Iteration 14 =====\n",
      "2023-05-16 17:03:44,853 - 53 - INFO - ====> Iteration 14 =====\n",
      "2023-05-16 17:03:44,853 - 53 - INFO - ====> Iteration 14 =====\n",
      "2023-05-16 17:03:44,853 - 53 - INFO - ====> Iteration 14 =====\n",
      "2023-05-16 17:03:44,853 - 53 - INFO - ====> Iteration 14 =====\n",
      "2023-05-16 17:03:44,989 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:03:44,989 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:03:44,989 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:03:44,989 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:03:44,989 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:03:45,190 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:03:45,190 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:03:45,190 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:03:45,190 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:03:45,190 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9992424076135541\n",
      "---------> Number of values that have not converged: 58\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0002518116254635\n",
      "---------> Number of values that have not converged: 34\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.0003170279376707\n",
      "---------> Number of values that have not converged: 10\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 1.0003278280627361\n",
      "---------> Number of values that have not converged: 8\n",
      "2023-05-16 17:10:44,840 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:10:44,840 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:10:44,840 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:10:44,840 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:10:44,840 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0002894790642574\n",
      "---------> Number of values that have not converged: 49\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9999678712234743\n",
      "---------> Number of values that have not converged: 9\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9998404528797508\n",
      "---------> Number of values that have not converged: 13\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.999777598657432\n",
      "---------> Number of values that have not converged: 1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====> Monte-Carlo Round 2500 - Average convergence rate = 1.0004889410449942\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 17:20:43,053 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:20:43,053 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:20:43,053 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:20:43,053 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:20:43,053 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:20:43,057 - 53 - INFO - ====> Iteration 15 =====\n",
      "2023-05-16 17:20:43,057 - 53 - INFO - ====> Iteration 15 =====\n",
      "2023-05-16 17:20:43,057 - 53 - INFO - ====> Iteration 15 =====\n",
      "2023-05-16 17:20:43,057 - 53 - INFO - ====> Iteration 15 =====\n",
      "2023-05-16 17:20:43,057 - 53 - INFO - ====> Iteration 15 =====\n",
      "2023-05-16 17:20:43,176 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:20:43,176 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:20:43,176 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:20:43,176 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:20:43,176 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:20:43,387 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:20:43,387 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:20:43,387 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:20:43,387 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:20:43,387 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0012561884652644\n",
      "---------> Number of values that have not converged: 64\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0000252925951771\n",
      "---------> Number of values that have not converged: 21\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.0000480423475933\n",
      "---------> Number of values that have not converged: 6\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 1.0002669373393405\n",
      "---------> Number of values that have not converged: 1\n",
      "====> Monte-Carlo Round 2500 - Average convergence rate = 1.00034073585415\n",
      "---------> Number of values that have not converged: 1\n",
      "====> Monte-Carlo Round 3000 - Average convergence rate = 1.0001260727771317\n",
      "---------> Number of values that have not converged: 2\n",
      "2023-05-16 17:30:37,952 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:30:37,952 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:30:37,952 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:30:37,952 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:30:37,952 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9976708005282375\n",
      "---------> Number of values that have not converged: 56\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9978493298321286\n",
      "---------> Number of values that have not converged: 17\n",
      "2023-05-16 17:34:51,057 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:34:51,057 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:34:51,057 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:34:51,057 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:34:51,057 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:34:51,061 - 53 - INFO - ====> Iteration 16 =====\n",
      "2023-05-16 17:34:51,061 - 53 - INFO - ====> Iteration 16 =====\n",
      "2023-05-16 17:34:51,061 - 53 - INFO - ====> Iteration 16 =====\n",
      "2023-05-16 17:34:51,061 - 53 - INFO - ====> Iteration 16 =====\n",
      "2023-05-16 17:34:51,061 - 53 - INFO - ====> Iteration 16 =====\n",
      "2023-05-16 17:34:51,173 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:34:51,173 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:34:51,173 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:34:51,173 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:34:51,173 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:34:51,383 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:34:51,383 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:34:51,383 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:34:51,383 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:34:51,383 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9990752641293531\n",
      "---------> Number of values that have not converged: 54\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9993354010330908\n",
      "---------> Number of values that have not converged: 19\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.0002318825474015\n",
      "---------> Number of values that have not converged: 10\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 1.0000867284544874\n",
      "---------> Number of values that have not converged: 3\n",
      "2023-05-16 17:41:41,324 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:41:41,324 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:41:41,324 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:41:41,324 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:41:41,324 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0007215833409246\n",
      "---------> Number of values that have not converged: 12\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9998856230182153\n",
      "---------> Number of values that have not converged: 8\n",
      "2023-05-16 17:46:39,717 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:46:39,717 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:46:39,717 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:46:39,717 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:46:39,717 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 17:46:39,723 - 53 - INFO - ====> Iteration 17 =====\n",
      "2023-05-16 17:46:39,723 - 53 - INFO - ====> Iteration 17 =====\n",
      "2023-05-16 17:46:39,723 - 53 - INFO - ====> Iteration 17 =====\n",
      "2023-05-16 17:46:39,723 - 53 - INFO - ====> Iteration 17 =====\n",
      "2023-05-16 17:46:39,723 - 53 - INFO - ====> Iteration 17 =====\n",
      "2023-05-16 17:46:39,816 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:46:39,816 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:46:39,816 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:46:39,816 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:46:39,816 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 17:46:40,036 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:46:40,036 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:46:40,036 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:46:40,036 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 17:46:40,036 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.999697529457171\n",
      "---------> Number of values that have not converged: 59\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.999382007382835\n",
      "---------> Number of values that have not converged: 20\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9995400601935717\n",
      "---------> Number of values that have not converged: 5\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.999920441667437\n",
      "---------> Number of values that have not converged: 2\n",
      "2023-05-16 17:54:18,870 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:54:18,870 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:54:18,870 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:54:18,870 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 17:54:18,870 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.001253262883729\n",
      "---------> Number of values that have not converged: 42\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9994680907944943\n",
      "---------> Number of values that have not converged: 14\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9996714880381051\n",
      "---------> Number of values that have not converged: 1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2023-05-16 18:01:03,087 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:01:03,087 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:01:03,087 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:01:03,087 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:01:03,087 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:01:03,091 - 53 - INFO - ====> Iteration 18 =====\n",
      "2023-05-16 18:01:03,091 - 53 - INFO - ====> Iteration 18 =====\n",
      "2023-05-16 18:01:03,091 - 53 - INFO - ====> Iteration 18 =====\n",
      "2023-05-16 18:01:03,091 - 53 - INFO - ====> Iteration 18 =====\n",
      "2023-05-16 18:01:03,091 - 53 - INFO - ====> Iteration 18 =====\n",
      "2023-05-16 18:01:03,187 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:01:03,187 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:01:03,187 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:01:03,187 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:01:03,187 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:01:03,385 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:01:03,385 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:01:03,385 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:01:03,385 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:01:03,385 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0003468158986095\n",
      "---------> Number of values that have not converged: 58\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0002496676689112\n",
      "---------> Number of values that have not converged: 21\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9999843120960553\n",
      "---------> Number of values that have not converged: 8\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 1.00033411407372\n",
      "---------> Number of values that have not converged: 8\n",
      "2023-05-16 18:08:22,388 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:08:22,388 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:08:22,388 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:08:22,388 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:08:22,388 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0003094712636236\n",
      "---------> Number of values that have not converged: 44\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9991048135816449\n",
      "---------> Number of values that have not converged: 17\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.999438652442874\n",
      "---------> Number of values that have not converged: 2\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9999071703154478\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 18:15:22,526 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:15:22,526 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:15:22,526 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:15:22,526 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:15:22,526 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:15:22,531 - 53 - INFO - ====> Iteration 19 =====\n",
      "2023-05-16 18:15:22,531 - 53 - INFO - ====> Iteration 19 =====\n",
      "2023-05-16 18:15:22,531 - 53 - INFO - ====> Iteration 19 =====\n",
      "2023-05-16 18:15:22,531 - 53 - INFO - ====> Iteration 19 =====\n",
      "2023-05-16 18:15:22,531 - 53 - INFO - ====> Iteration 19 =====\n",
      "2023-05-16 18:15:22,643 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:15:22,643 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:15:22,643 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:15:22,643 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:15:22,643 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:15:22,837 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:15:22,837 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:15:22,837 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:15:22,837 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:15:22,837 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.999832646376803\n",
      "---------> Number of values that have not converged: 61\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0005119780781755\n",
      "---------> Number of values that have not converged: 24\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9998304991473412\n",
      "---------> Number of values that have not converged: 9\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9999968012781524\n",
      "---------> Number of values that have not converged: 4\n",
      "2023-05-16 18:21:52,272 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:21:52,272 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:21:52,272 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:21:52,272 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:21:52,272 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.001392932315685\n",
      "---------> Number of values that have not converged: 48\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0009379546370003\n",
      "---------> Number of values that have not converged: 29\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 1.0004130723711713\n",
      "---------> Number of values that have not converged: 8\n",
      "2023-05-16 18:27:17,443 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:27:17,443 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:27:17,443 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:27:17,443 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:27:17,443 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:27:17,448 - 53 - INFO - ====> Iteration 20 =====\n",
      "2023-05-16 18:27:17,448 - 53 - INFO - ====> Iteration 20 =====\n",
      "2023-05-16 18:27:17,448 - 53 - INFO - ====> Iteration 20 =====\n",
      "2023-05-16 18:27:17,448 - 53 - INFO - ====> Iteration 20 =====\n",
      "2023-05-16 18:27:17,448 - 53 - INFO - ====> Iteration 20 =====\n",
      "2023-05-16 18:27:17,545 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:27:17,545 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:27:17,545 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:27:17,545 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:27:17,545 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:27:17,743 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:27:17,743 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:27:17,743 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:27:17,743 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:27:17,743 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.999436159393041\n",
      "---------> Number of values that have not converged: 57\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9996119021854016\n",
      "---------> Number of values that have not converged: 28\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9998939153976562\n",
      "---------> Number of values that have not converged: 7\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.999788531878607\n",
      "---------> Number of values that have not converged: 2\n",
      "2023-05-16 18:33:49,786 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:33:49,786 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:33:49,786 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:33:49,786 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:33:49,786 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9989854810735078\n",
      "---------> Number of values that have not converged: 9\n",
      "2023-05-16 18:36:56,731 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:36:56,731 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:36:56,731 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:36:56,731 - 85 - INFO - ---------> DeRDaVa done.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2023-05-16 18:36:56,731 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:36:56,734 - 53 - INFO - ====> Iteration 21 =====\n",
      "2023-05-16 18:36:56,734 - 53 - INFO - ====> Iteration 21 =====\n",
      "2023-05-16 18:36:56,734 - 53 - INFO - ====> Iteration 21 =====\n",
      "2023-05-16 18:36:56,734 - 53 - INFO - ====> Iteration 21 =====\n",
      "2023-05-16 18:36:56,734 - 53 - INFO - ====> Iteration 21 =====\n",
