code/07-Gradient-Boosting-Machines.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Gradient Boosting Machines\n",
    "\n",
    "Gradient Boosting Machines (GBM) are ensembles of _sequentially_ trained shallow decision trees. The basic premise is that of **boosting**, that is an ensemble of simple classifiers can perform better than a complex classifier. The next tree utilised the current error of the ensemble to better the performance of the model. It is named **gradient** boosting since it utilises gradient descent to minimize the error of the ensemble.\n",
    "\n",
    "The gif below illustrates the learning of the GBM in a regression setting. [More info here.](http://uc-r.github.io/gbm_regression) In the beginning the errors are large. With each iteration, the next tree tunes the predictions where the ensemble still lacks in performance."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"./misc/Gradient_Boosting_Machine_Viz.gif\" width=\"400\" align=\"center\"/>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Advantages of the gradient boosting machines are\n",
    "* Modern libraries, e.g. `lightgbm` allow the user to calculate feature importance for the different features. See the documentation for the libraries for the methods of calculation of features importance.\n",
    "* Abundance of parameters that can be tuned in order to optimse the fit to your sample.\n",
    "\n",
    "Whereas disadvantages include,\n",
    "* Risk of overfitting, since the model is trying to minimize even the smallest error. Risk can be alleviated by regularization. [1]\n",
    "* Slow considering that we need many weak learners to achieve high performance of the ensemble model. [1]\n",
    "\n",
    "In this notebook, we utilise Microsoft's implementation of GBM: `lightgbm`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comment on parameters\n",
    "\n",
    "As stated above, there is an abundace of parameters to tune. \n",
    "\n",
    "For `lightgbm`, some parameters to tune in order to deal with overfitting include [4]\n",
    "\n",
    "* `num_leaves`, i.e. the maximum number of leaves in one tree. Should be **small**\n",
    "* `min_data_in_leaf`, self-explanatory. Should be **small**.\n",
    "* `lambda_l1` and `lambda_l2`, regularization parameters. Hihger value means higher regularization.\n",
    "\n",
    "The authors also state that to achieve good performance one should modify the `max_depth` parameter.\n",
    "\n",
    "In order to decide which parameters to tune, we need to determine what we want to achieve with the tuning. This is evalauted after an intial fitting of the model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "References: \n",
    "\n",
    "[1] http://uc-r.github.io/gbm_regression\n",
    "\n",
    "[2] https://sefiks.com/2020/05/13/xgboost-vs-lightgbm/\n",
    "\n",
    "[3] https://lightgbm.readthedocs.io/en/latest/Parameters-Tuning.html"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Preprocessing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Utilise the custom function importData to load the data. We utilise `pyarrow.feather` to read the data faster than with a csv. See the docs for more information: https://arrow.apache.org/docs/python/feather.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# External\n",
    "import matplotlib.pyplot as plt\n",
    "from lightgbm import LGBMClassifier\n",
    "# Internal\n",
    "import utils.nicer_plots\n",
    "from utils.import_data import importData\n",
    "from utils.model_tester import ModelTester\n",
    "# Other notebook settings\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The file `data2.feather` was prepared beforehand and need to be for your case as well if you want to read the data in feather format."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of samples:    102400\n",
      "Number of features:   3851\n"
     ]
    }
   ],
   "source": [
    "file_data = \"data2.feather\"  ## Note: Change to data2.csv if .csv file instead\n",
    "file_labels = \"labels2.csv\"\n",
    "n_features = -1\n",
    "directory = \"/data/\"\n",
    "\n",
    "data, labels = importData(file_data, file_labels, directory, n_features, iterations=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "# Model Definition - Lightgbm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LGBMClassifier(verbosity=-1)\n"
     ]
    }
   ],
   "source": [
    "mt = ModelTester(data, labels, LGBMClassifier, verbosity = -1)\n",
    "print(mt.model(**mt.parameters))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lets run an initial test to know what direction to proceed in hyper-parameter tuning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "mt.training_loop(sampling='under')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Metrics on the Training Set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Results: \n",
      "\n",
      "variable                             average    std  95.0% PI            min     max\n",
      "---------------------------------  ---------  -----  ---------------  ------  ------\n",
      "weighted accuracy [%]                  99.94   0.11  (99.87, 100.02)   99.66  100\n",
      "duration [s]                           11.6    0.17  (11.48, 11.72)    11.37   11.89\n",
      "infeasible_percentage [%]              84.63   3.12  (82.4, 86.87)     76.62   87.85\n",
      "infeasible_guessed_percentage [%]      84.66   3.08  (82.45, 86.86)    76.78   87.85\n",
      "feasible_recall [%]                    99.89   0.22  (99.73, 100.04)   99.32  100\n",
      "feasible_precision [%]                100      0     (100.0, 100.0)   100     100\n",
      "infeasible_recall [%]                 100      0     (100.0, 100.0)   100     100\n",
      "infeasible_precision [%]               99.97   0.06  (99.93, 100.02)   99.79  100\n",
      "auc of roc                              0      0     (0.0, 0.0)         0       0\n",
      "\n",
      "Results saved to ../models/results/lgbm/train\n"
     ]
    }
   ],
   "source": [
    "mt.print_essentials(test = False, results_path = \"../models/results/lgbm\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Metrics on the Test Set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Results: \n",
      "\n",
      "variable                             average    std  95.0% PI          min     max\n",
      "---------------------------------  ---------  -----  --------------  -----  ------\n",
      "weighted accuracy [%]                  50.32   0.38  (50.05, 50.59)  49.82   50.98\n",
      "duration [s]                           11.48   0.17  (11.36, 11.6)   11.24   11.77\n",