      "2023-05-16 18:36:56,803 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:36:56,803 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:36:56,803 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:36:56,803 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:36:56,803 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:36:56,999 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:36:56,999 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:36:56,999 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:36:56,999 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:36:56,999 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9982665279452548\n",
      "---------> Number of values that have not converged: 46\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9990502397759635\n",
      "---------> Number of values that have not converged: 23\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9997344873761668\n",
      "---------> Number of values that have not converged: 5\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9999397194401937\n",
      "---------> Number of values that have not converged: 4\n",
      "====> Monte-Carlo Round 2500 - Average convergence rate = 0.9999708870071311\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 18:45:09,953 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:45:09,953 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:45:09,953 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:45:09,953 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:45:09,953 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.001704125449615\n",
      "---------> Number of values that have not converged: 62\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9994494088546856\n",
      "---------> Number of values that have not converged: 16\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9992879777884071\n",
      "---------> Number of values that have not converged: 2\n",
      "2023-05-16 18:50:38,446 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:50:38,446 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:50:38,446 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:50:38,446 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:50:38,446 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 18:50:38,451 - 53 - INFO - ====> Iteration 22 =====\n",
      "2023-05-16 18:50:38,451 - 53 - INFO - ====> Iteration 22 =====\n",
      "2023-05-16 18:50:38,451 - 53 - INFO - ====> Iteration 22 =====\n",
      "2023-05-16 18:50:38,451 - 53 - INFO - ====> Iteration 22 =====\n",
      "2023-05-16 18:50:38,451 - 53 - INFO - ====> Iteration 22 =====\n",
      "2023-05-16 18:50:38,565 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:50:38,565 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:50:38,565 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:50:38,565 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:50:38,565 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 18:50:38,767 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:50:38,767 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:50:38,767 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:50:38,767 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 18:50:38,767 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9993211025165942\n",
      "---------> Number of values that have not converged: 60\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.9998963130529265\n",
      "---------> Number of values that have not converged: 24\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9997861021028615\n",
      "---------> Number of values that have not converged: 8\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9998698235899433\n",
      "---------> Number of values that have not converged: 3\n",
      "====> Monte-Carlo Round 2500 - Average convergence rate = 0.9998952481087494\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 18:58:09,770 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:58:09,770 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:58:09,770 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:58:09,770 - 77 - INFO - ---------> Shapley done.\n",
      "2023-05-16 18:58:09,770 - 77 - INFO - ---------> Shapley done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 0.9977940881148163\n",
      "---------> Number of values that have not converged: 39\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 0.999815934092413\n",
      "---------> Number of values that have not converged: 5\n",
      "====> Monte-Carlo Round 1500 - Average convergence rate = 0.9999566794870328\n",
      "---------> Number of values that have not converged: 3\n",
      "====> Monte-Carlo Round 2000 - Average convergence rate = 0.9996860101373715\n",
      "---------> Number of values that have not converged: 1\n",
      "2023-05-16 19:06:03,481 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 19:06:03,481 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 19:06:03,481 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 19:06:03,481 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 19:06:03,481 - 85 - INFO - ---------> DeRDaVa done.\n",
      "2023-05-16 19:06:03,486 - 53 - INFO - ====> Iteration 23 =====\n",
      "2023-05-16 19:06:03,486 - 53 - INFO - ====> Iteration 23 =====\n",
      "2023-05-16 19:06:03,486 - 53 - INFO - ====> Iteration 23 =====\n",
      "2023-05-16 19:06:03,486 - 53 - INFO - ====> Iteration 23 =====\n",
      "2023-05-16 19:06:03,486 - 53 - INFO - ====> Iteration 23 =====\n",
      "2023-05-16 19:06:03,604 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 19:06:03,604 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 19:06:03,604 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 19:06:03,604 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 19:06:03,604 - 64 - INFO - ---------> Random done.\n",
      "2023-05-16 19:06:03,810 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 19:06:03,810 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 19:06:03,810 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 19:06:03,810 - 70 - INFO - ---------> LOO done.\n",
      "2023-05-16 19:06:03,810 - 70 - INFO - ---------> LOO done.\n",
      "====> Monte-Carlo Round 0 - Average convergence rate = 0.0\n",
      "---------> Number of values that have not converged: 100\n",
      "====> Monte-Carlo Round 500 - Average convergence rate = 1.0017637834093476\n",
      "---------> Number of values that have not converged: 50\n",
      "====> Monte-Carlo Round 1000 - Average convergence rate = 1.0009165868950363\n",
      "---------> Number of values that have not converged: 28\n"
     ]
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.utils import shuffle\n",
    "from statistics import mean, stdev\n",
    "from derdava.coalition_probability import IndependentCoalitionProbability\n",
    "from derdava.data_source import generate_random_data_sources, add_classification_noise\n",
    "from derdava.data_valuation import ValuableModel\n",
    "from derdava.dataset import load_dataset\n",
    "from derdava.model_utility import IClassificationModel, model_logistic_regression\n",
    "\n",
    "# Add logs\n",
    "import logging\n",
    "import sys\n",
    "logger = logging.getLogger()\n",
    "formatter = logging.Formatter('%(asctime)s - %(lineno)s - %(levelname)s - %(message)s')\n",
    "logger.setLevel(logging.DEBUG)\n",
    "\n",
    "fhandler = logging.FileHandler(filename='main4.log', mode='w')\n",
    "fhandler.setFormatter(formatter)\n",
    "fhandler.setLevel(logging.INFO)\n",
    "logger.addHandler(fhandler)\n",
    "\n",
    "consoleHandler = logging.StreamHandler(sys.stdout)\n",
    "consoleHandler.setFormatter(formatter)\n",
    "consoleHandler.setLevel(logging.DEBUG)\n",
    "logger.addHandler(consoleHandler)\n",
    "\n",
    "logger.info(\"Program has started.\")\n",
    "\n",
    "X, y = load_dataset('phoneme')\n",
    "one_indices = np.random.choice(np.asarray(y == 1).nonzero()[0], size=int(len(y[y == 2]) * 1.5), replace=False)\n",
    "two_indices = np.asarray(y == 2).nonzero()[0]\n",
    "indices = np.concatenate([one_indices, two_indices])\n",
    "X = X[indices]\n",
    "y = y[indices]\n",
    "y = add_classification_noise(y, noise_level=0.1)\n",
    "X, y = shuffle(X, y)\n",
    "X, X_test, y, y_test = train_test_split(X, y, test_size=0.2)\n",
    "\n",
    "NUM_OF_DATA_SOURCES = 100\n",
    "support_set = tuple(range(NUM_OF_DATA_SOURCES))\n",
    "staying_probabilities = {i:0.5 for i in range(NUM_OF_DATA_SOURCES)}\n",
    "coalition_probability = IndependentCoalitionProbability(staying_probabilities)\n",
    "model = model_logistic_regression\n",
    "\n",
    "random_orders = []\n",
    "loo_orders = []\n",
    "shapley_orders = []\n",
    "derdava_shapley_orders = []\n",
    "\n",
    "for N in range(50):\n",
    "    logger.info(f\"====> Iteration {N} =====\")\n",
    "    indices = np.random.choice(len(X), size=NUM_OF_DATA_SOURCES, replace=False)\n",
    "    X_train = X[indices]\n",
    "    y_train = y[indices]\n",
    "    data_sources = {i: (X_train[i:i+1], y_train[i:i+1]) for i in range(NUM_OF_DATA_SOURCES)}\n",
    "    model_utility_function = IClassificationModel(model, data_sources, X_test, y_test)\n",
    "    valuable_model = ValuableModel(support_set, model_utility_function)\n",
    "    random_order = np.arange(NUM_OF_DATA_SOURCES)\n",
    "    np.random.shuffle(random_order)\n",
    "    random_orders.append(indices[random_order])\n",
    "    \n",
    "    logger.info(\"---------> Random done.\")\n",
    "    \n",
    "    loo_values = valuable_model.valuate(data_valuation_function='loo')\n",
    "    loo_order = sorted(support_set, key=lambda x: loo_values[x], reverse=True)\n",
    "    loo_orders.append(indices[loo_order])\n",
    "    \n",
    "    logger.info(\"---------> LOO done.\")\n",
    "    \n",
    "    shapley_values = valuable_model.valuate(data_valuation_function='monte-carlo shapley',\n",
    "                                            tolerance=1.005)\n",
    "    shapley_order = sorted(support_set, key=lambda x: shapley_values[x], reverse=True)\n",
    "    shapley_orders.append(indices[shapley_order])\n",
    "    \n",
    "    logger.info(\"---------> Shapley done.\")\n",
    "    \n",
    "    derdava_shapley_values = valuable_model.valuate(data_valuation_function='012-mcmc robust shapley',\n",
    "                                                    coalition_probability=coalition_probability,\n",
    "                                                    tolerance=1.005)\n",
    "    derdava_shapley_order = sorted(support_set, key=lambda x: derdava_shapley_values[x], reverse=True)\n",
    "    derdava_shapley_orders.append(indices[derdava_shapley_order])\n",
    "    \n",
    "    logger.info(\"---------> DeRDaVa done.\")\n",
    "    \n",
    "logger.info(\"Random orders: \" + str(random_orders))\n",
    "logger.info(\"LOO orders: \" + str(loo_orders))\n",
    "logger.info(\"Shapley orders: \" + str(shapley_orders))\n",
    "logger.info(\"DeRDaVa Shapley orders: \" + str(derdava_shapley_orders))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "c5b2881a",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3172, 5) (3172,) (793, 5) (793,)\n"
     ]
    }
   ],
   "source": [
    "print(X.shape, y.shape, X_test.shape, y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "a989f60c",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.dummy import DummyClassifier\n",
    "\n",
    "def get_point_addition_performances(model, X_train, y_train, X_test, y_test, order):\n",
    "    performances = []\n",
    "    for i in range(1, len(order) + 1):\n",
    "        if i % 50 == 0: print(i)\n",
    "        X_train_curr = X_train[order[0:i]]\n",
    "        y_train_curr = y_train[order[0:i]]\n",
    "        \n",
    "        performance = 0\n",
    "        \n",
    "        for N in range(50):\n",
    "            indices = np.asarray(np.random.randint(0, 2, i) == 1).nonzero()[0]\n",
    "            #print(indices)\n",
    "            X_train_temp = X_train_curr[indices]\n",
    "            y_train_temp = y_train_curr[indices]\n",
    "            \n",
    "            if not len(y_train_temp):\n",
    "                performance += 0.5\n",
    "        \n",
    "            elif np.all(y_train_temp == y_train_temp[0]):\n",
    "                performance += DummyClassifier(strategy='constant', constant=y_train_temp[0]) \\\n",
    "                                  .fit(X_train_temp, y_train_temp) \\\n",
    "                                  .score(X_test, y_test)\n",
    "            else:\n",
    "                performance += model().fit(X_train_temp, y_train_temp).score(X_test, y_test)\n",
    "            \n",
    "        performance /= 50\n",
    "        performances.append(performance)\n",
    "    \n",
    "    return performances  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "428bfde9",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====> Iteration 0 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 1 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 2 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 3 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 4 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 5 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 6 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 7 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 8 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 9 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 10 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 11 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 12 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 13 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 14 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 15 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 16 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 17 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 18 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 19 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 20 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 21 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 22 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 23 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 24 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 25 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 26 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 27 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 28 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 29 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 30 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 31 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 32 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 33 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 34 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 35 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 36 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 37 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 38 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 39 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 40 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 41 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 42 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 43 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 44 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 45 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 46 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 47 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 48 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 49 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n"
     ]
    }
   ],
   "source": [
    "from urdava.model_utility import model_logistic_regression\n",
    "import numpy as np\n",
    "\n",
    "random_performances = []\n",
    "loo_performances = []\n",
    "shapley_performances = []\n",
    "derdava_shapley_performances = []\n",
    "\n",
    "model = model_logistic_regression\n",
    "\n",
    "for N in range(50):\n",
    "    random_performance = get_point_addition_performances(model, X, y, X_test, y_test, random_orders[N])\n",
    "    random_performances.append(random_performance)\n",
    "    loo_performance = get_point_addition_performances(model, X, y, X_test, y_test, loo_orders[N])\n",
    "    loo_performances.append(loo_performance)\n",
    "    shapley_performance = get_point_addition_performances(model, X, y, X_test, y_test, shapley_orders[N])\n",
    "    shapley_performances.append(shapley_performance)\n",
    "    derdava_shapley_performance = get_point_addition_performances(model, X, y, X_test, y_test, derdava_shapley_orders[N])\n",
    "    derdava_shapley_performances.append(derdava_shapley_performance)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "9a4d120e",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "from statistics import mean, stdev\n",
    "\n",
    "NUM_OF_DATA_SOURCES = 100\n",
    "support_set = tuple(range(NUM_OF_DATA_SOURCES))\n",
    "\n",
    "random_performances = np.array(random_performances)\n",
    "random_means = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "random_stdevs = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "loo_performances = np.array(loo_performances)\n",
    "loo_means = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "loo_stdevs = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "shapley_performances = np.array(shapley_performances)\n",
    "shapley_means = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "shapley_stdevs = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "derdava_shapley_performances = np.array(derdava_shapley_performances)\n",
    "derdava_shapley_means = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "derdava_shapley_stdevs = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "\n",
    "for i in support_set:\n",
    "    random_means[i + 1] = mean(random_performances[:, i])\n",
    "    random_stdevs[i + 1] = stdev(random_performances[:, i])\n",
    "    loo_means[i + 1] = mean(loo_performances[:, i])\n",
    "    loo_stdevs[i + 1] = stdev(loo_performances[:, i])\n",
    "    shapley_means[i + 1] = mean(shapley_performances[:, i])\n",
    "    shapley_stdevs[i + 1] = stdev(shapley_performances[:, i])\n",
    "    derdava_shapley_means[i + 1] = mean(derdava_shapley_performances[:, i])\n",
    "    derdava_shapley_stdevs[i + 1] = stdev(derdava_shapley_performances[:, i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "d8cb970f",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "random_means[0] = 0.50\n",
    "loo_means[0] = 0.50\n",
    "shapley_means[0] = 0.50\n",
    "derdava_shapley_means[0] = 0.50"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "id": "59fa45e1",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "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",