      "infeasible_percentage [%]              82.56   2.4   (80.85, 84.28)  77.25   85.45\n",
      "infeasible_guessed_percentage [%]      97.68   2.7   (95.74, 99.61)  91.41  100\n",
      "feasible_recall [%]                     2.84   3.03  (0.67, 5.0)      0       9\n",
      "feasible_precision [%]                 20.74  15.48  (9.67, 31.81)    0      50\n",
      "infeasible_recall [%]                  97.8    2.65  (95.9, 99.69)   91.5   100\n",
      "infeasible_precision [%]               82.66   2.33  (80.99, 84.33)  77.47   85.42\n",
      "auc of roc                              0.47   0.01  (0.46, 0.48)     0.44    0.5\n",
      "\n",
      "Results saved to ../models/results/lgbm/test\n"
     ]
    }
   ],
   "source": [
    "mt.print_essentials(results_path = \"../models/results/lgbm\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There is clear **overfitting** to the training data, with 99.94% in weighted accuracy. We thus want to reduce overfitting."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In order to reduce overfitting and improve performance, we utilise the information from the introduction to the notebook. The default parameters of these are given below"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'boosting_type': 'gbdt',\n",
       " 'objective': None,\n",
       " 'num_leaves': 31,\n",
       " 'max_depth': -1,\n",
       " 'learning_rate': 0.1,\n",
       " 'n_estimators': 100,\n",
       " 'subsample_for_bin': 200000,\n",
       " 'min_split_gain': 0.0,\n",
       " 'min_child_weight': 0.001,\n",
       " 'min_child_samples': 20,\n",
       " 'subsample': 1.0,\n",
       " 'subsample_freq': 0,\n",
       " 'colsample_bytree': 1.0,\n",
       " 'reg_alpha': 0.0,\n",
       " 'reg_lambda': 0.0,\n",
       " 'random_state': None,\n",
       " 'n_jobs': -1,\n",
       " 'silent': 'warn',\n",
       " 'importance_type': 'split',\n",
       " '_Booster': None,\n",
       " '_evals_result': None,\n",
       " '_best_score': None,\n",
       " '_best_iteration': None,\n",
       " '_other_params': {},\n",
       " '_objective': None,\n",
       " 'class_weight': None,\n",
       " '_class_weight': None,\n",
       " '_class_map': None,\n",
       " '_n_features': None,\n",
       " '_n_features_in': None,\n",
       " '_classes': None,\n",
       " '_n_classes': None}"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "LGBMClassifier().__dict__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Note:** Tried to suppress warnings using `verbosity=-1` but apparently a bug. Follow the issue here: https://github.com/microsoft/LightGBM/issues/4518"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0%|                                                                         | 0/27 [00:00<?, ?it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  4%|██▍                                                              | 1/27 [00:29<12:56, 29.86s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  7%|████▊                                                            | 2/27 [01:00<12:35, 30.22s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 11%|███████▏                                                         | 3/27 [01:32<12:27, 31.14s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 15%|█████████▋                                                       | 4/27 [02:03<11:56, 31.13s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 19%|████████████                                                     | 5/27 [02:37<11:43, 31.98s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 22%|██████████████▍                                                  | 6/27 [03:10<11:24, 32.58s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 26%|████████████████▊                                                | 7/27 [03:41<10:39, 31.99s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 30%|███████████████████▎                                             | 8/27 [04:13<10:05, 31.88s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 33%|█████████████████████▋                                           | 9/27 [04:45<09:38, 32.12s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 37%|███████████████████████▋                                        | 10/27 [05:39<11:00, 38.86s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 41%|██████████████████████████                                      | 11/27 [06:34<11:40, 43.78s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 44%|████████████████████████████▍                                   | 12/27 [07:36<12:19, 49.29s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 48%|██████████████████████████████▊                                 | 13/27 [08:29<11:45, 50.41s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 52%|█████████████████████████████████▏                              | 14/27 [09:26<11:20, 52.33s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 56%|███████████████████████████████████▌                            | 15/27 [10:25<10:52, 54.41s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 59%|█████████████████████████████████████▉                          | 16/27 [11:21<10:01, 54.70s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 63%|████████████████████████████████████████▎                       | 17/27 [12:13<09:00, 54.06s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 67%|██████████████████████████████████████████▋                     | 18/27 [13:07<08:04, 53.87s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 70%|█████████████████████████████████████████████                   | 19/27 [14:20<07:58, 59.82s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 74%|███████████████████████████████████████████████▍                | 20/27 [15:42<07:45, 66.53s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 78%|█████████████████████████████████████████████████▊              | 21/27 [17:03<07:04, 70.81s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 81%|████████████████████████████████████████████████████▏           | 22/27 [18:20<06:03, 72.67s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 85%|██████████████████████████████████████████████████████▌         | 23/27 [19:39<04:57, 74.43s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 89%|████████████████████████████████████████████████████████▉       | 24/27 [20:59<03:48, 76.18s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 93%|███████████████████████████████████████████████████████████▎    | 25/27 [22:13<02:31, 75.50s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.01, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.01\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 96%|█████████████████████████████████████████████████████████████▋  | 26/27 [23:29<01:15, 75.67s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.1, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████| 27/27 [24:47<00:00, 55.07s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "   Ranking    num_leaves    max_depth    lambda_l1    verbosity    balanced_accuracy    Standard deviation\n",