    "import os\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "x = np.arange(51)\n",
    "\n",
    "random_line, = ax.plot(x, random_means[0:51], color=\"gray\")\n",
    "random_line.set_label(\"Random\")\n",
    "ax.fill_between(x, (random_means-1/2 * random_stdevs)[0:51], (random_means+1/2 * random_stdevs)[0:51], facecolor=\"gray\", alpha=0.2)\n",
    "\n",
    "loo_line, = ax.plot(x, loo_means[0:51], color=\"C1\")\n",
    "loo_line.set_label(\"LOO\")\n",
    "ax.fill_between(x, (loo_means-1/2 * loo_stdevs)[0:51], (loo_means+1/2 * loo_stdevs)[0:51], facecolor=\"C1\", alpha=0.2)\n",
    "\n",
    "shapley_line, = ax.plot(x, shapley_means[0:51], color=\"C0\")\n",
    "shapley_line.set_label(\"Shapley\")\n",
    "ax.fill_between(x, (shapley_means-1/2 * shapley_stdevs)[0:51], (shapley_means+1/2 * shapley_stdevs)[0:51], facecolor=\"C0\", alpha=0.2)\n",
    "\n",
    "derdava_shapley_line, = ax.plot(x, derdava_shapley_means[0:51], color=\"red\")\n",
    "derdava_shapley_line.set_label(\"DeRDaVa\")\n",
    "ax.fill_between(x, (derdava_shapley_means-1/2 * derdava_shapley_stdevs)[0:51], (derdava_shapley_means+1/2 * derdava_shapley_stdevs)[0:51], facecolor=\"red\", alpha=0.2)\n",
    "\n",
    "ax.set_xlabel(\"Number of data added\", fontsize=30)\n",
    "ax.set_ylabel(\"Expected Accuracy\", fontsize=30)\n",
    "ax.tick_params(axis='both', labelsize='xx-large')\n",
    "\n",
    "ax.legend(fontsize='xx-large')\n",
    "\n",
    "plt.show()\n",
    "PATH_TO_IMAGES = \"./images/point_addition_removal/\"\n",
    "if not os.path.isdir(PATH_TO_IMAGES):\n",
    "    os.makedirs(PATH_TO_IMAGES)\n",
    "fig.savefig(os.path.join(PATH_TO_IMAGES, \"par_add_highest\"), dpi=300, bbox_inches='tight')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "id": "193d3afa",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "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",
    "\n",
    "fig, ax = plt.subplots()\n",
    "x = np.arange(51, 101)\n",
    "\n",
    "random_line, = ax.plot(100-x, random_means[51:101], color=\"gray\")\n",
    "random_line.set_label(\"Random\")\n",
    "ax.fill_between(100-x, (random_means-1/2 * random_stdevs)[51:101], (random_means+1/2 * random_stdevs)[51:101], facecolor=\"gray\", alpha=0.2)\n",
    "\n",
    "loo_line, = ax.plot(100-x, loo_means[51:101], color=\"C1\")\n",
    "loo_line.set_label(\"LOO\")\n",
    "ax.fill_between(100-x, (loo_means-1/2 * loo_stdevs)[51:101], (loo_means+1/2 * loo_stdevs)[51:101], facecolor=\"C1\", alpha=0.2)\n",
    "\n",
    "shapley_line, = ax.plot(100-x, shapley_means[51:101], color=\"C0\")\n",
    "shapley_line.set_label(\"Shapley\")\n",
    "ax.fill_between(100-x, (shapley_means-1/2 * shapley_stdevs)[51:101], (shapley_means+1/2 * shapley_stdevs)[51:101], facecolor=\"C0\", alpha=0.2)\n",
    "\n",
    "derdava_shapley_line, = ax.plot(100-x, derdava_shapley_means[51:101], color=\"red\")\n",
    "derdava_shapley_line.set_label(\"DeRDaVa\")\n",
    "ax.fill_between(100-x, (derdava_shapley_means-1/2 * derdava_shapley_stdevs)[51:101], (derdava_shapley_means+1/2 * derdava_shapley_stdevs)[51:101], facecolor=\"red\", alpha=0.2)\n",
    "\n",
    "ax.set_xlabel(\"Number of data removed\", fontsize=30)\n",
    "ax.set_ylabel(\"Expected Accuracy\", fontsize=30)\n",
    "ax.tick_params(axis='both', labelsize='xx-large')\n",
    "\n",
    "ax.legend(fontsize='xx-large')\n",
    "\n",
    "plt.show()\n",
    "PATH_TO_IMAGES = \"./images/point_addition_removal/\"\n",
    "if not os.path.isdir(PATH_TO_IMAGES):\n",
    "    os.makedirs(PATH_TO_IMAGES)\n",
    "fig.savefig(os.path.join(PATH_TO_IMAGES, \"par_remove_lowest\"), dpi=300, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "id": "988810a8",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====> Iteration 0 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 1 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 2 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 3 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 4 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 5 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 6 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 7 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 8 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 9 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 10 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 11 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 12 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 13 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 14 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 15 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 16 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 17 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 18 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 19 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 20 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 21 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 22 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 23 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 24 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 25 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 26 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 27 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 28 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 29 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 30 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 31 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 32 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 33 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 34 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 35 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 36 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 37 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 38 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 39 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 40 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 41 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 42 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 43 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 44 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 45 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 46 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 47 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 48 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "====> Iteration 49 <====\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n",
      "50\n",
      "100\n"
     ]
    }
   ],
   "source": [
    "from urdava.model_utility import model_logistic_regression\n",
    "import numpy as np\n",
    "\n",
    "random_performances_ = []\n",
    "loo_performances_ = []\n",
    "shapley_performances_ = []\n",
    "derdava_shapley_performances_ = []\n",
    "\n",
    "model = model_logistic_regression\n",
    "\n",
    "for N in range(50):\n",
    "    print(f\"====> Iteration {N} <====\")\n",
    "    random_performance_ = get_point_addition_performances(model, X, y, X_test, y_test, random_orders[N][::-1])\n",
    "    random_performances_.append(random_performance_)\n",
    "    loo_performance_ = get_point_addition_performances(model, X, y, X_test, y_test, loo_orders[N][::-1])\n",
    "    loo_performances_.append(loo_performance_)\n",
    "    shapley_performance_ = get_point_addition_performances(model, X, y, X_test, y_test, shapley_orders[N][::-1])\n",
    "    shapley_performances_.append(shapley_performance_)\n",
    "    derdava_shapley_performance_ = get_point_addition_performances(model, X, y, X_test, y_test, derdava_shapley_orders[N][::-1])\n",
    "    derdava_shapley_performances_.append(derdava_shapley_performance_)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "id": "23807fb5",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "from statistics import mean, stdev\n",
    "\n",
    "NUM_OF_DATA_SOURCES = 100\n",
    "support_set = tuple(range(NUM_OF_DATA_SOURCES))\n",
    "\n",
    "random_performances_ = np.array(random_performances_)\n",
    "random_means_ = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "random_stdevs_ = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "loo_performances_ = np.array(loo_performances_)\n",
    "loo_means_ = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "loo_stdevs_ = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "shapley_performances_ = np.array(shapley_performances_)\n",
    "shapley_means_ = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "shapley_stdevs_ = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "derdava_shapley_performances_ = np.array(derdava_shapley_performances_)\n",
    "derdava_shapley_means_ = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "derdava_shapley_stdevs_ = np.zeros(NUM_OF_DATA_SOURCES + 1)\n",
    "\n",
    "for i in support_set:\n",
    "    random_means_[i + 1] = mean(random_performances_[:, i])\n",
    "    random_stdevs_[i + 1] = stdev(random_performances_[:, i])\n",
    "    loo_means_[i + 1] = mean(loo_performances_[:, i])\n",
    "    loo_stdevs_[i + 1] = stdev(loo_performances_[:, i])\n",
    "    shapley_means_[i + 1] = mean(shapley_performances_[:, i])\n",
    "    shapley_stdevs_[i + 1] = stdev(shapley_performances_[:, i])\n",
    "    derdava_shapley_means_[i + 1] = mean(derdava_shapley_performances_[:, i])\n",
    "    derdava_shapley_stdevs_[i + 1] = stdev(derdava_shapley_performances_[:, i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "id": "8b8c108b",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [],
   "source": [
    "random_means_[0] = 0.50\n",
    "loo_means_[0] = 0.50\n",
    "shapley_means_[0] = 0.50\n",
    "derdava_shapley_means_[0] = 0.50"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "id": "218fb07a",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "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",
    "import os\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "x = np.arange(51)\n",
    "\n",
    "random_line_, = ax.plot(x, random_means_[0:51], color=\"gray\")\n",
    "random_line_.set_label(\"Random\")\n",
    "ax.fill_between(x, (random_means_-1/2 * random_stdevs_)[0:51], (random_means_+1/2 * random_stdevs_)[0:51], facecolor=\"gray\", alpha=0.2)\n",
    "\n",
    "loo_line_, = ax.plot(x, loo_means_[0:51], color=\"C1\")\n",
    "loo_line_.set_label(\"LOO\")\n",
    "ax.fill_between(x, (loo_means_-1/2 * loo_stdevs_)[0:51], (loo_means_+1/2 * loo_stdevs_)[0:51], facecolor=\"C1\", alpha=0.2)\n",
    "\n",
    "shapley_line_, = ax.plot(x, shapley_means_[0:51], color=\"C0\")\n",
    "shapley_line_.set_label(\"Shapley\")\n",
    "ax.fill_between(x, (shapley_means_-1/2 * shapley_stdevs_)[0:51], (shapley_means_+1/2 * shapley_stdevs_)[0:51], facecolor=\"C0\", alpha=0.2)\n",
    "\n",
    "derdava_shapley_line_, = ax.plot(x, derdava_shapley_means_[0:51], color=\"red\")\n",
    "derdava_shapley_line_.set_label(\"DeRDaVa\")\n",
    "ax.fill_between(x, (derdava_shapley_means_-1/2 * derdava_shapley_stdevs_)[0:51], (derdava_shapley_means_+1/2 * derdava_shapley_stdevs_)[0:51], facecolor=\"red\", alpha=0.2)\n",
    "\n",
    "ax.set_xlabel(\"Number of data added\", fontsize=30)\n",
    "ax.set_ylabel(\"Expected Accuracy\", fontsize=30)\n",
    "ax.tick_params(axis='both', labelsize='xx-large')\n",
    "\n",
    "#ax.legend(fontsize='xx-large')\n",
    "\n",
    "plt.show()\n",
    "PATH_TO_IMAGES = \"./images/point_addition_removal/\"\n",
    "if not os.path.isdir(PATH_TO_IMAGES):\n",
    "    os.makedirs(PATH_TO_IMAGES)\n",
    "fig.savefig(os.path.join(PATH_TO_IMAGES, \"par_add_lowest\"), dpi=300, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "id": "40f4b68e",
   "metadata": {
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAnUAAAHWCAYAAAARl3+JAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOydd3Qc1f23n5nZXrTqtuRu3G2K6SUUG5teTMdAgEAgPYGQ0MIbWgIJkPxSSELvEAgldjDFGNOLAQPuvUmWZPW6vcy8f1ztaiVLtspKu5buc849M7s7e+dqJc189lsVwzAMUoSqqiiKAsDkyZO55ppruPzyy8nLy0vVKSQSiUQikUgknaD0h6iLT6koChaLhXnz5vH973+fE088MVWnkkgkEolEIpEkkVJRN2zYMGpqasTEHcQdwLhx4/j+97/PlVdeyfDhw1N1WolEIpFIJJIhT0pFXTQaZeHChTz22GMsWbIEXdcTgi5Z4GmaxhlnnME111zDKaeckjhGIpFIJBKJRNI7UirqkiktLeXxxx/nySefpKysTJysE+vdyJEjufrqq/ne977HqFGj+mMpEolEIpFIJIOefhN1cXRd5+233+bRRx/ljTfeIBqNdmq9U1WVk046iWuvvZYzzjgDTdP6c1kSiUQikUgkg4p+F3XJVFVV8cQTT/DEE0+wdetWsYAkgRffHzZsGN/73ve4+uqrGT9+/EAtTyKRSCQSiWSfZUBFXTLvv/8+jzzyCAsWLCAUCnVqvVMUhRNOOIFrr72Wc845B7PZnI6lSiQSiUQikWQ8aRN1cerr63nmmWd4/PHHWbt2rVhUJ7F3eXl5XH755VxzzTVMnjw5bevd19B1nYqKCtxut0xIkUgkEolkH8EwDFpaWiguLkZV1W69J+2iLplly5bxz3/+k+eee243yx20CbwTTzyRX/7yl5xyyilpWee+RFlZmUxAkUgkEolkH2Xnzp2MHDmyW8dmjKjbtm0bjz32GE899RSVlZVdWpWSY++OPPJIHnroIfbff/+BXOo+RVNTE9nZ2ezcuZOsrKx0L0cikUgkEkk3aG5uZtSoUTQ2NuLxeLr1nrSKukgkwmuvvcajjz7KBx98QMelxB9PnjyZsrIyfD7fbhY8i8XCk08+yfz58wd28fsIzc3NeDwempqapKiTSCQSiWQfoTf37+45aVPMhg0buOGGGxgxYgSXXHIJ77//PrquJ143DAOr1crll1/Op59+yvr169m1axcPPfQQM2fOTFjrFEUhHA7zve99LxGPJ5FIJBKJRDIUGTBLXTAY5KWXXuLRRx/l888/B9pnusb3p0yZwg9+8AOuuOIKsrOzO53r3Xff5frrr2+XWHHVVVfx6KOP9v8Pso8hLXUSiUQikex79Ob+bernNbFixQoeffRRXnjhBZqbmwHaWdoMw8BsNnPuuefygx/8gOOPP36vc86ZM4dly5Zx0EEHsW3bNgzD4L333uvvH0UikUgkEokkY+kXUef1ennhhRd49NFH+eabbwB2K1FiGAb77bcf1157Ld/73vfIz8/v0TmcTic/+clP+OUvfwlARUVFCn8CiUQikUgkkn2LlIq6ZcuW8dhjj/Gf//wHn8+3WyFhwzDQNI2zzjqLH/7wh8yZM6dP55s0aVJiPxwO92kuiUQikUgkkn2ZlIq6o48+erfCwYZhYBgGo0eP5pprruHqq69m+PDhKTmf1WpNyTwSiUQikUgk+zr94n6NizlVVTn11FP54Q9/yKmnnpryjgZ2u50xY8akdE6JRCKRSCSSfZGUizrDMCguLubqq6/mmmuu6XYV5N5w9NFHs3379n6bXyKRSCQSiWRfIaWi7uSTT+YHP/gBZ555JpqmpXJqiUQikUgkEskeSKmoe+utt1I5nUQikUgkEomkm6Slo4REIpFIJBKJJLVIUSeRSCQSiUQyCEip+zUajfLZZ58lHk+YMIHi4uIez1NeXs7WrVsTj4899tiUZ85KJBKJRCKRDCZSKupee+015s+fD4Cqqqxfv75X8/h8PmbPnp2od7dw4ULOOOOMlK1TkiIMA6TYlkgkEokkI0ipqHv88ccTQuyMM85gwoQJvZpn0qRJnHbaaSxatCgxrxR1mUdDWQVVK9dhzc3F6nFjy/HgyM7CbLUkuojELay6rqPrOrFYrNP9eF1DVVXRNC2xnzySi1nHR/y9ySP5vMnrSB7x+XpK/Jzx9cfPF5+vs5EK4ueN/0yqKiMnJBKJRNKelIk6v9/PRx99lLiJXXzxxX2a75JLLkmIuqVLlxKJRDCbzX1epyR1xAIt+Ms34itXQDNjKCbQzKhOJyZPNubsHKweFzaXIyMteslirKN4jIumuPBM3vb0HIqioGlaYsSFa/JQFCUxfywWIxqNttvGBWQyyWvtTFjGz99x25ng7UyU7o3k9cT3u9p2PF9Xa5FIJBJJ70mZqFuxYgWhUAgQF+gTTzyxT/Mlv9/n87Fq1SoOOeSQPs2ZTGVlJffeey+LFi2ivLwcj8fD4YcfznXXXdfntZeXl/P3v/+dN954g5KSEnRdp6ioiMMOO4yLLrqIs88+e7f3dOeG9vLLL3P++ef3aW39gYIBsTAKYYgCoSai9RVEUPBrZjBZMGfnYs3Lw5abiznLBRlQx9AwjB6LtN6cI25li0QiXR6nqmrCEtdd4sf3988wUHQU151ZbZOFcU/QdZ1wOEwkEiEcDhONRncT25qmYTKZEvNLkSmRSPY1UibqNm7cmNgvLi4mPz+/T/MVFBRQXFxMRUUFABs2bEiZqFu1ahWzZ8+mrq4OgKysLGpra1m0aBFvvPEG99xzDzfffHOv5n755Ze5+uqraWlpAcDhcKAoClu2bGHLli1UV1d3Kuri5Ofnd1m42Waz9WpN6SIu9oiFiVZ5iVaV4kXDbLFg8WRjy83BklcALhcM8T6+PRV0g5G4yO6OSFVVNSHATCZTu31FUYhEIgkBF4lEiEajnc7T1fPAHq2r8cepEH7JrnUpJiUSSV9Imairr68HxLftwsLClMw5bNiwhKirqalJyZyBQICzzjqLuro6Zs6cybPPPsv06dNpbm7mrrvu4k9/+hO33norBx98MCeddFKP5n7zzTeZP38+sViMq666ihtvvJHJkycDUFdXx4cffsiOHTv2OMdXX33F2LFje/nTZT4KMaLhANGaAP6aXSgoWK0WTHYXhjsbIysXw2EHhxPFYkZVFBQFVAVURUFTFUyqvOkNdeKWt/6kOwIz2WW/p/jN+Jo7Gx3d6skWw45bVVXbid/Ohq7raJqG2WzGYrFgNpsTYlcikQxuUibqki+wqWoRljyP3+9PyZwPP/wwJSUluFwuXn/9dUaMGAEIa90DDzzA1q1bWbBgAbfcckuPRF1zczPf//73icVi3Hrrrfz+979v93peXh7nnntuSn6GwYSBQTAUglAIGusw2IZhsoBmRbc6iTmzidntrcMBmoYCaKqCpoBJU9CSxJ5JVXBaNEyavIFJ+p+4MEslexKT8WSh7qwrEokkrpuKoiQEXnwrWzlKJIOPlIm6vLw8QLgSqqurUzJnsnUuOzs7JXM+//zzgEjEiAu6ZH7961+zYMECvvnmGzZu3JiwtO2NJ598kl27djFy5EjuuOOOlKw1k2lo8PLKu+vIiVoYbwmTKg2lYKBEQxANoYaa0ZorMUw20KwYmpmYw03U5SbmchNyOAnFOj+xzaTismq4rRpWk8wUlQwOuiPounpfKBRKxD0DncYUdpbA09PzJFsNkxN99kayZXNvQyKRdE7KRF1BQUFiv6ysjOrq6j65YaurqyktLU1cVJLn7y0tLS18/fXXAJx88smdHnPkkUfi8Xhoampi6dKl3RZ1cbF4/vnnD4ks3W8eeYwf3nw9pZ5hLB8zg7LxE4lO2Y/9ihyMtMS6vBnEDINgTCeo6+22OmBRFayqikVVsCgqVlXBoqpYDD9KNACA5q/BUgsGGoaqEXI6CdschB0OoiYrMRRihkqDAbFYFD0WRdF1TIqOCR3FEDcYk8mExWLBarV2ujWbzei6jt/vT4xAINDusd/vJxKJ7DWz1WQyYbfbcTgciW18vzNria7rBIPBxPkCgQCBQIBgMIiqqol1xq0u8TXHR3ImbVdbVVVxOp3YbLZ+uUkahkEkEkms22q14nK5MJlSWkVJ0gcMwyAaje4xrrAr1zK0z2buiXjrK3Hh19n/XHJyTfzY5G1XdCyLlJy93Vkmd8fnksVo8r5EMtCk7Ap76KGHAm3ugVdeeYUf//jHvZ7v5ZdfbvcPc8ABB/R5jevXr0/MOX369E6PUVWVyZMn8+WXX7Ju3bpuzRsMBlmxYgUAM2fOZMOGDdx5550sXbqU5uZmiouLOfnkk7npppv2Gi934YUXsnnzZvx+PwUFBRxxxBFcddVVnH766d3+OQeC/I3LiCkKo5uqGL2qClYtBWB7ThErR09m59jR1O43hpw8B2ZFJ9Qq3iK9tDSYFQWzqhAzDGIG6IZBf6YWdNfN1VesVit2ux2bzUYoFEqIoIFCUZSE0HQ6nQnB6XA4sNlsXdYFTN7GRWfH0dkN3uFw4HK5cLvd7YbL5cJut2M2m/s19iscDtPQ0EBLSwtWqxWn04nL5cJisfTbOfdlMjGBpycJNR3pKPKSRVuq6cry2FF8piLbOrl+Znx0VqpJxlUOfhQjhX/RkydPZsuWLRiGwfDhw9mwYQNZWVk9nqepqYmpU6dSVVWFYRiMHTuWbdu29Xl9CxcuZN68eYCIgXO73Z0ed84557BgwQLOPfdcXn311b3Ou2HDBqZOnQrAb37zG/785z8TCARwOByoqorX6wVE3N7ChQs54YQTdpsj+Z/N7Xaj6zo+ny/x3AUXXMBzzz2315tPRxdLc3Mzo0aNoqmpqVe/i66oXvsx9W8/jHvFcmJb/Vh2BCiorEPr8Oe0LXcEG0aPo2rsMKrHjSKQlY2mgF1TsapqYqsqCmFdJ6wbhHUhAsO60W0RqIKIrVNErJ2mKKiahmo2oZrNqCYNVbOgmswoJjO6LqwUsWgUPRohFg0TCYcJh8PtbmSqqrYTOh2tbXGLXsdA9eTHcYtVsqUvEAjs9YZpt9vbDZvN1q40RygUIty65vhznf07d2bFiMViBAKBfheuJpMpIVj3VNKl43vMZnMiwD++H7dMdvxc4sNqtSb+j6LRKA0NDbuN+P9iR8xmM06nMzFcLldiGxedMgZN0p90Jr662u8s4aa75+iuizt5PZL00NzcnPAcdvf+nVJfyBVXXMFtt92GoihUVVVx9tln8+abb2K327s9RyAQ4Oyzz6ayshIQf8iXX355StaXLJL2tCaHwwHQ5Q2gI42NjYn9e++9l+HDh7NgwQLmzp2Loih89tlnfO9732PTpk1ccMEFbNy4kdzc3HZzXHHFFcyfP58jjjgiET+4YcMG7rvvPp588klefvllsrOzeeSRR/a4lnvvvZc777yzW+vuC4XTj0U1F1Dqehe+A5oewNuyHeuGdQQ21uDYVkdxZTXj68sZX18OK8T7duQVUTFuFMakEbj2n0Ese88uet0wiOgGIV0nahi7CzdFQWPv7pWY1UrM6STqcqJ3KA1jKCZQhTtXVTWsJg2TCi6nE1dWTr9c1OIxTnGBF3dRJgu4nrpv4i5PwzDa1Xfririb1+/34/P5EoIzvh8KhTotUJx841FVFZvN1qXQiocixH/elpaWxPB6ve0exy0vcZdgIBDo0c8ftzpqmpYoKdQZDoeDrKwswuEwPp8vITgbGxvb/S93nNvpdO5mYXS73Ynf1Z6KWcd/Nx1LrSQ/NgyDrKwsPB4PHo9HisghRn8k3XR2DuhZbcvuiMDO/valGEwPKbXU+Xw+xo0bl6j/BqLl10MPPcTxxx+/1/d/8MEH/PCHP2Tz5s2AuBHk5eWxbdu2Lq1qPeGFF17g0ksvBSASiXQZ33PppZfywgsvcNJJJ7F48eK9zvvZZ59xzDHHJB6/++67uxUwXr16NQcddBC6rvPHP/6RG2+8sdvrvvHGG7n//vtRFIX169fvMc5voCx1ALWbNlD6/rtdvm721qOtXUl4w06yt5UzumoXKu3/3MrzC2kcX4Rp6hhC+x9B1JXaNXaGbjIR9WQR8Xgw9hLjZTZpOB0OXE4XVrsLxbxv1QrcV4i71JKFTnxEo9HEftxF3XF0Vt7EarWSm5tLTk5Ou9Gx3mMkEsHn8yWG1+tN7HcUnAOFoii43W48Hg/Z2dlkZ2fj8XjIysoiFoslrLTx//f4iD/ndDrJz88nLy+P7OxsGd8lGXA6+zLYWdyhTITpmrRb6pxOJ48++ijnnXdeIlZh48aNzJ49mxkzZnDqqady6KGHUlhYiMvlwuv1Ul1dzfLly3nrrbdYs2ZNopdmvBfoo48+mhJBF19fnEAg0OW88TIALperW/MmHzd9+vROO1Lsv//+zJkzh3feeYelS5f2SNTdfvvtPPjggwQCARYtWrRHUWe1WrFmSCHfiCuXyBGz4AioB2qaGvGuWYuyaSfF23YyqaaUEbXVjKithi9XovM624tGUjZhIk3TDsA2eTx5diXlHcbUaBRLXT2WunqiLhfhnGz0Liy3kWiMxuYWGptb0BRwWk24HE7sLheKxQWajMVKBYqiJIoI98SyHyfuTg4EAkSjUTweD3a7vVvWArPZnBBOnWEYBoFAICHwmpub21kYI5HIHuvOJRN3J8eTcZLdyiBCTxobG4lEIjQ3N9Pc3MzOnTt7/Hkko2kaeXl55OXlkZ+fT35+Pjk5OdISKOlXkuNwe4LJZGqXuCZFXs9IeSra2WefzX333cevfvWrdsGoq1evZs2aNV2+LzkpIi7s/vSnPyVi4FJBcXFxYr+ioqJLcRQveFxUVNTjefckuCZPnsw777zT44u00+lkxowZfPXVVymJLUwXmicbzzHHwDHQbMCi2kZ8a7dg37yTKTs2M6luJ/vtEoOP3yNosvDtyCls3G8GtZOnYBlbzHh7hOGmKKmqP2zyejF5vegWC+GcbKJuN3RxEYkZ0ByM0hxsQqlvwmlWcFg1HBYTVotF9L9VTRiqGVRz61a4diX9i6ZpuFyubn8R6wmKoiRiKIcNG7bX4zsmlsSD1k0mU7duUHERGXcHx4VeY2MjXq8XTdMSX97iI34DjGdFt7S0UFtbS319PZFIhOrq6nalplRVxe127+YuT44ZjbvPO5Ym6fg47j7uaDVMthyGw2Hsdnun7mu3253ovLM34p+ldO0NXuLhF/FwKbPZ3C7Df0//Q8n/e0PVBdwv9QV++ctfMn78eK666ioaGxvbfbCdfYtN/ic1DIPc3FyeeuopzjjjjJSua8qUKQnRuHbt2k4FmK7riZZn06ZN69a8+fn5DBs2jKqqqm4dPxT/0DqiKlBckA0nHAonHEpdzODtylK0tevI2VzK2O07yfc2cdSOVRy1YxUshQabm6UTDuepSYeza8p0RrlggjXEfpYwhaZonyx6ajiMraoao6aWaFYW4WwPhsUChpEYimFA67dOxTDwBwx8ZhOGyYSqBHGaFOwmcJrBqiVl2KkahsmOYbZjmJyi7p4UeoOW5OtZb6xhySIy+QsjkPjC210Mw6CpqYm6ujpqa2sT21AoRFNTE01NTT1eX2/xer14vV527dq122uapuF0OtE0rV35nY778XjReLZ2V8Nut/fZwpMcoxqn42ef/HuWFqX+IR56EcdsNid6ZXdVigbE76Zj2aee/u/Ev7zE5+tqm0lu434rGjVv3jyOPvpo/vKXv/DQQw91GYAMbUIvNzeXH//4x/z85z/vc+/YznC73Rx66KF89dVXLFmypNMOD1988UXiQteZG7Ur5syZw/PPP9+uB25HNmzYANDjNmA+ny9h5Rw3blyP3tufWMwqbrOCN2LQ18BMq6ZQOGIMjBgDcw1qIhWES5cTXbsDdauX/JJ6coItnL9mKeevWYrfbOXDcYeweNJRPLLfYWA3M8HiZT9rhKNdEcZYu669tScUXcfc2Ih5D3+vAIqioqgiUSPqsBPNchNwufAbGnURWjN8wWlWsJsMrPgwwl5AFNQ2TFYMkx3d7MAwOcCUGS7z/iB+AZb0nZ5+IVQUJeFa3m+//QBxvY0LrOQ6iB3rIsYztJOLFMfblXV8vCfLYdx66Pf727mt48Pn8xGLxWhubu7WzxSLxRLv3RPJFp5kS09832Qy7TEuMTk2uTvEP4941nbyvtlsxuFwkJubm4jzlDUbe0d3s+gNw0hUBvB6ve1EXrLVT9f1hHWw4+guDocjZQ0S+kpKEyW6IhKJ8OWXX/Lxxx+zdetW6uvraWlpwe12k5uby8SJEzn22GM57LDD+v0P/S9/+QvXX389brebjRs37uZiPe+883jttdc45JBDWL58ebfnfeeddxIFjZcsWcKcOXPavZ6cKPHggw/yk5/8JPHa3r5933zzzfzxj39EURTWrFnTbQsi9C7QsrtEgwF8pZsxqnfhr6rC29SCrwcCT0FBUVVUVUFBQUfH0MHQdZJnsUZryfatwbNlDZbVlbDej6Wp7R87omp8PvoA3pl0FIsnHkmNK5eZWglnO7ZylDuCYRtBzJyDomriXKoKnRQZxUCswQB0A5Tdi6+qqta1RVBViLndxLKy0F1Okg+0aAoeqxDBVpOBoRvoRqv1QTfQUVBMNjDbElvVYkcx2VCS1tBZvbiOteOAvQYlJxeejY94bFhvid/IOo74hTPupotnfkqhl9nE/9b6290ZL9/k9XoTfWv3VNctHA4nsrQ7G36/f0BqTPaVrKws8vLyyMnJSYi9rKws6ckZIOJ/16m4DvWXqOvN/XtARF0mEQgEmDp1KiUlJRx88ME8++yzTJs2jZaWFu6++27uv/9+ABYvXrxb79f4P9vtt9/eaSuwk046iSVLllBcXMxTTz3FnDlzUBSFzz//nCuvvJJNmzYxbtw41qxZkyibAqIG3aRJkzjnnHM44IADEkHTGzdu5IEHHuCxxx4D4Morr+TJJ5/s0c/bn6KuI4a3hVh1OS1l5Xirq2jxBYi1fttRNAVVSRIYigp7uXYJr2dcsBiAASigR3FvX0vW8q9wrNiAeVdj4j068Pbko/m/Yy5lc8EYcmnmAu1DLrQsY7jdTtA2ipBtNEHbaGImT399FKBpxDxZRD0eDHv7bEuLppBlUfHYFBzmvZnsFWHFM1nB1DqPUL5Jw2j/WDW1HR/faha6VqNtdPzW2pmroeN+3ELR05tRcjmPcDjcLsM0lZelrlpgdeyykbwv2XfRdb2d1S05ri95G41GO7Uw2my2dha9uFuts84S8cfx2ML4l6OO20gkgtfrpb6+nvr6+j0WGO9Jm7TO/ufiz3W0liYPm83Wbl8mzfQNKerSzMqVKznxxBMTpVeysrIS3xIVReGee+7h5ptv3u19exN1dXV1zJo1i9WrVwPiF51cM6u4uJjFixczY8aMdu874YQT+PDDDwFxA/J4PIRCoXZ19c4//3yee+65Hme2DqSoa0c0it7YhLeiHH9lJb6mRvRIz+qOdRdzZSWur77Bufxr7Fu2AmAoCu9OOZI/HH05W/NHAXCcupLLtHeZrX6LSdGJallC5FnjQm9U/wg9k4ZusWBYzBjm+NaMbrZgtprxWIXAs5v60SKiqKAliUOzDcyZ7/btqjXTnga0F3I9/UzjVs/kpIDkMQQvmZIUEk+EiQu8+vr6RHHsdH2hiLuqbTZbu9GZSz3Zjd2b/6/BiBR1GUBlZSX33nsvixYtory8nKysLA4//HCuv/76LmPp9ibqQNSJ++tf/8qLL77I5s2bicVijB8/nrPPPptf/vKX5OXl7faed955h0WLFrFs2TLKysqor69HVVWGDx/OkUceyZVXXrmb1bC7pE3UdcDw+fBX1+It34m/pppo0A9G6i9gltKd5L62EPdXwnVuKAqf7H8Mdxx+KVvzhLgbrjRyofoe52sfMVqtbvf+qOZpteaNImDfj4B9PCj9GBKgKhhmM4bFguHJwp6bhdOm4bKo2EwDcLFUNCHuzPbW0Sr05IW6SzoKvY6X0K6C6bvqm5p8fLKFRwrIoUW8GHhn3SL21D2is7+ReDxZcqxgMBjsNEO5L39jiqK0E3rJMWsd95OLq3fsALOvI0WdZMDIFFHXjliMQF2jsOLtKifc3Aix3YvH9gVLSSl5ry3AtfwbQIi7b2d+hzsOu4RV2aMSx8201nCu5VtO531yIqUoHSICY6oNv2MKPtcMfM5p6FrqS2a0Q1NFXJ4nC9XtwmVRcJlVXBYFszZAF0BFbRN6JrsQeWa7zNgdYJLjHjsKvY4CseM2Ljzl5V2yJ+LiLxgMJkZcAMZHsts6eb+vf1uqqnZaSic/P5+ioqKMqbfaHaSokwwYGSnqOhDy+Wkuq8C/cwfh2irQe5e52hnWHSXkvroA1zffAkLcbT78O/zj8PP5n2M8Rmtgn001OC7LzxnOUg5X1uEIleDwb8QUa8uwM1AI2sbhc03H55xB2DK8Xy1ahtlEzOMh5snCsFqxaQrFbg2nJU2p85ql1XVrb78dJN+2BxudicJIJNIvLr49xXbFkYkxg4f431ZyjGJy+7vkhKjk/tTxrOruZBbHxV1xcTHDhw/fa9/zdDIkRV28PlJjY2OPU8UBjjvuuH5Y1eBnXxB1yQR9AZpLduAtKyFWVy2SAFKAddt28l5dgHPFysRzzRMn8t5hc/nb6OPZZrR1GxlujjE3J8TcHD9j9R04fWtx+tZgC5W3mzNiyiXgmEBMcxHTHMRUJ7rmIKY5iWlOdFXsG2rfL0aG3UbU40HPclPosVLgyKDCmooKKGK721BahyYsfe22avvHmqXLws+S1BFPhulY/61jV4y4CItnniZnoXbMSO0OySKgY2KMZGiR3AEmXkbH7/fj9XqprKzcrX6ioijtRJ7Vam0nJpP/ruLDYrEkuqjk5eX1q+VvyIi6xYsX8+STT/LZZ59RXl6+9zd0gaIoPaoZI2ljXxN1yfi9PppLtuMtLcForKXPxfAA69Zt5P5vEc5vVqDEm1s7HJQccQwvHHgKL9gn4tfbblIHOcOclBviWE8IV6whIfDs/k2oRvf+JmOqlbBlBCHbSELWEYSsIwlZi3oXq6eA7nRizc9hxAgPZvNgcokqbXF9ZgdYHNISOIjZk8jrqi5ePEuzY9Zy8pAWwX0fn8/Hrl27qKioYNeuXd2uYbgn3G53QuDFt3a7nWAwmKjb2LH/s9frJRwO4/F4yM3NJS8vL7FNFomDXtRt3LiRSy65hBUrVgB9L08QjxGR9Jx9WdTFMQwDX3MTTaU7CJRVYDQ19NlFqzU0kPXhx3g++AhzTW3ief+E/fjyiBN5ePwJLAu7E+5Zh6pzQnaYk3OCTHVEUY0QDv8mLKFdaLofLeZDjfnQYn5UXWy1mA+Fzm8wBhoh6/BWgTeSUGsWrqGau/0zqJpGYbEHd2E2dKiLN2hQVDA7hdizOESMn2aW8X2DkPgX975mVMZLjHQUel0Jv71ZIuNzdixp0tVzyetI3kajUcLhsBSdvSDeiaSiooLKykp0XW9XRLizEQwGqa2tpba2Fq/X2+m88dqfvcHpdCYEXlFRERMmTCA3N7cvP+ZuZISoW758ObNmzUoUgEz+0PbULqzjP3HHdh9S1PWOwSDqkjEMA78vSMvOCvzlZcRamiAWEYkWvUm20HUca9aR9d4HuL75FqX17yxmt1M/Y39W545hiXs8X2aPZUdOMTFVY5Q1ysk5IebkBMkz7+HfxzBQ9SCmaAPWUDnWUBnWYBnWUBmavnt5F10xE7SPw2+fRMAxkaBttHBL7oVsq0p+lgUlxwM5WWC39/xz2NdQVFDNrQLP1Lptfayool6fHgU91rqNtn/O0IVQtGaB1SX2JUOGuHt5oPuDxluPpTLhQLJ3gsEgdXV1iTCw2tradi7eeNs5l8uVaDUX7yVtNptpbGykvr4+MUdnInHChAlceumlKV132kVdY2Mj+++/P+Xl5Qkx5/F4OPXUUxk3bhz33nuvOKmicNlllzFixAjq6+tZu3YtX3/9NcFgMPEPNnXqVC644ILE3LfffnuqljmkGGyiriOBJi/eikp8lTVEAqFWcReBWAiirWKvC2tZR7SmJrI++oSs9z/EUlW92+sRzcS23JFszB/NpvzRbMkfjTFxHDNHuzgqK4xD6+a/kmFgitYnBJ41VIYtWNouKQOE2zZon4DfMRG/fRJha3Fr/Nru2DQY5jJh0RBWLJsFrFawWcFqAZsNLN23Ag45VBNYXGB1i63Ztvf3SCR9JC7y4gIv3gKrM4ufJLVEIhGCwWCinmxPCIfDCZEXrzM4depUjj/++JSuMe2i7u677+b2229PCLNTTz2VZ599lpycHIB234qWLFnC7NmzE+9taWnh8ccf56677qKxsRFFUbj00kt54oknZI+8PjDYRV0CwyBY34h35y6aKmto5+HQo22WvPjYk/tW17Ft3Ixt61asZeVYyiuwlJejhna3BEYVlZcOPIl/fWc+44tdnJAd4gh3mB4nqBoGlnAl9sBmHP7N2P2b0XR/u0NiqoOoKQtQMRQFEMkIBq2JCCiYNQ2TxQGW7A4jB2w5YHcKged0QHaWTEzoCtUsLHiapdUCaBLCT41bBuU1STJwdBR5yUktyQkvnQ0pCvufQRtTV1xcTFVVFYZhMHnyZFauXNkuDXlPoi5OeXk5Z5xxBitXrkRRFObPn89zzz2XqiUOOYaMqEsiVltH44YtNHqDdBm+okcgEhAj2o1OF7qOqbauVeSVYykrRykpw72zFICAycrjh53Nw0ecR8xu5xhPmFnZIWa6IvSqhrChYw2VY/dvxhHYhN2/FdXoedb4bphcQuTZCiH3ABh7NBSNBru0TPUMpVXkmdrq+CXKvOw79bUkg5+OIrCzLOe9FTiW7JlBKeo2b97M5MmTE6LtueeeY/78+e2O6Y6oA6ipqeHAAw+ksrISRVH497//zYUXXpiKZQ45hqKoAyAcJrp1G43V9TT6I3uujGLEhLCLBIXI60GnC9uGjeS/+DL2zVsAaLS7efDIC3j24DMImSx4NJ3jskNcVBBgmKUPF0ojhjVUjqoHwTBQiLVuRb9XkZBhoBg6asyPKdqYNJowRRtRjUgnEyvgGgeFB8N+x8Ho6SD7QPYNRWsv8sx2YeEj3qPX6Hw/nv0rRaEkzcSTQJLFX/xxcku+rl7reNxgZ1CKupdeeikh4jRNo7GxEafT2e6YZFG3ePFi5syZ0+V8jz/+ONdccw2KojB9+nRWrVqVimUOOYasqItTVUWkdCeNvhBNgWg3yt4ZIh4vHICoX8Tn7fUtBs6vvyXvpVewVlQAUJ+dx1++cxnPTZ2NrmqYFYNz8wPMLwzg7G7sXSoxDFQ9kBB6tsAOXL41WENl7Y+z5kPRoULgjT1MJB5IBpZ2otAh4vtMdukql+yTJAu9rrKR9/VSNINS1P3jH//gZz/7GYqiMG7cOLZs2bLbMcmibsGCBZx55pldzuf1esnPzyccDqMoChs3bmTChAmpWOqQYsiLOoBAALZtI+L10xAI09wtcddKLARhP0R8ey+jEouR9fGn5L66AHN9PQCNxSN56LhLeKr4cIJmGx5N57Jhfs7IC/bOLZtiTBFRe88TWIPFuwklufaeZoXs8TBsCuRPFCNnrBR66cJka+3LGy/2rHS9VU0iHlCzyF6+kn2CeCma5HZk+0rVi0wSdSmL9k1OD+6saT2IH9zv96MoSpd1Y+K4XC7GjRvHxo0bAfjqq6+kqJP0Drsdpk7FXF5OYXU12fYYDf4ILcFuiDvNCnYr2HNaBZ4PIv7OBZ6m0XzCcbQcfSSeJUvJXbiI7Ioybn7xPm5UNTYMH8/nxdNYPmIqN4+bxLmTbByVFU7r/TZqzqEp+zs0ZX8HRQ+RHdpEln8t5uY1KJFmqFsvRhzVBDnjhMDLmyCG3dNWMNhs7zJDV9JHokExeoNqFuJOM4u/ac3Sum+RnTwkGYGiKJhMJkwmEw6HA2ir7dddkacoCqqqoqpqQiQOBfdvMikTdfak2lhdfYhZWVn4/SKjr6ysrNNjkklWvhWtbi2JpFeoKowaBR4Plh07GGbSyHHEqPcJcdctEgIvF2JBYcGLhhCtLgwRH2UYGBYrjaefSvMJx5Gz6C2yPv4EU0Mj0yo2M61iM1cvXwjATs8wtoyZTP7+48masR/hUSPTalExVCsN9v1psO+PKV8nhxqyouWo/p3gax2xANRtFqMr4m5DS5LQi7sTTdaut67hMGyaeI8ktegRCO8hlCDepi1Z6HXcl9Y+yQDTlcgzDCMh3pJHZ/Vu4+7deB/k+H6y4Os4R/JjEHXu4uVmMp2Uibpk61xLS0unxxQVFVFZWQnAunXr9jpnTU1NYj8c7kVhWYmkI1lZMH061NdjqatjuMlHjjNGQ0/EHYBm22vGqO7Rqbt2CnXf/zmmynLs61ZjW78Wy4YN2EpLGdVUxahVVbDqIwDq95uI76JzCU6f2pefMCVEDZUahtFgHkbOsEPxWFUUDAjVg78UQrsgVAneMgh7hfXSaI2JiVuUAvU9P7GiQv5kKNofhh8Iw/cH2xANGxhI4slCe8oE70rsaVbp4pUMCHGR112SrX8de7/GY/2607vY7XYTjUYT/WozuW1pykTdlClTAPFB7dy5s9NjDjzwQL799lsMw+D999/f43ylpaVs27Ytobzjte4kkj6jaVBQIEYggLWujuH19eQEgj0Xd3si7obUNKIjxtEyYhwtc88CQPX7CK5ezdbPV5GzeR2Hla0jd+tmcu/5I5snHUDdheeRN3VMatbRB6I61Ph1GoI6eXYNtzUPxZYHzBQHqCpYLGA2gUkBJQJKFIgA4fYWzWhwD9sA1G8HbxXUrBdj1X/EOXLGQVGrwCuYDPZs0TpMioiBZY9dW5RWq6tVJHUkLLA26dqVZCSKovSom4jJZMLtduN2u4lEIgmBl2lxfylLlPD5fOTk5BCNRlEUhc2bNzN+/Ph2xzzxxBN8//vfFydWFJ5//nkuvvjiTue78soreeaZZxLHvvfeeymv1jwUkIkS3cQwoLkZ6uoI1dZR3xLGGxqYb2NbvDEWr63ksEUvcMGKxVha4/U+nnwEq886n2lThzHCmhmZYRYNcm0abms3L4aaCaxmyHJDjgfMe0myaKmEylWwazVUroTG0s6PUzRhwbN5xLB62h6b7UIoxusQRgLCkpi8jUWgcAqMPgpGHQGO1PZslCQRt+RBUgmXzsq6GKIbitra2ze59Vu8JmC8DVxHOt7GFEUcJ4W/pJ+J9/O12VJf6zPtxYePPvpoli1bhqIoPPzwwwkBF6epqYnhw4cnfOJZWVk88sgjXHDBBQnF3NzczC233MK//vWvRKux7OxsKisr2xUylnQPKep6QTQKDQ2EKqtpqmvsWbZsHwjEDNZuLKP4pWf4ztdL0QwdHYWF045n4YkXMnm/XE7KCZFlSn/gr02DXLuG09LDm6bLJcSdx909C06gASpXw65VQuw1lvY+WWBPFEyB0UeKkT9JJnsMFhS1dWit/YK1tuc0c2tbuG7+LUokA0zaRd1dd93FHXfcgaIonHHGGSxcuHC3Y2688UYeeOCBhGBTFIW8vDwmTJhAOBxm7dq17ZocK4rC//t//4877rgjVcscUkhR10eCQaI1tTSVVdLU5CM2QAazyPYdmJ98nIlffwyIdmT/OeAkXj72XK472EZxhlju7CaFAqeGtaf1ilUNcrIgJxsc9r0e3o5oCIJNEGoW22Bz+8dhv6jtlkjSSErWiD82dCj/Bko/h9pNHX6oHBgVF3gTwJEvCwIPZhS1te9vlrD2yt+1JENIu6iLd5UA4X/eunUro0aNaneM3+/nyCOPZM2aNQlhB7TbT358xBFH8NFHH2Hem9tG0ilS1KUOvbmFprJdNFXUEAkPjGvWumUTWU8/RvY3X4o1oPD5+JnY552B+9ijQFNbXVix1uxbHXQDiD+OtT7W21xdPeiY0V2yrSq5dhWtNwYPq7XVepcF1jRY4/11UPqFEHjly4V7drc1ZoEzXwxHPjgL2h67i8EzQrgHJfs+Jpv4fVvdYkgXriRNpF3UAbzxxhsEg8I9cuihhzJmzO7B3nV1dZx//vl8+OGHYhGdpCEDnHbaaTz//PN4PJ5ULnFIIUVd6jFiMZor62gs20W4rnFAzmlbuwr3C8+QvWJ54rlAdh6Bk0+l6eTTiQ4r6tmEsZDIWg35gNRY/TQF8hwqHmsfXFlWq4i/87hFdvFA31BjYeHqLV0mBF7zLvFZ7Q3VDNmjRFJHzljIbd26i4RVUrKPorTF9iXH+akmES+qmsTj+BcrPQp6fBtNehwTFkGTpTXG0NIWayj/PiRdkBGirie8/vrrvPjiiyxbtoyqqioMw2D48OEcffTRXH755cydOzddSxs0SFHXv/jqm2hatxFf8x5KQaSQ8M4y1r2wkBOWL6HA3wiAoSj4Dz6MplPOxHf4UdCDlH8MXSQPhFq6J166gU2DfKeGva8tMzQTZLmEyHM70xP3ZBjis/HXgq8GfLVtw9/6uKms6zg/zQo5Y0QNvkQ8l5IU45UU86WZRQ1EZ16rNTAPHAWi3p9k8KJorYWhk8rEqKakpJG4eJRxf0ONfU7USfofKeoGgFiM0PYdNJdVdrO/bN8IxgzuXd2C58vPuHTF23ynZEXitWhuHs1zT6Vp7mlEi4p7NnEsBOGWVutd33+ILKtCnl3DlIp7kaKAywluFzgde60ROKAYOrRUQcN2aNghSrM07IDGkj2UAOkBZrsQeY484e71jITsMUIsekYKISAZ/Chqm2Uw2dIXbwUn2/cNOqSok+yGFHUDSEMDse07aPYGaQyEifZj+aKobvDAphBLq6OMbtjFfaXvcuiydzA1NiSO8R94ME0nn47vqO9gWHoQ/G3EWq133j5b71QFcu0q2VY1tZ5UVQOnXQi9uMjLtNgnPQYtu4TI89ftHvtoxECPxzrqIgEkUN9qFawFX53oObwnFFW4eOMiL3t0m2XQni2zeIcSitq+3y8klXpJLiOT9FxyJ5FEv2CzFIgZQlpF3ebNm3nrrbcSj+fMmcO0adNSMbWkD0hRN8CEw7BjB0ZzM95QlEZ/hGCkfzJVdcPgX1vDLKgQ7WuuKIYf7PoKzztv4vh2OUrrv3bM5aZ59kk0n3w64bHj9zRlJxhtRWejob0UoO0aswZ5Palv11NUFRwOIfTcrp5n1GYqET/461vdvbXgrYbGncIK2FgiehF3haK1Wvfykix9BW2P3cNlzJ+kCzrEEiZCBzopDRMvD2Oyyb+lFJNWUff3v/+d6667TkyqKGzdurXTJAnJwCJFXRowDKiqgooKMAxaglGqW4Lo/aDtDMPg+dIIT5cIoXVmkZmfTrBgqakia8lbZC15C3NNdeL44KSpNJ10Gt7jZ6M7nL08aay90NN1Etm18f0usGmQ79Cwm/vZqmYyiWSLLLew5mWaFS8VGIaw7DW0CryGElHHr7FUWAa740LXzOAZ3WrhG9tq7Yu7daW1RtJDVHMXvZ1liEBvSKuou/fee/nNb34DwMiRIykt7aISvGRAkaIujfj9sG0bhEKEozEqm4KEov0T7fC/iggPbglhAAdna5w9wsxhORpmQ8fx7XKy3nkD17JPUVpb2uhWK96jjqVl1lz8Mw8RSQkpw0hyM7a6GGl1/bS+7LSo5DnMWE1Km1soGhKWJyPF5WJUrU3gpSvhYqDRo8LC569rdefWtVn7/HUi6aO5omurq6JC1giRxZs3URRkzp8oO29IeoeiihhA2MsXLKXN6peIF7Sl+Pq079Cb+3fKPqmCggJAWOmKi3sYoC2RDEYcDpg2DSorsdTXM1JVqW4Jpa63bBJnFZtxm+CPG0N80xjjm8YYWSaYVWhm7uRDmXTI4ZiaGnG/9w6exW9gKSsl64N3yfrgXaLZObQcN5uWWXMJTZycAquWAooJFAAzdOKR8QE+P7htJvKcFszJ2RTRkCggHPZB1C8sgH1Bj0FDoxiqKtyz8bIpg1XgqSZwFYrRFXpM9Npt2CGsew072qx+ET807RRj+0dt73HkC3EXF3n5k4RLdzBaQiWpw9BFf+fuEOnkuXiGsMkqxKHZLgpGD1GxtydSZql78803OeOMM1AUhUmTJrF+/fpUTCvpI9JSl0H4/dDYSGN5FbW1zf2SJVvi13m7MsJ71VHqw20nGGVXmDvMzImFJgqtCtbNG3G/vwT3h0sxNTUmjguPHEXzrLm0nDCX6PAe1r7rJYoCHoeJXIcVTe0gDgxDFAMO+0XSQDRIKjJzgbauFrnZYB8kMXipwDCEJa+xBOq2Qu1m0XWjqYxOP3tbtmizVjAZCqeKfXv2AC9aMiTRrKLkj8XV1jlmEH3BSKv7tb6+nqKiIiKRCFarldraWpzOXsbsSFKGFHWZib+phfIduzDqGlCCqe9lGjMMvmmI8W5VlE/rooRajV0KcKBH49QiE7MKTCixmHDPvr8E57JPUENt2a6B6ftTd9lVBA6YmfL1dYaqQq7DQrbDvFtB8nYkZ4wmZ5Git72WqAcXn0dpywRVVMAQQjHUAnpYZM/m5kB2Fmgy2LtTIv72Iq92syjj0pkl1T0cCqZC4RQh8vInipuuRNKfKCqYnULoJVy41n3Wopf2kiZnnXUWixYtQlEUnnjiCa644opUTS3pJVLUZS7RmE5pvR9/kw+toR5TbXXKjFDJ+KIGH9dGebcqysqmtjorh+Zo3DDJSn5rBwjF78f12Udkvb8E+8pvEtmzLcccR+1VPxowy51Jg3yXFbdtgAL1I0Eh7kItQFS4ZnOzRYKFZM9EQ1C/Fao3QE3raOwknlpRRTJGIj5vghB6FtfAr1ky9IjH9JniBZ5b91Vz6xc/pe2LIK1fBhOP00faRd3nn3/Od77zHQCKior49ttvE7F2kvQgRV1mYxgGVc0halpCKIEAlrJSlED/daeoCuosrozwUlmEsA4uE/xsgpXZhe0FlFZbQ+5/nsfz1v9QdB3dbKbxnAupv+BSDMfAWFxsZoU8lxWHZQC/ZUeDEPSKIsyaIfrRZrule7YnhL1Qs1EIvOoNULNeJGl0RlZxW3xe3kSxtecM7Holkj2haMKta3YIC6DZ0VYHsJ9Ju6gDuO+++7j55ptRFIVp06bxn//8h6lTp6byFJIeIEXdvkFLMEJNSwhfMIqpqhJTTVW/WO3ilPh17tsQZJNXuM5OKDDxswlWsjqUGrHs2EbBIw/iWPkNIDpW1F5xDS2zTxqwJAOnVSPPZcFqGmC3aDQkRF4sCiYVsuzgsoNV63vyxlDDX9fmso27b71VnR9rzxWWvNz9IG+C2M8aKWugSTIH1dQaw+fo16SNjBB1AI888gjXX389gUAAi8XCxRdfzMUXX8zhhx9Obq5MiR9IpKjbtwhGYtT5wjTWNmIuKUEJ9a2jw56I6gb/3hnhuZIwOpBrUbhhkpXDcztcnAwD57JPyX/8n1h2VYh1TppCzTU/JThtRr+tryNum4kcp3ngxV1HrFbhos1yCoEXDQvxFxeBRj+2EhlMBJuEwKvb0ir4NkFTOZ1+m9GskDse8lqFXt5+QvTJvriSTMCRJ8ILUkzaRd348W3V6uvq6mhpaREnSfJLu1wusrKyMJu7Hy8TL2Ys6TlS1O2bxHSD+pYgzVtL0Cu7sGikiI0tMf64IcjOgLgUnF5k4gfjrdi19lY7JRLG87/XyP33M2gBPwAtx59Iy3dOIDx2HJFhRQOSZGAxKbisJlw2U2YIvOxs8HjA5RIxOLFIe5EXCYp6cIk2YdLK1yURv2irVrdFJGXUbYH6ba1Zz52QVQy5E9ose/kTwFm4eyyUYYAeEb+LaFBkVCuAZ5RspSbpO4NV1KmqiqIoGIbRTsj19RSKohCLyW+/vUGKun0bwzBorm2kZcNmAi39F2sXihk8sSPMa+WiSFSRTeHGyTZmeHYXTVpDPXnPPEbWkrcSyRQgChqHR48lPGYcoTHjCI8dT2jMOGK5ef0WcGzWFFw2E06rht2c5gw3k0mIO48HsvaSRasnCTw91r4vbHJ/znaPDXFs2CvGUEGPiULJ9VuhdgvUtwo+X03nx1vdoj5fNNQq4gJCxHUmph35MOZoGHMMjJgpguglkp4y2EVdKokLRCnqeocUdYMEXSe4vZSKLaXEYv0XbPdtQ5T7N4WoCRkoCKvd98buHmsHYN2yCc+i/2LdthXLzh2o4c67E8TcWTSfeDJ1370aw2brt7WbNHBZTWTZM8BFqyjgdgsrXnY29MAz0S1iUQg1CxdmqHloWv+CTW3WvLqtQuw1lIhuGntCNYPZJtzmsaTwBrMdRh4OY4+BUUeCTV4vJd1ksIq6sWPHplzUxdm+fXu/zDvYkaJucOH3BSjdUIJWVwvR/vmi44sa/GNriCVV4uboNsH3xlo5rciE1tX/dyyGuXIXlh3bsJZsx1KyHeuObZgrylBam96Gi0dS9ctbCE6d3i/rjqMokOeykOPIIKuL2w1FRWKbagxDlGOJC7yuWn8NBWIRUTTZXy/qlJntrVtb22O11aIbDUHFCij5BEo+a+2X24qiwvADYPSR4n2RgHANt9sG2qyAzgJReDlefNkiy+EMKQarqJNkHlLUDT6agxFKa7yoDQ2Yaqr7LZliZWOUB7eE2eEXomyCS+Wn+1mZ3olLtiuUcAjHt19T8I8/Y66rxVBVGs69iPrLvodh7l/R5bRqFLqtmLQMiplyuWD4cOGi7S8iASFY9JiwWrUb8eciQ9O61xWGDjWb2gRe/bY+TKaIG3xc5BVOFUke6r5ZAFfSDaSokwwUUtQNThp8YcoaRIyd2tyEqa4WtTUxKZXEDIPXKyI8tSOMr9UwOKfQxPfHWcizdl8sqd4WCh55kKyliwEIjRlL1S9vJTRhUsrXnIxJg+FZNuwDWeuuOzgcQtzlpLEmWywi+uvGLU9hn8zcjdNcIcTdrhWAChY7mOxtdcridcvMdhGH11QG1etFTb6Wyt3n0yytGbutmbt541uzd6VFb1AgRZ1koJCibvBS3RKkqqnNSqcEAphqq9EaG1Je464hrPPEjjBvVwqXrEOD746xMK/YjKljv9Y94Pz8Ewof/BOmxgYMTaP+4u9Sf+FlIsmgn1AUyHWYyXFa+i08pNfYbMItm5OT9ur1gEgsiPhbxV6rq7E/CyYORgINrQJvg9hWr+86scVd1FaeJW8/UXzZNTwz/hYk3UeKOslAIUXd4KaiMUCdt30MlRIIYK4oQ/X5Un6+Dc0xHtwaYmOLcN2Ndij8epKNKVndd8mqTY0U/uP/cH/6IQDB/SZRdcMthMeMS/l6k7FbVIZl2TBnkjs2jtUKBQUiqcI6MNXqu4Wui+4a8TZqXZUWkXSNYQhLXt3m1oSOrXvO3nUWQNGBUHQQFB3QWnZFiryMRoo6yUAhRd3gp7TOT1MgstvzWn0d5sqKlCdU6IbB4sooj+8I0RQBswI/n2jllOE9yPA0DFwfvUfhP/+C5m1BN5lpPO8imuaeRrSoOKXrTUZTYViWDac1w9yxyVitoiSKxyMSKwaoc0e3iIaF1SnULETe3jJNJV0TbBKxe3VboG5bm9jr6AK357aKvAPENmesrK2XaUhRl34qKyu59957WbRoEeXl5Xg8Hg4//HCuu+46TjzxxD7NXV5ezt///nfeeOMNSkpK0HWdoqIiDjvsMC666CLOPvvsTt+n6zqPPfYYTz75JOvXrycWizFhwgQuueQSfvGLX2Cx9DywXIq6wY9hGGyv9eELdSLeolHMlRVo9fUpP29LxOBPm4J8WifOe1axmR+Nt/TIHavV1zHsb/fj/GpZ4rnAtBm0zJpLy7Gz0N398zeb4zSTl4nu2I7ES6NkZYmRaT1ow34h7uK182TyRd+IBKB6HexaJeL5qteJ2MdkrFlQMBnyJ4ttwWRh3cv0v+XBjBR16WXVqlXMnj2bujqRwp6VlYXX60XXdRRF4Z577uHmm2/u1dwvv/wyV199daKbhsPhQFEUfK2usBNPPJF33313t/dFIhHmzZvHm2++CYDFYkHTNAKtzd0PO+ww3nvvPVwuV4/WI0Xd0CCmG2yr8RKMdH5TVX1ezOVlKMHUus90w+D50gjPlAgX8P4elf831UaOpQeWBMPA9cmHZC1ehGPlN4kSKIbJhO+wI2medRL+w49MebaszaIyPFPdsV1hNgtLnsXSNpIfp9OqZxgiFi/shVC8QPKQu72klmgIajYKgbdrFVSt6dwFbs+B/EltYi9/ohAasl/uwDBYRd1VV12VqqnaoSgKjz/+eErmCgQCTJ06lZKSEmbOnMmzzz7L9OnTaW5u5q677uJPf/oTiqLw9ttvc9JJJ/Vo7jfffJOzzjqLWCzGVVddxY033sjkyZMB0Tbtww8/ZMeOHfzyl7/c7b033XQT9913HzabjYceeojLLrsMVVV54403uOKKK6ivr+eSSy7h+eef79GapKgbOkRiOltrvESiXfxLGwam2hpMVbtAT+3N9rO6KH/cEMQfgwKLwu3TbUx29/yGotXV4v5wKe73l2DbtiXxfMzpwnvsLFpOOJHAtP1T1o5MU2G4x4Yj07Jje4vJJMSdxwN5eemNz9N1iPiEwAu1iH1J39Cjwl1bswlqNwrBV7+966xli1NY9qzuTkaWSNTIHQtZI2TJlb4wWEXdvtBR4i9/+QvXX389LpeLDRs2MGLEiHavn3POOSxYsICDDz6Yr7/+utvzNjc3M2XKFHbt2sWtt97K73//+26/t7KykrFjxxIKhfjrX//Kz3/+83avL1y4kHnz5qEoCitWrOCAAw7o0bqkqBs6BCMxyhr8BMJdu8GUcFhkyTY1QiR1MVGlfp071gbYGTAwK3D9JCtzh/W+k4Jlxzbc7y/B/cG7mGvbgsqjWR58RxyN74hj8M88NCVdKvKcZnJdGZSgkCqcTiHucnL6NcO4W0QC4KuFQL1006aSaEjE49VsEiKvdiM07OjZZ6yaIWc05IxrHWMhdxy4h8v4ve4gRV3ndFxKch/ZVIm6ww47jOXLl3Pttdfy8MMP7/b6Z599xjHHHAPAhg0bEpa2vfHXv/6V6667jpEjR7Jt2zbMPWgL9I9//IOf/vSneDweKisrsXVyk5o8eTKbNm3ipptu4g9/+EO355aibmjS4AtT2RwkupeWYorfh9bUiNbUhNJFm6+e4Isa/GFDkGX14v/1nBFmrh3Xszi73YjFsK9Zhfv9d3B9/gmat60en2614p95KL4jv4Pv8KOIebJ7fRqHVWNYphUrThWK0ma983jSG3+lx0TZD1+t6MggST16tDVjubnVUtrclsEcbBau8WCTyMpt2NF1VrPJBp6R4BomBJ5ruNi6h4l9q1vG8kFGibqUf3Xrq0aMi0LDMPo8V0daWloS1reTTz6502OOPPLIxIe4dOnSbou6uFv0/PPP75GgA3j//fcBOO644zoVdAAnnXQSmzZt4r333uvR3JKhSY7TgsduptYborolRFf/SobDSdThJFo0AiUQEAKvuanXsXdOk8Kd0208UxLm+dII/y2PsM2rc9tUG9mWXl78NY3AgTMJHDiT6p/dgH3NapxffILr808wV1fhWvYprmWfYqgqwakz8B59LE0nn45hd/ToNP5QjJ1Rf2YWK+4rhgGNjWJoGuTmitHDGN2UoGrgzBcj5AV/LQQakfF3KUQ1iTg7ezeKWxs6tFRBw3bhym1oHY2lQuzVbRGjM8x2Ifg8o6H4IBh5iNiXQi9tpPTK1Zv+rH6/n5qaGr766iteeuklli9fjqIoTJgwgccff5zRo1OnftevX58QitOnd95/UlVVJk+ezJdffsm6deu6NW8wGGTFihUAzJw5kw0bNnDnnXeydOlSmpubKS4u5uSTT+amm25i7Nixu70/fp6u1gQwbdq0dj9DVxbRUChEKKltVHNzc7d+BsngQ1UVCrNs5DgtVLeEqPfu2RJn2O1E7Xaiw4tQQiG0+lpMNV3U0trTeRWFK8dameDSuG9jkJVNMb7/tY/vj7Ny0jATal8u+JopIfBqr/kplu1bcS37BOeyT7Ft3Yx97Srsa1eR8/IL1F1+Nc1zTu1R/F00BuWNQfKcFnKcGdQ7NpXEYlBTI4bFIlyzubmiy8VAY3WJkRURvVeDTa0FjyUDhqJCVpEYY45ue16Pis4azeVC9HmrRLcMb6XYBhqES71hhxg7PhLvc+TDiENax8FCvEsGjIzLfn3ttde45ppraGxspKioiHfffZcpU6akZO54bBoIsePuorl2PK7u3HPP5dVXX93rvBs2bGDq1KkA/OY3v+HPf/4zgUAAh8OBqqp4vaKaeFZWFgsXLuSEE05o9/6cnBwaGxv529/+xs9+9rM+rf2OO+7gzjvv3O156X6VBCMxKpuCtAS7H0entjRjKSvtdexdiU/nd+uDif6xM7JUfj7Ryjhn6rPyTNWVOJd9Svb/XsWyqwKA0Njx1F79I/wHH9bj+VxWjYLB6o7tDJutTeClIE6x18Si7d2F+u41GCUZQDTUJvTqNkP511C5evcSLNljhMArOkDU3LM4wOwUSRwWx+BI0Mgg92vGiTqAr776ihNOOIFAIMDYsWNZsWJFSgTJCy+8wKWXXgqIEiKmLgKHL730Ul544QVOOukkFi9evNd5ly1bxlFHHQUIS9/w4cN58sknmTt3Loqi8Nlnn/G9732PTZs2kZ+fz8aNG8nNzU2832KxEIlEePTRR/n+97/f6TmWLFmSyMatqKigqKio0+M6s9SNGjVKijpJgpZghPLGQNdZsh2JRrHsLEVt6Z3VN6ob/LdclD0J6qAC5400890xFuxaP7hpIhGy31hA7r+fScTf+Q45nNqrf9TjrhUmDQrdGV6suD9wONoEXi/qY6aUSKB9LJh002Yu0ZAou1L+tRg1m9jr70uztgo8J1hcIkGjcBoMmyYE4b5QliWDRF1GXqkOO+wwfvWrX3H33XdTUlLC7bffzv/93/+le1ldout6u/1nnnmmXQHjo48+mldeeYWDDjqI2tpaHnvsMW688cZ+WYvVasWaSW2GJBmH22ZmvwKN7bU+Ql3UtWuHyUR43Hi02hrMu8p7fE81qQoXjLJwfIGJf24N8WldjJfLInxQHeVHE6x8J09Lbda82UzjvAtoPvFkcl98luxF/8X59Zc4vl1O80mnU3fZ94jl5O59HoQ7tqIxSLbDRJ7TitqXhI99Cb9fjPJykViRn5++BAuzvTV2q1CUSQk1gb9eWPMkmYXJ2uZ6BSHEd61oFXgb2+oYhv0QazU+xEIQCImsaICa9bBR1GvFbIeCKVA4VQi9wqlCQEm6JCNFHcAPf/hD7r77bgzD4NFHH+Wee+7B3sdq6k6nM7EfCAS6dGH6/SKmo7uFfpOPmz59eqcdKfbff3/mzJnDO++8w9KlS9uJOqfTSWNjY6LQ8J7W1JN1SSRdYdZUxuc72V7r67JgcUdi+QXoTieW0hKUJGtwdym0qdwx3c4XdVEe3BqiMmhw17ogh+dq/HQ/K0X21Lo5dXcWtdf8hKbTzyb/yUdwffYRnrdfx/3hu9SffwmN887HsHXvmtLojxIIxxjmsWE17QOWg1TS1CSGySTEXV5e+tyzqtqWABCLCHEXqJc9aTMVWxaMO06MjuhRIe7CPlHDMOyHYGNraZZ1UL1BxFdWfCtGHNcwGL4/DD8Aig+UiRkdyFhRV1RUxLhx49i+fTuBQICPPvqoy4zV7lJc3NZTsqKiosvM1oqKisQaejrvnrJlJ0+ezDvvvMPOnTt3e39jY2PivHtak8vl6lKMSiQ9waSpjMt3sqPOTyDcvZJBht1BaOJkzBVlvW49dkSeiQOzNf5dGuY/ZRG+rI/x/UY/l462cOFIc9/Kn3RCpHgku35zF7Y1qyh4/J/YNm0g/9nHyV74Co1nn0/TGfPQXXv/nwpFDXbWBwZ3EsWeiEahslIMl0sIvJyc9HWx0MyitIZ7mBAG/noRvN9pIV5FWJE0S+vWKhIEQk3CtSvr5g08qkmIPlsHt2JcAOoxaCyB6vWiXVr1epGd662CLVWwpbUzkz1HCLyiA0UG7hDvjZuxog6goKAgkVG7YcOGPou6KVOmJGrfrV27tlMBpus6GzduBNoyTvdGfn4+w4YNo6qqqlvHd3Q1TZs2jXXr1rF27dou3xPPkI0nZEgkqaBN2Pnwd9Y7tjNUlcjI0cRcWVjKSyHW8xuiTVP43jgrc4aZ+duWECsaYzy5I8wntVF+Pbl/EimCMw5g55/+ifuj98h99gkslRXkP/s4Oa++SNMZ59A47/y91rkzDKj1hvGFowzb11qMpRKvV4ydO4Ww83hEj9oUdfroMfGYLM9IYe2JBIRwM1nEVjN3bs1x5rW5dAONwqXbbYGniDpuZpsozSITOlKLqkHueDGmnC6eC/uhZgPsWilG9Toh5Ld/KAaIThnD94dhMyBvPzHsuUPGmpfRoi6eNQrt3Y+9xe12c+ihh/LVV1+xZMkSzj333N2O+eKLL2hqagLo1I3aFXPmzOH5559PCMLO2LBhA8BuZU1mzZrFK6+8wscff0wwGOy0Vt2SJUt6vCaJpDtoqsK4PCcl9X68PciM1bOzCbpcqH4fqt+PGvCj+n09EnmjHCr37W/j3eoo/9oaYrNX5yffBPjuGAsXjjKjpfpCrKq0nDCHlmNPwP3R++T853mspTvI/c9zZC98haZTzqDh3IuI5RfscZpAWKe03k+h24bbltGX0f4lFoPaWjEURQi7rCwh8tLholWU7tdni5Ps0tV1IeyCTWLErX6K1hbbZ3YIMWeytVkpdV3U2/NWCbeipH+wOESZlBEHi8exsBB5Fa0ir2qN+P2VfCpGHFt2m8DLbd1mjxFif5CRkdmvAI2NjRQWFiY6STz44IP86Ec/6vO88TZhbrebjRs37uZiPe+883jttdc45JBDWL58ebfnfeeddxKWxCVLljBnzpx2r69evZqDDjoIXdd58MEH+clPfpJ4LblN2N///nd++tOftnvv66+/zllnnSXbhEn6FV032NngpznQh5uSYaAEgwmBp/r93S5kXBvS+evmUKIbxWS3yq8n2Rjj7EdrmK7j/OJTcl96Dttm8YVMN5lpmXMK9efPJ1pUvJcJIN9tIccxBN2xe8NqFeLO4xHu2nS5aXuLYQi3rmYWLtvuIMVdetGjULtJiLzaTaJoclMZnWZ3KRpkFYuM20SZFUeb1dfcum91C5euZ2TX5VcyKPs1Y0XdrbfemmiHpSgKb7zxBqecckqf5w0EAkydOpWSkhIOPvhgnn32WaZNm0ZLSwt33303999/PwCLFy9OlBCJE3eb3n777dxxxx27zX3SSSexZMkSiouLeeqpp5gzZw6KovD5559z5ZVXsmnTJsaNG8eaNWtwdCj0edNNN3Hfffdht9t5+OGHueSSS9A0jTfffJPLL7+curo6LrnkkkTniu4iRZ2kJxiGiBtrCqTQlRSNYt2+FWUPiUDJ53+3Oso/toTwxcCswBVjLZw/sh+sdu1PjOPb5eS89ByONSvFU6qK74hjaD7xZHyHHbnH3qmFWVY89sH3rT9laJqIwSssTH+JlIFAirvMIRqE+h1QvwXqtkL9NiH2wr6ezaNZhLjLm9De4md1S1G3J3Rd57777uM3v/kNIC7yNpuNurq6Pme/xlm5ciUnnngidXV1gCgK7PV60XUdRVG45557uPnmm3d7395EXV1dHbNmzWL16tUAOBwONE2jpUXUyiouLmbx4sXMmDFjt/dGIhHmzZvHm2+KVG6r1YqmaQm382GHHcbSpUt7nCQhRZ2kpxiGQVlDgEZ/CoVdLCaEXTfDKGpDOv+3OcSXrVa7qW6VX022MdrR/9Ye29pV5L70HM6vv0w8F83yCLftiScT2m9ip/E5wz1W3DYp7PaIoojad8OGQYqu5xmNFHeZiWG0Fk7eJeL04tm38W3YJzJvwz6RXV2/vesMa9cwUWplyhlw+DUpXWbaRd1HH33Uq/cFAgGqqqr49ttv+e9//8vOnTsT7bwUReHHP/4xf//731O1TEC4PO+9914WLVpEeXk5WVlZHH744Vx//fVdxq3tTdSBKP7717/+lRdffJHNmzcTi8UYP348Z599Nr/85S/Jy+u6xo6u6zz66KM89dRTrFu3jlgsxsSJE5k/fz7XXXcdll58w5WiTtJbqpqDVDf3vHRJl8RiWHZsQ/V17xuyYRgsrhKxdv5Wq92VYy2cM8KMeQDqxVl2bCPr3bdxv78EU2ND4vnQmHE0n3gyLbPmEstt+39WFBjuseEaaoWKe0tWFgwfLuLwBjvxWL24UAj7kEWU9yEMXbRMq9sK9VuhdovYepOSI/e/AM57LKWnTbuoU1W1z0VEk8WcYRhMmDCBL774gpycHgS+ShJIUSfpC4FwjLIGf7dr2e2VHgo7gOqgsNotbxBWu2KbwlXjrByXn+KixV0Ri+L4ZjlZSxfjXPYJakRYMA1VxX/wYTSdcia+I48BRUFRoDjbhsMihV23cTiE5S4nZ8hkKGIYSQLPK7bSkrfvEWoRQq+lEkYeChPnpnT6jBF1vZ0yfoGOv3/mzJm89tprjBkzJlVLHHJIUSfpK4ZhUN0SoqYlREquFrouhF1Sdnt31rC4KsoT28M0RMQiprhVfjDeygzPwJXRUFtacH3yPllL38G+fk3i+cDUGdRe9UOC02agKDAi24ZdCrueYbGIwsY5OUPDNduRaEh0YPDXymLK+xqDNaZOTVF208yZM7nmmmu45ppr0NJV92iQIEWdJFUEwjHKG/0Ewimw2uk6lpLtqK3xpt3FHzV4pSzMy2URgq3LOCZP4+pxVkYNQLxdMubyMrIWv0H2otdQWztstBxzHHVXXEts1EhGZNuxmeX1q1fYbELcDVWBF/IKcRdoRLpp9wEGq6i78847e/weRVGw2WxkZWUxZswYDj74YIYNG5aqJQ15pKiTpBLDMKhpCVGdCqtdL4UdQF1I55mSMG9XRtEBFTi9yMR3x1jIsQysuNPqasl7/kmylryFousYmkbTqWfReOkVFI0rGnptxVLNUBZ4sagI1PfVtvVKlWQeg1XUSTIPKeok/UEwEqOsIdDt9mJdoutYSnagtvSuOXuJT+ex7W217ewaXDRSFC4eiGSKZCw7tpH/5CM4ly8DIGZ30HTRJbiu/C4Wl3Mv75Z0C5tNZM/m5oo6eEOJUIsQd8EmpPUuw5CiTjJQSFEn6S/isXZ9zpA1DCw7tvda2AGsbIzy8LYwm73CJzsjS+W302wDbrUDsK/8hvzHH8K2dRMA0bx8uPZaTGedCWZZ8iRluFxC3OXk7LGG4KAjFoVwixB5Ia+04GUCUtRJBgop6iT9TXVLkKqmvgs7rbEBU21Nt4oUd4ZuGLxfHeVvW0QJlHyLwh3TbUx2p8H9qeu4P3qPvKcfxVwtyh7o+QWoF18E554rynlIUoOiiM8zL090r9jXOlf0lWioVeA1C5Fn9NF6Luk5UtRJBgop6iQDQXVzkKoU1bRT/D5MdbVojQ298jLt9OvcvjbAzoCBWYHrJ1mZOyw9FjIlHMLzxkJyXnsJU70odm7Y7Shnnw3z58OIEWlZ16BF0yA7Wwy3WzweSsRLpYS8EA0IwRcNijprkv5DijrJQCFFnWSgSKWwAyAaxVRfh1ZXixLpWXcLX9TgjxuDfF4nrBbnjDBz7TgLpgGOs0sQieD+6D1yXnsJ645t4jlVhdmz4bLLoJMuM5IU4HAIced2C3ftUBN5cWIRIe7iIi++jYXTvbLBwWAVdYFAgJdffjnx+KijjmLixIk9nmfTpk0sW7Ys8Xj+/PmYZSxKr5CiTjKQpLwLBYBhoLY0Y6qr7VGmrG4YPFsS5rlSIQgP8mjcNs2Gx5zGAreGgWPF1+S89hKOb75qe/7AA4W4O/bYoRUfNpAoyu4ib6i5ajsSi7YVP463xpJJGD1nsIq6xx9/nGuvvRYAi8XCtm3bKCoq6vE8FRUVjB8/nkjrt/PnnnuO+fPnp2qZQwop6iQDTWVTkJqW/gneVkIhtIZ6tIb6blvvPqmNct/GIIEYDLOKOLsJrvRbbKw7tjFs0StYl7wD0dZuAnl5cNppcOaZMH58ehc4FLDZhNCz28XW4RjaolrXW/uf+tq6XUjX7d4ZrKLu2GOP5dNPP0VRFC655BKeffbZXs91ySWX8OKLLwIwd+5cFi9enKplDimkqJOkg/4UdnFUbwtafR1acxPoe76M7fDFuH1tkIqggVWFGyZZmVWYGdb/rJZGCt5egLpwITS09Zhl+nQh7k4+eWj0R80UzOY2ged0iuSLoYphCDdtJNDa1swvYvWk0GvPYBR1TU1N5Ofno+vil71gwQLOPPPMXs/3v//9j3nz5gFgtVqpr6/HPtQKT6YAKeok6WJXU4DalgGI2YnF0JoaMdXXofj9XR7WEjG4d0OQr1p7yF462swVYywD0z92L1hMCsMdGtYvlsH//geffgqx1ixGiwVOOAHOOgsOO2zoxoWlC6cTRo4U7lrJ7kIvvh3KQm8wiroPP/yQWbNmiUkVhfr6ejx9+IbT1NREbm4uhmGgKAqffPIJRx11VCqWOqSQok6STioaA9R5By4YWwkGMdVUoSVbvJKIGQZPbA/znzLhuj1tuImfT7SiZYCwUxQocFvx2M1QVwdvvQWvvw5bt7YdNGwYnHGGEHgyc3ZgyckRn/lQK3rcHfSYKIrsrxc19IYaGSTqUhYlumnTpsT+6NGj+yToADweD6NGjUo83rhxY5/mk0gkA09xtp08l2XAzmfYbERGjSEychR0otM0ReGa8VZ+PsGKCrxZGeWudUFCsfQHhxsGVDeH2NUYJJaTKxInXnwRnn0WLrhA1GKrqoLHH4ezz4Yf/xjefhtCsvjsgNDQAGvXQllZmxVVIlA1cORC/gQonAbuItCk+E0HKRN1jY2NgLDS5efnp2TOgoKCxH59fX1K5pRIJANLcbad0bkOtAEsJxLLzSO030Qwdx70fmaxmf83zYZZgc/qYty8OkBLJP3CDsAbirKz3k8gEhXmu6lT4aabhIC791448kjx/Jdfwm23wamnwgMPwObN6V764McwhLBeswaqq+l7A+RBiMkK7uEwbBrkTRRWLEWGDAwUKRN18Vg6EO2DUkHyPCH5bVQi2WfxOMxMHObCbRu4zELD4SQ4YTK6s/O+q9/JN/GH/e04NVjTrPPLlQFqQpkRFxSJGZQ3BGnwJbmuLRaYOxcefFDE3V1zjXDHNjcLi978+XD55fDqq+D1pm/xQ4FoFHbuFJa7hgYp7rrC6hJuyWEzIHc8ZI0E1zCw54DFLax5yhAvK5NiUnaFjVvnDMOgsrIyJXNWVVUl9nNyclIyp0QiSQ9mTWVsvpMGX5iKpgD6QOgns5nw+AmYK8rQ6up2e/mAbI3/O8jOzauD7PDr/GJFgHv3tzPGkf4bjWFArTdMIBKj0G3FpCWtqagIfvAD+P73hcVuwQL48ENYt06MBx6AY46Bk04Ste9kkln/EArBtm0iYzYvTwybLd2ryjxUFWx7CMmKRUGPiKLIYS8Em2VP216SMlFXWFiY2N+1axelpaWMHt37wMHS0lLKy8sTmWnJ80skkn2XHKcFp9VEWYMfX2gAYpMUhciIUeh2B+bynbvVVh3n1PjrQXZuWR2gLGBw/Qo/v5thZ1pWZriMfKEYO6N+hmXZcFg6XLI1DY46SoyGBnjjDWHF27YNPvhADJsNjjtOlEY56ihh8ZOklkgEKivFcDiEuMvNHdo173qCZhLDbAd7NngQAi/Y3NrTtgVZFLl7pCz7tbq6ul2h4d/97nfccsstvZ7vnnvu4bbbbgNEnN6OHTvaJU5IuofMfpVkMrXeEJVNwQHzXil+H9aS7RCJ7vZaU8TgtjUBNrToWFX4f1NtHJGXWTflHKeZXIcFdW/xiVu2wOLF8M47UF7e9rzLBbNmCYF36KFSdPQniiKSW/LyRK27od69oi/oepu4CzVnXnuzDMp+TWnx4YMOOojVq1djGAYej4e1a9dSXFzc43nKy8uZMWMGzc3NGIbBlClTWLduXaqWOaSQok6S6QQjMcoaAgTCA5RRGIlgKd2B6vPt9lIgZnD3OlHLTgV+OsHKGUWmjKhlF8ekQZ7TitvWjXUZhnDHvvMOLFkigvvjDBsGl14K55wj3bP9jabB8OHiM8+gv6V9lkiwTeCFfWCkORs5g0RdSr86XHvttYm6ck1NTZxyyilUJ19EukFVVRWnnHIKTU1NibnircckEsngw2bW2K/ASYF7gEogmM2E95tIZNTo3bJj7ZrCXdNtzC00oQN/2xLit2uDNIQzI4ECIBqDquYQZQ1+/OHdLY7tUBTRmeL662HRInjkETj/fMjOFlmcf/6zqHv32GMi4ULSP8RiwmK6bp38nFOB2QauAsjbD4bvL7Js3UVgcdFpLaMhREotdZFIhIkTJ7Jz505AJE3k5+dz//33c8kll2A2d92WJxKJ8Pzzz3PjjTdS1xrQbBgGo0aNYvPmzVhkHEivkJY6yb5Ekz/Czgb/wCUT6jqm6ipMtdXtWo0ZhsEr5RGe3B4mYkC2WeH6SVaOzjB3LIDTqpHnsmA19SAGMBwWIu+ZZ0TdNRCxYOedJ6x3KSpLJemCnBzRpULe11KPHhPJFiGvsOJF/PR7PF4GWepSKuoAPv/8c2bPnk04LHzecWtbXl4es2fP5tBDD6WwsBCXy4XX66W6uprly5fz3nvvUVdXlzjeMAxsNhvvvvsuRx99dCqXOKSQok6yrxGMxCip8xOODpx1TAmHMVVWoLXW24yz3RfjDxtCbPOJtZwy3MSPxltxmDLLGqAokGU3keuwtM+S3RvRKCxdCk891VbnzmwWPWcvv1wID0n/oKoii1m6ZPsXXYeITwi8+Ei1u3YwizqAF198kSuvvJJIRLTiiZ9iT/EfyccYhoHVauWZZ57hggsuSPXyhhRS1En2RWK6wc56Py3BvbgXU4zq82LeVdGuh2xYN3h6R5iXyyIYQJFN4abJNqZ7MiM7NhlVhVyHhRxnDy1AhiH6zT75JKxc2TbZnDlw4YVw4IFSePQXNhuMHg1ud7pXMjSI966NW/H0aGtJldbRG8E32EUdwFdffcWFF15ISUlJOzHX2ek6vr7ffvvx0ksvcfDBB/fH0oYUUtRJ9mWqmoNUNw98vSqtoR5zZUW7LNmVjTHu2xikOmSgAheNMvPdMRbMA9gpo7vkOc3kunoZo/jtt0LcffZZ23MTJohYvFNPFQ3uJalHumQzA11vE3h6RAi+aLB1hDqvnzcURB1AOBzmqaee4q9//Svr16/f6/H7778/v/jFL/jud7+7x/g7SfeRok6yr9MUiFDW4B+YYsXJ6DpaYwOmulqUQAAAX9TgH1tCLKkWYm+CS+XmKbaMKFbckWFZVrLsfbiObtwI//lP+/6yDgecdpoQeBMmpGahkjYURZRAGT4crLJ3akai6+1FXjQo6uu5h6f8VBkn6pIpLy/n448/ZuvWrdTX19PS0oLb7SY3N5eJEydy7LHHtqtzJ0kNUtRJBgPBSIzSej+hSHqyUBW/D1NdLVpTI+gGH9VE+cvmIC1RcGrw5wPtjHdlljtWUWC4x4bL2sfkjuZmUdT4lVegpKTt+ZkzRWLF7NnSutQf5OYKcSfLzQxZMlrUSdKDFHWSwUJMNyhr8NMcGNg4u/aLiKE11GOqq6W+OcBd64Osa9bJtSj85UA7RfbMstgpCozIsWE3pyBr1zBg+XJ4+WXRkizWGnuUnQ1HHy26VRx5pHAjSlJHdrZIqHA40r0SyQAjRZ1kN6Sokww2alpCVDUPXBeKrlC9LQQqa/j1x5Vs9+mMsCv834F2ciyZJew0FUbm2LH0pOTJ3qiuFv1m//tfqKlpe15RYMoUIe6OOgoOOEB2rUgVHo+w3Llc6V6JZICQok6yG1LUSQYj3lCUnfV+orH0X74aKmv59aKNVAUNJrpUHjjAnnElT8yawogcO+aelDvpDtEorFgBn38uxqZN7V93OkU7sqOOkkkWqcLlEuLO40n3SiT9jBR1kt2Qok4yWInEdErr/fhDaW4RBFRUNXHjgjU0RQwOztb43QxbxmXF2swKI7Ide+8b2xdqa2HZsraRXPevqAhuv12IPEnfsdmgsFAkVsi+soOStIu6xsZG/va3vyUez5s3jwMOOKDH86xcuZKFCxcmHt9www045Te8XiFFnWQwYxgGlc1BalvS3+B7064mfrNgDcGYwQkFJm6ZYkXNsNpuDqtGscc2ML1sdR02bBAWvIULoaJCPD9/PvzkJ0KUSPqOyQQFBWLIqhGDirSLuv/7v//jhhtuQFEUXC4XJSUlZGdn93iehoYGxowZg6+14fY///lPfvCDH6RqmUMKKeokQ4EBby/WBd+UNnD3orVEdZhXbObH+1kGRkD1ALfNxHDPAAsqnw/+8hcRgwcwZgzceSfMmDGw6xjMxMuhDBsmBfMgoTf375TabJ9//vnE/iWXXNIrQQeQk5PDxRdfjGEYGIbBc889l6IVSiSSwYjHYWZCoQurOb1uqINH53DdnMkALKiI8O+dkbSupzNaglFqvQNc0NnphN/8Bv76V9FXtqQErroK/vlPiGTeZ7RPYhjC/b12rYhtLC+HujohqGPpD1GQDAwps9TV1NQwfHhb8b133nmHE088sdfzvfvuu5x00kkAmEwmamtrpaWpF0hLnWQooesGZQ0BmgLpFQr/W1nBox9vA+D6iVZOK8o8t1i2w0S+yzrwlsSmJrj/flHUGGDSJLjrLlnMuL+xWIQFLz7sdlEmRcbjZSxptdStXLkyYVlTVZWjjjqqT/MdffTRqK1/bLFYjBUrVqRglRKJZDCjqgqj8xzku9NbDPesA4u54JCRAPx1c4hPa9NYW68LGv1RyhsDRGMDXNDZ44Hf/Q7+8Aexv2kTfPe78NRT0qLUn4TDopB0dTWUloqOIStWiG1FBbS0MPBtWySpJmWibuvWrYDo4zp27FgcfSyU6HA4GDt2bOLx5s2b+zSfRCIZOhR57BRnpzeu6LtHjmHutGHowO83hPg6lHltnwJhnZ0NfoKRNIipOXPgpZfg2GOFC/bBB4W4+/xz0h4cOVQwDPB6YdcuIa5XrBDbXbvE8/L3sM+RMlHXmJS6npubm5I58/LyEvsNDQ0pmVMikQwN8lxWRuc5SFeegqIo/OSECRw5PpeIbvDbrxtYnjWCaEEBpLIQcB+JxqCsIUBLMA3WxPx8+POfRakTl0sIip/9DH78YxEbJhlYDENY7Coq2ix55eXSgrcPkTJRpyb55cPh1JQXSJ4nGs0894VEIslsPHYz4wucaGmqGaepCjeePIVDxuQQjurcuXgzq3ATnDKdyKjRGBnS19MwoLIpOPAJFCCyNs88U3SouPRSUZbjq6/giivgppva95uVDCy6DpWVsH69SLiQZDwpE3X5+fmAqBu1a9eulMyZPE+y1U4ikUi6i8NiYr9CZ9oyY82ayq2nTuWgUdkEIzp3/G8tm2p8xHJyCU2cTGjCRGLZ2ZABlU8afBEqGgPE9DS43bKz4frr4bXX4IwzhNhbuhQuvBB+//v27cgkA0swKGoOlpVJq12Gk7KrXFFRUWK/pqaGDRs29Gm+jRs3Ul1dnXg8bNiwPs0nkUiGLlaTxvh8Jw5retyeFpPKb06byv4jPAQiMX67cA1bqr0AGA4nkdFjCU6ZLlyzae5E4QvFKGvwE4qmKWmhqAjuuAP+/W8RbxeLifp28+bBP/4hynTIWK/0UFUlrHZeb7pXIumClJU0aWpqIj8/H71Vxf/qV7/ij3/8Y6/nu/HGG3nggQcA4drdtWsXBQUFqVjqkEKWNJFI2jAMUfKk0Z+ekieBcIzbX1/L+l3NuK0mfn/ODMbld2jQHothqq3BVFcjAt7ShKrC8CwbTqspbWsARFzX3/4Gq1a1PedwCPEXH8XF7fezs0lbMOVQobAQRoyQJVH6kbR3lDj66KP54osvMAwDm83G8uXLmTZtWo/nWbNmDYcffjihkIjvmDlzJsuXL0/VMgGorKzk3nvvZdGiRZSXl+PxeDj88MO57rrrelVf74MPPmDWrFl7Pa6mpibhqk5m7NixlOwlduT+++/nV7/6VY/WJUWdRLI7u5oCaWst5g9H+e3CtWysaiHLZuKec/ZnTF4nbRB1Ha2+DlNNNUqaCvQqCgzLsuK2pbnOnmHARx/Bww+LZIq9kZMDp5wi3LiTJ/f/+oYqVqvoDuJ2p3slg5Le3L9T+hXsZz/7GcuWLUNRFILBICeffDKLFy/ukbBbu3Ytp556KqFQCMMwUBSFn/70p6lcJqtWrWL27NnU1dUBkJWVRW1tLYsWLeKNN97gnnvu4eabb+7V3Kqq7tGiqO7lW01OTg4WS+c1tmT/W4kkNRR57BgG1HkHXtg5LCbuOGs6/2/BGrbUeLlt4RruOWd/RuV0KAOlqsTyC4jl5aM1NmCqrkIJDWwig0igCBHVDXIcaaz9pyhw/PFihEIieH/XLpGluWtX+/3aWmhoEO7bf/9bFDc+4wwh8lJUmUHSSigkRHZBgbDaaZmT1T1USamlzjAMDjzwQNa2pqLHLXY33XQTP/rRjygsLOzyvdXV1fzzn//kvvvuIxgMJqqcT506ldWrV6es6nkgEGDq1KmUlJQwc+ZMnn32WaZPn05zczN33XUXf/rTn1AUhbfffjvR0aI7xC11Y8aMYceOHT1eV9xS9/7773PCCSf0+P1dIS11EknXlDX4afClxwrWEozwmwVr2F7rI9dh4d5z96c4e8/ZsGpjI5ZdZRAZ+GoAOU4z+a7Mq7W3G+GwyJ5dtAg++KCtDZmmwXe+IzJtjzlGZNlKUofFAqNHi4LSkpSQdvcrwIYNGzj66KNpamoCSFjbTCYTM2fO5NBDD6WwsBCXy4XX66W6uprly5fz7bffEo1GE8cbhkFOTg6ffvopU6ZMSdn6/vKXv3D99dfjcrnYsGEDI0aMaPf6Oeecw4IFCzj44IP5+uuvuz2vFHUSyb5HumPsmgIRfvPf1ZTU+8l3WbjrrBmMyt1L4fZoFPOucrQ01O7MspsodKehtVhvaWqCJUvg9dfb173LzhaWu1mz4MADwZTmuMHBRG4ujBolP9MUkBGiDuD999/n/PPPTxQkTj5FZxeDjq8bhkFubi6vvfYaxx13XErXdthhh7F8+XKuvfZaHn744d1e/+yzzzjmmGMAIVAndzMeQ4o6iWTfxDAMSuv9NAfSUwuzwR/m1v+upqwhgN2scf3cSRw1fu8lnNTmJizlOwfcaueyagzLsqGmOUu3x2zbJqx3b74pXLRxsrKE5e644+Coo0QRZEnfMJmE1S4nJ90r2adJa+/XZGbNmsXXX3/NEUcckRBsiqK0E3QdhVz8NcMwOPbYY/n2229TLuhaWloS1reTTz6502OOPPJIPK3m46VLl6b0/BKJJPNQFIXRuQ5ctvRYFnIcFu49Z3+mF2cRiMS45831PPdFCfpevm/rWR6Ck6YSG+A4MW8oRnlTmmrZ9YXx4+HnPxfC7q9/hdNPF67C5mZ46y245RY48UTRzeLFF0UnBUnviEaFiN66tc39LRkQ+i0XeezYsXz22We8/fbbnHLKKdhsNgzDSAyg3WO73c6ZZ57Ju+++y4cffsioUaNSvqb169cnzj19+vROj1FVNWGdW7duXY/PUVNTw8EHH4zT6cTpdDJp0iSuvfZaVq9e3a33X3/99RQUFGCxWBg+fDinnXYaL7zwArFuNroOhUI0Nze3GxKJZM8oisKYXEfa6thlOyz87uwZnHGAqPf50lc7+d0b6/CF9mKF0zQiI0cTHjceYwBjxIJhnfIGP5HYPliI1mQSlrk774R33oHHHoPLL4exY0VNvC+/hAcegLPPhosvFsfIuni9o7FRuL2TLaOSfqVf3K+dEY1GWb58OVu3bqW+vp6Wlhbcbje5ublMnDiRgw8+GFM/++AXLlzIvHnzAGHWdHeRhh2Pqzv33HN59dVXuzV3x5ImOTk5+Hy+RKszTdP4wx/+0GVJkuSSJg6HA03TaGlpSbx+/PHHs2DBArKzs/e4jjvuuIM777xzt+el+1Ui2Tsx3WB7rY9AOH314Zaur+IfH2whEjMYkW3n1tOmMnpvcXYAsRjmygq01qz+gcCkQXG2HWsG9bLtE6WlonTKRx/BypVC5IGw4N18s3Qn9gW3W8TaZUhrvH2BjImpy1ReeOEFLr30UgAikUiXIvLSSy/lhRde4KSTTmLx4sXdmnvFihW88MILXHTRRUyfPh2bzUYsFuPTTz/llltu4bPPPgPg+eef55JLLtnt/ddddx3HHXccxx9/fKIlWmlpKX//+9/585//jK7rzJ07l3feeWeP6wiFQon6fiD+KEaNGiVFnUTSTaIxne21PoKR9FmhNle1cM9bG6j1hnoUZweg+rxodbVozU0wAC5STYWibBt28yALjG9qEm7YJ54Q4i43F269FVIY8zwkKSgQBaJlIsVeGZSirq6ujueff56f//znfZ6rP0XdngiHw8yePZtPP/2UkSNHUlJSstd6dcn885//5Cc/+QkAixcv7lGpFZkoIZH0nGhMZ1utj1AahV2jP8wf397AmgoRQnHRYaO45PDRqN3NPNV1VG8LWlOjEHj96CoVRYptuNMUl9ivrF8Pt98uYsQATj0Vfv1rkWAh6R2aJrp/FBbKzh97IGMSJfpKLBZj4cKFnHPOOYwYMYLrr78+JfMmF+8NBAJdHuf3+wFwpSgLymKxcPfddwNQVlbGt99+26P3/+hHP2Ls2LEAvP766ylZk0Qi6RqTpjIu35m25AkQcXZ3nz2DM5Pi7O5etA7v3uLs4qgqepaHyKgxBKftT3jceJFU0Q+uUlGkOEhTYBAGxU+dCs89B1dcIVpivfUWXHQRfPppule27xKLQVkZrFsnLKKSlJFRou7bb7/luuuuo7i4mHPPPZf//e9/iZi0VFBcXJzYr6io6PK4+GtFRUUpO/cRRxyR2N8W/8bXTRRF4bDDDuvVeyUSSe8wtwq74mxb2owJJk3l2uP24/o5E7FoKstLGrjupW/ZXNWy9zcnoyjo7iwiI0cTnDqD8Pj90PuhtVN1c4g678B2vRgQLBb42c9EUsXo0VBTA7/4Bfzud7K5fV8IBmHLFti8WexL+kzaRV11dTV//vOfOfDAAzn00EP5+9//Tk1NDf3hFZ4yZUqidMra5EKUSei6zsaNGwF61bdWIpEMLvJcViYOc2G3pC8ZYPaUYfzxvAModFupag5x46urWLSqonfXSUVBd7kJj9uP8OixKc+arfdFqGoO9ss1PO0ccAC88ALMny8eL1ggMmTffhtaPTySXtDcLKx2O3e2JadIekVaRF0kEuHVV1/lzDPPZOTIkfz6179m9erVu9W0iz8+8MADU3Jet9vNoYceCsCSJUs6PeaLL75IdMM48cQTU3Le+Lxxxo0b16P3GobBV1991av3SiSSvmM1aexX4GSYx5o2q92EQhd/vWgmR47PJaobPPzRNv64eOPey57sAT07m9DkqUQLCiGFP1dzIMqupiD6vlbLrjvYbHDDDfDww6LfaWUl3HYbzJ0rYu3efht8vnSvct/DMKC6GtasESVQBuOXggFgQBMlvvrqK55++mlefPFFGlpb3CQLueTHkyZN4uKLL+biiy/ulzZhbrebjRs37uZiPe+883jttdc45JBDWL58ebfnjbc364xIJMLs2bP55JNPKCoqoqysrF2ixJ7eC/DQQw/xox/9CIA333yTU089tdvrkokSEklqCUZi7Kz3py071jAM/reygic/20FMNyjy2Ljx5ClMKOxbDLASDGLeVY7a0kPX7h6wmVWKPDZMWtqdQv2D3w/PPAOLFwsrUxyrFY4+WpRCOfZYSIrnlnQTh0OUQBnCHT4yMvt1165dPPvsszz99NNs2LAB6LxtmGEYjBo1iosuuoj58+czc+bMfllPIBBg6tSplJSUcPDBB/Pss88ybdo0WlpauPvuu7n//vuBzrNM42u9/fbbueOOO9q9Nn36dH74wx9yyimnMGHCBBRFIRaL8fnnn3Prrbfy8ccfA/DUU09xxRVXtHvvz372M1RV5YILLuCQQw7B3lrHZ+fOnfzjH//ggQceIBaLMWvWLN57770e/bxS1EkkqccwDKqaQ9S0pC9+bGNlC39cvIGalhAmVeGaY8dz6ozhfe7LqjY2Yt5VjpKiTgAWk0KRx4ZlsNSy6wzDgE2b4N13YelSUe8ujsUiBN6ZZ4pWZDLbs2fk5gqLqMWS7pUMOBkj6kKhEAsWLOCpp57i3XffRdf13SxyceJWqo8++ijRc7W/WblyJSeeeCJ1rUU6s7Ky8Hq96LqOoijcc8893Hzzzbu9b0+iLvnnslqtuN1umpubE4keJpOJ3/3ud9x00027zXvllVfy9NNPA6KjhcfjIRaLtesGcfzxx/Paa6+R28OWQFLUSST9hy8UZUedDz1NlU9aghH+unQzX2yvB+DYifn8dNYEHJY+Zu3qOqaqSky11ZCCO4RJg5HZDsymQWqxS8YwROD/u++KkSzwDj9cuGhlGE3PUFUYPhyGDRP7Q4S0i7rPP/+cp59+mv/85z+JuLSu3Kvjxo1j+/btide62wYrVVRWVnLvvfeyaNEiysvLycrK4vDDD+f666/vMpZuT6LukUce4dNPP+Xrr7+murqahoYG7HY7Y8eO5fjjj+dHP/pRl4kXy5Yt49VXX+Wzzz6jtLSUuro6dF2nsLCQQw89lPnz53Peeef1qLZdHCnqJJL+xReKsr3Wl7YQIMMwWLiigqc+F+7YYo+Nm0+dyrj8FLj8IhFMdbWY6msh2rdrtFlTGJFjxzxYXbGdYRgiu/PNN+GllyAcFjXaLrkEvv996ZbtKVYrjBwJe+msNFhIi6grKyvjmWee4ZlnnmHz5s3AnuPkLr30Ui677DK2bdvG3LlzE8cNtKgbKkhRJ5H0P43+MDvru659ORBs2NXMHxdvpNYbIstm4oELDqTIk6KWTIaB1tiAqbYGZQ81PveG1aRQnG0fvDF2e6KsDP70J2gNxSE/H667Dk4+Wbpke4rdLix3OTmD+rMbMFEXDAZ59dVXeeqpp3j//fcxDKNLIVdYWMjFF1/MZZddlsg8BVi6dKkUdQOAFHUSycBQ3RKkqim9NdpaghF+u3AtW2q8jMi2c//5B+C2pbZkierzotXWojU19ur9NrNKcbYdTR28N+M98skn8MADQuQBHHww3HgjTJiQ3nXti1itwiWblzco3bL9Luo+/vhjnn76aV555ZVEs/lkMRffdzgczJs3j8suu4y5c+eiabsHyEpRNzBIUSeRDBxlDX4afOntqlDvC3PDyyup9YaYXpzF3WfP6BeXpxIOo9XVYmqo67Fr1m5RKfbYUYeqsAuFRJeKJ54Q+5oGF1wAP/iBaHwv6Rlms2g5VlAgPstBQr+LOlVV24m35H1N05gzZw6XXXYZ8+bNa9eSqzOkqBsYpKiTSAYOwzDYUefHG+x97bhUsKPWx42vriIQiTFrcgHXz5nU56zYLjEMlIAfzetF9bag+rzdSq5wWjWKPLb+W9e+QGUl/PnPEK9qkJsLP/6xyJQdROJkwNA0IewKC4XQ28cZMFEHbRa6Qw45hMsuu4z58+dTWFjY7cVKUTcwSFEnkQwsMd1ge62XQDhNKbGtfFPawJ2vr0U34JLDRzP/8NEDc2JdR/UJgad5vXuMwXNZTQz3WIe2sANYtgzuvx9KSsTjyZPhV7+CfirtNehRFCguFq7Zffhvqzf3717Z5ONlSI499ljuu+8+fvGLX/RI0EkkEslgRVMVxuQ5MZvSezM5eHQOPz5BxGm98GUp722oHpgTqyq6O4to0QhCEycTnDaDyIiRnXas8IaiVKex1l/GcOSR8OKLcP31otjuxo1wzTVwyy2wa1e6V7fvYRhQXi4+x9DQ+vvqlaiLf6v65JNPmDNnDqNHj+bmm29m9erVKV2cRCKR7IuYNZWxec60x26fPH045x08EoC/v7eZ1WWNA78Ik4lYXj7hcftBJ7F9zYEoNS2ymTtmM1x6Kfz3v3DuuSLwf8kSOP98eOgh6EPW8ZDF5xM9ZWtq0r2SAaNHl5zDDz+8XaYrCKtdWVkZ999/PwcddBAHHngg999/P+Xl5SlfrEQikewr2Mwao3Mdaff+XH7UGL4zIZ+obvD7t9azsyE9jed1l5vQ+Ilg3r0wcqM/Sp13aFlUuiQnB269VSRSHHKIsDQ99hicdx689ZbsidpTdF0UgN68WdQJHOT0SNQtW7aM9evXc9NNN1FcXNxpGZPVq1dz8803M2bMGGbPns2TTz6ZyJSVSCSSoYTbZmZEdopqxfUSVVG4bs5Epgx34wvFuOv1dTQF0pOha9jthMZPxOik5VO9L0JZg59INL2xiBnDpEnCQnfffSI+rLoa/t//g6uugrffhqC0bvaI5mZhtauvT/dK+pVeFx82DIMlS5bw1FNPsXDhQgKtpuGO2bEg2madccYZXHrppZx++umYTCaZKDFAyEQJiST97GoKUNuSXitBUyDCr15eSWVzkCnD3fxu3gys6erHGo1i3b610yQKRYF8l4Vsx9Dr9dkloRC88IIogRL/zJxOmD0bTj1VWPRktmz3ycmB0aPB1Md2ev1M2tqENTc389JLL/H000/z2WefiYk7ZMnGH+fk5HDBBRcwduxYbrnllsRrUtT1D1LUSSTpxzAMttf68IXSe50ra/Dz61dW4Q1FOWp8Hr8+eXL62nbFYlhKd6B24cmxWVSGua1Y0iU8M5HaWnjlFeGGTQ5xKiyEU06B006TRYy7i9kssmNtNjEslozLlE1771eALVu28NRTT/Hcc89R2trIuCuBF39Oirr+Q4o6iSQziMR0tlR7icbSGxO1uryJ3y5cQ1Q3OHCkh1tOnYrTmiaLhWFg3lmC1tjY6cuKAnlOC9kOsyx7koxhwMqVoqfskiWQLIwnTRLi7vjjRZ9U+bl1D1UVHSriIi95pCnjKSNEXRzDMHj//fd58skn+e9//4vfL4JzkwVe3FWrKAonnHACF198Meeffz45OTn9saQhiRR1Eknm4A9H2VbjS3us+zclDfzh7Q0EIjHG5jm448zp5LmsaVuPuaIMrba2y9dtZpXCLGv63MWZTDgsWo+9+abYRpMKXxcVwRFHiHHYYZCdnbZl7rOYzcJVm4bPLqNEXTJer5f//Oc/PP3003zyyScJIdcZZrOZOXPmMH/+/G51ppDsGSnqJJLMos4boqIx/UHuW6q93LVoLQ3+CPkuC3ecOZ0xeem73pqqqzBVdl2TTVEg12km15k+8ZnxNDUJy92SJcKSlyzwFEUUNY6LvAMPFJYpSffIy4NRowY0djFjRV0y27dv56mnnuLZZ59lx44dYhFduGdtNhunn3468+fP55xzzhnIZQ4apKiTSDKPnfV+Gv3p7RELUNUc5I7X11LWEMBp0fjNaVPZf2R22tajtjRjKSuFSNdt1vKcZnLTaFXcZwgE4Jtv4Isv4MsvYcuW9q9brULYHXIIHHooTJ+e8YkDacdshjFjwOMZkNPtE6IumQ8//JAnn3ySV199FZ/PJxbUQeCBaE8Wjaa3l+K+ihR1EknmoesGW2u8BCPpL9/REoxw9xvrWb+rGZOqcP2cSRw3qSB9C4pEsJSVdplAAVCcbUtfHOC+Sm2tEHdffCFGR3e33Q4HHdQm8qZMkSKvK/LzRbxiP1vt9jlRF8fn8/HKK6/w9NNP8+GHH7Zzz8pEir4hRZ1EkpmEojG2VHvR06/rCEd1/rRkI59trQPge0eP5ZyZI9KanGCqrsJUtQs6uUNpKozKcWA2pbllx76KYcD27bB8uRhffy1ct8k4nULkzZ0Lp58uEy46YrEIq10/3lf3WVGXTGlpacI9u3XrVkCWPOkLUtRJJJlLczBCSW16Ojx0RDcMHv9kO/9bWQHAmQcUcfV3xqOp6buZK34flpIdKJHdXdU2s8KIbAdqGtc3aNB12Lq1TeR98037jNpZs+C3vwW3O31rzFQKCoTVrh8yZAeFqEvmk08+4cknn+SVV16hqeO3CEm3kKJOIslsqpqDVDdnTousBSvKefyT7QAcNT6PG06alN6s01gMy85S1Obd7wFum4nhHlsaFjXIicVEW60PP4QnnxQJFyNGwL33wrRp6V5d5lFQIDJkU8ygE3VxAoEAdnt6W+3sq0hRJ5FkNoZhsKPOjzeYOXHDH2+u4c9LNhHVDSYNc3Hb6dPISXOHB62uFnNF2W7u2AK3hexslwhit1jENhKBhob0LHSwsXYt3HILVFSIGLvrroOLLpLu2GTy84UrNsUMWlEn6T1S1EkkmU80prOlxkskmjmX47UVTfz+jfW0hKIUuq389oxpaS15AqAEAqg+L4bJjGExY5gtKGYT4wtdOCwdgvqDQdi1a9D3+hwQWlrgrrvg