      "---------  ------------  -----------  -----------  -----------  -------------------  --------------------\n",
      "        1            31          500        0.001           -1             0.506215            0.00747652\n",
      "        2            31           50        0.001           -1             0.506215            0.00747652\n",
      "        3            31           -1        0.001           -1             0.506215            0.00747652\n",
      "        4            20           50        0.01            -1             0.503425            0.00447985\n",
      "        5            20          500        0.01            -1             0.503425            0.00447985\n",
      "        6            20           -1        0.01            -1             0.503425            0.00447985\n",
      "        7            31           50        0.01            -1             0.502841            0.007407\n",
      "        8            31           -1        0.01            -1             0.502841            0.007407\n",
      "        9            31          500        0.01            -1             0.502841            0.007407\n",
      "       10            20          500        0.1             -1             0.501981            0.00472375\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'num_leaves': 31, 'max_depth': -1, 'lambda_l1': 0.001, 'verbosity': -1}"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sampling = 'under'\n",
    "\n",
    "## More info here https://lightgbm.readthedocs.io/en/latest/Parameters-Tuning.html\n",
    "parameters = {\n",
    "    'num_leaves': [10, 20, 31],\n",
    "    'max_depth': [-1, 50, 500],  ## -1 refers to unlimited depth\n",
    "    'lambda_l1': [0.001, 0.01, 0.1],\n",
    "    'verbosity': [-1]}\n",
    "# Run the hyperparameter tuning\n",
    "mt.hyperparameter_tuning(n_train = 8, n_test = 2, n_epochs = 5, score = \"balanced_accuracy\", update_parameters = True, sampling = 'under', **parameters)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Re run the training loop using the best hyper parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n",
      "[LightGBM] [Warning] lambda_l1 is set=0.001, reg_alpha=0.0 will be ignored. Current value: lambda_l1=0.001\n"
     ]
    }
   ],
   "source": [
    "mt.parameters = {\n",
    "    'num_leaves': 31, \n",
    "    'max_depth': -1, \n",
    "    'lambda_l1': 0.001, \n",
    "    'verbosity': -1\n",
    "}\n",
    "mt.training_loop(sampling = 'under')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And print the metrics using those parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Results: \n",
      "\n",
      "variable                             average    std  95.0% PI           min    max\n",
      "---------------------------------  ---------  -----  ---------------  -----  -----\n",
      "weighted accuracy [%]                  50.62   0.75  (49.69, 51.55)   49.93  51.87\n",
      "duration [s]                           13.87   0.61  (13.12, 14.62)   13.25  14.99\n",
      "infeasible_percentage [%]              81.47   2.67  (78.17, 84.78)   77.25  85.06\n",
      "infeasible_guessed_percentage [%]      96.62   3.05  (92.83, 100.41)  91.5   99.95\n",
      "feasible_recall [%]                     4.36   4.19  (-0.84, 9.57)     0     11.5\n",
      "feasible_precision [%]                 18.16  11.49  (3.89, 32.42)     0     32.5\n",
      "infeasible_recall [%]                  96.88   2.78  (93.42, 100.34)  92.23  99.94\n",
      "infeasible_precision [%]               81.69   2.49  (78.6, 84.79)    77.64  85.04\n",
      "auc of roc                              0.47   0.01  (0.45, 0.48)      0.45   0.48\n"
     ]
    }
   ],
   "source": [
    "mt.print_essentials()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We notice two major things:\n",
    "\n",
    "* The prediction interval of the balanced accuracy contains 0.5, meaning that with 95% confidence we can say that the model did not perform better than chance.\n",
    "* The percentage of points guessed as infeasible is still high, even including regularization\n",
    "\n",
    "Lets plot these results as well"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Weighted accuracies:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 450x300 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test accuracies: [0.506215  0.5093722 0.5082378]\n",
      "Confusion matrices:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1050x900 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Percentage infeasible on test and train data:\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1cAAAFLCAYAAAA6SXY5AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8/fFQqAAAACXBIWXMAAAuJAAALiQE3ycutAACeVklEQVR4nOydeXhTVfqA35Om+0IpZSu0FpCtIIIosqkIgiujoAgIKm4MorjhzKA4LiMibozjKPjTUZlBWUQddGRGBcUNUBR0RmQVBFrKVkpbujfJ/f2RhaRN2rQ9IU36vc+Tp8ldv7znNCdfzrnnKsMwEARBEARBEARBEBqHKdgBCIIgCIIgCIIghAOSXAmCIAiCIAiCIGhAkitBEARBEARBEAQNSHIlCIIgCIIgCIKgAUmuBEEQBEEQBEEQNCDJlSAIgiAIgiAIggYkuRIEQRAEQRAEQdCAJFcCSinD8cjUdLyuSqn3lVLHlFI2x7H76ji2bpRSjzrim1Jt+ef1daKUmuLY51E/z7HIsXxYQ+PXhVIqUSl1WCn1QbBjqQ9KqY5KqXKl1EvBjkUQhKaHtG/SvjW19k2nG6XUS0qpMqVURuMjE3QhyZWgFaWUCXgH+A2wA3gT+DuQH8y4QomGNHwaeABoAzxaLRavDeqpQik1zHH+Rd7WG4aRA7wGTFVKdT+lwQmC0KyQ9q3xSPumnScBBcwJdiDCSczBDkAIOzoBfYCvDMM4P9jB+MGLwDLgoIZj/RP4Bsjzc/sHgHnAfg3nbjBKqTbAPcDHhmFsDmYsDeQpYBrwJ2B8kGMRBCF8kfZN2jcdaHNjGEaOUmoxcItS6inDMH5udHRCo5HkStBNB8ffPUGNwk8Mw8jD/8airmMVAoX12P4gehq9xnILEIv9F9iQwzCM/UqpL4AxSqm2hmEcDnZMgiCEJdK++b+9tG8+CICbxcCtwG+BuzQeV2ggMixQ8Imju3yvUsqklLpfKbXdcX1LrmOcb5LbtplKKQP4wrHoRrex7ouqHfdcpdQKpdRBpVSlUipHKfU3b2OGlVLJSqk7lVIfOWIpV0odV0p9ppS62kfcUUqp25VSG5VSRx3jkfcrpT5RSk2rtq3X8eLVtpmslNqklCp1HO8fSqmOXrar1xCD6uOu3Rxe4NjkVzeHhmObDx2vL/NxzEjH2PJKxy92dcWggNuAEuD9aus+B95wvHzEPRYv4+t7Ot5PtlKqwhHDMqVULx/nvVQp9bGj7CscdeEbpdRcpVSc0w+w1rHLjdXO/2i1Q74FRAJTEARBqANp31zbSPvmo31TbsPSlVLtHOWYo5SyKKXucWzTXin1O6XUWrf276jjvVzojxu35XvdXNyklPrBUb55Sqk3lVJpPt7qV9h7wSYrpWLq8iIEHkmuBH94E3gM2A18hP1L7HTgfceHF0Ax9l+GPna83u14/Xfga+eBlFLTgfXAWOwfBiuBY9h/XfpeKdWz2rkHAn8FegK/YB+a8D/gPOAdpdRDXuJ9C1jg2Oc7xzn2AGcB99fzvd8P/MPx/t4HTgDXA996aywbidOhs+flXU46dP7q9rLj720+jnEV9rHl7xuGccSPc2ZhH+ryrWEYpdXWfQSsczz/b7VYfnFupJS6CvgBuBF7WX4A/ApcC2xUSnkMn1FK3QH8GxiBvVzecxy/DfbhEimOTb/Ge336O/BjtVidSdjlfrxnQRAEJ9K+Sfvms31z0Bq768uBDcB/AOfxrgSext6r+TP2MtwNXAZ8qpS6wY84PVBKzQP+D3vdWQWUA5OAz7wlT4ZhOBP/lsDg+p5PCACGYcijmT8Aw/HI9LF8J5DutrwN9g9zAxhWbZ9hjuWLvJxnIGABcoGzq627xbHfN9WWdwIGezlWF2Cf43inVdvecKxLrbZPJHBetWWPOrafUm35547lVcBlbsvNwCLHug+r7TPFsfxRP8/hPE51h85zZ3p53ybHe6sC2nlZ/4lj35F+lv00x/ZP+ljv9T25rc/E3mgWA6OqrbsEqMT+JSPKbfk+wAYM8HK8c4F4f+qTl32PYm+EYoL9PyUPecijaTykfZP2rRHtm7O8Dew/AtZoW4AzgDO8LD8bKHA84qut8+Vmr2P5EaC32/IE7Ne71fDsts2djvWPBft/Th6G9FwJfnGXYRjZzheG/RejBY6XF3jfxSuzgAhgmmEY37uvMAzjNew9Hucqpfq5Lf/VMIz11Q9kGMZu7LPjRGCfuclJquPvZsM+3tx9nyrDML6qR7wAbxuG8W+3Y1iAe7EnE5cppTrV83iNwjAMG/AK9kbwJvd1jlguwv7FYI2fh+zj+Lu9gSHdA8QDDxqG8Um1WD8CFgLpePYopQKFhmFsrH4wwzC+NQyjpIGx7ACisf+iKwiC4A/Svp08hrRv3qkAZhiGUV59hWEYPxmG8ZOX5d8DLwEtgOH1PN/DhmFscTtWMfCs46WvOul8j33reS4hAEhyJdRFFd4/yHY4/voaA+yBsk9hOwJ7V/p/fGzmbBgGVN9XKTVCKfWwUmqhUuoNxzj3axybdK0WVwlwuVJqplKqnT/x1cKy6gsMwziOfUiBAoY28vgN4TXs5XKr27AVsF/QqoC/GYb9pyw/cI5bb+hUwqMcf9/1sd5bmW4Ckh3j17MaeF5vON9Da43HFAQhfJH2rRrSvnlls2EYB3ytdFwHN1op9bhS6v8c11Qtwt7zBZ5l6A/e6lBddVLavyaEzBYo1MUhx69Z1Tnh+Bvt53FSsXdtA1R6fmZ63RYApVQH7L/4nVXL9onOJ4ZhFCmlbgb+hv2XnmeVUruxj0deahiGv794OdnnY/lex1+/Gl+dGIZxSCn1PvbGdzj2cd3OX/osnLxI1x9aOP6eqHUr32Q6/ub4W6bYr2dYiX2ozC1KqSPYx76vBJYZhlHZwFiKHH+TG7i/IAjNC2nfvLPX8be5t29OfE6ZruyTNn0AdK5l/8Ra1nkj28uyuuqktH9NCEmuhLqwaTqOs5e0GN+9HE7c79PwN+wNz0rs9zPaARQZhmFVSo3CfoGxR0tmGMbbSqlPsQ+nGIm9G/1m4Gal1BLDMCY18r00BRZib3xuAz4FrgDaA+8ZhnGoHscpcPxNqm2jWnCWa13T3H7rfGIYxhZHgzQKuBR7+YxxPP6glBpiGEaB16PUjrMhbci+giA0P6R9a5o0lfbNSY3hgG6swJ5YvYo97t1AsWEYNqXUVOwTU9SabVfHMTyyvkj714SQ5Eo4VeRh/4CyATf5062vlIrH/gX8MHCNYRjWapuc7mtfwzCOYf+F6w3HsYZi/xC8Tim1yDCM1X7GfRr22Zu8LQf7xcunHMMwPlNK7cR+b6dUYKpj1Sv1PJRzxqWUWrfyTQ72i69nOpz7hWEYFcC/HA+UUp2xX+R7HvA7YHYDYmnp+Hu0AfsKgiA0FGnfNNKE2rdaccz+2BPYZBjGVC+b+CzDACDtXxNCrrkSTgmOoRefY/8FaYSfu7XAXkcPeml4ACbU4/xfA0sdL3v7ux/26cQ9UEq1wD4THpycylUnzmFxdf348X9AFPAQcDH2oRz+NqpO/uv426OBsTjPN6ae5/XAMIw9nLxg1718/HUB9vdQAWxrTCyCIAj1Qdq3ehFK7VtdOBOaGsP4lFJR2KfkP1U4J3L68RSeU/CBJFfCqeQJ7L/svVH95nkASqlWSqk7lFKxjkWHsXdx91ZKnee2nVJKPYC9l6P6Mfoppa52fLC5L4/j5Cw73sYz+2K8Uupit+OYgT9jH0P9H0dSoBvnr4Xd69huEfZfS+/G/r/8twYMJ3DeFHOAj/V1xfIcUIZ97H+Nm14qpRKUUjcqx00plVJxSqm7lFItq2+L/b4g4Fk+frlQSnUBWgEbvc3oJAiCEGCkffOPUGrf6uIX7GU+XCnlSuCUUpHA89hHdZwqnO/xi1q3Ek4JMixQOGUYhvG1st9A9kVgrVLqZ+z3GDEDGdhv+BeJ/SaJZY5x508BTzq2/xx7l/dZ2D+0ngNmVjvNacA7wAml1PfAQewXGg/FPjRgA9Xu1F4HrwD/UUp9if2D+Fzs46sPYZ+YIRB8gP2GvEuUUp8AhQCGYdzqvpFhGPlKqbeBG7Bf6Pt6fU9kGMYOpdQu7FMExxk1b7T4DfahFdc4/O/B3pi8bhjGesMwflFKTQSWYL/p5S/Ye44s2Mu0JxAH9MM+hDAK+Av2ZOwHx/HMjvVdsJfvc27x7VVK/Q84Wym1Efv1ClbgA8MwPnCL80LH31X1dSAIgtBYpH3zm5Bp3/w4/hGl1KvAb4EflVKfYZ94YhD2H/texH7/qYDimFXxAuzJep1xC4FHeq6EU4phGC9jv7ne37E3Cpdj/4XODCzGfuFqodv284DJ2Lu6B2IfHvAr9ilOP/Ryim+wX6/zLfbxzldj/6DbBcwARhiGUVWPkJ/FPktRC+xD31pgbxzPNQxjbz2O4zeGYbyH/V4jOcBoHLPq+dj8M8ffDw3DONjAU76KPQG6ykss5djLaDX2+2dMccTSzW2b97HfT2QBjhs8Yr8fSQL2hvRaYKtj82LgDsfyFOzvbxT23q+ngDMNw/i1WhhXY7/guzP2hvYWas6udR326XsX+f+2BUEQ9CHtW92EWvvmB3dg713bib3cLgS+w96TtKmBMdeX87DfT3KxjNxoGij/bxcgCEJTQym1CvtwuksdN+1tyDFSsU/J+5VhGJfUtX1TQymVgf0LyTuGYYwPdjyCIAhC45H2zT+UUn/DPmPkGYZh/FzX9kLgkZ4rQQhRlFL9sU9lvgv7lL0NwjCMPOzj7C92HDPU+AP2oRwPBzsQQRAEofFI++YfjuuZJwNvSmLVdJCeK0EIMZRS87CP4b8M+zCO8YZhvN3IYyZivzh3o2EYoxsf5anBcRPO3djHyAfqGgFBEAThFCDtW/1QSr2IfShjd8MwfN7sWDi1SHIlCCGGUmov9vHV+4H5hmH8NbgRCYIgCELjkfZNCAckuRIEQRAEQRAEQdBAk5mKfe3atUZUVFTdG9aC1WolIiJCU0SC+NSL+NSL+NSLDp/l5eUnRowYkaQppLBF2rumh/jUi/jUi/jUS6DbuyaTXEVFRZGVldWoY+Tn55OSkqIpIkF86kV86kV86kWHz82bN+doCieskfau6SE+9SI+9SI+9RLo9i6sZgu02ep7826hNsSnXsSnXsSnXsRnaCHlpRfxqRfxqRfxqZdA+wyr5Kq4uDjYIYQV4lMv4lMv4lMv4jO0kPLSi/jUi/jUi/jUS6B9hlVy1aZNm2CHEFaIT72IT72IT72Iz9BCyksv4lMv4lMv4lMvgfYZVslVfn5+sEMIK8SnXsSnXsSnXsRnaCHlpRfxqRfxqRfxqZdA+2wyE1rowGKxBDuEsEJ86kV86kV86kV8hhZSXiepqKigvLy8UccwmUwUFhZqikgQn3oRn3rx16fJZCIhIQGlVL2OH1bJVVpaWrBDCCvEp17Ep17Ep17EZ2gh5WWnpKQEgKSkpHp/AXInMTERkymsBvMEFfGpF/GpF399VlZWUlxcTGJiYr2OH1YllZubG+wQwgrxqRfxqRfxqRfxGVpIedmxWCzEx8fXK7Eqq7LWeJworXA9FxpPVVVVsEMIK8SnXvz1GRUV1aCZBcOq5yoyMjLYIYQV4lMv4lMv4lMv4jO0kPJqOFf+/X+1rv/k1n6nKJLwpTG9iEJNxKdeAu0zrHquWrRoEewQwgrxqRfxqRfxqRfxGVpIeQlNmYiIiGCHEFaIT70E2mdY9Vzl5eWRkJAQ7DDCBvGpF/GpF/GpF/EZWkh5NZz3b+wDQHmVjfFLtgCw+JpuJMXHBDOssMJisUhCoBHxqZdA+wyrnqv6XnAm1I741Iv41Iv41Iv4DC2kvBpObGQEsZERxESe/AoUHx3pWi54cuaZZ7JmzRqv6+677z6efPLJGst9fXF94oknOP3008nMzNQZogeDBg3i888/B2DevHnccsstPrcdN24cixcvBmDJkiWMHDnStS49PZ1ffvlFW1yLFy9m5syZDdrXl8/58+dzxx13+HWMDz/8sFYXzYlAJ6phlVwJgiAIgiDUlwpL/S9a95f333+fiy++mPT0dLp06cKwYcN44YUXGj19fFNg/vz5PPDAA35tm5OTw1//+lfWrVvH3r17AxbThg0bGDZsmF/brlixguuvv97ruuzsbE4//XQA7rjjDh577LEGx2SxWHj66ae55557XMtSUlLYuXNng48J9uT2pZde8mvbyy+/nJ9//pmff/65Ueesi/fff59+/frRsWNHxo4dW+vkO7t372bs2LFkZmbSp08fV6LrJCUlhY4dO5Kenk56ejrjxo3zeawlS5aQmppKeno6GRkZnHvuuSxdutTjWI317S9hlVydOHEi2CGEFeJTL+JTL+KzcVSfLS2voMjjtdC0kfrfOKw2g9e/O/ml74YV23nl2xysNkPreRYuXMj999/P1KlT+fnnn9m9ezcvv/wy2dnZHDhwQOu5mhJWa83PkJycHJKTk2nbtm0QIgouH330EZmZmaSnp/u9j/u97Lz5rC9KKa655hoWLVrU6GP5YufOncyYMYM///nP7Nq1iy5dujB16lSv21osFq677jqGDh3KL7/8wj/+8Q/++Mc/sm7dOo/tPvvsM7Kzs8nOzmbFihW1nr9fv35kZ2ezb98+Zs+ezV133cWOHTtqbKfDZ22EVXKVmpoa7BDCCvGpF/GpF/HZOK78+/88HjPWFni8Fpo2Uv9rYhiG12nWvT1e/iaH/+w45tq3ymbwwdY8Xv4mx+9jGEbtiVhRURFz587lmWee4eqrryYpKQmAHj168Mwzz9ClSxegZq/Izp07SUlJ8TjOjBkzyMrKolevXjz22GOuL9579uxh9OjRnHbaaZx++unccMMNLhezZ8+me/fuZGRkMGTIELZssV9fVlFRwSOPPEKfPn3o1q0bd999t+t+YWBPCHv16kXXrl3585//XOt7dI/966+/pkePHrzyyiuceeaZdO3alfnz5wOwZs0axo4dy9GjR0lPT2fKlCkAfPfdd1xyySV06tSJoUOHsnbtWtexly5dysCBA8nIyKBfv3689tprrnXHjh1jwoQJdOrUic6dO3PppZe6pteuPoyxsrKS2267jYyMDM4//3x++OEH17rRo0fzxhtveH1vzp6ORYsWsWLFChYsWEB6ejpXXnklK1euZOjQoR7bL168mEsvvdTrsT755BOGDBnien355ZcDMHz4cNLT01m8eDH79+8nJSWFN998kzPPPNM1RHH27Nn079+fjIwMhg0b5pF8uA97dO6/fPly+vbtS6dOnfj973/vUU+HDh3KJ5984jXGDRs2cPbZZ7te33jjjTVe/+1vf/O6r5O3336bESNGMGzYMGJjY3nwwQfZuHEjv/76a41td+3aRXZ2Nvfccw9ms5m+ffty+eWX89Zbb9V6Dn9QSvGb3/yG5ORktm/fXmP9//73P0aMGEFGRgbdu3dn1qxZjT6nO2E1oUVhYaFc4KsR8akX8akX8Sk0Z6T+16TcYmvUDwOVVoP3t+bx/tY8v7Z//8Y+tV6jtXHjRsrLy11fpBvK9OnTSUlJcR3vuuuu429/+xvTpk1j7ty5DBs2jPfff5/Kykp+/PFHwP5r//r169m4cSNJSUns2rXLdZ3eo48+yp49e1i7di3R0dFMmzaNJ554grlz57J27VqeffZZ3nvvPXr06MHDDz9cr3uqHTt2jAMHDvDdd9+xe/duLr74Yi699FIuuugi3n77bW699VbXl93c3FzGjx/PwoULueiii1i3bh033ngj69evp3379qSmprJ06VIyMzNZv3491157Lf369eOss87ipZdeIi0tzTXMa9OmTT5vCvvRRx/x8ssvs2DBAl599VUmT57M5s2biY6O9us9TZkyhe+++442bdrwyCOPAPYEdebMmWzZsoXevXsD9sRi/PjxXo+xdetWLrjgAtfrVatWkZKSwmeffUa3bt0Ae3IE8MUXX7Bu3TrXdUF9+/ZlxowZpKam8vLLL3PTTTfx448/EhcX5/Vcn3/+OevWrSMvL4/hw4czcuRIV6LWvXt3srOzKSgoIDk52WO//v37c/DgQXJzc2nfvj3fffcdZrPZ9XrDhg384Q9/qNXVtm3bOOuss1yvW7ZsSceOHdm2bRudOnXy2NaZ9FX/kaL6sMWrrroKq9VK3759efTRR+nZs2etMYC9Z+pf//oXhYWF9OrVq8b6Bx98kKlTpzJ+/HiKi4vZunVrncesD2HVcyU3WdOL+NSL+NSL+Gwc79/Yh/dv7MPy63q7li2/rrdrudC0kfrf9MnPz6dVq1Ye9ySbMGECmZmZdOjQgeXLl9d5jCNHjvDJJ58wd+5cEhISSE1NZfr06bz33nuA/X5n2dnZHDx4