/ffFY+mObU8GiToZZSqRSCRpxqSpjM93YTNnziV5erGHBy44kGKPjeqWEDe+uopvS9Nr/TLsdmL5BejZ2RgOJ5jNGCiU1vuJxjpknNhsMG6cKNWRm5ueBQ8W3G647z741a+Ete799+Gyy2DdunSvTNKBzLmCSCQSyRDGYlIZX+DCZcucMhLF2XbuP/9Aphdn4Q/HuOP1tby9pjLdy9qNSNSgtN5Pp46nZHGXlzfwixssKApcfDE8/jgUF4ves1ddBS++SNpbpEgSSFEnkUgkGYKmKozNc5DrSm/8WjJZdjN3nz2DWZML0A34xwdbePyT7egZdiP3hWJUNu+hU4fNBmPHwowZQtzJmLDeMX06PP+8cMFGo/DAA3DjjdBF/17JwCJFnUQikWQQiqIwItueUVmdZk3l+jmTuOwIkeG3YEU59761nmAks0pN1baEadpbpw6rtU3cDR8uC+z2hs7csRddJLJlJWlFijqJRCLJQArcVkbnOjLGoKQoChcdNppfnTQZs6awbFs9t7y2mnpfON1La8fOBj/+cDcyiS0WUabjgAOEe1b2Ge8ZcXfsE0+Iz6+uDm64Ae64o32NO8mAIkWdRCKRZCgeh5nxBc60FgDuyPGTCvjdvP3JspnYUuPlhpdXUlLnS/eyEhgGbK32UdkURNe74SJWFJFIMWUKTJ0qMhn7oZDsoGXaNHjuOfjud8VnuWiREHvLlqV7ZUMSWdJkkCNLmkgk+z6haIySOj+hDOgVG2dXU4A7X19HeWMAp1XjttOmMWPEwDQ67y5Ws8rIHPvu5U72RiwmLE/V1RDKnMLQGc+KFXDnnbBzp3h87rnwi18MfiuoLGkikUgkku5iNWnsV+DCaU1jZ4cOFHnsPHD+gUwrysIXivHb/63h0y21e3/jABKK6AmrXY/sF5oGhYXCCiUzZrvPQQfBCy/AhReKx6+9BpdcInrLSgYEKeokEolkH0BkxjrJsmdOYL/LZuKus6dz5PhcIjGDP769gTdW70r3snajpiXE5mpv92LtklFVkVQxerTMlu0udrvIhv3Xv6CoSJQ++cEP4E9/Aq833asb9EhRJ5FIJPsIqqowOtdBtsOc7qUksJo0bj5lKqfOGI4BPPThVp75fEfPLGMDQCiis62mF1Y7EL09J00SLcgk3eOww+Df/4ZzzhGP//1vmDdPlEORLu1+Q8bUDXJkTJ1EMjgpbwxQ782czFPDMPjP8p0890UpAHOmFvKTEyZg0jLPdmAzq4zKdWAz99CdHYnA1q3gy5zEkH2Czz4TlrqSEvF42DC45ho444zBUVImg2LqpKgb5EhRJ5EMXiqbgtS0ZJbVY/HaSv75wRZ0Aw4dk8NNp0zpuXgaAFQVxuY5cVp7KCoMQyQC1NT0z8IGK9GoyIx99FGoqhLPjRkDP/oRzJ69b2ccS1EnGSikqJNIBjfVLUGqmjJL2H25vZ4/Lt5AOKozaZiL354xHY8981yXigKj8xxk2XqxttpaKC2VLbJ6SigEr7wi6ts1NYnnpkyBn/wEjjxy34xdzCBRtw9LY4lEIpEUum0UZWdO9wmAw8fl8vt5M3BbTWyq8nLjKyup2lMLrzRhGFBa5997F4rOyM+HyZNlnF1PsVrh0kth4ULhgnU4YMMG+NnPRELFqlXpXuE+jbTUDXKkpU4iGRo0+MKUNQTSvYx27Gzwc8f/1lLdEmJYlpU/nnsAeS5rupfVKSNy7OQ6e9FzNxqFhgZhdWpulpa7ntLQAE8+CS+/LGIWAY47Dn78Y5gwIb1r6y4ZZKnrlqi76667+ry4vvLb3/423UvYJ5GiTiIZOjT5I+xs8GeUrqjzhrjlv6vZ1RRkZI6de8/Zn2xHL8TTADDcY6PA3QfRqetC2DU1iRHphQVwqFJZCY88IuLudF24YU85RVjvRo5M9+r2zL4m6lRVRUmznzsWy6zG0fsKUtRJJEOL5mCE0rrMEnbVzUFuem01td4Q4/Od/H7e/rhsmZn1WJhlZVhWitzZfr8Qd42NYl+yd3bsEDXuli4Vj00mURbl6quFeMpEhoKo62razubZ07GGYaAoihR1vUSKOolk6FHnDVHRmFkxbBWNAW56bRWN/giTh7m56+zpPW/fNUDkuSwUZ9tTO2lLi8iaDWSWizxjWb8e/vGPth6yNhvMnw+XXw5ud3rX1pEMEnXdTpQwDKNHA4QoSx5dzbOnYyUSiUTSM/JcVnJdmeXiLM62c/dZInliY1ULv39jPaFoZn5Zr/OG2VnvT+09yO2GqVNh1CjRhkyyZ6ZOhQcfhIcegv33h2BQxN6ddZYoYBztYXeQIUK3LHUffvhhjyatqqri5ptvpqSkJPFPcdBBB3HCCSew//77k5eXh9PpxOfzUVdXx6pVq/jwww9ZsWKFWJSiMHbsWP7whz9QWFgIwPHHH9/DH00C0lInkQxVDMNge60PXyizhNOmqhZuW7CGQCTGoWNyuPW0qZgzsEAxgMduZnSeI/UTR6OifVZtZvXKzVgMAz78EP75T9i2TTw3dSrcdpvIQE43GWSpS3n265o1azj11FOpqKjAMAzmzp3L/fffzwEHHLDX965cuZJf//rXvPvuuyiKwsiRI3n77beZOnVqKpc4pJCiTiIZusR0gy3VXsJRPd1Lacea8iZuf30t4ajOMRPy+fVJk9HUzKxPVuC2MtzTTyVj/H5R6052qOgesRj873/wt78Jd7amwWWXidIotjSW9ckgUZfSr0d1dXWccsoplJeXA3DbbbexePHibgk6gAMPPJB33nmH3/zmNxiGwc6dOzn55JOpr69P5TIBqKys5Be/+AX77bcfNpuNYcOGceaZZ7I0HpzZQz744IPdXMidjdq9fDN7+eWXmT17Nnl5eTgcDqZOncptt91GS0tLr9YlkUiGLpqqMCbPkXHF+meM8PCbU6diUhU+3VLL39/bjJ6h4TY1LSHqff3Ujs3hEIV3x46V9e66g6aJpImXX4Y5c4TIe/ppEWu3fHm6V5cRpNRS94Mf/IBHH30URVE499xzefnll3s91/nnn89rr72Goihce+21/Otf/0rVMlm1ahWzZ8+mrq4OgKysLLxeL7quoygK99xzDzfffHOP5vzggw+YNWsWqqpSUFDQ5XHr1q0jNze309euvfZaHn30UQBMJhM2mw2v1wvA+PHj+fjjjykuLu7RuqSlTiKRtAQjlGRYRizA51tr+cPbG9ANOGP/Iq49bnzaKy10hqLAmDwH7t50nugusRhUVEB1df+dY7DxwQdw331tn9lZZ8F118FA3+sGo6XO6/XyzDPPJB73tbZd/P2GYfD000/jS5F5OhAIcNZZZ1FXV8fMmTNZs2YNTU1NNDQ0cMMNN2AYBrfeeivvvPNOr+YfNWoUlZWVXY6uBN2//vUvHn30UVRV5f7778fr9dLS0sKnn37KmDFj2LZtGxdeeGFffnSJRDJEcdvMFPWXC7EPHLVfPtfNmYQCLFq9iyc+3Z6RFjvDgNJ6P8FIP8YnappIopg0CSyZleSSsZxwAvznP3DBBeLx//4n9t99d8gWgU6ZqPv4448JhUIoikJxcXGf4+CmTZvGyNaCg6FQiI8++igVy+Thhx+mpKQEl8vF66+/zvTp0wFhrXvggQeYN28ehmFwyy23pOR83SEUCnHHHXcA8Itf/IJf/epXWK2iAObRRx/Nf//7XxRF4dNPP+X1118fsHVJJJLBQyZmxALMmlzIj07YD4AFKyq4f/HGjIsBBFEPd0edj0isn9fmdsP06bAHj48kCZcLbroJHnsMxo2Dujq4+Wa44YYhmYiSMlG3efPmxH5PXYRdUVRU1On8feH5558H4JJLLmHE/2fvvMOiuLo4/Jvtld6VqgJWROy99xY1thhLEjXNGDUaTTS2RJOoKcaYaNSo0dg1auwNG/ZewEpVEKQvdcv9/tjsfCzswgILu8B9n2cemJ1bzvQz595zTq1aRbbPnDkTAHDz5k08evTILH2WxMmTJ5GYmAiGYTBjxowi24ODg9G9e3cA/5efQqFQSouHrQhSofWF0+jTyB3TutcDj8PgwtPXmPvPPaTnWF82BqWKIDo5GxpNBVuBOBzAy4ta7UpD06baUCcTJ2oDFp87B4wYAZRx1K2qYrbIjwWHR9PS0szSZsF2zDH8mpmZiRs3bgAAevXqZbBM69at2THsU6dOIaAS3KXPnDkDAGjUqJFBRRPQynvixAmcPn26wuUpCCEESqUSGo31fTlTKBUNh8MBn8+3ynleZYFhGHg5SPAsKcvqrGFdA13hJBNiyeFwhCdkYubuO1gwoKH5gwCXk5x8NeJScyom1Elh5HKgQQNt+JOkpIrvr6ojEGjTinXrBsyfDzx6BHzxBXD6tNZ6Z2dnaQkrHLMpdU7/pe8ghCAyMhKvX79mfysLr1+/RmRkJPswLU9bOsLDw9m4ebph18JwOBwEBATg6tWrePjwYan7SEpKQrNmzVgrX61atdC5c2dMmTIFjRs3NlhH148xmQDtcLSu/fIeW1PIz89HYmIisrOzaTYPSo2Gy+VCIpHAxcUFgmpgNeFxOfB2lOB5UhbUFW1xKiVNatth2bAgLDj4APHpufhs1x182a8+GnrYWlo0PdJzlEhIz624UCcF4XK1Vjt7e20KrfwK8sStTtStq/WK3bABWL9eO8fu5k2tgte5s6Wlq1DMptT5+fkBAJvS6/fff8fcuXPL3N7q1auh+i9iNMMwbPvlIT4+nv2/uCFi3baC5U0lOzsbt27dgr29PbKysvDkyRM8efIEGzZswLfffovPPvvMqFymyKQrb0ypy8vLQ15eHruekZFRpn2IjY0Fl8uFvb09xGIxuFxutbFWUCimQAiBWq1GTk4O0tPTERUVhdq1a0MiqQQLTQUj4nPh5yxF5OssqNTWpdh5Okiw/M0gLP73IZ4kKjD3n/uY1t0fHf2ta45ZUmYeBDwOHKSVpOjrrHYJCdp8sjTdWPHweMCkSUDHjsBXX2mDFn/2GdC3r/ZvNY0GYTalrlOnTnBwcEBqaioIIfjmm2/QoUOHMmWCCA0NxZIlS9jcr/b29mbJKFFwCFcsNm7S1z20deFETMHOzg4zZ87EiBEj0LBhQ4hEIqjValy8eBFz5sxBWFgYZs6cCQ8PD4wePdqgXKbIVJJcS5cuxcKFC02W2xCvX78Gn8+Ht7c3uDSdDaWGI5PJ4ODggOjoaLx+/RpeXl6WFsksFFTslCrrUuzsJQIseaMxfjjxGJeeJ2PZ8Ud4lZGLYSG1rerj8mVaDgQ8DmTCSsphy+UCtWppF6USyMj4/0LTZhkmMBDYsgVYswb46y/g8GHg2jVg3jygbVtLS2d2zOYowePxMHHiRBBCwDAM8vLy0KdPH/z0008mD9+p1Wr8+OOP6Nu3L5RKJdvWpEmTwONZZ+JnHU2bNsX333+PkJAQiP6LbM3lctGxY0ecOXMG7dq1AwB8/vnnFTo/bc6cOUhPT2eX2NjYUtVXqVTIysqCg4MDVegolP/gcrlwcHBAVlYWO4JQHRDyuPBzkkHAs7LoxNAqnZ/3DsSgIO0oxebL0fjlzFOoKtr7tBQQAkS9zkJatgWGRPl8wNFR6/EZFKRNm+XhofUGtSLF1yoQCIApU7Qesl5e2vmJn3yinXd36ZI2r2w1wax38vz581G3bl0A2iHT3NxczJgxA76+vpg1axYOHz6MmJgYZGVlgRCCrKwsREdH49ChQ5g1axZ8fX3x2WefIbfAAa5Tpw6++uors8gnlUrZ/3OKMV1nZ2cD0H6hmwOBQIDFixcDAOLi4nDr1i2DcpkiU0lyCYVC2NjY6C2lQffC0oVUoVAoWnT3RHVS6gBAwOPAz1kKEd/6FDsuh8F7Hfzwfkc/cBjgxMNX+PpwuFXNBSQEiE3JQWxKJXjFFodEAri7a3OhBgVpg+EWeOdRADRpAvz9tzYDBQAcOqRV9rp0Ad5/H/jzT+DhQ20g6CqKWc1fIpEIx44dQ9euXREdHc0On8bFxWHFihVYsWJFsfV1Tgy6et7e3jh69Chr+SovBeelvXz50qhn68uXLwHoh1QpL61atWL/f/78OUJCQvTkun37NttvcTKZWy5jWNMQB4ViDVTne4LP5cDXSYqo5Czk5FuPJUxHvyYecJaL8P2xCNyITsWmS1F4p52vpcXSIy1biex8NbwcJBALLDzKweVqsxw4OWmtUMnJ2kVpfWFiKh2RSBvDrls3bbDiK1eAV6+0acauXwd+/RWwtQWaNwdatQJat9ZaQKsIZv808/X1RVhYGHr16sUOn+oehoQQowsAvXLdu3fHhQsXzOIgoSMwMJDt48GDBwbLaDQa1nNV53Fa0ej6MSYT8H8PWWdn5wr3fKVQKDUPHpcDXyeZ5RUSI7T0dcC07v4AgH23XuDiU+sLLJuv0uBZkgJJmXklF64sRCLtHLzGjbVeofb2dHgW0Ma1++or4N9/gT17gFmztJ6xUqnWEeXUKWDJEm3qsUWLtPMWqwAVYm93d3fHkSNHsH37drRq1UpPcQOgl+Beh65Mq1atsG3bNhw/ftxozLayIpfL0bx5cwDAiRMnDJa5cuUK0tPTAQDdunUzW99Xrlxh//f11f/C7NKlCwCtUmfM41aXtsycMlEoFEpBuBwGfk5SqwxQDADt6jrhjWDte+HnU08Qm5pdQo3KhxAgIT1X64BiRfP/wDBaC5Sfn3YY0stLu17T504zjHaoevhwYPlyrTK3YYN2ODY4WLu9CqUfYwipeAkjIiIQGhqKq1evIjIyEmlpaVAoFJBKpbCzs4Ofnx9atGiBzp07lzu9WEn89NNPmDZtGuRyOR49elRkKHPo0KHYu3cvQkJCcP36dZPb1VklDaFUKtG1a1dcuHAB7u7uiIuLA4fzf306Ly8PXl5eSExMxIwZM7B8+XK9+nfu3EGzZs2g0Whw8OBB9O/f32S5SpsQODc3F5GRkfD19TXbsDeFUh2oSfeGRkMQk5KNzFzrmz+o1hDM/ece7r/MgKe9GCvebGq11kUuh4GngxhyEd/SohiHECA7G8jM1C4KhTYnGkXL7dvAN98AkZHa9Y4dtWnJXF3/X8bJSasYmpnSvr+BSlLqrImcnBzUr18f0dHRaNasGf766y80aNAAmZmZWLx4MZYtWwYAOHbsGHr27KlXV6e0zZ8/n83VqqNhw4Z4//330bt3b9StW5eN13fp0iV88cUXOH/+PABg48aNGDduXBG5fvvtN3z44YfgcDj4/vvv8fHHH0MoFOLSpUt46623EBkZiXbt2uHChQul2l+q1FEo5qGm3RuEEMSl5iAt2/rmYaVm5+PTHbeRkpWP9nWdMKtXgFXPeXSSC+BmI7JqGVkKK3lVZNixQsnP1zpR/PmnNnSMVAp8/DEwdKg2pRtV6izLnTt30K1bNyQnJwMAbGxsoFAooNFowDAMlixZgtmzZxepV5xSV/BmFQqFkMvlyMjIQP5/0b95PB6+/vprfP7550blmjRpEv744w8AAJ/Ph1AoZGPS+fn54fz586XOq0uVOoo50F3fkZGR8PHxsawwFqKm3huvFXlISM+1ulGn8PgMzNl3D2oNwbvtfTG4qXmn65gbsYADTwcJhDzrtCoaJT1da6Wqwh6hZuPZM63V7u5d7XqTJsDcuUDLllaj1FmfD3slEBQUhPv37+OTTz6Bn58f8vLy4OjoiH79+uHEiRMGFbqSWLNmDcaOHYuGDRvCxsYGaWlpEAqFaNy4MT7++GPcuXOnWIUOANauXYsdO3agS5cukMlkUKlUCAwMxJdffonbt2+XWqGjVD7jx4/XmzOqW+RyORo2bIgPP/wQ4eHhlhaTQjEZJ5kQPk5ScDnWZWWq726Dd//zgP3zYiQevEy3sETFk5OvwZNXCqRmVbE0X7a22jAp1SBFXrmpU0cb627WLG0Imbt3gbfeAn76CcizDueYGmmpq0lQS13lMn78eGzatAl8Ph8ODg4AtMNYr1+/ZoNOCwQCbNmyBW+++aYlRS0V1FJH7418lQYxKdYV8oQQghUnHuPs4yTYS/j4aURw5aXtKgd2Ej487MRWpygXi0oFPH0KFMjMVKNJSAC++w74b2oV2rQBLl40q2ex1VrqkpKSsHfvXnzzzTeYMWMG3nvvPbzzzjuIiIiojO4plEqnbdu2SEhIQEJCAl69eoXc3FwcOXIEPj4+yM/Px4QJE5CUlGRpMSkUkxHwOPBzksFOYj2T/hmGwcdd6sLbQYLUbCW+OxphVRknjJGWrcTTRAWy863PEcUoPB7g7w/897Fa43FzA374QRv2xMkJGDfOKkLFVGjurf3792PFihW4ePGiwe1jxoxBYGBgkd9nz56Nq1evAgC8vb3x559/VqSYFEqFw+fz0bt3b2zduhXt2rVDVlYW9uzZg/fff9/SolEoJsPhMPB0kEDE186zswZEfC7m9KmP6btu42F8Bv4Mi8LEDuaLb1pR5Ks0eJ6UBRcbIZxlwqrhRMHhaNOSicXAixeWlsbyMAzQsyfQrx/QqJGlpQFQQZa65ORkDBgwAEOGDMHFixf14tQVjllniNatWyM0NBShoaHYvHlzsUF5KZSqRJs2bdg0b7qA0joyMzOxceNGDB8+HI0aNYKdnR3EYjHq1q2LSZMm4cmTJ0bb1c3di4qKQkxMDCZOnIjatWtDKBSy6fcyivFi02g0+OWXXxAUFASxWAxnZ2cMGDAAly5dMmm/bt26hTFjxsDT0xNCoRBOTk7o1asX9uzZY7SOj48PGIZBaGgo4uPj8f7778PT0xNisRj169fHjz/+qJcnedeuXejQoQPs7OxgY2ODfv364f79+ybJRzEvznIhfJwkVjN8WMtezAYmPnDnJc4/qRpWcEKAV+l51hfTriTc3LTx7jg1clp+UWxtreZYmF2K169fo02bNjh8+HAR5a24WG4FGTRoEDw9Pdn1v//+29xiUigWQ3dfqAt5k23atAkTJkzArl27EBERAS6XC41Gg2fPnuGPP/5AcHAwTp48WWzbd+7cQXBwMNatW4eMjAxoNBpERUVhxYoV6NatG5QG0gSpVCoMGTIEn3zyCe7evQuVSgWVSoV///0XHTt2xN69e4vtc+3atWjevDm2bt2KuLg4SCQSpKWl4fjx4xg2bBjefvvtIvtakMjISDRr1gxr1qxBRkYGlEolIiIiMH36dEydOhWA1no/fPhwXLp0CRqNBpmZmTh8+DA6dOhQrLJLqTjkIj7quFhPztjWfo4Y1qw2AOCnU0/wz+0XVWIoFgCy8tR48kqBnPwq5GFqb691oOBbz3A8xcxKHSEEAwcOxNOnT9nfZDIZpk2bhiNHjuD+/fslWukArdVh6NCh7PqxY8fMKSaFYjHCwsKQ9d9E48Ip8JycnPDll1/i6tWryM7ORnJyMnJzcxEeHo633noLWVlZGD16NFvfEOPHj0fTpk1x7949ZGRkQKFQYP369RAKhbh+/TobMqcg3333Hfbv3w8Oh4Nly5YhPT0dqampeP78Obp374533nmn2P354IMPoNFoMGzYMMTGxiI1NRVpaWn4+uuvwTAMtmzZgqVLlxptY9q0afD19cWdO3eQnp6OjIwMLF68GADw66+/YsmSJfjhhx/w008/sdvv3buHgIAApKWl4csvvyz2mFMqDiGPCz9nmdVkoBjT2hvNve2Rr9Jg/YVIfLrjNu6/sG6vWB1qDUFcarZJ70irQSIB6tfXxm2jWAVm9X7VWRp01rg2bdpgz549cC0QeZnD4bDbT5w4ga5duxps69SpU+jRowdbJzU1FXK53Fyi1hgqwvuVEGLQ4lMV4fP5Zp3LovN+7dSpE0JDQ9nflUolTp8+jffffx9RUVHg8/l4/vw5ateubVK7hBD07NkTJ0+eNBjAWrcPDRs2xI0bNyAUCvW2T5kyBatWrUKXLl1w+vRp9vesrCy4u7sjMzPTYPzFvLw8NGvWjB0qLuz92q1bN5w+fRrt2rXD2bNnwS2UcuiLL77A0qVLIZPJ8OLFC71r0MfHB9HR0bC3t8fz589hZ2enV1fXNgAsXLgQX331ld728+fPo2PHjhAKhcjIyICggkMu1HTv1+LQaLSBitNzLP9c0BCCk+GvsDEsis2I0TnAGRPa+lYJz1hXGyFcbKrY9UUIEB+v9QitSkqpubCi4MNmdZT4/vvvAWhfQHXr1sXhw4dNFqQwISEh7P+EEDx8+BCtWrUyi5yU8qFUKou1vFQl5syZUyHKQFhYGNzc3AAUDWnC4XCwZs0akxU6QKu09evXDydPnsTFixcNZiUBgOnTpxdR6ABg8ODBWLVqVZE5aMePH0dmZiaEQiGmTZtWpJ5QKMRnn31m0FqXkpKCM2fOANAex8IKHQB8/vnn+PHHH6FQKHD48GGMHDmySJn333+/iEIHAN27d8fp06chEAgwffr0ItvbtWsHkUiE3NxcPH36FA0aNChShlI5cP5Lh8VLZ5CssGwcNg7DoGcDN7Txc8Rfl6Nx9H4CQh8l4crzFLzVygv9m3hYzVxAQyRm5sFGzIeIbx3WT5NgGMDDQzu3LDLSamK21UTMNvwaFRWF8PBwdsL2kiVLyqzQAYCdnZ1eXtbHjx+bQ0wKpVJQKpV49eoVXr16hcTERFahc3BwwJUrVzBhwgSD9eLi4vD5558jJCQEdnZ24HK57D2lU7pevnxptN8WLVoY/L1WLW20/dTUVL3fb968CQBo2rQpbG1tDdbt1KmTwd9v3brFzpM1VsbW1pb9QNP1VZjGjRsb/N3FxQWA1qKncy4pCIfDgZOTE4Ci+0WpfBiGgYedGK62RT8qLIFcxMeHnetixZtB8HeVIUepxroLkfh0xy2rHpIlBIhLzalaw7A6pFLtcOx/9yWl8jGbpU4XgoQQArFYjIEDB5a7TQcHB8THxwOgD21rgs/nY86cOZYWwyzwK2iSb8Hh17y8PERERODrr7/G7t278e677yI0NBT29vZ6dc6ePYv+/fuzqeEArVKkG+rLyclBRkZGsXPqjE1R0LWhUunHxdLFyisuW4lOISyMrq6tra1BpUuHziJpLC5fwY+3gugsf8a2FyxTXaYDVAdc5CLwORy8SMuxipG4eq5yLBsWhBMPX2HTpShEJWdjzr576BzgjPc71oFUWKGRvcpETr4ayVn5cJJZh4JcKrhc7VCkrS0QHa0NWkypNMx2NScmJgLQfq35+fmZZUir4IuiuBcZpXJhGKbC5y9VJ4RCIYKCgrBz50706dMHx44dw+TJk7Fz5062jFKpxJgxY6BQKNC9e3d89dVXaNGihd7crfXr1+O9996zui/4PDrUQimEvVQALpdBTHK2VSh2HIZBr4b/H5I99kA7JJuQnouFAxtCIrA+xS4hPRc2Ij4EPOvwLi41dnZay11UFFBMOCWKeTHb1VLQulDcV3tpyMzMZP+XUu8aShWHYRisXLkSXC4Xu3btwtmzZ9ltly5dQlxcHBwcHLB//3506NChyGT8V69emV0mZ2dnAMUP6Rrbpqubk5NTbHaMuLg4vfKUmoGNiA8/Z+vKGWsj5uOjLnXx/dAmkAl5iEjIxKJ/HyJXaX2hRAgBXqTlWFqM8sHnA/XqAZ6eVhPHrbpjtqPs6OjI/p+SkmKWNl8UiFjtRMfoKdUAf39/jBgxAgD0QnHoFB9/f39IJBKDdUuKUVcWmjVrBgC4ffu20eDEBZXPggQHB7NetzqHicKkp6fjxo0ben1Rag4SAQ9+zlLwedaj2AFAoLsNFg9qBKmAiwcvM7D40EPkqaxPsVPkqpCaZVnHE7Pg4gI0bAg4O1tFKq3qjNmUuoKefpGRkcjOzi5Xe3fv3kV6+v8ns/r6+parPQrFWvjss88AABcvXmTn3emcFJ48eYLc3KLpl44fP25UcSoPPXv2hI2NDfLy8vDzzz8X2Z6fn48VK1YYrOvg4IAuXboA0Ma6K5j9Qcd3332H3NxcyGQy9O3b17zCU6oEIj4Xfk4yq1Ps6rrIsGBgQ4j5XNyNS8eSw+HIV1lfsOL49NyqlW3CGAIB4OUFNG4MuLpSy10FYbaj2rp1a9ZLT61W49ChQ+Vqr2C+V6lUipYtW5ZXRArFKggODkb37t0BAF9//TUAbXgOiUSC5ORkjB07lnUQysnJwYYNGzB06FA9a7i5kEqlmDVrFgBtLLgffvgBOTnaIZ+oqCi88cYbiI2NNVp/8eLF4HA4uHnzJkaOHMlaHBUKBZYsWYJvv/0WgDYjRHm84SlVGwGPA18nKXhc61LsAt1sMH9AAwh5HNyMScO3R8OtToFSawji06wjz65Z4POB2rW1yp27u9axgmI2zKbUOTs7IyQkhJ3EvWTJkmJTAxXHo0eP8Mcff7BKYteuXQ3GwKJQqio6RerUqVO4fPky7Ozs2Nh/u3btgoeHB5vj9N1330XdunUxf/78CpHl888/x6BBg6BWqzFjxgzY2NjA3t4evr6+OH78ODZs2GC0btu2bbF69WpwOBzs2rULXl5ecHBwgJ2dHb788ksQQvDWW29h9uzZFSI7peog5HHh62Rdc+wAoKGHLb7q3wACLgfXolKx7Ngjq0svlp6jtIrAzmaFx9PGtmvcGKhVS7tOKTdmtX9+/PHH7P93797FlClTSt1GXFwcBg8ejOzs/6dLmTlzptlkpFCsgR49eiA4OBgA2JRYn3zyCfbu3cta7VQqFQIDA7Fw4UKEhYVVWEYVHo+HPXv2YOXKlWjSpAl4PB64XC769euHs2fPYsiQIcXWnzx5Mq5du4bRo0fD3d0dCoUCtra26NGjB3bt2oUtW7bQjzIKAO1QrK+T1OpG3prUtsOX/eqDx2Fw6Xkyfjj5GGqNFbjtFuBlWo7VyWQWuFzAzU2r3Hl6aodpKWXGrGnCCCEIDg7GvXv32KCkffv2xQ8//IB69eoBMJ4mLD8/H3/++Sfmz5+v50nXq1cvHD582Fwi1jgqIk0YhVITofeG+cjOV+F5UpZVhDspyLWoFCw5HA6VhqBzgDM+7eZvVZZFeykfte0NO1JVGwgBUlK0KccMzC+2SqprmjCGYbBr1y60a9cOycnJAIDDhw/j8OHDaNmyJZo3bw4ArMK3YcMG7N27F0+fPmUTneu2EULg5eWFv/76y5wiUigUCsXCSAQ8+DhJEfXauhS7Fj4OmNU7EN8djUDooyTwORx83LUuOFbisZmapYSdRAWZFQZMNhsMAzg6ape0NG1O2XI6XtYkzGqp03HlyhUMGjQIiYmJrIKms84V7K5gInXd77rynp6eOHToEBo1amRu8WoU1FJHoZgHem+Yn8xcJaKtJEBxQS48fY1lxyKgIUCfRm74oFMdvfeVJeFyGNRxkULIq0FTGjIytJa7ArFrrQorstRVyMyGVq1a4c6dO+jVq1eR6Pc654fCN0hBpa9Xr164fv06VegoFAqlGiMX8eHpILG60GXt6zphWnd/MACO3E/AuguRVpPJRa0hiE7Orp7z64xhYwP4+wOBgdpMFRSjVNh0VVdXVxw5cgSXLl3C8OHDYWNjA0KI0UUsFqN///44e/Ysjhw5QqPPUygUSg3AVsxHbXuxpcUoQucAF3zSVTsX/MCdl9h0KdpqFLs8pQYxKdlWI0+lIZUCdeporWLW9iVgJVT4wHyrVq2wfft2EEJw7949hIeHIzk5GWlpaZBIJHBycoKvry9atmxZYcnVKRQKhWK92EkE0BDgRap1pcXq3sAVSo0Gq0OfYc/NOAh5HIxq6WVpsQBos03Ep+fCw876FOIKx8kJEImA588BZTUL9VJOKm22JcMwaNKkCZo0aVJZXVIoFAqliuAgFUCtIUhIty6Pxz6N3JGv0mDdhUj8fTUGPC6DN0M8LS0WACBZkQ8hjwNHmdDSolQ+Mpl2OPbZM+pIUQArixZEoVAolJqKs1wIFxvrU1AGNa2FcW18AACbL0Vj/+0XxVeoROLTc5GZW0OtVQIBEBCg9ZSlADCzpe7cuXPs/0FBQWw+y7KQnp6OO3fusOsdO3Ysl2wUCoVCsX5cbURQawiSFdaVyH5YSG0o1Rr8fTUG6y5Egs/loG9jd0uLBUKAmJRs1HGWQcSvQR6xOjgcwMcHkEiAYlIa1hTMqtR17tzZYGDhsnD9+nX07NkTgHboVqVSmUVGCoVCoVg3HnZiqDUEadnWZYEa2cIT+SoNdt+Mw29nn4HPZdCjgZulxYJGA0QnZ6OOsxQ8bg0dgHNx0c6zi4wEarC+YPazb05vnILesRQKhUKpOdS2F8NGbF1BdhmGwdg23hgY5AEA+OX0U4Q+SrSwVFryVRpE10SP2ILY2Gjn2YlroPPIf5hdqbOWAI0UCoVCqbowDANPewmkQusaUmQYBu+190WfRm4gAH48+RhnrESxy85T40WadXkQVzpCoVaxc3KytCQWoUrYaamiSKFQKDUPDoeBt6MUYoH1KXbvd6qD7vVdoCHADyceY8XxR1DkWn7YLzVLicRM6/IgrnQ4HG0su3r1tM4UNQirVeqysrLY/8U12JRKoVAoNRkuh4GPowQivnW9rjgMg4+71MPw5p7gMEDo4yR8tO0mrkenWFo0vErPw8u0nJo9FAtoh2MbNKhRVjvruksKEBERwf5vR9OCUCgUSo2Fx+XAx0kKAc+6XllcDoO3W3vj+6FBqGUnRkpWPhYefIhVp58gO9+yVrtkRX7NSydmCC63RlntrOsO+Y/09HRs2LABgNbMHRgYaGGJKBQKhWJJ+FwOfJwk4HKsbzpOgJscP49syjpQHHv4ClO23cK9F+kWlSszV4XnSQrkqzQWlcMqqCFWu1K7Fi1atMikcps3b8aFCxdMbpcQguzsbERGRuLUqVNIS0tjt7Vv3760YlIoFAqlmiHkceHpIEbUa+vLICDkcTGxgx9a+zrgp1NPkJiZhy/23cPAIA+MbeMNIc8y8wJzlRo8S1LA21ECicC6vIkrHZ3Vzt4eiI4G8q0rFqI5YEgpB905HI5Rx4WCTZXHuYEQAoZhQAiBQCBAeHg4fH19y9xeTSYjIwO2trZIT0+HjY1NieVzc3MRGRkJX19fiESiSpCQQqka0HvDekjMyMWrjDxLi2GU7HwVNlyMwrEHCQCAWnZiTOvujwA3ucVkYhjA014CWwnNsQ4AUKuBV6+A5OTyK3dOTlpl0cyU9v0NWOnwq06hYxgGK1asoAodpcowfvx4MAyDzp07l7ruiRMnMGHCBNSrVw9yuRwymQx169bF+PHjcfz4cYu1RaFYGy42IshF1mt1kgh4+LhLXSwY0BAOUgFepOVg1p472BgWZbGhUF3miRrvGauDywU8PIBGjbTz7RwctJpvFadMd4Upxr3yeN3IZDJ069YN06ZNo+nBKNWelJQUvPXWWzh69Cj7m0QiAcMwePbsGZ49e4ZNmzahV69e+Pvvv+Hg4FApbVEo1oyngwRPEjOhVFmvI0CItz1+HdUMa84/Q+ijJOy5GYerkcmY2s1yVrtX6XnIU2pQ215Mw4UBWkXOxka7eHkBKSla612BCBxViVIrdWfOnDH4OyEEXbt2ZS+SZcuWISQkxOR2ORwOpFIpHBwc4O3tTS82So0gLS0N7du3R3h4OIRCIWbOnIl3330XPj4+AICYmBhs2LAB3333HY4dO4b27dsjLCzMoEe4OduiUKwdLoeBt4MUz5IUsObIHTIRDzN6BKB9XSf8euYpYlO1Vrs3gmthdEtvi3j0pmUrka/WwN1WROfZFYTLBZydtUtOjla5S06uUmnHSn02O3XqZFK5pk2bmlyWQqmpTJw4EeHh4RCLxThy5EiRe8bLywsLFixA165d0bt3b4SHh2PSpEnYuXNnhbZFoVQFxAIuPOzEeJFq/VkUWvk6ooG7Ddaef/6f1e4FrkSm4FMLWe2y89R4lpgFsYALJ5kAtmI+NaYURCwGatcG3N2B+HggMRFW/fXwHxWS+7XGBzykUEzg+vXr2L17NwCtV3lxH0EdO3bEwoULAQC7du3CjRs3KqwtCqUq4SAVwF5aNSb/y0V8zOgRgLn96sNewkdcqm6uXaTF5trl5KsRm5KDiIRMJGbmQqWm4U/04HK1yl39+oBMZmlpSsSsSp1Go2GXrl27mrNpCqXasWbNGgDa4NofffRRieU/+ugj2Nra6tWtiLYolKqGh60YYoFV+v0ZpJWvI34d3QydA5yhIcCemy8wdcctRCRkWEwmlZrgVXoeIhIyEZeajVyl2mKyWCViMRAQAPj4ADzrHbKuOncBhVLNCA0NBQD07NnTpFR4EokEPXv21KtbEW1RKFUNDoeBp4MEnCr0RjNktZu99x7uxqVZVC5CtPljn7xS4HmSAnkqqtzp4eio9Zh1dra0JAapQrcAxWogBMjPqh6LhaYKKJVKPH36FAAQFBRkcr0mTZoAAJ48eQLVf5N3zdkWhVJV0QYmllhajFKjs9q18nWAWkOw7NgjJCusIwZf1n/z7nLyqWKnB5er9ZStXx+QSi0tjR5mtSGqVCqEhYWx63Xr1oWHh0ep23nx4gWePXvGrnfo0MHsEzgTEhKwdOlS/Pvvv3jx4gVsbW3RsmVLfPrpp+jWrZtZ+lCr1WjVqhU7Z2n+/PlYsGCBwbKm7N+uXbswbNgws8hWLpTZwJLSn1er5IuXgKDyb8qUlP8n/XZ0dDS5nlOBFDcpKSlwcXExa1sUSlXGRsSHi40QiVYcmNgQchEfn/UMwMzddxCVnI1lxx/hm8GNrSIlmlpD2IwUclHVmLtYaUgkQGCg1lPWSjCrUrd3716MGjUKgDZESXh4eJnaycrKQteuXVmHi/3796N///5mk/Pu3bvo2rUrkpOTAQA2NjZ4/fo1/v33Xxw6dAhLlizB7Nmzy93PL7/8UupJ6E5OTuByDaeToVHsKRQKpXhc5EJoCEGyIr8qOCuyiPhczO5dH9N23saDlxnYfCkKE9pZR+B9QoDo5GzUthfDTiKwtDjWhwlTXioLsyp169evZxWx/v37o27dumVqx9/fH3379sW///7LtmsupS4nJwcDBw5EcnIygoOD8ddff6Fhw4bIyMjAokWLsGLFCnzxxRdo1qwZO+eoLMTFxWHevHnw9vZGbm4uXr16ZVK9a9eusXHFrBa+RGvhqg7wLTNcUzDor+7jwhRev35dpA1ztkWhVHUYhoG7rRiOUiESM3ORmqW0tEgmU8tejKnd6uHboxHYe+sFAt1t0MbPdOt7RUIIEJuSA5WGwEkmtLQ4FCOYbU5ddnY2zp07B4ZhwDAMRo4cWa72Ro8ezf5/6tQpKJXmuTHXrFmD6OhoyGQyHDx4EA0bNgSgtdYtX74cgwcPBiEEc+bMKVc/U6ZMgUKhwMqVK6ufhY1htEOW1WGxUFwmPp+POnXqAADu3Lljcr27d+8CAOrVqwfefx5Y5myLQqkuCHgc1LaXoJ6rDDbiqnN9t6vrhEFB2uktP598jPh06xnaA4D4tFwkpNNUY9aK2ZS627dvIy8vj7XUlXdeWsH6WVlZ7AuovGzduhWAVmmsVatWke0zZ84EANy8eROPHj0qUx8HDhzAP//8g/79+2PgwIFlF5ZSrenSpQsA4Pjx48gxYU5GdnY2m7e1cBw6c7ZFoVQnRHwuvB2lqOMihVRoeGqLtTG+rQ/qu8mRla/Gt0cirM4DNSkzD7Ep2TQmrRViNqWuoALk4eGhNwm7LDg7O+s5WURERJSrPQDIzMxk57j16tXLYJnWrVuz8btOnTpV6j6ysrLw8ccfQywW45dffim7sJRqz6RJkwBo03v9+uuvJZb/9ddfkZ6eDgCYPHlyhbVFoVRHJAIe/Jxl8HGSWH1MOx6Xg1m9A2Ej4uH56yysPffc0iIVIS1biejkbGg0VLGzJsx2Zes88BiGMZsXnaurK/t/UlJSudsLDw9nvyx0w66F4XA4CAgIAAA8fPiw1H3MmzcPsbGx+OKLL8o0N2748OGwt7eHUChE7dq1MXToUBw6dKjU7VCsnxYtWmDIkCEAgK+++grnzp0zWvb8+fOYP38+AGDo0KFo3rx5hbVFoVRn5CI+6rrI4WZr3dNinGRCzOwVCAbA8YevcDLctHnZlUlmrgrPX2fRLBRWhNmUuvz8fPZ/Y96bpaVgO9nZ2eVuLz4+nv2/uFArum0Fy5vCrVu3sHLlSvj7+2PWrFllkvHatWtQq9Xg8/l48eIF9u7di/79+2P48OF6x9gYeXl5yMjI0FsolY9SqcTr16+LXZRKJdatW4eAgADk5OSgZ8+e+OqrrxATE8O2Exsbi4ULF6Jnz57IyclBQEAA/vjjD4N9mrMtCqW64ywXwllu3RP+m3raYXQrLwDAb6HPEPk6y8ISFSUnX42oZKrYWQtmU+p08bEIIUhMTDRLmwWtc3Z2duVuLyvr/zdEcVH3JRKtR6RCoTC5bY1Gg8mTJ0OtVmPVqlUQCErn9j1u3DgcPXoUqampyMjIgEKhQHh4OCZMmABAG6Pu448/LrGdpUuXwtbWll08PT1LJQfFPISFhcHZ2bnY5eLFi7C3t8fFixfRo0cP5OXlYfHixfD29oZMJoNcLoeXlxcWLFiA3NxcdO/ena1jCHO2RaHUBNxsRbCTWHfsteHNPdHMyx75ag2WHglHVp71BQrPydcgklrsrAKzKXXOBVJmxMXFlVuxS0xMRExMDBuU19lKU3Lo+PXXX3Ht2jUMHz4cPXr0KHX9jRs3olevXnrKa2BgIDZs2MA6b6xbt65E5405c+YgPT2dXWJjY0stC6VycXR0xPHjx3H06FGMHTsWfn5+IIRAo9HAz88Pb7/9No4cOYITJ06UGFzYnG1RKDWB2vZiyEXW6x3LYRjM6OEPJ5kQ8em5WHn6iVU6KOQqqWJnDZjtStbNy2EYBoQQ7N69Gx9++GGZ29u1axd74TIMw6Y0Kg/SAuk8cnJyIJfLDZbTDfXKZDKT2n358iXmzp0LuVyOH3/8sdxyFmb+/PlYtWoVcnJy8O+//7Jz/gwhFAohFFr3kEJ1ZuPGjdi4cWOZ6vbq1cuoA48l26JQqjMMw8DLQYLnr603HZaNmI/ZvQMxe+9dhD1Lxl+Xo/F2a2+zZ1oqLzrFztdJCh7Xup1RqitmO+qenp6oV68eAO0Q7Ndff13m+Vzp6en45ptv2AvW29u7WEXGVArOo3v50njwXN02d3d3k9qdM2cOMjIyMGvWLNjY2EChUOgtOuU0Pz+f/a00SKVSNGrUCADw/Ln1eUFRKBRKVYbDYeDjKIGAZ72KSICbHJM7auNR7roRh21XY0qoYRlylRo8f50FJbXYWQSzXsHjxo0DIQQMw+DVq1cYNGiQSTGzCpKTk4NBgwYhISGBbWvs2LFmkS8wMJBVFB88eGCwjEajYYc4GzRoYFK70dHRALSer3K5vMiim6y+dOlS9jcKhUKhWA88Lgc+ThLwuNZl/SpI70ZueLe9NnXYtmux2H7NOhW7vP8sdlSxq3zMqtRNnTpVLz7duXPn0KxZM5w9e9ak+qGhoQgODsb58+dZ5cvBwQEzZswwi3xyuZwdJj5x4oTBMleuXGHjd5U3gLK5yMrKwv379wEAvr7WkQuQQqFQqhtCHhc+jlJwrNdgh8FNa2FCWx8AwNYrMdh53TrnTVPFzjKY9dKVSqX4448/WIWMEIJHjx6ha9euCAoKwuzZs7F7926cO3cON2/exLlz57B7927Mnj0bQUFB6NatGx4/fszW5XA4+OOPP8xq2dKlH9u6davBkCXLly8HAISEhJg85BsaGgpCiNHF29sbgHZunO63gpQ06XXx4sXIyckBwzDo27evSTJRKBQKpfSIBdoMFFY2XU2PIc1qY2wb7Xvlr8vR2HMzzsISGYYqdpWP2V1+Bg0ahO+//x6fffaZnnJ379491tpkiIJOEbph1xUrVmDw4MFmlW/y5Mn46aefEB0djf79++Ovv/5CgwYNkJmZicWLF2Pv3r0AgCVLlhSpq9uf+fPnY8GCBWaTafjw4fD398cbb7yBJk2asOFQHj16hOXLl2PdunUAtMPbpg4JUygUCqVsyIQ81LYXIzbFuvKuFuTNEE9oCLDlcjQ2hkWByzAYHFw09aWlyVNq8DwpC37OUvCp80SFUyF+3NOnT4efnx/eeecdpKWl6XnoGLJKMQyjpwA6ODhg48aN6N+/v9llE4vF2L9/P7p164abN2+iYcOGrHODRqMBwzBYsmQJevbsafa+jZGUlITdu3djyZIl4HK5sLW1RV5enl5cvWHDhuH333+vNJkoFAqlJmMnEUCtIXiZZr3J60c094RGQ/D31RisvxgJDofBwCDjgfUtRb5Kg6eJCng7SiARWG/4mOpAhanNgwcPRkREBGbPng1bW1uDw446dNvs7e0xd+5cREREVIhCpyMoKAj379/HJ598Aj8/P+Tl5cHR0RH9+vXDiRMnMHv27Arr2xBffPEFpkyZghYtWsDFxQVZWVnQaDTw9fXFqFGjcOzYMezatYuGKqFQKJRKxFEmhKeD2KqHYke28MSI5tog83+cf45Dd41HdrAkKjXB86QsJCvyLC1KtYYhlRDFUKlU4urVqzh//jyePXuGlJQUZGZmQi6Xw8HBAfXq1UOHDh3QokUL8HhUizcnGRkZsLW1RXp6OmxsbEosn5ubi8jISPj6+kIksu7ciBRKZULvjZpLZq42eb0VxvwFoDWMbL4Ujd3/za37sHMd9GlkWkguS2An4aOWnRgcjhVry1ZAad/fQAUNvxaGz+ejXbt2aNeuXWV0R6FQKBSK2ZCL+KjjLPsvx6n1aXYMw2BsG29oCMHeWy+wOvQZMnKUGBbiCa4VKk5p2UrkqdTwcpBadWzAqgg9mhQKhUKhlIBYwIWfs/UqIQzDYHxbHwz6b07dlisx+Gz3HUQnZ5VQ0zLk5Gvn2WXmKi0tSrXCOq9OCoVCoVCsDCGPizrOUogFXEuLYhCGYfBue19M6+4PqZCLp4kKfLrjNnZej4VaY30WRrWGIOp1NhIzrdcZpapBlToKhUKhUEyEx+XAz0kKucg6538zDIOugS74dVQztPCxh0pD8NflaHy2y3qtdq/S8xCdnGWVimdVo8KvSkIIrl+/jvPnzyM8PBwpKSlIT0+HRqPBihUrEBwcXNEiUCgUCoViNjgcBt6OErxIy0FqlnUOHzrKhJjXrwHOPErC2vPP8DRJa7Ub1dILQ5vVtrq5dhk5KjxRZsJGxIdUyINMyLM6GasCFabUKZVKrF69Gj///DObG1WHLrhwamqqwbqTJ09m03h5e3vjzJkzFSUmhUKhUCilhmEY1LaXgM/NRWKGdYbp0Fntgmrb4tfQp7gWlYq/Lkfj0rNkfNq9HrwdpZYWUQ+liiBZkY9kRT4A7TxGuYgHqZAHqYCrF/OWYpgKGX59+vQpWrdujenTpyMqKkovPp0pEVRGjBiBqKgoREVF4dy5c7hy5UpFiEmhUCgUSrlwtRHBz1kKId96ZzPprHbsXLv/rHZbrkQjO19lafGMkpOvRmJGHiKTsvDgZQYiX2chKTMPKpp2zChmvwqfP3+O9u3b4/bt26xFDvh/gGFTNO2uXbvq5V3dsWOHucWkUCgUCsUsSIU81HORwdVGaLWBinVWu9WjQ9DSxwEqDcGOa7GYvOUGjtyPt3pFiRBAkatCQnouIhIyEZ+eQ3PKGsCsSl1+fj769u2LxMRE9jdvb2/8/PPPCA8Ph0KhMFmxGzp0KPv/sWPHzCkmhWKVLFiwQBuWYPx4S4tiMqGhoWAYBj4+PpYWhUKxKAzDwMVGhLouMkiE1ukdCwAOUgHm9quP2b0D4W4rQlq2EqtDn+Hjbbdw6dlrk0bTLA0hwOvMfDxKyMSLtBzkq6hyp8OsSt3q1avx+PFjVml744038ODBA0yZMgUBAQGQSCQmt9WnTx8AWgtfREQEkpOTzSkqhVKhqFQqbNy4Eb1794a7uzsEAgHs7e1Rv3599OvXD99++y2uXr1qaTEpFIqZEfG5qOMsg4edCBwrHZFlGAbt6jrh19HNMLmjH2xEPLxIy8GSIxH4fM9dhMdnWFpEkyAESFHk4/GrTMSmZCNPpba0SBbHrI4SP/30E6vQNWvWDNu3bwefzy9TW02bNgXDMOxXw8OHD9GhQwezyUqhVBRJSUno27cvrl+/zv4mEolACMGjR48QERGBw4cPw9bWFmlpaZYTlEKhVBiOMiFsxHy8TMtBRo51zlvjczno38QDXQNdsOfmC/xz+wXCEzIxa89dtPFzxLg2PqhlL7a0mCVCiDZLRVq2EnYSPpzlQoj41mstrUjM9h0RHh6OmJgYVgn75ptvyqzQAYBUKkWtWrXY9WfPnpVbRgqlMhgzZgyuX78OuVyO77//HvHx8cjJyUFaWhrS09Nx4sQJfPjhh7Czs7O0qBQKpQLhcznwdpTCy1ECPs9KJ9sBkAh4eLu1N9aOCUHPBq7gMMCl58n48O8b+P3ssyqV9SEtW4knrxSIT8+pEkPJ5sZsSt2tW7fY/2UyGbp3717uNu3t7dn/qUWDUhWIiIjA8ePHAQAbNmzAzJkz4ebmxm6Xy+Xo3r07fv31V0RERFhKTAqFUonYivnwd5HDSS6wWkcKQGtdnNK1HlaODEYLH3toCHDoXjze33IDxx8mQFOFlKTXmfl4lqSocUOyZlPqkpKSAGjH6n19fcExw2SCgnPwsrOzy90ehVLR3Lt3j/2/f//+xZYViUTFbt+0aRNatWoFuVwOGxsbdOnShY3faIhz585h6tSpaNWqFTw8PCAQCODi4oLevXtj9+7dRuuNHz8eDMNgwYIFyM3Nxfz58xEYGAixWAwXFxeMGjUKjx8/LlbW4oiKitKbVyuXyxESEoLvvvsOWVn6Ee7PnTsHhmEgFAqLnUf7/PlzcDgcMAyDR48elVk2CqWy4HAYuNuKrd6RAgC8HaX4qn9DfD24ETwdJMjIVeGX008xa/ddPE1UWFo8k9Hll03Lzre0KJWG2ZS63Nz/524TCoVmaTM9PZ3938bGxixtUiiVxYsXL8pc97333sP48eNx48YNcDgcZGZmIjQ0FL1798aePXuKlFcoFOjUqRNWrlyJq1evQqFQQCwWIykpCceOHcObb76JyZMnF9tnXl4eunTpgkWLFiEyMhICgQBJSUnYvn07goODce7cuVLvx969e1G/fn2sWrWKdaLKy8vDzZs3MXv2bLRp0wavXr1iy3fs2BH+/v7Iz8/H33//bbTdP//8E4QQtGvXTi/8EYVi7egcKWrbi60+Y0JQbTusHNEU77bzhZjPxaNXmZi+8zZWhz6tMkOyGg0Qm5KDuNRsaGpAGjKzKXXOzs7s/69fvy53e2q1GrGxsey6k5NTudukUCqakJAQ9v+PPvqItWCXhv3792Pr1q347bffkJGRgfT0dDx//hwdO3aERqPBlClToFLpT7zmcDgYNmwY9u3bh+TkZLZeamoqVq1aBZlMhrVr12LXrl1G+/3tt99w9+5dbN68GQqFAunp6bh16xaaNWuG7OxsDB8+3GgWGENcu3YNI0eOhEqlwpdffom4uDhkZWUhJycHYWFhaN68Oe7du4exY8fq1Xv33XcBaBU3Q2g0GmzatAkA8M4775gsD4ViTdhLBQhwk8NRJrC0KMXC43IwOLgWfnurGTr5O4MAOHI/AZO33MCxB1VnSDY1S4mnSQrkKqv3cKzZlDp3d3cA2hAk0dHRela2snD9+nW9oRl/f/9ytUcxH4QQZOerqsVi7om0fn5+rJJy7Ngx1K5dG927d8fcuXOxf/9+k5S8tLQ0rFu3Du+//z47BcHX1xfbtm2DQCBAfHw8wsLC9OpIJBLs2rULgwcPhoODA/u7nZ0dPvroI6xevRoA2L+GSE9Px9q1a/H222+zTk5NmzbFsWPH4OjoiFevXuHXX381+VhMmzYNSqUSq1atwtdff806PnG5XLRp0wbHjh2Du7s7jh8/rucpPH78ePD5fNy6dQt37twp0u7JkycRGxsLmUyG4cOHmywPhWJtcDkMPOy0Q7JigXUPyTrKhPisZwCWvNEYXg4SZOaqsOrMU8zcfQdPXmVaWjyTyFNqh2NTskwfjq1qzhZmC2nStm1b8Hg8qNVqEEKwe/du9ou7LKxZs4b9387ODs2aNTOHmBQzkKNUo8FX1SMg9MNFvSARmDcF8h9//AEnJyesWrUK+fn5OHXqFE6dOsVub9GiBaZOnYrRo0cbDMTt5eWF0aNHF/ndw8MDLVu2xIULF3D//n107NjRZJkGDBgAALh8+TLUajW43KIvEG9vb4P9Ojk5YfLkyViyZAl2796NuXPnltjfs2fPcPHiRdjZ2Rl9Djg4OKBPnz7YsGEDTpw4gebNmwMAXFxcMGDAAOzduxcbNmzAzz//rFdPZ8F78803IZPJSpSFQrF2xAIu6rrIEJ+eg9eZ1j3/q3EtW/w8oin+vRePv6/E4PErBT7bfQeTO9ZB38bulhavRAgBXqTmQJGrgrNcCJVGA5WaQKnRQKkmUKn/+/vf7zwuAx9HaZUJkWI2S52trS3atm3LpgNbunRpmZ0bLl++jC1btoBhGDAMwwYiplCqAgKBACtWrEBsbCx+//13jBo1CvXq1WMVuGvXrmHMmDEYMWIENJqikdCbN29uNOuKztplaBhUpVJh/fr1bMBjoVDI3kM6T/Lc3FyjQ6idOnUy2m+nTp0AAPfv30d+fskvHZ0lUaFQoHbt2nBzczO46FIAFpxqAWjnFALA1q1b9fpLTU3FP//8AwDl+mikUKwRd1sx7KVlDwVWWfC4HAxuWgu/jwlBu7pO0BDgt7PPsOFiZJUZjk3PUeJpogJRr7MRl5qDV+l5SFHkIyNHhZx8NZQqAkIApYrgaaKiyswhNKuJYsaMGaz3WmRkJN566y3s3LmzVPHqbt++jaFDh7IWPw6Hg9mzZ5tTTEo5EfO5eLiol6XFMAviCvz6cnFxweTJk1kHhVevXuHgwYNYtGgRYmNjsWvXLrRr1w5Tp07VqyeXy422qfOYVSr1HzAKhQK9evXSG5YVi8VwdnZmPdF1DglZWVkG56gWjAtpbJtarUZqaipcXV2NlgWA+Ph4AFpFs6AjhDEKfwD26tULnp6eiI2NxcGDB9m0gX///Tdyc3MREBCAdu3aldguhVLVqGUnZoPpWjsOUgE+7xWAnY4SbLkSg323XuBVRi6m9/CHkFc1LFumQAgQnZwNd1sRHGXmcQStKMyaxGTAgAHo1KkTOwZ94MABtGrVSm/oyRgvXrzAnDlz0L59e8THx7M5Yt966y00atTInGJSygnDMJAIeNViMSUPsblwdXXFe++9h5s3b7JK0YYNG8zS9uLFixEWFgYnJyds2rQJr169QnZ2NhITE5GQkKDniVsZc0R0FsigoCDWel/csnHjRr36HA6HdYIo6DCh+3/ChAkVvg8UiiVgGAa17cWQi8w7LaSiYBgGI1p4YUYPf/A4DMKeJePLfferXRgRQoCXabmIT8+xtCjFYvbMdDt37tRL7n379m307NkTtWrVYuf16F4qS5YsQf/+/REYGAgvLy98//33yM7OZl+0jRs3xu+//25uESkUi+Lk5IRBgwYBQLnivxVE59X6yy+/YOzYsXBxcdHbboq17OXLlyVu43K5ekHBjaFTWgsPq5aGd955BxwOB0ePHkV8fDzu3r2LGzdugMvlFvGYpVCqEwzDwMtBAqmVx7MrSOcAFywa1AgyIQ+PXmXis913EJda/eLLvs7MR0yy9YZHMbtS5+zsjGPHjiEgIIC1thFCEB8fj8OHD7PlCCE4c+YMjhw5gsePH7Nf7LrywcHBOHToEMRi6887R6GUFqlUCkA7/84cxMXFAQCCg4MNbj958mSJbZw9e7bEbY0aNTJJ5jZt2gAAUlJScOXKlRLLG8LLyws9evSAWq3G5s2bWStdnz59WG97CqW6wuFoJ+hbu1dsQRrXssWyYU3gZiPCq4w8zNx9F/dflC8ShjWSnqPE89dZUKmLzom2NGZX6gCgXr16uHHjBiZPnsx62emsb7qJ2wWHvQquc7lcTJ48GRcvXix2jg+FYo1ERkaWmKc4OzubnezftGlTs/Rra2sLQD+jhQ6FQoFvvvmmxDaioqKwbdu2Ir+npKRg7dq1ALQep6YQGBiI1q1bAwBmzZpVZA5gQXJycpCXl2dw28SJEwFoh6m3bt0KgDpIUGoOHA4DXycpRPwKeVVXCLXtJVj+ZhACXOVQ5Kkwb/99hD5KtLRYZicnX41nSVlWF/euwq4UsViM3377Dc+fP8esWbPYeXHG5tTUqVMHH3/8MR49eoTffvutxBRKFIo18uDBAwQEBGDIkCHYuXMn6zAAaB0UDh48iA4dOiAyMhIAijhJlJUePXoAAKZPn46zZ8+yUxyuXbuGbt26FZtyS4etrS0mTpyIrVu3ssGN7969i169eiEpKQkuLi748MMPTZZp5cqVEAqFOHfuHLp164YLFy6wc+3UajXu3buHRYsWwc/PT+84FWTgwIFwcXHB48ePWRlKSr9GoVQnuBwGPk5SCKuQYmcr5uObNxqhbR1HqDQEK048xvZrMVUu5ltJ5Ks0eJakQFaequTClUSFz8SsXbs2vv32W3z77bdIS0vDo0ePkJycjLS0NEgkEjg5OcHX15da5SjVAj6fD7VajX379mHfvn0AtB84AoFALyA3l8vFokWLMGTIELP0+/XXX+PEiROIjY1F586dIRKJwOVykZWVBbFYjH/++Qe9ehXvsfzBBx/gzJkzGDNmDN59910IhUJkZGQA+H9wY1Pm0+lo0aIF9u3bh1GjRuH8+fPo0KEDhEIhZDIZMjIy9Kx3xhxW+Hw+xo4di+XLlwMA3n77bfB4VWMCOYViLvhcDnwcpXj+WgGlqmooRkIeF5/3DsSmsCjsvfUCW6/E4MHLDHzUuS7cbKuP0UajAVKz8yEVWsdzqVKlsLOzQ6tWrSqzSwqlUunVqxcePXqEgwcPskGCX7x4AYVCATs7O/j5+aFjx45477330LBhQ7P16+fnh6tXr+Krr77C8ePHkZqaCkdHRwwePBhz5swxqS+hUIjQ0FAsXboU27dvR0xMDJydndGtWzcsWLCgTDlW+/Tpg8ePH+OXX37B4cOH8fTpU6SlpcHOzg4BAQHo2LEj3nzzTXh7exttY8iQIaxSR9OCUWoqAh4Hvk5SPE/KgkpdNRQ7DsNgQjtfuNuK8cf557gdm4aPtt3EmFZeGBhUy+pz31ZFGFLd7KEUPTIyMmBra4v09HTY2NiUWD43NxeRkZHw9fWlQ+A1hPHjx2PTpk2YP38+FixYYGlxivDNN99g7ty5aNWqFS5fvmwxOei9QbEG8lRqRCdnI09pfZP0i+NlWg5+PfMUd/9znKjrLMOUrnXh51z1s8LYS/mobS8xe7ulfX8DFTinjkKhUMqLWq3GunXrAACTJk2ysDQUiuUR8rio4yyDrIrEsdPhYSfG14Mb4ZOudSEVcvE0SYFpO29jY1gU8lTW5WxQlbGIUpeZmYn4+HhkZlaNJMAUCqXy0Wg0WLRoEaKiouDq6opRo0ZZWiQKxSrgchj4OErgKDNPSKTKgmEY9Gjght9G/z+92J6bcZiy7RbuxKVZWrxqQYWr+pmZmdi2bRtCQ0Nx9epVREdH6+W75HA48Pb2RsuWLdGpUyeMGjXKZDMjhUKpfly+fBkjR45Eamoq66ixZMkSGrOSQikAwzDwsBNDyOMgPj0XVWkilb1UgNm9A3H5eTJ+P/sM8em5mPvPffSo74qRLTzhYkOnN5SVClPqMjIy8OWXX+LPP/9ETo42rYah6XtqtRrPnz9HZGQkduzYgRkzZmD8+PH45ptv2NhbFAql5pCbm4vo6Gjw+XwEBgZi+vTp1EGCQjGCo0wIAY+DmJRsaKrWNDu09nNEk9q22HQpGofvxeNE+CucCH+Fei4ytKvrhLZ1HOFuSz/mSkOFOEqcPXsWo0ePRkJCAqvImZJjs2BZNzc3bNmyBV26dDG3eDUK6ihBoZgHem9QrJlcpdaBIl9VxTS7/3jwMh1br8Tg/ot0FFRK/JykaPufgudZAc4I5sCaHCXMrtQdPXoUQ4cOZa1zurRfOmxsbODg4ACpVIqsrCykpKSwQyyFy4tEIuzZswd9+vQxp4g1CqrUUSjmgd4bFGtHpdYgOiUb2XlV1/EgNTsfl58nI+xZMu7GpaFgilUvBwna1nFE+7pO8HaUWk7IQlRbpe7ly5do0KABMjIyWMscIQRdu3bFuHHj0LlzZ3h6ehapFxMTg7Nnz2LTpk04ffq0Xl0bGxs8fPgQHh4e5hKzRkGVOgrFPNB7g1IVIITgZXouUhT5lhal3GTkKHElMhkXnyXjTmwaVAU0PB9HCTr6O6NjPWe4WngOXrVV6kaPHo3t27ez1jYXFxds3rwZPXv2NLmNY8eOYdy4cUhKSgIhBAzDYMSIEfj777/NJWaNgip1FIp5oPcGpSqRp1IjKTMPadnKKuVEYQxFngpXI1Nw8elr3IxJ1VPw6rvJ0cnfGe3qOsFOUvkewdVSqcvIyICrqyvy8/NBCIG9vT2uXr2KOnXqlLqtp0+folWrVkhLSwMhBEKhEK9evaJesWWAKnUUinmg9walKpKv0uC1Ig8pWfnVQrkDgMxcJcKeJePckyTci/v/HDwOAwTVtkMnf2e0qeMIiaByYvlZk1Jntjh1oaGhyMvLY61r3333XZkUOgCoW7cuvv32W3ZuXX5+Ps6cOWMuUSkUCoVCqREIeBx42IkR4CaHs1wIE3wWrR65iI9eDd3wzeDG+HN8C7zX3hf+rjJoCHArNg0/nXqCdzZdwz+3XkCprpqOI2XFbEpdbGws+79IJMLo0aPL1d5bb70FsVjMzq+Li4srV3sUCoVCodRU+FwO3GxFqO9uA1cbITjVJJ+Uo0yIQU1rYcWbTbFmTAjGtPJCLTsxsvLUWH8xEh/9fROXnicbDKlWHTHbadV5sDIMA19fX0gk5TNFSiQS+Pr6sieioIcshUKhUCiU0sPlMHCxESHQzQYOVSwjRUl42IkxooUXfh3dDFO61oWdhI/49FwsORyOuf/cx/MkhaVFrHDMptQ5ODiw/5trvknBduzt7c3SJoVCoVAoNR0uh0EtOzF8naUQ8KqJ2e4/uBwGPRu4Yc2YELwZUht8LoO7L9Lx6Y7bWHnqCVKyqr5nsDHMdiYbNWoEQOtOHRMTY5Y2o6Oj2f8bN25sljYpFAqFQqFokQl5qOciq3J5ZE1BIuBhbBsf/P5WCDrWcwYBcCL8FSZvuY4d12ORp6q68fyMYTalrnXr1nB2dgYAJCcn4/z58+Vq79y5c0hOTgbDMHByckKbNm3MISZLQkICpk6dijp16kAkEsHV1RUDBgzAqVOnzNaHWq1G8+bNwTAMGIbBggULii2v0Wiwdu1atGnTBnZ2dpDL5QgODsayZcuQn199vywoFAqFYjk4HG0eWT9nKYT86mW1AwAXGxFm9grAsmFNEOAqR65Sgy2XozFuw1VM2XYT8/bfx4rjj7Du/HPsvhGHkw9f4XpUCp4mKpCaXbXevWbz9+VyuZg+fTrmzJkDAJgxYwYuXboELpdb6rZUKhVmzJjBrs+YMQMcM87qvHv3Lrp27Yrk5GQA2iwXr1+/xr///otDhw5hyZIlmD17drn7+eWXX3Djxg2TyiqVSgwePBiHDx8GAAgEAnC5XNy+fRu3b9/Grl27cPr0achksnLLRaFQKBRKYaRCHuo6y5CYmYekzDxLi2N2At1ssGxYE5x78hqbLkUhKTMPWcnZQHJ2sfVa+NhjRHMvBLjJK0nSsmNWlXzGjBlo27YtCCG4ceMG3njjDWRlZZWqjaysLLzxxhusMtSuXTs9Ba+85OTkYODAgUhOTkZwcDDu37+P9PR0pKamYsaMGSCE4IsvvsDx48fL1U9cXBzmzZsHb29vuLq6llh+7ty5OHz4MEQiETZu3Ijs7GxkZWXh4MGDcHBwwLVr1zB58uRyyUShUCgUSnFwOAzcbEWo41I9rXYMw6CTvzPWjAnBzyOaYuHAhpjW3R8T2vpgSHAtdA1wQTMve/g5S+EgFYABcC0qFZ/tvoN5++/j/ot0S+9CsZj1jPF4PBw6dAjdu3cHIQSHDh1CQEAA1qxZg7S0tGLrpqWl4ffff0dAQABrrerZsycOHTpUJmufMdasWYPo6GjIZDIcPHgQDRs2BKC11i1fvhyDBw8GIYS1OJaVKVOmQKFQYOXKlSU6jiQkJODnn38GAHz33XcYN24cuFwuGIZB//79sWHDBgDAtm3bcPfu3XLJRalYxo8fzw636xY+nw9HR0fUrVsXgwcPxpIlSxAZGVkh/W/cuLFI/wzDQCwWw8fHB8OHD8eJEyeKbaNz585F6gsEAjg7OyMwMBDDhw/HDz/8gISEBLPLv2DBAjAMA3t7e+TlmWYpOHToECtnRESE2WWiUGoiEoF2rp2XgwR2Ej543GoQ4K4AfC4Hfs4yNPOyR9dAFwxpVhsT2vliWg9/LBzYED+PCMamCS3x21sh6F7fBVwOg9uxaZiz7x5m772LmzGpVhkmxaxpwhYtWgRAO5ds8+bNrKOD7sXWpEkTNGrUCI6OjpBIJMjOzsbr169x//593Lt3D0qlkj1IPj4+GDt2bKmGXb/66qsSy7Ro0QLXr1/HpEmTsGbNmiLbw8LC0K5dOwBAREQEAgICTO5fx4EDBzBo0CD0798fBw8ehI+PD6KjozF//nyD8+p+/fVXfPzxx7C1tUVCQoJBJTAgIACPHz/G559/jm+//dZkWWhGicpl/Pjx2LRpE/h8PusRTghBRkYGcnNz2XIMw2DYsGFYvXo1nJyczNb/xo0bMWHCBADQsxCnpaXpKUnTpk3DDz/8YLCNzp074+zZsxCJRLC1tQWgne+ZkZGh1waPx8N7772H5cuXQyo1T3Lt58+fo27duiCEYNeuXRg2bFiJdUaMGIGdO3eiZcuWuHLlilnkMAS9Nyg1nZx8NTJzlcjIVSEnv/o5GRTHq4xc7LkZhxMPX7EpyvxdZRjR3BM9G7rC08E8z8CClCWjhFlzaOi+snXo/ieEID8/H9evXzc4x6ygXqmrEx0djcWLF5eq/5KUuszMTLb/Xr16GSzTunVr9iCeOnWq1EpdVlYWPv74Y4jFYvzyyy8m1dFly+jYsaPRl0XPnj3x+PFjnD59ulTyUCxD27ZtERoaqvdbWloaLl++jI0bN2LXrl3YtWsXwsLCcPnyZdSuXdvsMhS0pGk0Gjx48ABTp07FmTNn8OOPP6J79+7o27ev0fojRozAxo0b9X5LTEzExYsXsXbtWhw9ehS///47wsLCcP78ebOk8fPz80P79u1x/vx5bN68uUSlLj09HQcOHAAAjBs3rtz9UygU44gFXIgFXLjYAEq1BopcFTJzVcjIrR75ZYvD1UaEDzvXxYjmnth76wWOPkjA41cKLD4Ujm3XYjGtuz/6NXG3tJjmHX41RsFhnLJsLwlTjY3h4eFsWd2wa2E4HA6ryD18+LDUssybNw+xsbH44osv4OPjY1IdXT/GZAKABg0aANDfB0Pk5eUhIyNDb6FYB3Z2dujduze2b9+OQ4cOQSQS4cWLFyZZo8oLh8NB48aNsW/fPtZLffPmzaVux8XFBW+88QaOHDmCDRs2gGEY3L17FxMnTjSbrDrl7OjRo3j9+nWxZXfu3Inc3FwIBAKMHDnSbDJQKJTi4XM5sJcK4OUoQV0XGcSC6jf/zhCOMiEmdvDD+rHNMaxZbYj5XDxNVGDn9diSK1cCZj8LhJBKX0wlPj6e/d/Dw8NoOd22guVN4datW1i5ciX8/f0xa9asUstlikwKhQIKhfGo2EuXLoWtrS27eHp6miwHpfLo3bs3li9fDgC4cuUKDh48WKRMfn4+Vq1ahQ4dOsDBwQFCoRDe3t545513EB4eXqZ+bW1t0bJlSwBl+2gpyIQJE1gnpl27dhWZ76lWq3HkyBFMnjwZISEhcHV1hUAggIeHB9544w2jVuc333wTYrEYSqUS27ZtK1YGnWI6YMAAdrj75s2bmD17Ntq3bw8vLy8IhUI4Ojqic+fOWLduHdTqmjVsRKFUNCI+F35O1TPWnTHsJAKMa+uD9eOa4512PpjavZ6lRQJg5uFX3TCitVLQE1csFhstp0txVpzyVBiNRoPJkydDrVZj1apVEAhMv7h1cpkik04uudywa/WcOXMwffp0dj0jI4MqdlbKxIkTsWjRIiQmJuLvv//GgAED2G3x8fHo06cP7ty5A0BraZNKpYiJicGff/6Jbdu2YevWrRgyZEip+9V9CJlDuZk5cyZWrlyJ/Px8bNu2DU2aNGG3hYeH6w3v2tjYQCAQID4+Hv/88w/++ecfLFmypIhTko2NDd544w38/fff2Lx5M6ZMmWKw78jISFy8eBGA/tBrz5492XBFEokEEokEKSkpOHv2LM6ePYt9+/Zh//794PHM+vijUGo0ulh3UiEPcanZ0GgsLVHlIBfx8U57X9S2L19qVHNh1qdap06dzNlcleLXX3/FtWvXMHz4cPTo0cNicgiFQgiFQov1TzEdgUCAbt26Ydu2bXrBupVKJQYNGoQ7d+6gW7duWLx4MZo3bw4+n4/4+Hh8//33+Omnn/D2228jKCgIderUMbnPtLQ0XL16FYB2/lp5cXFxQUhICC5dulQk4LhAIMA777yDESNGoHXr1uycu8TERKxduxYLFizAl19+ia5du6JVq1Z6dceNG4e///4b169fR0REBAIDA4v0vXnzZhBC4OLigj59+rC/9+zZEwMHDkTnzp3h5uYGQPvhtG/fPsycOROHDx/Gjz/+iJkzZ5Z7/ykUij62Yj7EfDliU7ORnUet4pVNzRgE/4+CHno5OTlGy2VnawMRmhro9+XLl5g7dy7kcjl+/PHHMstlikylkavCIATIyqoei4Vn9+rS37148QJKpRIAsGnTJly7dg0dOnTAkSNH0KZNG/D5fACAu7s7fvzxR0yePBnZ2dkmX2+EENy7dw9Dhw5l56mNGTPGrPtQOEyLv78/1q9fj549e+o5Ubi4uGDu3LmYP38+CCH4/fffi7TZvXt31KpVC4DxuX9//fUXAGD06NF6Vre///4bI0eOZBU6QHuPjRkzBjt37gQArF69uiy7SqFQTEDA48DPSQpnOTUwVDY1SqkrOGft5cuXRsvptrm7m+bJMmfOHGRkZGDWrFmwsbFh573pFt1wV35+vsE5cTq5TJFJJpMZHXqtNLKzAZmseizZxUcSr2js7e3Z/1NSUgBolToAmDp1KqvMFeatt94CgGJjzrm5ubGLWCxGkyZN2Hls77zzDoYPH27WfdDJbyq64WbdEGpBOBwOq3Ru3bq1yNzZixcv4tmzZwCAsWPHmtxnhw4dYGdnh6ioqGLvNwqFUj4YRhvE2MdJAi6nesW4s2ZqlFIXGBjIetg+ePDAYBmNRoNHjx4B+L/HaUno4vHNmzcPcrm8yBITEwNA68Sg+60gun6MyQT8f1J7/fr1TZKJUjVRqVTs8OjkyZP1FLOCi24uXWyscY+rV69esYsuvhyHw8Eff/yB9evXl9nbvDTk5OTgxx9/ROfOneHi4gI+n896ugcHBwMw/jEzfvx4AEBMTEyR8DA6613jxo3Zdgqya9cuDB48GF5eXhCLxXoe9rpA6FSpo1AqHrmIj3quMshFdA5rZWDWo5yfn18qBwFTOXHihFnmqcnlcjRv3hzXrl3DiRMnDE4yv3LlCtLTtWlAunXrVu4+TaFLly7YvXs3zp8/j9zcXIOx6nQWmcqSqVgkEqAUTiRWjcSyk1tTU1PZ/x0cHJCSkoL8fG0Cad1k/+Iobsi+oENEbGwsNmzYgG+++QafffYZgoODERISUk7ptej2Qed9qiM+Ph6dO3fG48eP2d+kUins7e3B4XCgVqvx+vVro6kEAwMD0bJlS1y9ehWbN29Gly5dAGjD9uiGUQvHplOpVBg+fDj27dvH/iYUCuHk5MRmpklKSoJGoyl1CkMKhVI2+FwOfJykyM5XITEjD5m5KkuLVG0xq6WuVatWePLkidnaU6vV+Pzzz/UmQZeX0aNHA9AO6RgKWaILMxESEmJy4OHQ0NBiQ654e3sDADuHqPBQ0pAhQyAUCpGWloZ169YVaf/gwYN49OgRGIbBqFGjSrW/FQLDAFJp9VgqwVpVHPfu3QMA1K5dG3w+H5oCLmO3bt0yS0gfLpcLHx8fLFq0CIsXL0Z6ejqGDx9uNqVGtw+FHS8+/fRTPH78GH5+ftizZw9SUlKgUCiQmJiIhIQEXL58ucS2dUrbnj172HmlBw4cQFpaGrhcLjsMreOPP/7Avn37IJFIsHLlSsTGxiI3NxdJSUlISEhAQkICO93BGlP8UCjVGYmABx8nKeq6yGAjppa7isCsSt2dO3cQEhLCTmAuD9HR0Wjfvj2WLVtm1ofv5MmT4e3tjczMTPTv358d1szMzMSsWbOwd+9eAMCSJUuK1NUN3xhK9VUe3NzcMHXqVADArFmz8Ndff7HhJg4fPsymfRo1apReyAhK1SY/Px+nTp0CoJ3rBQCOjo6sRUk3bG9OZs6cCT8/Pzx//pz9gCkPiYmJbJYW3T4A2n3bv38/ALChVwrOHwS0w8MlMXLkSAgEAmRmZuKff/4B8P+h1169euk5QwDaYVdAOxViypQpRTJ16KyDFArFcogFXHg7SlHPVQZbseF5w5SyYfY5dQqFAuPHj8e4ceP0PDZLw65duxAcHMzOLTInYrEY+/fvh6OjI27evImGDRvC1tYWdnZ2WLZsGRiGwdKlS9GzZ0+z910cX3/9Nfr27YucnByMHTsWUqkUUqkU/fr1Q3JyMlq0aGHQS5BSdfnjjz+QmJgI4P+OD3w+H82bNwcAHDlyxOx98vl8zJ49GwCwYsUKveHfsrBs2TLk5+eDYRjWCg4Ar1+/ZufxGZrzBgAnT54ssX0HBwfWoWLz5s1ISkrC0aNHARhOCxYXF1dsnxcvXtTLwUuhUCyHiM+Fl6ME9VxlsJNQ5c4cmF2pYxgGhBBs2bIFISEhbPBUU8jNzcWkSZMwcuRIdjIzUHxQ3rIQFBSE+/fv45NPPoGfnx/y8vLg6OiIfv364cSJE+xLrzLh8/k4ePAgfv/9d7Ru3RpCoRAMw6Bp06b47rvvcOHCBct7vVLMxrFjx9g4aW3atEG/fv3YbToHgY0bN5Z4/5RFKRs7dixcXV2RmZmJn3/+udT1dWzcuBErVqwAoLWoNWrUiN0ml8tZRwzd8GxB4uPjTc6NrFPeTp48iRUrVkClUsHOzg4DBw4sUtbW1tZonyqVCnPnzjWpTwqFUnmI+Fx4Okjg7yaDk1xAvWXLgVmVuk8//RSEEPZh/ujRI7Ru3RqrVq0qse7Dhw/RokULrF+/nm2DEIKGDRtWiMXOzc0NP//8M549e4bc3FwkJibi33//LdYRQTeHqbTDr1FRUSbV43A4mDx5Mi5duoT09HQoFArcunULs2bNqhAHFErlkp6ejmPHjmHUqFGsVdbT0xO7d+/WK/fuu++idevWyM3NRdeuXfHHH3/o5fBNSEjA1q1b0alTpzIpZUKhkM3SsHLlSmRmZppc9/Xr1/jnn3/Qt29fTJgwAYQQNG3aFGvXrtUrJ5fL0bp1awDa8Cm3b98GoPUuP3XqFDp16mTytIo+ffrAxcUFarUay5YtAwCMGDHCoEORzqFq8eLF2L9/PzuNISIiAgMGDMDVq1f14lVSKBTrQcjjwt1WjPrucng5SqjHbBkwq1L3ww8/4MCBA3BwcGAVs7y8PEydOhVvvPGGnvWtIGvWrEGLFi3w8OFDPYVu4sSJuHbtmsmhRSgUayEsLIwNP+Lq6gqJRAI7Ozv07t0b27dvByEEw4cPx82bN4vk/OXz+di/fz/atWuHlJQUTJo0Cfb29nB0dIRMJoO7uzvGjBmDc+fOlTksyYcffgiZTIbU1FSjH107duxg98HFxQUikQjOzs544403cOTIEfD5fHzwwQe4ePGiwYDYP/74I8RiMe7du4fg4GDIZDLIZDJ0794dycnJWL9+vUmy8ng8dmhX50hiaOgVAD777DPUqVMHGRkZGDx4MMRiMWxtbVG/fn2cOHECv//+O5ycnEzql0KhWAaGYWAr5sPHSYpAdzlcbYUQ8GpUBLYyY/aj1L9/f9y+fRsdOnTQU9AOHDiApk2bIiwsjC2bkZGBN998Ex9++KFeaAYbGxvs2LEDa9asMfg1TqFYO0qlko0Rl5ycDJFIBD8/PwwcOBDffPMNnj9/jh07dhhVMFxcXHD27Fls3boVffv2hbOzM2tRCwwMxNixY7Fz584yTxWwt7fHe++9B0CrfBma/5qbm8vuQ1paGqRSKfz9/TFs2DD88MMPiI2NxerVq/XyEhekVatWuHTpEgYPHgx7e3solUq4uLhg8uTJuH37NoKCgkyWt6ASV69ePbRp08ZgOQcHB1y+fBkffPAB6yQhFosxePBgnD17lh3aplAoVQM+lwMXuQgBbnL4OkthJ+HT4dliYEgF+fVrNBosXLgQ33zzDTvMQggBj8fDggUL0LlzZ4wZMwbR0dF6yl+rVq2wfft2NgwIpXxkZGTA1tYW6enpeqmajJGbm4vIyEj4+vpShZpCKQC9NygU6yFfpUFOvhrZShWy89XIyVdbLOujvZSP2vbmj3la2vc3YObgwwXhcDhYuHAhunTpgjFjxuDly5dgGAYqlQrz5s0DUDRO1KxZs/DNN9+wIR0oFAqFQqFQCiPgcSDgcWALrdcsIQS5Sg2y87VKXmauCmpNzYtFWeGD1J07d8bt27fRp08fPYtcQYcKFxcXHD16FN9++y1V6CgUCoVCoZQKhmEgFnDhKBPC00GidbZwkEBWw5wtKmXmoZOTEz755BPWg1MXxFen2K1evdosacAoFAqFQqFQGIaBrYQPXycp/N1kcLERgset/nPxKlyp02g0+OKLL9CvXz8olUq9bTrFbuTIkVi6dGlFi0KhUCgUCqWGIeRx4WojQqBb9Q+VUqFKXWxsLDp27IjvvvuODUVACIGvry88PDxYS50uKGjPnj1NSh1EoVAoFAqFUhoKhkoJcJPDzVYEmYhn6RTgZqXClLr9+/cjODgYly5d0ptLN3z4cNy6dQt37txBv3799LadPHkSQUFBOH78eEWJRaFQKBQKpYYj4HHgLBfC10mKBu428HaSwFEmgJBftePhmV16pVKJTz75BEOGDEFKSgr7u0gkwpo1a7B9+3bI5XI4ODjg4MGDWLFiBfh8PjvPLjExEX379sWsWbPYaPAUCoVCoVAoFQGHw8BGxIeHnRj+rnIEuMlRy14MGzGvysXEM+vA8tOnTzFixAjcvn1bzwJXv3597Ny5Ew0bNixSZ9q0aejYsSNGjhyJZ8+egWEYaDQarFixAufOncP27dvh4+NjTjEpFAqFQqFQDCLgceDAE8BBqnXuVGsIlGoNVBoCpUoDpUYDpVr7v0qjsSrFz6yWumbNmhVR6N555x1cv37doEKnIyQkBLdu3cLo0aP16l69ehXBwcHYuXOnOcWkmEAFxaSmUKos9J6gUGomXA4DEZ8LmZAHe6kALnIRatmJ4eMkRV0XOdxtxZYWkcWsSp