kJiYGAYOHOhaXlxczI4dOzAMg27dutG+fXsMw2Dx4sXMmTOHVq1akZCQwH333cc///lPAN59910mTpzImWeeSXR0NA8//HC97glkMpmYPXs2UVFR9O3bl969e/PTTz953XbFihUMHz6ciy++mIiICM4//3wGDBjAxx9/DMDIkSPp1KkTSimGDBnChRdeyPr16wEwm80cPnyY/fv3ExkZycCBA31OUnDGGWdw9dVXExkZye233w7gOk5DiY6OZuzYsa4yzMnJYfPmzYwZM8br9gUFBX7/GPKHP/yBhIQEYmNjAfukGy1btsRsNnPnnXdSVVVV67VDDzzwAPHx8Zx22mmcd955/Pe//3Wtc8ZQUFBQY7+oqCjOPvts1q9fz/bt2+nQoQPDhw93vTYMo87EpqSkxNVD66RFixYUFxfX2LZr1660b9+e5557jsrKSr7//ns+/PBDysrKXNv861//4scff+T777+nT58+XHPNNRQVFfk8/48//khmZibdunXj+eef55VXXnFdN+eO2Wxmz549rllXBwwYUOv7qi9h1XOVlpYW7BDCCvHZOKpft5LSup3HMpmVqnFI/Wwc3upfTKRJ6mWIIPW/JjFmk18/DFRYbExa9jNV1prD+qIiFG9O6EW0ue7fnmPq2CYlJYVjx45RVVXlSrCWLVsG2BMHf677yM7Oxmq1unpHAGw2m2tY6KOPPsqTTz7JqFGjSExMZNq0aUyZMoXzzz+fW2+9lVmzZrF//34uvfRSHn/8caqqqigtLfWYFc8wDKqqqqiqquLQoUMe54qPj/cYolgXLVu2JCoqCpvNPkFIXFycx5DD6u9t1apVHjMHWiwW+vWz38R59erVPP300+zevRubzUZZWZmrl2fGjBk89dRTjBs3DpvNxuTJk5k5c6bXRLBDhw6u50opOnTowKFDh/x+T76YOHEikydP5tFHH2X58uWMGjXK5/3nkpOTvSYY3nCPF+DFF19k8eLFHDp0CKUUJ06c4NixYz72hjZt2riex8fHe/h3xlC918rJkCFD+PrrrykoKGDo0KH06tXL9XrQoEF1Jtrx8fE1rgctKirymlhGRkby5ptv8sADD7iSoIkTJ3oM43MOpYyKimL27NksX76cjRs3ctFFF3k9f9++fVm9enWtMQK88MILPP300wwaNIiOHTsyc+ZMrrjiijr38xe/kiul1PnATKAvkAE8ZhjGo37sNwmYDXQBDgAvGIbxfANjrZPc3NyATu/Z3BCfjaOu4Smf3NrvFEUSnkj9FAKBtHehi1LKrx8HYiMjuDIrlQ+25lHplmBFRyhGZ6WSHBtZy97+M2DAAKKjo1m1ahVXXXWVz+3i4+M9fq0/cuSI63nHjh2JjIxk165dHj1gTtq2bcvzzz8P2IchXnXVVQwYMICsrCx++9vf8tvf/pa8vDxuvfVW5s+fz6OPPkpsbCxfffWV18kV2rVrR05Ojut1SUkJ+fn59X7vVVVVdQ67c84m5222u4qKCqZMmcKLL77IFVdcQWRkJJMnT3YNIUtMTGTOnDnMmTOH7du3c9VVV9GnTx9GjRpV41juE4cYhsGBAwdo165dvd9Tdc466yySk5P5/PPPefvtt3n88cd9bpuVleX3tO7uCcyGDRt4/vnneeedd+jTpw8mk4lOnTrVeb2fL3bs2EHHjh1rTa7uueceCgoKmDx5MllZWTz99NMUFBR4XDPmi549e7qu7QN7D1lOTo7PHq+ePXuycuVK1+tbbrnF4zqv6phMpga/d3cyMjJ49dVXsdls/Oc//+Hmm29mx44dPr3UF3+HBSYAW4HfA36l+0qp0cBi4B/AmcBc4Cml1LQGxOmV6heXWomQ2a40YjaHVcemEGZI/RQCRJNs76oj9b9x3HJOBy7t3sr1OsqRWN1yToda9qofSUlJzJo1i9///ve89957FBUVYRgGu3bt8uh5OOOMM1i9ejV5eXkcP37clSyBPXkaMWIEs2fPdu2/d+9evvrqKwBWrlzpSh4SExNRShEREcHmzZv5/vvvqaqqIjY2lqioKCIiIjCZTFx//fXMnj3blcTl5ua6fu0fM2YMy5Yt46effqKiooLHH3+8QV9m/RlKeM0117BmzRpWr16N1WqloqKC9evXk52dTWVlJRUVFaSkpGA2m1m9erXHZBcff/wxe/bswTAMEhMTXe/NG1u2bOGf//wnFouFV155BbDfC6s+tGnTxuv08RMmTODRRx+loKCA4cOH+9x/5MiRNYYitmnThn379tV63uLiYiIiIkhJSXFN596YmULXrVvnNQF1cvbZZ3PgwAG+/PJLBg4cSFpaGjabjc8++6zGBB7eGDduHGvWrOHLL7+krKyMJ598knPOOafG9VZOtm7dSllZGeXl5bz11lt8+eWXTJ8+HbBfv/Xf//4Xi8VCaWkpTz31FOXl5VqG8L377rvk5eVhMplITEzEMAyt977yK7kyDOPfhmE8YBjGcqDCz2PPAv5pGMY8wzC2G4bxN+AlwL8bIvhB9dmubl+TJ7NdaaQ+QwGEmsg1LYFF6qcQCJpqe1cdqf8Np6zKSqXVxnV9T/ZevDKmO9ef1Z5Kq977Xd155508+eSTLFy4kKysLLp27crUqVO5+eabufLKKwG49tpr6d+/P/379+eSSy6pMTxp4cKFWK1WBg8eTGZmJtdff71rkokff/yRUaNGkZ6ezsSJE3nsscfo3r07J06c4N5776Vz58706dOHxMRE7r33XgAee+wxunXrxsUXX0xGRgZjxoxxTVc9YsQI7rnnHiZMmEDv3r1p06ZNg4ag+vNFtWPHjrz11lu88MILdO3ald69e/P8889jtVpJTExk3rx5TJ06lU6dOvHuu+9yySWXuPZ13h8pIyODUaNGMWnSJJ9DxS6++GJWrVpF586defPNN/nHP/5BTExMvd7P5MmT2bNnD506dfK4rmr8+PFs376dq6++utYfPC699FJ+/fVXj17BWbNmcdddd5GZmelzhrzhw4czatQohg4dyplnnklMTEyNYYP+YhgG77zzjmumRm9ER0fTv39/unTp4hrKN3jwYCIjI8nKygLsiamvKdG7d+/OCy+8wN13383pp5/Orl27XAkt2O+L5n6vqnfeecc1M+WyZct47733XJ9teXl5TJ061XUPrO+++4533nnH59DL+vDFF18wePBg0tPT+d3vfscrr7yi9cbsqr6/SCil9gKLahsmoZSKBEqA2w3DeM1t+QhgDZBpGIZHur5u3TrDWXD+MupvP9S6XoZdNY69e/fKsBMNlFVZXcl+XbNLCf4j9VMPOuvn5s2bt40YMaJ+H+RNmKbU3lVH6r+dwsLCen/Zku8OgaeiosLv2fhCmaqqKrp3787KlSvp06f2H03ffPNNfvjhB5577rl6n0eHz1WrVvHOO+/4nHq+OVEfn74+Y2pr7wI1riAViAQOVlvuHGKRBng0Nvn5+fTv35+IiAhsNhvjxo3jnnvucV102LZtWw4dOkRsbCwxMTEcP36cxWMyKS8vp7CkjPu/LATg2fNbkNkhjYOHDnLo0CHi4uLIz8+nRYsWWCwWSkpKaNu2LXl5eVitVtLS0sjNzSUqKoqkpCTy8vJISkrCZrNRXFxMmzZtyM/Px2KxuLaNjIwkOTmZo0ePujLdEydOkJqaSlFREZWVla5tzWYzKSkpHDlyhPj4eMxmM4WFhbRq1Yri4mIqKipc25pMJtq0aeOKOyoqioKCAlJSUigtLaW8vJz27du7nLRr146DBw96OElOTqayspLS0lLatWvHkSNHsNlsrnNER0eTkJDAsWPHajg5duyYx/usqqqiuLi4hpPWrVtTUFBAVVWVVycJCQmYTCaKiop8OmnVqhWHDx9usJOWLVtSXl5OWVmZVycxMTGusm+oE29l715P3MvemxNn2WcfPOyq5/v376dD29Y1nERERJCamlrDiXvZe3PiXvbuTtq1a8fhw4f9cuLt/8HdiT//Dy1atKjhJDU1lcLCQq9Ojhw54lFPvJW9uxNv/w9lZWVUVFTU+hlR3Un1snd30lw/I3JyPetnUlyM358R1Z00U05Je1e9LpeXl7N3795mX5crKiqIiYkhMjLSNYOi87lziJzFYiEiIgLDMFwTLdSGxWJxbWs2m7Fa7fe08nZck8mEUgqr1epzW7PZTFVVVZ3busfuvq177L62NZlMNd5nfZzU9T6dPVG1vU93J2D/AltfJzabrdb3qcOJs3x1OFm0aBGZmZn06tWLysrKWp1MmDCBcePGYbVaG+TEZrM1ysmoUaO45JJLqKioCKiT+tSThvw/6KgnJpOJiooKn+/T3YnNZmPv3r31au8C1XOVhv2C3ssNw/i32/JewBZgkGEY37jv05hf8qRnIDDk5+fL0BMNSP0MDFI/9SA9V75piu2dE6n/dhrSc+XtmmyrxUqE2V735TO68VgslrC/LrB///5UVFSwaNGiWidh0EFz8HkqqY/PptRzdRSwAO2rLXcOcK7+C5/QBCkqKpLGW2iySP0UmghBae+k/jccb8lThc1CtCRV2nD2MIQzmzZtOmXnag4+TyWB9hmQmwgbhlEFbAQurbbqMmB/9fHnQtOkVatWdW8kCEFC6qfQFAhWeyf1Xy/yxVUv4lMv4lMvgfbpV3KllEpQSvVVSvUFooB2jtdZjvUDlFLblVLu8yPOA8YopX6vlOqulLoZuAN4UvN78KDConemn+aMvze8E4RgIPVTCASh0t5J/deLPzf0FfxHfOpFfOol0D797bk6G/jB8WgP/Nbx3Dm+PA7o7vgLgGEY/wJuAG4CfgL+CMwyDONlLZE7sNoMXv8u1/V60rKfeeXbHKy2xt9krLlTUeHvLMSCcOqR+ikEiCbb3rkj9b/hWErKvDxKXc+FxqPjRq/CScSnXgLt069+McMwPgd83hHO13rDMN4CvE/er4nXvjvAf3acvBlfldXgg615lFfZuOmcNMwmRVSECZPy76Z2wkkacm8LQThVSP0UAkFTbu/ckfrfcNZ0GVHr+ksOra91vVA3kZGRwQ4hrBCfegm0z5AexFlusfHB1jwqrZ4ZaKXV4MPtx/hw+8mkSwHmCEWkSREZYSLSpDBHKEfypTCbTJhNisgIx8Nkqrbevk2k4xhmx3Zmk8ltveexI13HM7nWu85VbX1EE0z+cnNz5T4qQpNF6qfQnJH6LzRlqqqqmsV9rk4V4lMvgfYZ0snV8dIq/O3YM7D3alVZDahqmtdlOZMtsys5q5bMOV57JHIe6z0TuZPrTm4XWT2RdCWKJo/zR5pMlFkVpZVW17GbWvIXilRYbDLNryac99MQhOaI1P+Gc9HuTwGwlpax9owrABi6+Z/EtAjfe7Xt37+fvn37uu6JVp1Bgwbx5JNPMmzYsDqPVVZWxi233MK6desYPHgwS5curbFNY78v5OTkMGjQIH755Reio6MZPXo0Y8eO5aabbvK6fXp6OmvXruX000/njjvuoE2bNjzyyCNs2LCBO+64g82bNzcqHnemTZvGqFGjGDt2rLZjjhs3jt/85jdcf/31Xte7+5w1axbdu3f36UKom0B/nw3p5KplXKTPsRtREYplk84gQoHFZk+q7H9tVNkMLFaDKtdymz3xci23VVtfcz/ndifXO/arvo/Hes/l1RPDKpt9m6ZFvuuZ2SP5ckvk3HvhqvXKmZ3JXfVkz329W6JodjuOz3NVSySdCaKpCSZ/3q4JvDIrlVvO6UCEqenFG0qkpqYGOwRBCBpS/xuOOT62xrKISLPX5bq4+uqr+fbbb9m2bZvrZsy18fXXX3Prrbeyffv2gMXkzoYNG/ze9oMPPiA3N5dffvnF5/Cqxs7G1rFjR7Kzs/3e3te2gwYN8kiszjzzTJ577jkuuuiiBsW1fft2Nm/ezIIFCwB95bRixYpa17v7nDFjBqNGjWLSpElERUU16ry18cQTT/DGG29gsVi4+uqrmTdvns/yXrNmDY899hh79+6lc+fOPPvss5xzzjmA3dGVV15JXJzrMlXuvfde7rvvPq/HuuOOO3jnnXeIiooiIiKCnj178vjjj3P22Wdr8x3o2QJDOrmKMZv4TVZqjaGB0RGK0VmpJEQ17V/2rDZnomarkXxZbAaVjsTMeyLn2M+xvtLx2nP9yUSx0vHaI5F0Sy4rrc71JxPJ6nmexbFPUyVCgTni5DBNZy9d3T2CnslblFsC6K1HMCri5JBSbz1+kW7r3/rhEB9VuybwX1vzAJh6bsdgqQoLDh8+LMOihGaL1P/GYVit7Jx7cr6Rr86+moybr6b7Q9NRmnsFc3Nz+eKLL2jRogUrV6702TsRKmRnZ9OlS5dar1sJ12FsixYtYuzYsZhM/t/JSMcNgN19dujQgW7duvHvf/+bq666qlHH9cU//vEP3nvvPT777DPi4uKYMGECzz77LA888ECNbX/99VduvvlmFi9ezNChQ1myZAkTJkxg8+bNrpvvtm7dul4J0fTp03nkkUeorKzkT3/6EzfccAM///yztvcnwwLr4JZzOlBlNXjf8YU1ypFY3XJOhyBHVjcRJmXvvTAH5HZjjebQ4cO0bNXakejZTiZ61XrpqtwSQovVLSl0S9SqXImizS3R896T6EwIPXv8TiaX9kTQRrVL7bAaYLXYqLAEx5e/VFgN3vnpKN/uLyIm0kR0hIlos4kos4kYsz05tP+1L3