1CoWD/l8lk2Lp1K9atWwexuOQdlslk2LJlC9avXw+xWMwOx6anp2PUqFHmFJNSDByO9pKgQ98Uij66e0J3j1AoFIq1YfanEyEEzZo1w82bN8ukjE2YMAHXr19Ho0aN6JexBeDz+eDz+XoKOoVCATIzM9n7g0KhUKwRsyt1U6dOxaVLl1CnTp0ytxEYGIirV6/i/fffN6NkFFNgGAZyuRzp6enIycmxtDgUilWQk5ODjIwMyOVyNhMOhUKhWBtmdZTYv38/BgwYYJa2hEIhVq9eje7du2PixIlmaZNiGk5OTsjJyUFMTAxsbGwgl8vB5XLpy4xSoyCEQK1WIzMzExkZGRAKhXBycrK0WBQKhWIUhlSBMc6YmBh4eXlZWowqSUZGBmxtbZGeng4bGxuT66nVarx+/RqZmZlFMoFQKDUJPp8PuVwOJycnmpuaQqFUGmV5f1eJXBlUoat8uFwuXF1d4eLiAqVSyWYEoVBqEhwOh42jSaFQKNZOlVDqKJaDYRgIBAJLi0GhUCgUCqUEqG8+hUKhUCgUSjWgVJa6AwcOsP+3a9cOjo6OZhdIx7179zB//nwAWmvRnj17KqwvCoVCoVAolKpOqZS6wYMHs3NLTpw4ga5duxZbvjyKWWJiIv755x+2LoVCoVAoFArFOKWeU6dL42UKVDGjUCgUCoVCqRxKPaeOKmcUCoVCoVAo1gd1lKBQKBQKhUKpBlCljkKhUCgUCqUaQJU6CoVCoVAolGoADT5czdFlgcvIyLCwJBQKhUKhUExF994uTTZXqtRVczIzMwEAnp6eFpaEQqFQKBRKacnMzIStra1JZalSV83x8PBAbGws5HK52T2XMzIy4OnpidjYWJOTDVPKDj3elQs93pULPd6VCz3elUtZjjchBJmZmfDw8DC5H6rUVXM4HA5q165doX3Y2NjQh0IlQo935UKPd+VCj3flQo935VLa422qhU4HdZSgUCgUCoVCqQZQpY5CoVAoFAqlGkCVOkqZEQqFmD9/PoRCoaVFqRHQ41250ONdudDjXbnQ4125VNbxZkgpfGU5HA472f7EiRPo2rVrseVPnTqFHj16aDtiGKjVapMFK09dCoVCoVAolJpGmR0l7ty5Ax6v+Op37tzRWz9//rzJ8VYK16VQKBQKhUKhGKdMljpCiMnhMQo2X5aQGrq+qKWOQqFQKBQKxThlstTpFDtTy+ooTVTkwnUpFAqFQqFQKMYp8/BrWRQuqqRRKBQKhUKhVAyl8n718vKq9MXb2xteXl4Vtf+UMpCQkICpU6eiTp06EIlEcHV1xYABA3Dq1ClLi1alyMzMxIEDBzBv3jz06dMHTk5OYBgGDMMgIiKixPoajQZr165FmzZtYGdnB7lcjuDgYCxbtgz5+fmVsAdVi5iYGPz0008YMGAAvLy8IBQKIZfLERQUhNmzZyM+Pr7Y+vn5+fj+++/RtGlTyGQy2NnZoU2bNli7dm2pRyFqAtevX8e8efPQu3dv1K1bF7a2thAKhahVqxYGDRqEf/75p9j69HiXH4VCAU9PT/a5snHjRqNl6fEuHRs3bmSPq7FFJpMZrV9hz29CoZSCO3fuEEdHRwKAACA2NjaEw+EQAIRhGLJ06VJLi1hl2LdvH3scCy/h4eHF1s3Pzyd9+/ZlywsEAiIWi9n1Fi1akMzMzEraE+snJiaGMAyjd4xtbGwIl8tl1+3t7cnp06cN1k9PTychISFsWYlEQgQCAbvev39/olQqK3mvrJvJkyfrHW+ZTEZEIpHeb0OHDiX5+flF6tLjbR6mTp2qd7z//PNPg+Xo8S49f/75JwFA+Hw+cXV1Nbj4+fkZrFuRz2+q1FFMJjs7m3h7exMAJDg4mNy/f58Qon0gzJgxg1Xsjh07ZmFJqwb79u0jLi4upG/fvmT+/Plk7dq1Jit1s2bNIgCISCQiGzduJCqVimg0GnLw4EHi4OBAAJDRo0dX0p5YP5GRkYRhGNKvXz+ya9cukpKSQgghJC8vjxw+fJj4+vqyil58fHyR+sOHDycAiIODAzl48CDRaDREpVKRjRs3sorKF198Udm7ZdVs3LiR/Pjjj+TGjRt6L6iYmBgyc+ZM9lpftGhRkbr0eJefGzduEC6XS1q1alWiUkePd+nRKXWdOnUqdd2KfH5TpY5iMj/++CP7xR0XF1dk++DBgwkA0qxZMwtIV/VQqVR665GRkSYpdfHx8UQoFBIA5Oeffy6y/Z9//mEV7Dt37phd7qpIWloauX37ttHt4eHh7MtrwYIFettu3rzJnpf9+/cXqfvTTz8RAEQsFpNXr16ZXfbqypgxYwiAItYMerzLj1qtJs2bNydcLlfveBpS6ujxLhtlVeoq+vlNM0pQTGbr1q0AgNGjR6NWrVpFts+cORMAcPPmTTx69KhSZauKcLncMtXbs2cP8vLyYGtri0mTJhXZPmjQIPj7+4MQgr///ru8YlYLbG1tERQUZHR7YGAgWrduDQC4ceOG3jbdMQwICMDAgQOL1J00aRJsbW2Rk5ODvXv3mlHq6k2LFi0AAC9fvtT7nR7v8vPLL7/g+vXr+OCDDxAcHFxsWXq8K5eKfn5TpY5iEpmZmezLrlevXgbLtG7dGra2tgBAnSYqkDNnzgAAOnbsCJFIZLBMz549AQCnT5+uNLmqOo6OjgBQJCam7njrjmlhxGIxOnToAIAe79IQFhYGAPD19dX7nR7v8vHixQvMmzcPrq6u+Prrr0ssT4935VLRz2+q1FFMIjw8nPWAatiwocEyHA4HAQEBAICHDx9Wmmw1Dd2xNXYeAKBBgwYA9M8bxTgqlQoXL14EADRq1Ij9nRDCeiKbcrzpdV88CoUCd+/exUcffYQdO3YAAD7++GN2Oz3e5WfKlCnIzMzE8uXL2Y9sY9DjXX4ePHiAhg0bQiwWQy6Xo1GjRpg2bRoiIyMNlq/o53eZ49RRahYFwz14eHgYLafbVlJ4CErZ0R1bU86DQqGAQqGAXC6vFNmqKr/++isSEhLA4XAwbtw49veMjAxkZWUBoNd9WYmLi4Onp2eR30UiEb788kt8+OGH7G/0eJePgwcPYt++fejcuTPGjBlTYnl6vMvP69evkZycDHt7e2RkZODBgwd48OAB1qxZg3Xr1mH06NF65Sv6+U0tdRST0N34gNYcbwyJRAJAezFSKgbduTDlPAD0XJTE3bt3MWfOHABaq5HuKxmg17054HK5cHV1haurKwQCAQCAx+Nhzpw5+Oijj/TK0uNddrKysvDxxx+Dz+fj119/NbmODnq8S4eHhwcWLlyI+/fvIzc3F8nJyVAoFDh06BAaNGiAnJwcjBs3DufOndOrV9HPb6rUUSiUGkt8fDwGDx6MnJwchISE4LvvvrO0SNUOd3d3JCQkICEhATk5OXj06BHGjh2L+fPno2nTpnjw4IGlRawWfPXVV4iJicG0adP0PkwoFUPPnj3x1VdfoWHDhuzHilAoRN++fREWFoa6detCpVJh9uzZlSoXVeooJiGVStn/c3JyjJbLzs4GgGIjaVPKh+5cmHIeAHoujJGSkoKePXsiMjIS9erVw6FDh4pMXKbXvXnhcDjw9/fH+vXrMX36dMTExODtt9+GRqMBQI93Wbl9+zZ+/vlneHp64quvvjK5Hj3eFYOtrS2++OILAMDly5fx+vVrdltFP7+pUkcxiYLj/4VDEBREt83d3b3CZaqp6M6FKedBJpPR+XQGSE9PR69evXD//n14eXnh5MmTcHV1LVLOxsaGfQjT6968TJkyBQBw69Yt3Lp1CwA93mVl6tSpUKvV+Oabb0AIYedi6RYdeXl5UCgUrNJAj3fF0apVKwBaZ5SCThMV/fymSh3FJAIDA8EwDAAYHS7RaDRsfDpq/q84dMe2uGErnYdV/fr1K0WmqkRWVhb69u2L69evw83NDSdPnjSaX5phGPYYmnK86XVvOgVjXT579gwAPd5lJTo6GgAwduxYyOXyIouO999/H3K5nD1u9HhXPhX9/KZKHcUk5HI5mjdvDgA4ceKEwTJXrlxBeno6AKBbt26VJltNo0uXLgCA8+fPIzc312AZ3Tmi50GfnJwcDBgwAGFhYXB0dMTJkydRr169Yuvojrex6z43Nxfnz58HQI93aShovSg4xESPd+VCj3fFcOXKFfZ/Hx8f9v8Kf36XOgcFpcaiSxMml8vJy5cvi2wfMmQIAUBCQkIsIF3Vpyxpwn755Zci2w8cOEDThBkgLy+P9O7dmwAgdnZ25MaNGybVK5hG6eDBg0W2r1y5kqZRKoQul2VxvPfeewQA4fF4bC5eQujxrghgYpowerxNo6RrOz09nfj7+xMApGXLlnrbKvr5TZU6islkZ2cTb29vNr/rgwcPCCGEZGRk6CXoPnbsmIUlrTokJSWxS8GH66VLl/S2qdVqvXq6hNBisZhs3ryZzSN76NAh4ujoWK6E0NURlUpFhg4dyn6UXLp0qVT1dQnPHR0dyaFDh9g2N23aRMRiMU14XojIyEjSrFkzsn79ehIbG8v+rlarya1bt8jo0aPZa33atGlF6tPjbV6KU+oIoce7tERGRpJWrVqRdevWkejoaPb3vLw8cuTIEdKoUSMCgHA4HHLq1Kki9Svy+U2VOkqpuH37NnvRASA2NjaEw+GwXxZLly61tIhVCt1xLGmJjIzUq5efn0/69u3LbhcKhUQikbDrLVq0IBkZGZbZKSvk7Nmz7LERiUTE1dXV6NK8efMi9dPT00lISAjbhkQiYb+2AZD+/fsTpVJpgT2zTgpanXXH3MnJSe+YASDjx483eNzo8TYvJSl19HiXDkPXt6OjI+Hz+XrHcPPmzQbrV+Tzmyp1lFITHx9PPvnkE+Ln50eEQiFxdnYm/fr1IydPnrS0aFWOsip1hGitHr///jtp3bo1sbGxIVKplDRt2pR89913JC8vr/J3xoo5c+aMycfa29vbYBt5eXnk22+/JUFBQUQqlRIbGxvSunVrsmbNmhKHY2oaeXl5ZMeOHWTSpEmkadOmxNXVlfB4PCKTyUiDBg3Iu+++Sy5cuFBiG/R4m4eSlDpC6PEuDdnZ2WTlypVk+PDhJCAggDg4OBAej0dsbW1J8+bNyeeff06ioqKKbaOint8MITQxJIVCoVAoFEpVh3q/UigUCoVCoVQDqFJHoVAoFAqFUg2gSh2FQqFQKBRKNYAqdRQKhUKhUCjVAKrUUSgUCoVCoVQDqFJHoVAoFAqFUg2gSh2FQqFQKBRKNYAqdRQKhUKhUCjVAKrUUSgUCoVCoVQDqFJHoVAoFAqFUg2gSh2FUg5CQ0PBMAy7LFiwwNIiUUpBQkIC5s+fj/bt28PFxQV8Pl/vfIaGhlpMto0bN+rJsnHjRovJQqFQrA9DzwiepYWiUCgUS7Bjxw5MmDABOTk5lhaFQqFQzAK11FVhfHx89LR0hmFQp04dKJXKcreVlpZmfoEpFCvhwoULGD16NFXoKBRKtYIqddWM58+fY8OGDZYWg0KxaqZPnw6NRsOud+jQAX/88QeOHDmCEydOsEtQUJAFpbQ+FixYYDXD0xQKpSh0abN5AAAAIRBJREFU+LUasnjxYowbNw4ikcjSolAoVkdERASuXbvGrnfu3BmnTp0Ch0O/cSkUStWGPsWqIS9evMBvv/1maTEoFKvk0qVLeusTJ06kCh2FQqkW0CdZNWXp0qVQKBSWFoNCsToeP36st96wYUMLSUKhUCjmhSp11Yi2bduy/yclJeGnn36ynDAUipVS2AnIxsbGMoJQKBSKmaFKXTXi66+/BsMw7Pry5cuRmppqQYkoFOsjNzdXb73gPUOhUChVGeooUY0IDg7G0KFDsXv3bgBAeno6li1bhiVLllhYsoojOTkZYWFhePHiBVJSUuDk5ISWLVuiadOmJdZ9+vQpLl++jJcvX4JhGHh4eKBz586oVauW2eRTqVS4fPky7t+/j5SUFNjY2MDT0xOdO3eGra2tWfpIS0tDWFgY4uPjkZSUBJFIBGdnZwQHB6NBgwZm6aMg9+/fR3h4OOLj46FQKODq6oqxY8eCz+ebvS+NRoOrV6/iyZMnSExMhFqthouLC3x9fdG2bdsy9UkIMbucpSE7OxuhoaGIjo5GSkoKbG1tUb9+fbRr186szk1xcXF48OABIiMjkZ6eDgBwcHBArVq10KZNG9jb25utr/IQGRmJBw8eICYmBunp6eDxeHBwcIC3tzdat24NmUxmaREBAGq1GpcvX0ZUVBTi4+OhVqvRsGFD9O/fv8S6jx49wp07d5CUlIT09HQ4ODjAw8MD7du3h4ODg1nlVCgUuHDhAmJjY5GUlARbW1s0a9YMrVu3LvEDJi4uDmFhYYiJiYFarYabmxvatWuHunXrlluu7OxsPblEIhFcXFzQsGHDKudlHhMTg+vXr+PVq1dITU2Fra0te6zc3NzK3X65nhGEUmXx9vYmANglNTWVPHz4kHC5XPY3qVRKEhISytSWMf7880+9sn/++Wep5C5Yt1OnTsWW7dSpk155HeHh4eTNN98kAoFAb7tuCQ4OJmFhYQbbPH36NGndurXBegzDkIEDB5Lo6GiT9uXMmTN69efPn08IISQ/P598++23xNnZ2WA/QqGQjBw5ksTFxZnUjyEOHDhAOnbsSHg8nsE+ABAvLy/yww8/kNzc3HLtj1KpJCtXriQNGzY02E9x10tZePXqFfnoo4+Io6Oj0X2Ty+Vk7NixJCoqqlT7ZOpy5swZs+5TSkoKmTx5MpFKpUb3Z9asWUShUBBCSn+fKZVKcuTIEfLuu+8WuZ8NXedt2rQh+/btIxqNpth2y3r8jL1ecnJyyO7du8no0aOJm5tbsfW5XC7p2bOn2c+FIebPn2/w/CcnJ5PPPvvMoKxBQUFG28vMzCQLFiwgvr6+xe5f586dyblz50yWc9y4cXptREZGEkIIiYuLI++88w6RyWQG+6pTpw45ePCgwTZv3bpFevbsSRiGMVi3Q4cO5N69eybLWJDbt2+TQYMGEZFIZPQ41KpVi8ybN49kZmYW21Z8fLze+83Ly6vE69cQoaGhev3369evxDp5eXnkp59+Ig0aNCj2vmrevDnZv39/qWUixDzPCKrUVWGMKWKFb/pPPvmkzG0ZwtJK3T///EPEYnGJLxQ+n0/27t2r196CBQuMPrgKLm5ubiQiIqLEfTGkBKWmppJ27dqZ9NKzsbEhhw8fLtXxe/XqFencuXOpXq7+/v7k6dOnZdqflJQU0r59+2LbN6dSt2PHDiKXy03eN6FQSFatWmXyPpm6mFORuHbtWokKjG6pX78+iY2NLfV9NnTo0DLt55AhQ9iXhDmPn+5+LUxISEiZ2vroo4+IUqksz2koFkNK3a1bt4iHh4dRmYwpdQcPHiQuLi6l2r9JkyaZtH+GlLqwsLBiP4B0C8MwZOXKlXrtrVu3jvD5/BLrymQycuHCBZOPp0ajIbNmzSIcDsfkY+Dm5kbOnz9fbLt9+/bVq3P69GmTZdIxYcIEvTZ27txZbPnLly8Xq5wbWgYMGFDsfVUYcz0jqFJXhTGmiD1//lzvJhUKhSQmJqZMbRnCkkrdhQsX9KxzIpGI1K9fnzRv3tygVUwikZDHjx8TQghZunSp3jZbW1vSpEkT0rRpU4Nftw0aNCD5+fnFylf4hTdv3jzSvXt3vd8cHR1JcHAwqV+/vsGvVZFIZPKD6fHjxwYfLgzDEB8fH9K8eXPSuHFjg/vj7OxMHj16VKr9+fLLL4sodPb29qRJkyakSZMmxNbWtsTrpTSsXbvW4EtAJpORhg0bkqZNmxI7OzuDD7qFCxeatE+mLuZS6u7du0ccHByKtC8QCEhAQAAJCQkpogA0btyYrF69ulT3Wb9+/Qye8wYNGpBWrVqRoKAg4uTkZHBfu3XrRtRqtVmPH2D49WLI2uvh4UEaNWpEWrduTRo3bsxeV4WXCRMmlPd0GKWwUrd58+Yi56VWrVokJCSE+Pv7E5FIZFCpW7NmjZ41qeCzqH79+qRly5akbt26Bq/zQYMGlWh5KqzUHT9+XO+e4PF4JCAggLRo0YLUqlWrSB8cDoe1DP799996H7lSqZQ0bNiQhISEEHt7e4PXU0pKSonHUqPRkLFjxxo8h05OTqRZs2ZGn4disZgcPXrUaNs7duzQKz9+/PgS5SlIVlaW3kejvb19sSMZBw4cMGhE0N2/LVu2JIGBgQZHTFq2bElycnJKlMmczwiq1FVhilPEPvzwQ71t7733XpnbKowllTofHx/2JbBx40aSlZXFltVoNOTYsWPEy8tLr86wYcPIhQsX2IdoixYtyMmTJ4lKpWLr5ubmktWrVxOhUKhXt/BXbWEKv/AKHsdmzZqRM2fO6D2kMzMzyZo1a4ooJm5ubiUqRllZWaR+/fp69Xx9fcmaNWuKPGh1Q3HNmjXTK9+0adNiH2CF98fV1ZX9v3v37uTixYt6L3+NRkNOnDhh8vBucdy4caOIxcDLy4vs3LlTr32VSkWOHj1qUDkwZPVMSUkhJ06cYJeePXvq1dmyZYvedt1iysurJPLz80njxo31+pPL5eTnn38maWlpemUvX75MOnbsWORaN/U+69evH3FyciIffvghOXToEElKSjJY7smTJ2TOnDlFXqg//PCDwfIFj9/bb7+tV2f58uUGj51uMUTDhg2Jl5cXmTFjBjl16hRJT08vUkaj0ZA7d+6QDz74oIiCVNj6bi4KK3W6a5/L5ZIPPviAPHnyRK98dnZ2kX08efJkEWVtwIABJDQ0tIgVLjk5mXz33XdFrNLffvttsXIWVup014mdnR356aefijxHrly5UuQabN68OXn69CmRSCQE0Fry9+7dS/Ly8th6KpWK7Nixo8izavr06SUey5UrVxa5Nzt06EAuXrxY5Hn4xx9/FFFoHBwcyIsXLwy2nZOToyeTXC7Xew+UxJYtW/T6ev/9942WvX//fhGFrkOHDuTff/8toqxlZGSQtWvX6j0zS2qfEPM/I6hSV4UpThF7+fKl3sXI4/FYi1Vp2yqMJZU6ACQgIMDoDU8IIY8ePdJTzrhcLqlXrx6r4BVnfSu8b8XNmSHEuBWjb9++xfbz+PHjIl9eH374YbF9vf/++0VeFiXNQcnNzSVvvPGGSS/v4vbn008/Lbaf8qLRaIo82IKCgoq9DnNzc0nXrl316ri7u5Ps7Oxi+zI2J6ki+O677/T6sre3L3ZukkajKaI4mXqfhYWFmWQV0HHr1i29l2mtWrVKHP4zNu+sNJw7d07vg6okjh8/rnc/t2zZstR9mkLhfdM9N3ft2mVS/dTUVL3hMw6HQ9avX19ivQcPHuiNMggEAhIfH2+0fOHrV6eAPnjwwGidpKSkIs8b3Qdihw4dSEZGhtG6p0+f1qvn7Oxc7LMtNja2iCI0bty4Yi2QkZGRRayKgwYNMlp+8uTJemX/+usvo2UL06NHD726ly5dMlhOqVSSRo0a6ZVduHBhiZbUuLg49n2jW27evGm0vLmfEVSpq8KUpIh99tlnettHjRpV5rYKYkmljs/nkzt37pTYR2EFCACpW7duiV90Go2GBAQE6NV7+fKl0fKGlCB3d/diH5I6jh49qldPLBYX+TLTERMTo2feb9KkicnWsaysLL0vOm9vb6MvVUP707Zt2zJNRi4Nx44d0+tTIpGY5KySlpZWZB7KunXriq1TWUqdSqUq8qLavXt3ifXy8/MNTsYu7X1mCuvWrdPr49ChQ8WWN4dSVxbmzp2r129xCkxZMaTUzZ492+T6had3LF261OS6hw4d0qv75ZdfGi1rSKk7cuRIiX18++23Reo5ODgUq0DqKGzdNuaERgghc+bM0SsbFBRk0lzBsLAwvaFghmGMGiLCwsL0+ujRo0eJ7ROiVbgKWlL9/f2Nlt22bZteH5MnTzapD0IIuXv3rl4/b731lsFyFfGMoHHqqjGzZ8+GXC5n13fs2IF79+5ZUKLy8+abb6JJkyYllhs4cGCR32bPng2JRFJsPYZhitS9detWqWScO3eu3nE3Rq9evdC1a1d2PScnB9u2bTNY9tdff4VKpWLXly1bBqFQaJI8EokE06ZNY9ejo6Nx/fp1k+oCwKJFiyo8ltu6dev01qdNmwYvL68S69na2mLBggV6v61du9acopWZ48eP48WLF+x6ixYtMHTo0BLr8fn8SgtDNHLkSHC5XHY9LCysUvotLWPGjNFbrww5JRIJ5syZY1JZtVqNX375hV338vLCjBkzTO6rb9++CA4OZtf37Nljct22bduid+/eJZYz9Ez88MMPTQrBMWjQIL11Y89EQgjWr1+v99vy5cvB45UcPa1NmzYYMWKEXluFnwsFy/r7+7Prp06d0rvXjLFlyxZoNBp2fdy4cUbLFgzeL5FIsHTp0hLb19G4cWO9Y7Z//36o1eoi5SriGUGVumqMo6Mjpk+fzq5rNBrMmzfPghKVnzfffNOkco0aNdJbZxgGQ4YMMalu48aN9dZjYmJMEw7am23kyJEmly/8UAkNDTVY7vDhw+z/bm5u6N69u8l9AEDPnj311s+fP29SPVdXVz3Fs6I4d+6c3vqECRNMrjtq1CiIxWJ2/caNG8jOzjabbGWl8LkcO3asyXX79esHZ2dnM0tUFKlUChcXF3a9tB8wlYWvr6/eemXI2a9fP5Ozjdy5cwcvX75k10eOHFnqGIoF79GIiAi8fv3apHqmPhP9/f0hEAj0fhs2bJhJdU19JkZERCAxMZFd9/LyQrdu3UzqAwDeeecdvfXCz4WCFLyfNBoNtmzZUmL7mzdvZv/ncDh4++23DZZLTk7G1atX2fX+/fuXOq5jwfOpUCgMXrMV8YygSl01Z/r06XB0dGTX9+/fr3exVjVCQkJMKldwnwHtS8HUm7Jw3YyMDNOEA9CkSZNSBRPt3Lmz3rqhc5Oamor79++z682aNSt1AvrCVq/w8HCT6jVv3rzCrXRRUVF49eoVu+7t7Y06deqYXN/GxgbNmzdn19VqNa5du2ZWGctC4XNZ+FwXB4/HQ7t27crc94MHD7Bw4UIMGjQI9erVg5OTEwQCARiGKbLEx8ez9UxVJMzF1atX8cUXX6Bv377w8/ODg4MD+Hx+ERkLW6UrQ86WLVuaXLbwR1LB69FUynqPmvpM5HK5egHP+Xx+kY9fY5j6TLxy5YreepcuXUr1/OjYsaOeVe/WrVvIz883WHbs2LF6z8G//vqr2LavX7+Ohw8f6snm6elpsOyFCxf0gpRX1PmsiGcEVeqqOTY2Nvj888/1fvvyyy8tJE35MdV6UXiYtTRWj8J1c3JyTK5r6kNSh5eXl541IDo6ukjGg0ePHun9dvjwYYMv5+IWqVSq12ZKSopJ8hW2kFQE0dHReuumDK8XpnBE+tJYVyuKyMhI9n8ul4vAwMBS1S9sHTGFe/fuoVOnTmjUqBEWLFiAAwcO4OnTp0hOToZSqSyxfuG8uBXF+fPnERQUhFatWmHp0qU4cuQIIiMjkZqaqjfNwBiVIWdprv3CL+zhw4eX+h796KOP9Now9R4t67PNwcFBb+jd1HqA8Wdiee9loVCod5/k5eXpffAVxNPTE126dGHXHzx4gBs3bhhtu6CVDih+6LXw+Zw1a1apz2e/fv302jB0PiviGUGVuhrAxx9/DHd3d3b95MmTRof5rJ2yplEqT/qlwkpWcRT+ojWFgpY9jUZT5Cs4OTm51G2WhC5tVElURrL7wvmJnZycSt1G4TrWkPO4oOJha2tr0ryigpT2Wvr333/RvHnzYoesSiIvL6/MdU1lzZo16NSpE+7evVvmNipDztJc+5a8R63pmVjZ93JhxWzTpk0GyymVSr35yjKZrNjpOJV1PiviGUFzv9YAxGIx5s6dq/cl+OWXX+LixYsWlKp6UpIjhiEKW9EUCoXeMElFWCUKThYujorI51oYhUKht174eJhC4TqZmZnlkskcFNwvc1wXxfH48WMMGzZMb6iKYRi0bNkSbdu2hZ+fH9zc3CASiYq8zMeMGWPUGmJuzpw5gw8++EBPKeDxeGjfvj1atWoFb29vuLi4QCQSFRly7dGjR6XIqKM0174l71FrorLv5SFDhuDDDz9k+92+fTtWrFhR5NwdPnxYb8h+2LBhxcpWWeezIp4RVKmrIbz33ntYtmwZoqKiAGi9xw4dOlTEREwpH2WZoJ+VlaW3XjiBeeGbvUuXLvjiiy9KL1wBrCWZO1B0fwsfD1MoXMcU7+OKRiqVslZXc1wXxTF79mw961XLli2xadMmk4ZzKnrOZEFmzJihp9D169cPv//+O2rXrl1svcqwzJWHwvfot99+a/JcN2M0bNiwXPUtQWXfy1KpFMOGDcPGjRsBAElJSThy5EgRT9/CFrzihl6Boufz008/Lfe70s/Pr8hvFfGMoEpdDUEgEGDBggUYP348+9vcuXPRt2/fUj/Uy/MSsAavxIqkLBO4C8614HA4RYZ9Cg9HiESiUnu/WjOFFcyyDH0UPu7WoLTa2dmxD+z09HQolcpSWX9MPQ4KhQKHDh1i111dXXH06FGTj0FlDVU/fvxYzwOwUaNG2Lt3bxGPTEOYOr/MUhS+R319favVPWoqlriXx40bxyp1gFaBK6jUpaSk6N0f3t7e6NSpU7FtFj6f7u7uFXI+K+IZQefU1SDGjBmj9+V++/Zt7N69u9TtFB6+KY0jQVJSUqn7q0oU9FI1hejoaL05dN7e3kWU5sITtp8+fVp2Aa0Qb29vvfU7d+6Uuo3CdQq3aQkKfpmr1WpERESUqr6pc85u3rypN+w6atQokxW6p0+fVpoV7PLly3rr7733nkkKHaCdBG/NVPd71FTKey/n5eXh0aNH7LpQKISrq2uxdTp16qTX77///qv3obJt2za9+2Ps2LElGiYq63xWxDOCKnU1CC6Xi0WLFun99tVXXxkMilgchS1JpZmPYw2hJiqSe/fulcqqcPbsWb11Q2EUateujbp167LrT548QWxsbNmFtDJ8fHz0HtzR0dF4/vy5yfUzMzP1ginzeLwyhSAwNy1atNBbL3yui0OlUpk857Xw/RcQEGByP6dPnza5LIAioXRK40RUmXJWNgW9MAHrl7eiaN26td56aGhoqa6R8+fP63lpN2vWrETFn2EYvfhu+fn52L59O7te2OvVlFhwlXU+K+IZQZW6GsawYcP0IpdHRESUGN+nMIW/xkoTCHTHjh2l6quqoVQq9R4oJVF4roexYYHCEeNXrVpVeuGsmML7XXA4pSS2bdumZy1u3rx5mSYdm5vCMacKv1yK49ChQyYP5Rd+aRqL62Wo3m+//WayTEDRidmlmU5RVjnz8vKwYcMGk/uxBC1bttSzjp4+fVovJlpNISAgoMgH2pkzZ0yuX/g8lzRMqsOYF2xERIReLLh27drpfSAbo1atWnrhqZ49e4YjR46YJEtpqIhnBFXqahgMw+Drr7/W+23hwoUmxbDSERAQoPfSPHHihEneQteuXcO+fftM7qeq8vXXX5vkfXns2DG9L0CxWIxRo0YZLDtt2jQ9d/dffvkFN2/eLL+wVsJ7772nt/7DDz8gLi6uxHoZGRlF0oRNnDjRnKKVmV69eqFWrVrs+rVr10xK/6RUKk1OTwWgSJqnCxcumFTvt99+w+3bt03uB0CRwNoF42yVRFnlnDdvXqV555YVPp+PTz/9lF0nhGDy5Mmleq5WBxiGwbvvvqv328yZM00aDbp69areBzHDMEWeC8aoU6cO2rdvz65fuXIFjx8/LlVsusLMnDlTb/3TTz81OcyMqVTEM4IqdTWQvn376kWijoqK0ktxUxJcLhe9evVi13NycjBr1qxi6zx79gzDhw8v9VBvVSQ+Ph4jR44s9oH+9OnTIsMA48aNg52dncHyfn5+eg/LnJwc9O/fH5cuXSqVbKdPn8akSZNKVacy6N69u16g0qysLAwaNKjYh2h+fj7efPNNvYwI7u7uGD16dIXKaipcLhcff/yx3m8TJ04sdt4lIQTvvfeeydkEAG1GgYJDVHv37i0xN+q///6rl0LQVAp7ZJY2R2lBfv/99xLnKq1ZswbLly83XUALMnXqVD0r1YULFzBs2LBSKQJZWVlYuXJlkfypVYkPPvhAL23fzZs38f777xc7DBsTE4Nhw4bplRk0aFCpMssUVtg2btyolzpMJBJh+PDhJrf31ltv6V3vjx8/Rp8+fUr1rlQqldi0aRO+++47g9sr5BlBKFUWb29vAoBdUlNTTa4bGhqqV7fwUlJbhw8fLlJnwoQJ5MWLF3rlkpOTyY8//kgcHBwIAFKnTh29Op06dSq2n06dOumVLw2l6acgZ86c0as7f/58k8sWPCchISEkNDSUaDQatrxCoSBr164l9vb2evVcXV1JSkpKsXJlZ2eT4OBgvXo8Ho+MHz+eXLp0iSiVyiJ1MjMzyfnz58kXX3xBAgICWBnNse/m5saNG4TP5+v17+PjQ3bv3k3y8vLYcmq1mhw/fpw0bty4yDV45MiREvsZN26cXp3IyMgK26e8vDzSoEEDvf7kcjlZuXIlSU9P1yt75coVvevdx8dHr96ff/5ptJ9Ro0bplbWxsSFr1qwhOTk5euUeP35MPvjgA8LhcAgA4uLiQhwdHfWu3+LIz88nTk5Oen116dKF/Pbbb+TQoUPkxIkTekth2rRpo1fX3d2d7Ny5s8i1e/v2bTJ8+HC2XP369ct8P5vK/Pnz9fo4c+ZMqds4d+5ckWu4Vq1aZPny5SQ6OtpgnZiYGLJr1y7y1ltvERsbmxLvu/JcvwWfTyWd64JERkbq9Tlu3Lhiy//yyy9F7s3OnTuTS5cu6ZVTKBRk3bp1Ra4pBweHIu+SkkhPTydisZhtQyQS6bU5cuTIUrVHCCERERHE1tZWrx17e3vy1VdfkUePHhmsk5CQQA4ePEgmTZpEnJ2dSzxe5n5GUKWuClMepY4QQnr06FFmpY4QQvr371+kHsMwxN/fn7Rs2ZLUrVuXfXkAIFKplNy4caNaK3Xz5s0rclwdHR1Js2bNSIMGDfQeOrpFKBQafAEaIjY21qAyozu+gYGBpFWrVqRx48akdu3ahGGYIuWsVakjhJC1a9fqXTMFH3KNGzcmTZs2LaIQ65aFCxea1EdlKnWEEHLnzh2DMguFQhIYGEhCQkKIi4uL3rYmTZqQ1atXm6zUPX36lFUICi4ikYg0adKEtGjRgtSuXVtvG5fLJYcPHy71i37RokVGnxuFl8JcvHixiNIDgMhkMhIcHExCQkKIq6trkev65s2bZb6fTcUcSh0hhGzbtq2IQqFb3N3dSVBQEGnZsiUJCAgwei1XdaVOo9GQt99+2+C+OTs7F/s8FIvF5OjRoybLVpDCHzcFF1M++Axx+vRpo+fJycmJNGrUiLRq1YrUr1+fVeIKLyUdL3M+I6hSV4Upr1J39epVozeAKW0lJyeTFi1amPRwd3BwIGfPniWElE7ZqmpK3fz580laWhpp3769ScdFLpeTgwcPlmq/FAoFGTNmjEGFzZSlQ4cOZtn3imLHjh1ELpebvD9CoZCsWrXK5PYrW6kjRPuFXfihbGypX78+iY2N1T6gCz2wi+PYsWNEJpOZ1IdIJCLbt28nhJT+Ra9SqciYMWNM6scQGzduNKjYGVrs7e1Z5aqs97OpmEupI4SQ69evE39//zLdn1wul6xZs8Zo21VBqSNEq9jNnDnT4EeascXNzY2cP3/eZLkKc/ToUYPturu7E5VKVeZ2nz59avK7rvDCMAyZO3duiX2Y6xlB59TVYFq0aIHBgweXub6DgwPOnDmDuXPnFokkroPH42HMmDG4d+8eOnbsWOa+qhK2trY4ffo0li5dajT3oUAgwPDhw/Hw4UP079+/VO1LpVL89ddfuH37NkaNGmV0Hl5BAgMDMXXqVISFhZUrN2hlMHz4cDx9+hQfffRRkYn5BZHL5Rg7diwiIiKKJEO3Nlq2bInw8HBMnDjRqGeuXC7HzJkzce3atRIzLBiiZ8+euHbtGgYMGGC0DI/Hw7Bhw3Dnzh2MGDGi1H0A2nlAf/31Fy5cuICPPvoILVu2hJOTU5G0XsYYN24czp07V+zzQCQS4Z133sGDBw+KeAhWBUJCQvDw4UNs3rwZrVu3BpfLLba8UChE165dsXz5csTGxlrlvNfSwjAMvv/+e9y8eRMDBw4s9vrw8PDAvHnz8OTJEz2Hh9LSvXt3eHh4FPn9rbfeKvEcFEedOnVw9epVHDhwAF27di0xzAqXy0WbNm2waNEiPH36FIsXLy6xD3M9IxhCShFEhkIxQn5+Ps6fP48nT54gOTkZIpEIderUQadOnawiur+lUKlUuHTpEu7du4fU1FTY2Nigdu3a6NKli0nK2P/au5cVhWEwDMPfThAPXdhegAqxunEdvHcXKhQXXoBuPGFBIqJgRMQyq4FZzqHMDOF9biCBlvBCIP9nFEWhxWKh5XIp55yu16uq1aqiKFKn01G/31ccx6Ws9dter5fm87lWq5WOx6OKolAcx2q327LW/sps2rLdbjeNx2NtNhudz2c1m02laarRaPSjIesf5XmuyWSi/X4v770ajYa63a6staX9d2VYr9eazWbK81yPx0NRFMkYI2vtv3iWpiyXy0VZlulwOMg5p+fzqXq9riRJ1Ov1ZIwp7dv/V957TadTbbdbOedUqVSUJIkGg4GGw+Ffb+9LvPfKsky73U6n00n3+121Wk2tVkvGGKVp+q25t+9+ckYQdQAAAAHg+hUAACAARB0AAEAAiDoAAIAAEHUAAAABIOoAAAACQNQBAAAEgKgDAAAIAFEHAAAQAKIOAAAgAEQdAABAAIg6AACAABB1AAAAASDqAAAAAkDUAQAABICoAwAACABRBwAAEIA3jNF37i+Td+oAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import os\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "x = np.arange(51, 101)\n",
    "\n",
    "random_line_, = ax.plot(100-x, random_means_[51:101], color=\"gray\")\n",
    "random_line_.set_label(\"Random\")\n",
    "ax.fill_between(100-x, (random_means_-1/2 * random_stdevs_)[51:101], (random_means_+1/2 * random_stdevs_)[51:101], facecolor=\"gray\", alpha=0.2)\n",
    "\n",
    "loo_line_, = ax.plot(100-x, loo_means_[51:101], color=\"C1\")\n",
    "loo_line_.set_label(\"LOO\")\n",
    "ax.fill_between(100-x, (loo_means_-1/2 * loo_stdevs_)[51:101], (loo_means_+1/2 * loo_stdevs_)[51:101], facecolor=\"C1\", alpha=0.2)\n",
    "\n",
    "shapley_line_, = ax.plot(100-x, shapley_means_[51:101], color=\"C0\")\n",
    "shapley_line_.set_label(\"Shapley\")\n",
    "ax.fill_between(100-x, (shapley_means_-1/2 * shapley_stdevs_)[51:101], (shapley_means_+1/2 * shapley_stdevs_)[51:101], facecolor=\"C0\", alpha=0.2)\n",
    "\n",
    "derdava_shapley_line_, = ax.plot(100-x, derdava_shapley_means_[51:101], color=\"red\")\n",
    "derdava_shapley_line_.set_label(\"DeRDaVa\")\n",
    "ax.fill_between(100-x, (derdava_shapley_means_-1/2 * derdava_shapley_stdevs_)[51:101], (derdava_shapley_means_+1/2 * derdava_shapley_stdevs_)[51:101], facecolor=\"red\", alpha=0.2)\n",
    "\n",
    "ax.set_xlabel(\"Number of data removed\", fontsize=30)\n",
    "ax.set_ylabel(\"Expected Accuracy\", fontsize=30)\n",
    "ax.tick_params(axis='both', labelsize='xx-large')\n",
    "\n",
    "ax.legend(fontsize='xx-large')\n",
    "\n",
    "plt.show()\n",
    "PATH_TO_IMAGES = \"./images/point_addition_removal/\"\n",
    "if not os.path.isdir(PATH_TO_IMAGES):\n",
    "    os.makedirs(PATH_TO_IMAGES)\n",
    "fig.savefig(os.path.join(PATH_TO_IMAGES, \"par_remove_highest\"), dpi=300, bbox_inches='tight')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}