c9IlS17Rx/zcp1DOfD3Ax6xdx/BROE5obU/5oYhoG1tNyvbXfOfZmcN993vbZVVLL/jfewVVro9uA0v44RERfj1/Ci5cuXk5WVxaWXXsrSpUs9kqvc3FwefPBB1q9fj8Vi4ZJLLuGZZ57h2muvpaKigvT0dADWrl3LO++8w65du3jttdcAKC8vJy0tjR9//JGMjAxWr17NnDlz+PXXX0lKSuK6667jgQce8CtG9x6defPmsXXrVlq2bMnKlStp1aoVTz/9NBdddBFz5szhr3/9K4ZhkJ6ezkMPPcRvf/tb3nrrLV544QWOHDlC3759efrpp+natSsAs2fP5v3336eoqIjOnTvzxBNPMGTIEMB+I97f//737Nq1i9jYWMaMGcO8efO8DmPct28fI0eOZMeOHQwaNIiXXnrJ1YObkpLCN998Q7du3Tzel3tPx7Rp08jJyeGGG24gIiKC6dOnk5+fT0REBPPmzXPtM23aNNLT05k9e3YNT6tXr+Yvf/kLACUlJV7LaePGjbzxxhsMGjSIJUuWcO211zJr1iymTp3Kpk2bsFgsnHPOOcyfP5+OHe0/sroPe1yyZAlvvPEGF1xwAW+88QYxMTHMmjXLo94MHTqUTz75xGty9cwzz5CTk8Nf/vIXDMOgW7duXHbZZR6vP/jgA3r27OmzPixZsoTp06eTkZEBwP3338/999/vNbn69NNPOfvss7ngggsAuP7663nuuef48MMPmTRpks9z+ENUVBSTJk1iwYIF5Ofn11i/dOlSnn76aY4dO0bLli154IEHmDBhQp3HrU9y3BBCPrmKMCluPifNlVy9OaEXybEyq4oOYqKj7V/WgXiaXi+gzXAbplk9kauevLklbh49gM6k0Eui6N5D6Lm+5lBSb+eqq5cvuzBwUymbFF4SNOU1kbMnbie3cyV3rqTObf9q2wYzkQvkcAhBaOpI/a+JtbS8zpkAa8NWXsH+11aw/7Xah2g5uWj3p34NJVy2bBmTJk3isssu49lnn2XPnj107twZq9XKxIkTGThwIJs2bSIqKopNmzYRHx/P22+/Xe/hT/Hx8SxcuJAePXqwbds2xo4dS69evbjyyiv9PoaTTz75hFdffZX58+fz2muvMWPGDH7++WceeughzGazR5K3atUqnn32WZYsWULXrl3561//ym233cbatWtRStG3b19mzpxJUlISL7/8MjfddBM//vgjcXFxPPDAA0ydOpXx48dTXFzM1q1bfca0fPly3nnnHTp16sQdd9zBzJkz+fvf/+73e3r55ZfZsGGDx7DAzZs3M2HCBObMmYPZbKakpIRVq1axdu3aGvuXlJTw66+/cvrppwP4LKeNGzfy448/MmbMGLZv305VVRWVlZVcd911vP7661itVmbMmMH999/PsmXLvMb63//+l3HjxrFjxw5Wr17NLbfcwuWXX05KSgoA3bp1Y+XKlV73HTp0KDNmzABwDUNdv3696zVAjx49anW1bds2evXq5Xrdu3dvcnNzKSoqIinJ89pEm81WY2pzwzA8epry8/Pp0aMHUVFRjBgxgocffpiWLVvWGgPYf0B488036dixY40bp5eUlPCHP/yBTz/9lK5du3LkyBHy8vLqPCbINVf1JrqJ9gKFIgUFBSQnJwc7DJ+YlCLKrGiqXzHKqqyMe/OnGrNZgn3ykkdGdgLss15WWgz7X6uNCov9UWl1LHO8rnCtM9yee27nPJPNgNIqG6WnYPIWk8KPBE259cD52K5aEneyZ+5kImh2m1ilqddPQQgkUv9Dg40bN7J7926uueYa0tLS6NevH0uXLmX27Nls3ryZnJwc5syZ4xpiN3jw4Aafy33fXr16MXbsWNavX9+g5Oqcc85h9OjRAEyaNIlZs2Zx8OBBOnSoOSpo0aJF3H333a6ekLvvvpv58+fzyy+/0LVrV8aNG+fa9s477+S5555j586d9O3bl8jISPbs2UNeXh6pqakMGDCgxvGdXHPNNa4v/LNnz2bgwIFUVFQ0anjXWWedRWpqKp9++ikXX3wxH374IT169HAlUO4UFhYCkJCQUOdx27Rpw+23345SCrPZTGxsLL/5zW9c62fOnMnll1/uc//27dszdepUAC699FLi4uLYuXMnAwcOdMVQUFDgdd/+/ftz8OBBDh486Cr/VatWuV4PGjSozuSipKTEI4lyDu87ceJEjeTqwgsv5LHHHuPTTz/l/PPP56233iInJ4eyMvs947p27coXX3xB9+7dOXjwIDNnzmT69OleJ0Jx8vLLL/PGG28QHR1NVlYWb73l/S4XJpOJbdu20bFjR9q0aUObNm1qfV9OrFZrQK+7CrvkStCH8xcSoWHERkbUek3ggPQWWs9nGPYeswqrZ7LmStCsjsTMI1GrLZEzHOtsbomf4drHPZErq7JRdooTuUgTxH6/7WTvmytBU27Jm4+hk14SuerDL90TueaCzGYZOsjnc00i4mJcMwHWhrWsnC/6j8FWUVljnSk6igs2/ZOI2Joz6nk7X10sXbqUIUOGuO5Lds0117BgwQIefPBBcnJy6NChg7Z77nz//ff86U9/Ytu2bVRWVlJZWVnrF/jacP+SGh8fD9i/cHsjOzubP/7xjzz66KOuZVarlYMHD9K1a1defPFFFi9ezKFDh1BKceLECY4ds1+L/MILLzBv3jwGDRpEx44dmTlzJldccYXX87gndunp6dhsNo4cOeIaktdQJkyYwLJly7j44otZtmyZz2FlzgSjuLi4zgSrffv2Hu1HaWkpDz74IJ9++qkrSSsuLvaZHFZPEuLj4ykuLna9Li4u9vnjSlRUFGeffTbr1q3j66+/ZvLkyRw/ftz1eujQobXG7jzfiRMnXK+LiooAvE7G0rVrV1555RUeffRRDhw4wMUXX8wFF1zgqvNt27albdu2gL0Mn3rqKfr3709paanP4c3Tpk3jkUceqTPG119/nRdffJG77rqLc845h8cee4ysrKw6359MaFELZVVWAMrdvtS5P5cvCY2jtLS0xi8UQv04ldcEKmdPntkEAb6O2DDsQyIrLP4lcpUWZ4JmeCR1Hr1xDUnkSgN3gZ0zkXMlaO5JnI9ErsbwS7fkrrbr6IKVyMlslqGLfD7XRCnl1zA9c3wsGTdfbb/Gqvzk8GxTTDQZN40lOrXu4Ur+UF5ezsqVK6mqqnINw6qqquL48eN8+eWXdOjQgQMHDlBVVVUjwfL2eRAfH+/qDQA4cuSIx/qpU6dy88038/bbbxMTE8MDDzxQY5tA0LFjR+6++24mTpzoWuZ8Txs2bOD555/ngw8+oEePHphMJjp16uQaRtalSxdeffVVbDYb//nPf7j55pvZsWOH1/McOHDA9TwnJwellN89FU68eR0/fjzPPPMM27dvZ+PGjbz++ute942Pj6dTp07s3r2bdu3a+Tyet+UvvfQSu3btYs2aNbRt25affvqJCy64oMZwOn/ZuXMnZ5xxhs/1gwcP5uuvv+a7777jxRdf5Pjx43z99dds2LCB3/3ud3Uev2fPnvz888+ce+65AGzZsoW0tDSfnzmXX365K5G3WCz07dvXNTSxOiaTCcMwGvze3Rk+fDjDhw+nvLycefPmMWPGDD79tO4fWGw2W0BvZxHSydWVf/9fjWXjl2xxPf/k1n6nMpywo7zcvwuDBd+E6zWBStlnTYyKOHWJXGW1IZF7sw+Q0rqtX4ncycSv5vDL6r10zkvl3BO5wsC+RY9ELrpaL1vtCZpnIld9WfVEMMps8rhn3WvfHeA/MptlSCKfz42j+0PTsVVaXNdXmaKjyLhpLN0fmq7tHKtWrcJqtbJ+/XqP3on777+fpUuX8tJLL5GWlsbDDz/M7NmziYyMZNOmTQwePJjWrVtTUFDgMfyzT58+zJ8/n71799K6dWueeuopj/MVFxfTokULYmJi2LRpE++++y7nnXeetvfjiylTpvD4449z5plnkpWVRVFREatXr2bMmDEUFxcTERFBy5YtsVgsPP/88x49Im+//TbDhw8nNTWVxMREDMPw+aX33XffZeLEiWRmZvLkk09y+eWX13tIYOvWrdm3b5/HsrZt2zJ48GBuu+02RowYUeu1QCNHjmTdunWuCTm8lZM3iouLiY6OJikpiePHj/P000/XK+7qicjXX3/NDTfc4HP7oUOHMnHiRLp3705CQgJDhgxh5syZREZG+tWzM3HiRF588UVGjhxJXFwczz77bK2TU/zwww/06dOHEydO8MQTT3Daaadx4YUXAvDVV19x2mmnkZ6eztGjR5k1axYXXnihq0e0oRw5coRNmzZx/vnnExsbS2xsrN8TVdhsgR1pE9LJlRBY2rdvH+wQwg65JrD+uCdyCW7taFp8uvapVH0lcrUmaNV62WpL5FzDLx3LvCVygcakIMpxy4KiCmuN9RVWgw+25nFD/zRipL42WeTzueFYSuy9P13uudGVXA3+8i1iUlOwlldqu9/V0qVLue6661wzrjmZNm0a48eP5+mnn2bp0qU88MAD9O3bF8MwuOSSSxg8eDDdunVj3Lhx9O/fH6vVyqeffsoFF1zA+PHjufDCC0lKSuLBBx/0uG7lmWee4Y9//COzZ89m8ODBXHXVVa7hd4HkiiuuoKSkhNtuu43s7GySkpIYMmQIY8eOZfjw4YwaNYpzzz2X+Ph4br/9do/hfZ999hkPPfQQZWVlpKWl8corr5CYmMjx48drnOfaa6/l7rvvZseOHQwcOJCXXnqp3rHee++9zJo1izlz5jBt2jT+8Ic/APZk4uabb/Y6G547U6ZM4cYbb+R3v/sdSimv5eSNadOmMXXqVLp160a7du244447WLVqld9xu/eE5ebmsnPnzlqHfJ599tlYLBbXEMC0tDRat25NVlYWSinXTZo3bNjgmrHQnRtuuIGcnBwuvPBCqqqquPrqq7n//vtd68eNG8egQYNc96p6+OGH+e9//4vJZOKSSy7hH//4h2vb//3vf9x+++0UFBTQokULhg8fXueQP3+w2Wy89NJLTJs2DaUUWVlZPP/8837tq2sori+Ujm45Haxbt87wJ5t2xzks0El2drbH2FsZFtg49u3bx2mnnRbsMEKesiqrq5f1/Rv7SL3URKjXT8OwzyhpT9A8h0vWlsh57X2rlgjWlsj5Q2SE4m9X96R9Uv2S182bN28bMWJE/T7ImyENae+qE+r1XxeFhYWua2H85aN2tU8accmh9Y0JSYBGTzRxqvnhhx+49tpr2bp1a51fvG+//XZGjhzJ2LFjT1F0nj4feOABTj/9dG655ZZTdv5woz7109dnTG3tXUj3XFX/khodoeSLqyAIIYFS9ptPR0aYqHvuqcbhTOTcJy4pKrdw74e7vN4yQAEt40J/+KogCEJdWCwWFi5cyHXXXedXj8bChQtPQVS+efLJJ4N6fqFuQjq5qo5zNhJBD+JTaMpI/fSfk4kcxEfZf4BqnxTNVb18z2YpQwKbNlL/G463GQVthg2Tkjqvi0APu9LF7t27ueCCC+jRo0eNa9iaEqHiM1QItM+w+iQ5dOhQsEMIK8Sn0JSR+tl4bjmnA5d2P3ljxkDOZinoRep/wzHHx9Z4GJFm13Oh8VRVVQU7BL/o0qULOTk5rFmzxq+b2gaLUPEZKgTaZ1j1XMXGyoeiTsSn0JSR+tl4wnU2y+aA1H+9+DvLmOAf4lMv4lMvgfYZVslVTEzdN/QT/Ed8Ck0ZqZ+NwzljmtVixVxpv9ePubICi81+7zD5Bb9pI/X/JIZhNPo+cc3thuGBRnzqRXzqxV+fDZ30L6ySq+PHj9d71iDBN+JTaMpI/Wwca7qMcD2/y/H36z+dXC8zpjUtiguLPV7nZucSwckJnBJaBHpalKZJfHy8azavxnwBtVqtmM1h9ZXolGK1eM7eXFVVheJkeUSYZbKxxiD1s3E0tH6WlpY26IessCqppjxeNhQRn43DeauAcrd7F7k/l5ktG4fUT6E58XX3UTWWud8Ktbkmw2azmYSEBIqKihp1nMrKSqKiojRF1fw4+M/Vta5vP2bkKYokPJH62TgaWj/NZnODbikQVslVeXm5/JKtEfHZOJz3tnJn/JItruef3NrvVIYTdkj9bBzOGdNKiorZ0O9KAAb98D7xSc2zB0QIXcxmc6M/Cw4fPiyfJ41gw6z5ta7vMeWaUxRJeCL1s3Gc6voZVslVWVlZsEMIK8Sn0JSR+tk4nNdURbgNl4iIk9nSmipDd3wCQElhKT8MuAqAfhtXEt8iLohRhQ/yedI4pH4GFqmfjeNU18+wSq7atWsX7BDCCvHZON6/sY/H61C7Y31TR+qn0Jzwdk1VfIu4ZnutlW7k86RxSP0MLFI/G8eprp9hNbfjkSNHgh1CWCE+G0dsZITHoyg/z+O10DikfurHWlYe7BAEISjI54nQlJH6GVqEVXJls9nq3kjwG/GpF/GpF/HZeAyrlV/nveJ6/c2517D9sb9iWK217CUI4Yd8nghNGamfoUVYJVdpaWnBDiGsEJ96EZ96EZ8NxzAMbBWVbPvjX8hd8sHJ5RWV7H/jPXbMWRDE6ATh1COfJ0JTRupnaBFW11zl5uaSmZkZ7DDCBvHZOJw3aXWSs38fGRmnuV7LxAGNIxTrp2G1YquowlZZibWi0vXc5nxeUYGtssrxutLjudX5urzSc59Kt23L7ftbnftXO4b7c1/YyivY/8Z7dP3dbUTEyY1qheZBKH6eCM0HqZ+hRUgnV9W/vEYansvky2vjaMiN04STuN+k1clOt+fN9b40uvC3fhqGgWGx2hOXiipHgmF/bk9YKh0JS5XXbWyObVzJkHMfj+eVjuNVYFRWYXUmQK5kxn48wxIiw+0Mg4qjx4g7rUOwIxG8UGU2UxaXQOmJUpkwQBPS3ulD6qd+pH7q41TUz5BOrrx9ef3J7bl8eW0ccXEyharQcJxDz7z2nvjVw+IjAXI8rywtI9tqq7HOWl5pT3Dc9sMwgq2jJkphio7EFB2NKSrS/oixP4+IjsLkfERFen/ueB0RE4Upyrku2nFMx7KYKI/jKbfnhs3Gl+eO896LpRTRrVudeidCrViqLPzr/udJBOJKi9k4cByll1/C2Bd/hzkypJvzoCPtXeOR+hk4pH42nlNZP6W2Cz7Jz88nKSkp2GGELBft/hTDamXnEwvJ/vs/ATBFR9Fx8pV0/cNtATuvzWJpWA9L9eFmFSfXWWusq3v4mlFZFbD32BiUOcIj8XAmIhEx9uTD+dyVrERFOraN8ngeER2FinYkK1FRnklNbQmQ47mKNKOUCqqLjJuvZv8b72Err3AtM8VEk3HTWBkS2AR5785nSPhoNWaLxbUsdtVHvGOp4or59wUxstDnQPYBTOlhdRn6KefD++aT8PGnUj8DgNTPxuO1fv77I967E679vwe0niukk6uLdn8KgLW0jLVnXAHAhT99SEScDAfUgdwNvH4YVqs9CXH1xlSw+8+LyP3natc2topKsv+xktJ9B2h3xYVeh5959NC4kqHqw9e8JE6OdU11pjdfPS/O5xFee2mq9cS4PS+3VpHYMtnn8dz3d0+AVIRMg++k+0PTqSqv4sAb7wD2Msq4aSzdH5oe5MiE6pScKCVu1UeYLZ4/WkRaqkhe9RFfr/ooSJGFD/uCHUCIk+xlmdRPfUj9bBzJXpZFVlWhVn1EyYm7iU/U1zsY0smVt2uqIuJi5VorTVjcsvumjscQNPceFOejvBJrRYVnr0xFhX2b8so69q3wnEzAcRyr27FtlZV+X09jVFWRt2Y9eWtOwbBVk8meWMS496hEnnxerSfG1zrXfs6enOhIIqr36nhJbiIcw9xUVKT2XpqjR4/SunVrrcdsbqiICDrNmupKrs799h1atEsNclSCN47uOww0weGtgiAIIY3B0X2Hie/dSdsRQzq5EgJLSUmJX19eXRMGVJ5MVKyOpMR92JjV/XVFpaOXp8ItaXFLgJwJj3MigfJqiU95tWFptcx+FgyUOQIVGYnN101ZlaLlwL5EJid6DEszRUVWS2aqXztj74lRbs89h6JV298cvv/i/tZPwTvOyX+spW4TAxmGa7n8SNW0aH1aW8D7DxQWcyQDvllBnMZfXpsb2dnZpKenBzuMkKX0RCkbB17jMeTKidTPxiP1s3HUVj9BOT5f9RF237ysZeXN7kuBYbO5roOxVktMTiYxbj0u7jOjuSdDbr0wtopKqkrLyKuyuvZzv17Hs0eoEprSDe6UwhQT5TYUzJGcuIafRXsmL9HR9sQm2vPhMalATPX9ok9u49r3ZK+OyWymIu84X/Qf4zXxM0VFcvZbz8l1LY2gbVu9H4bNDW8TAm3od6XruUwI1LSIT4yj9PJLiP33R0RWnRwaWBUZSdlll9CmY5sgRhf6ZEZlEhvbvL476CShRQKll18q9TNASP1sHHXVT51DAiEMkivDamXn3Jddr7/oP4aMm6+m+0PTT8m1FYZhYFRZHElLhefF/c7hZNUSF8+emArPKZ7dE6AKtySp3PdQtaY2cYByzHxmTzyiPZMVj2FmUW5JzcnEJcL9tXuSFFMt8XEe2zEULcLxuilMFACwtvflPtfZKiolsWokeXl58kue0KwY++LvWDHdSuJ/VmMfIqgou8w+25XQOOTzpPFI/QwcUj8bz6msnyGfXO2Ys4Cctz5wvbZVVLL/9XepOJpPxo1jPSYLsLr11NS4rqb6UDW3hKd6T0314WpNappn92tsPHpq3CcSqNZT45a4RLj11BwvLqZ1+3Y1EiDXa7deoZMJUSTKJDPaCIHH2kQn7ggVhu74BIDyKhs3vL0VgH9cm0VMpPz/NlXMkWZGP3sPa//zCWVxCQz7bJH0CGhCPk8aj9TPwCH1s/GcyvoZ0smVtbSc/a+/W2PYla2ikoPvfMzBdz4+5THVmLksJtrL8LKaw9Ui3HpqnLOcVR+q5pwc4OSxoj2SIVNUlP1aH029NmmVlURFRWk5VnPEOZsl2Ierlhw8THz7tkTESo+VDtLS0oIdQkgzdsWuky+iogG4buVu16JPbu13qkMS/CTSYiGyqECuYdGIfJ7oQ+qnfqR+6uNU1E+/kiul1CjgSaAXcAx4A3jEMAyfqbRSKhZ4HLgGaAccBt4DZhuGUdrIuAGoOHqs1vUxHdpiToiz96i430yz+lC16g+PnpiT1+ZEOI/ja9+o8Oq1yc3NJTMzM9hhhCzu1/6Z42M5WFxISmrLIEYUXkj9FAJBU23vhMAinydCU0bqZ2hRZ3KllOoHfAi8BEwGsoDXHPvOqmXX54AxwM3AVsd+rwPxwNRGRe0gunUrn+tMMdGc99VSua6lEUivlV7Ep17EZ+N4/8Y+Hq8PHTxEu/btghRN06Apt3dCYJHPE6EpI/UztPCn5+p+4CfDMO51vN6mlOoAzFNKzTEMo9jHfucBSwzD+I/j9T6l1FJgZONCPklEXAwZN1/N/jfew1Ze4Vpuiokm46axklg1kqSkpGCHEFaIT72Iz8YRG+k54U/rlBY1ljVDmmx7JwQW+TwRmjJSP0MLf5KrIcDiastWAX8B+gNf+NjvK+BypdTLhmHsUkp1BS4DVnrbOD8/n/79+xMREYHNZmPcuHHcc889HDp0CLBPu3zo0CFiY2OJiYnh+PHjtGzZkuTbrqbwWD7H37ZfX6WiIulww5VET76cQ4cOERcXR35+Pi1atMBisVBSUkLbtm3Jy8vDarWSlpZGbm4uUVFRJCUlkZeXR1JSEjabjeLiYtq0aUN+fj4Wi8W1bWRkJMnJyRw9epTExEQATpw4QWpqKkVFRVRWVrq2NZvNpKSkcOTIEeLj4zGbzRQWFtKqVSuKi4upqKhwbWsymWjTpo0r7qioKAoKCkhJSaG0tJTy8nLat2/vctKuXTsOHjzo4SQ5OZnKykpKS0tp164dR44cwWazuc4RHR1NQkICx44dq+Hk2LFjHu+zuLiYzMzMGk5at25NQUEBVVVVXp0kJCRgMpkoKiry6aRVq1YcPny4wU5atmxJeXk5ZWVlXp3ExMS4yr6hTryVvXs9cS97b07cyz4hIYGjR48SGxvr1UlERASpqak1nLiXvTcn1f8fnE7atWvH4cOH/XLi7f/B3Yk//w8tWrSo4SQ1NZXCwsJanTjribeyd3fi7f/h8OHD9OjRo87PCHcn1cve3Ulz/4ywWq3k5eXV6zOiupMwoMm2d7GJLSgvL+d43nHXcX7Zu5/Op3fm4KGDtIiPlbos7V3Q2jsrEbRMacn+vfs96mfH0zpyougEylYl7Z20d0H7jDh+wv4+9+/L8aifySktiE9IoKTwuNb2Thl1zHSnlKoA7jUMY4HbsnigGLjOMIylPvaLBOYDdwIW7Inc/wG3G15Oum7dOiMrK6vWWHxhKSlz3bPlwi2riJbrWrSQn59PSkpKsMMIG8SnXsSnXnT43Lx587YRI0Y07IO8CdCU27vLFnwDgLmygunzHgBgwawnsTgmI/n39IH1Op7giXyeNA6pn4FF6mfjCET9rK29a+hsgUa1v96YBowGxgPbgJ7An7FfIDy7geetE5mJTR+2pnRj4DBAfOpFfOpFfPqkSbR3d/3pvhrLnF8S7C/kps+NQep/45D6GVikfjaOU10//Zna7iDQvtqy9m7raqCUigGeBR4yDONtwzB+MgzjbewXBP9BKaVl/kNLSRmWkjKspWWuZdbSMtdyoXEUF/u6vEBoCOJTL+JTL+ITaMLtnRBYpP4LTRmpn6GFPz1X64BLgT+6LbsMKAM2+dgn0vGoPnWtFVCOR6NxDgV0Z+0ZV7ieX3JIfilpDG3ayM3/dCI+9SI+9SI+gSbc3rnfNw+grKyU2FjJ23Qh9b9xSP0MLFI/G8eprp/+9Fw9B/RRSj2nlOqplBoL/Al4wTlzklJqgFJqu1JqAIBhGCeAT4EnlFKXKqUylVKXAXOBjwzDKAnM2xF0kp+fH+wQwgrxqRfxqRfxCTTh9s4cH+vxKCwv83gtNA6p/41D6mdgkfrZOE51/ayz58owjM1KqdHYb6p4J/Yx5AuAh902iwO6O/46mYi9UXoZaIv9por/qrZfo6ieie7fv4+MjNN0Hb7ZY7FYgh1CWCE+9SI+9SI+m3Z7Vx0pL72IT72IT72IT70E2qdfE1oYhvEx8HEt6z+n2tAHwzDygOmNCa4uqmebHbt0xiw3WtNGWlpasEMIK8SnXsSnXsSnnaba3lVHyksv4lMv4lMv4lMvgfbpz7DAkCE3NzfYIYQV4lMv4lMv4lMv4jO0kPLSi/jUi/jUi/jUS6B9hlVyFRkZGewQwgrxqRfxqRfxqRfxGVpIeelFfOpFfOpFfOol0D7DKrlKTk4OdghhhfjUi/jUi/jUi/gMLaS89CI+9SI+9SI+9RJon2GVXB09ejTYIYQV4lMv4lMv4lMv4jO0kPLSi/jUi/jUi/jUS6B9hlVylZiYGOwQwgrxqRfxqRfxqRfxGVpIeelFfOpFfOpFfOol0D7DKrkSBEEQBEEQBEEIFmGVXJ04cSLYIYQV4lMv4lMv4lMv4jO0kPLSi/jUi/jUi/jUS6B9hlVylZqaGuwQwgrxqRfxqRfxqRfxGVpIeelFfOpFfOpFfOol0D7DKrkqKioKdghhhfjUi/jUi/jUi/gMLaS89CI+9SI+9SI+9RJon2GVXFVWVgY7hLBCfOpFfOpFfOpFfIYWUl56EZ96EZ96EZ96CbTPsEqu0tLSgh1CWCE+9SI+9SI+9SI+QwspL72IT72IT72IT70E2mdYJVe5ubnBDiGsEJ96EZ96EZ96EZ+hhZSXXsSnXsSnXsSnXgLtM6ySK7PZHOwQwgrxqRfxqRfxqRfxGVpIeelFfOpFfOpFfOol0D7DKrlKSUkJdghhhfjUi/jUi/jUi/gMLaS89CI+9SI+9SI+9RJon2GVXB05ciTYIYQV4lMv4lMv4lMv4jO0kPLSi/jUi/jUi/jUS6B9hlVyFR8fH+wQwgrxqRfxqRfxqRfxGVpIeelFfOpFfOpFfOol0D7DKrmSMal6EZ96EZ96EZ96EZ+hhZSXXsSnXsSnXsSnXuSaq3pQWFgY7BDCCvGpF/GpF/GpF/EZWkh56UV86kV86kV86iXQPsMquWrVqlWwQwgrxKdexKdexKdexGdoIeWlF/GpF/GpF/Gpl0D7DKvkqri4ONghhBXiUy/iUy/iUy/iM7SQ8tKL+NSL+NSL+NRLoH2GVXJVUVER7BDCCvGpF/GpF/GpF/EZWkh56UV86kV86kV86iXQPsMquUpLSwt2CGGF+NSL+NSL+NSL+AwtpLz0Ij71Ij71Ij71EmifYZVc5ebmBjuEsEJ86kV86kV86kV8hhZSXnoRn3oRn3oRn3oJtM+wSq5MprB6O0FHfOpFfOpFfOpFfIYWUl56EZ96EZ96EZ96CbTPsCqtNm3aBDuEsEJ86kV86kV86kV8hhZSXnoRn3oRn3oRn3oJtM+wSq4OHToU7BDCCvGpF/GpF/GpF/EZWkh56UV86kV86kV86iXQPsMquYqLiwt2CGGF+NSL+NSL+NSL+AwtpLz0Ij71Ij71Ij71EmifYZVcRUVFBTuEsEJ86kV86kV86kV8hhZSXnoRn3oRn3oRn3oJtM+wSq4KCgqCHUJYIT71Ij71Ij71Ij5DCykvvYhPvYhPvYhPvQTaZ1glVykpKcEOIawQn3oRn3oRn3oRn6GFlJdexKdexKdexKdeAu0zrJKr0tLSYIcQVohPvYhPvYhPvYjP0ELKSy/iUy/iUy/iUy+B9hlWyVV5eXmwQwgrxKdexKdexKdexGdoIeWlF/GpF/GpF/Gpl0D7DKvkqn379sEOIawQn3oRn3oRn3oRn6GFlJdexKdexKdexKdeAu0zrJIruQ+AXsSnXsSnXsSnXsRnaCHlpRfxqRfxqRfxqRe5z5UgCIIgCIIgCEII4FdypZQapZTapJQqV0odUErNUUpF+LHfaUqpt5RSRx377lRKXdXoqH3Qrl27QB26WSI+9SI+9SI+9SI+7Uh71zwRn3oRn3oRn3oJtM86kyulVD/gQ+BLoB9wF3An8EQd+3UAvnGcYzTQA7gNyG5cyL45ePBgoA7dLBGfehGfehGfehGf0t41Z8SnXsSnXsSnXgLt0+zHNvcDPxmGca/j9TZHQzJPKTXHMIxiH/vNBfYahjHRbdnehodaN7GxsYE8fLNDfOpFfOpFfOpFfALS3jVbxKdexKdexKdeAu3Tn2GBQ4B/V1u2CogF+nvbQSllAq4CNiilliqljiilflJKPaSU8iehaxAxMTGBOnSzRHzqRXzqRXzqRXwC0t41W8SnXsSnXsSnXgLt058P/vZA9f4z5zQbaT72aQ0kAdOBF4CLgSzgr0Ai8IfqO+Tn59O/f38iIiKw2WyMGzeOe+65xzWjR9u2bTl06BCxsbHExMRw/PhxWrZsSXl5OWVlZbRr1459+/aRlJREWloaubm5xMTEEBcXR35+Pi1atMBisVBSUkLbtm3Jy8vDarW6to2KiiIpKYm8vDySkpKw2WwUFxfTpk0b8vPzsVgsrm0jIyNJTk7m6NGjJCYmAnDixAlSU1MpKiqisrLSta3ZbCYlJYUjR44QHx+P2WymsLCQVq1aUVxcTEVFhWtbk8lEmzZtOHToEHFxcURFRVFQUEBKSgqlpaWUl5fTvn17l5N27dpx8OBBDyfJyclUVlZSWlpKu3btOHLkCDabzXWO6OhoEhISOHbsWA0nx44d83ifxcXFRERE1HDSunVrCgoKqKqq8uokISEBk8lEUVGRTyetWrXi8OHDDXbiXvbenLiXfUOdeCt793riXvbenLiXfUJCAkePHuX48eNenURERJCamlrDiXvZe3NS2//D4cOH/XLi7f/B3Yk//w8tWrSo4SQ1NZXCwsJanTjribeyd3fi7f/h8OHD9OjRo16fEdXLXj4jTjqxWq0cP368Xp8R1Z2EAdLehUFdlvZO2jtp76S9C2Z7pwzDqH0DpSqAew3DWOC2LA4oASYahrHMyz5pwAHgO8MwBrgtvx94FEg0qp143bp1RlZWVq2x1EVBQQHJycmNOoZwEvGpF/GpF/GpFx0+N2/evG3EiBGN+yAPItLeNV/Ep17Ep17Ep14C3d75MyzwIPZf89xp77bOG0eBKmBrteU/A/FAqh/nrTeVlZWBOGyzRXzqRXzqRXzqRXwC0t41W8SnXsSnXsSnXgLt05/kah1wabVllwFlwCZvOxiGUQV8C3Svtqo79l8A8+oXpn+UlpYG4rDNFvGpF/GpF/GpF/EJSHvXbBGfehGfehGfegm0T3+Sq+eAPkqp55RSPZVSY4E/AS84Z05SSg1QSm1XSg1w2+9J4Fyl1MNKqdOVUr8BZgN/qT5EQhdyHwC9iE+9iE+9iE+9iE9A2rtmi/jUi/jUi/jUS9Dvc2UYxmbs9+24EPgReBFYgL3hcBKH/Ve6OLf9/g1MBK4FtmBvtP4MPKIn9JocOXIkUIdulohPvYhPvYhPvYhPae+aM+JTL+JTL+JTL4H26dc0sYZhfAx8XMv6zwHlZflyYHlDg6svNpvtVJ2qWSA+9SI+9SI+9SI+7Uh71zwRn3oRn3oRn3oJtE9/hgWGDGlpvmbKFRqC+NSL+NSL+NSL+AwtpLz0Ij71Ij71Ij71EmifYZVc5ebmBjuEsEJ86kV86kV86kV8hhZSXnoRn3oRn3oRn3oJtM+wSq6io6ODHUJYIT71Ij71Ij71Ij5DCykvvYhPvYhPvYhPvQTaZ1glVwkJCcEOIawQn3oRn3oRn3oRn6GFlJdexKdexKdexKdeAu0zrJKrY8eOBTuEsEJ86kV86kV86kV8hhZSXnoRn3oRn3oRn3oJtM+wSq5atGgR7BDCCvGpF/GpF/GpF/EZWkh56UV86kV86kV86iXQPsMqubJYLMEOIawQn3oRn3oRn3oRn6GFlJdexKdexKdexKdeAu0zrJKrkpKSYIcQVohPvYhPvYhPvYjP0ELKSy/iUy/iUy/iUy+B9hlWyVXbtm2DHUJYIT71Ij71Ij71Ij5DCykvvYhPvYhPvYhPvQTaZ1glV3LBn17Ep17Ep17Ep17EZ2gh5aUX8akX8akX8akXmdCiHsiYVL2IT72IT72IT72Iz9BCyksv4lMv4lMv4lMvcs1VPUhLSwt2CGGF+NSL+NSL+NSL+AwtpLz0Ij71Ij71Ij71EmifYZVc5ebmBjuEsEJ86kV86kV86kV8hhZSXnoRn3oRn3oRn3oJtM+wSq6ioqKCHUJYIT71Ij71Ij71Ij5DCykvvYhPvYhPvYhPvQTaZ1glV0lJScEOIawQn3oRn3oRn3oRn6GFlJdexKdexKdexKdeAu0zrJKrvLy8YIcQVohPvYhPvYhPvYjP0ELKSy/iUy/iUy/iUy+B9hlWyZVk9noRn3oRn3oRn3oRn6GFlJdexKdexKdexKdepOeqHthstmCHEFaIT72IT72IT72Iz9BCyksv4lMv4lMv4lMvgfYZVslVcXFxsEMIK8SnXsSnXsSnXsRnaCHlpRfxqRfxqRfxqZdA+wyr5Kp169bBDiGsEJ96EZ96EZ96EZ+hhZSXXsSnXsSnXsSnXgLtM6ySq4KCgmCHEFaIT72IT72IT72Iz9BCyksv4lMv4lMv4lMvgfYZVslVVVVVsEMIK8SnXsSnXsSnXsRnaCHlpRfxqRfxqRfxqZdA+wyr5CotLS3YIYQV4lMv4lMv4lMv4jO0kPLSi/jUi/jUi/jUS6B9hlVylZubG+wQwgrxqRfxqRfxqRfxGVpIeelFfOpFfOpFfOol0D7DKrmKjIwMdghhhfjUi/jUi/jUi/gMLaS89CI+9SI+9SI+9RJon2GVXCUnJwc7hLBCfOpFfOpFfOpFfIYWUl56EZ96EZ96EZ96CbTPsEqujh49GuwQwgrxqRfxqRfxqRfxGVpIeelFfOpFfOpFfOol0D7DKrlKSEgIdghhhfjUi/jUi/jUi/gMLaS89CI+9SI+9SI+9RJon2GVXJlMYfV2go741Iv41Iv41Iv4DC2kvPQiPvUiPvUiPvUSaJ9hVVpFRUXBDiGsEJ96EZ96EZ96EZ+hhZSXXsSnXsSnXsSnXgLtM6ySq9TU1GCHEFaIT72IT72IT72Iz9BCyksv4lMv4lMv4lMvgfYZVsmVZPZ6EZ96EZ96EZ96EZ+hhZSXXsSnXsSnXsSnXqTnqh5UVlYGO4SwQnzqRXzqRXzqRXyGFlJeehGfehGfehGfegm0z7BKrtLS0oIdQlghPvUiPvUiPvUiPkMLKS+9iE+9iE+9iE+9BNpnWCVXubm5wQ4hrBCfehGfehGfehGfoYWUl17Ep17Ep17Ep14C7dOv5EopNUoptUkpVa6UOqCUmqOUivD3JEqpm5RShlLq8wZH6gdmszmQh292iE+9iE+9iE+9iE870t41T8SnXsSnXsSnXgLts87kSinVD/gQ+BLoB9wF3Ak84c8JlFJnAHMd+weUVq1aBfoUzQrxqRfxqRfxqRfxKe1dc0Z86kV86kV86iXQPv3pubof+MkwjHsNw9hmGMa7wMPAXUqpWm9xrJRKBFZgb6B+bXS0dXD48OFAn6JZIT71Ij71Ij71Ij4Bae+aLeJTL+JTL+JTL4H26U9yNQT4d7Vlq4BYoH8d+74KfG4YxooGxFZv4uPjT8Vpmg3iUy/iUy/iUy/iE5D2rtkiPvUiPvUiPvUSaJ/+DDpsDxystuyQ46/P6TaUUncAWcAAfwLJz8+nf//+REREYLPZGDduHPfccw+HDtlP1bZtWw4dOkRsbCwxMTEcP36cli1bUl5eTllZGe3atSM/P5+SkhLS0tLIzc0lJiaGuLg48vPzadGiBRaLhZKSEtq2bUteXh5Wq9W1bVRUFElJSeTl5ZGUlITNZqO4uJg2bdqQn5+PxWJxbRsZGUlycjJHjx4lMTERgBMnTpCamkpRURGVlZWubc1mMykpKRw5coT4+HjMZjOFhYW0atWK4uJiKioqXNuaTCbatGnDoUOHiIuLIyoqioKCAlJSUigtLaW8vJz27du7nLRr146DBw96OElOTqayspLS0lLatWvHkSNHsNlsrnNER0eTkJDAsWPHajg5duyYx/u0WCzExsbWcNK6dWsKCgqoqqry6iQhIQGTyURRUZFPJ61ateLw4cMNduJe9t6cuJd9Q514K3v3euJe9t6cuJd9QkICZWVl7N2716uTiIgIUlNTazhxL3tvTmr7f3D+MlOXE2//D+5O/Pl/aNGiRQ0nqampFBYW1urEWU+8lb27E2//DwUFBSQlJdXrM6J62ctnxEkn8fHx7N27t16fEdWdhAHS3oVBXZb2Tto7ae+kvQtme6cMw6h9A6UqgHsNw1jgtiwOKAEmGoaxzMs+/YA1wBDDMLY7li0CMg3DGObtPOvWrTOysrJqjaUu9u7dS2ZmZqOOIZxEfOpFfOpFfOpFh8/NmzdvGzFiROM+yIOItHfNF/GpF/GpF/Gpl0C3d/70XB3E/mueO+3d1nnjAqAlsEUp5VxmApRSygKMMgzjMz/OXS/kgj+9iE+9iE+9iE+9iE9A2rtmi/jUi/jUi/jUS1OY0GIdcGm1ZZcBZcAmH/ssBvoAfd0eHwDfO55/W99A/aG4uDgQh222iE+9iE+9iE+9iE9A2rtmi/jUi/jUi/jUS6B9+tNz9RzwjVLqOeBvQE/gT8ALhmEUAyilBgD/AG4wDGOjYRjHgGPuB1FKFQDJhmFs0Ri/BxUVFYE6dLNEfOpFfOpFfOpFfALS3jVbxKdexKdexKdeAu2zzp4rwzA2A6OBC4EfgReBBcBst83igO6Ov0EjLc3n9cZCAxCfehGfehGfehGf0t41Z8SnXsSnXsSnXgLt059hgRiG8bFhGGcZhhFtGEaaYRizDcOwuq3/3DAMZRjG57UcY4qvi3t1kZubG8jDNzvEp17Ep17Ep17Epx1p75on4lMv4lMv4lMvgfbpV3IVKphMYfV2go741Iv41Iv41Iv4DC2kvPQiPvUiPvUiPvUSaJ9hVVpt2rQJdghhhfjUi/jUi/jUi/gMLaS89CI+9SI+9SI+9RJon2GVXDlvJCboQXzqRXzqRXzqRXyGFlJeehGfehGfehGfegm0z7BKruLignp9cdghPvUiPvUiPvUiPkMLKS+9iE+9iE+9iE+9BNpnWCVXUVFRwQ4hrBCfehGfehGfehGfoYWUl17Ep17Ep17Ep14C7TOskquCgoJghxBWiE+9iE+9iE+9iM/QQspLL+JTL+JTL+JTL4H2GVbJVcuWLYMdQlghPvUiPvUiPvUiPkMLKS+9iE+9iE+9iE+9BNpnWCVX5eXlwQ4hrBCfehGfehGfehGfoYWUl17Ep17Ep17Ep14C7TOskquysrJghxBWiE+9iE+9iE+9iM/QQspLL+JTL+JTL+JTL4H2GVbJVfv27YMdQlghPvUiPvUiPvUiPkMLKS+9iE+9iE+9iE+9BNpnWCVXch8AvYhPvYhPvYhPvYjP0ELKSy/iUy/iUy/iUy9ynytBEARBEARBEIQQIKySq3bt2gU7hLBCfOpFfOpFfOpFfIYWUl56EZ96EZ96EZ96CbTPsEquDh48GOwQwgrxqRfxqRfxqRfxGVpIeelFfOpFfOpFfOol0D7DKrmKiYkJdghhhfjUi/jUi/jUi/gMLaS89CI+9SI+9SI+9RJon2GVXMXFxQU7hLBCfOpFfOpFfOpFfIYWUl56EZ96EZ96EZ96CbTPsEqu8vPzgx1CWCE+9SI+9SI+9SI+QwspL72IT72IT72IT70E2mdYJVfJycnBDiGsEJ96EZ96EZ96EZ+hhZSXXsSnXsSnXsSnXgLtM6ySq8rKymCHEFaIT72IT72IT72Iz9BCyksv4lMv4lMv4lMvgfYZVslVaWlpsEMIK8SnXsSnXsSnXsRnaCHlpRfxqRfxqRfxqZdA+wyr5EruA6AX8akX8akX8akX8RlaSHnpRXzqRXzqRXzqRe5zVQ+OHDkS7BDCCvGpF/GpF/GpF/EZWkh56UV86kV86kV86iXQPsMqubLZbMEOIawQn3oRn3oRn3oRn6GFlJdexKdexKdexKdeAu0zrJKrtLS0YIcQVohPvYhPvYhPvYjP0ELKSy/iUy/iUy/iUy+B9hlWyVVubm6wQwgrxKdexKdexKdexGdoIeWlF/GpF/GpF/Gpl0D7DKvkKjo6OtghhBXiUy/iUy/iUy/iM7SQ8tKL+NSL+NSL+NRLoH2GVXKVkJAQ7BDCCvGpF/GpF/GpF/EZWkh56UV86kV86kV86iXQPsMquTp27FiwQwgrxKdexKdexKdexGdoIeWlF/GpF/GpF/Gpl0D7DKvkqkWLFsEOIawQn3oRn3oRn3oRn6GFlJdexKdexKdexKdeAu0zrJIri8US7BDCCvGpF/GpF/GpF/EZWkh56UV86kV86kV86iXQPsMquSopKQl2CGGF+NSL+NSL+NSL+AwtpLz0Ij71Ij71Ij71EmifYZVctWnTJtghhBXiUy/iUy/iUy/iM7SQ8tKL+NSL+NSL+NRLoH2GVXKVn58f7BDCCvGpF/GpF/GpF/EZWkh56UV86kV86kV86iXQPsMquZIxqXoRn3oRn3oRn3oRn6GFlJdexKdexKdexKde5JqrepCWlhbsEMIK8akX8akX8akX8RlaSHnpRXzqRXzqRXzqJdA+wyq5ys3NDXYIYYX41Iv41Iv41Iv4DC2kvPQiPvUiPvUiPvUSaJ9+JVdKqVFKqU1KqXKl1AGl1BylVEQd+/xOKbVBKXVcKVWglPpaKXWJnrC9ExUVFcjDNzvEp17Ep17Ep17Epx1p75on4lMv4lMv4lMvgfZZZ3KllOoHfAh8CfQD7gLuBJ6oY9fhwOvAhcC5wLfAh0qpIY0JuDaSkpICdehmifjUi/jUi/jUi/iU9q45Iz71Ij71Ij71EmifZj+2uR/4yTCMex2vtymlOgDzlFJzDMMo9raTYRiXVls0Uyl1KTAWWFefICsqKigvL69zu5KSEqxWa30OLdSC+NSL+NRLfX3GxMQQHR0dwIhCm7y8PBISEoIdRrAJenvnL1JeehGfehGfehGfegm0T3+SqyHA4mrLVgF/AfoDX/hzIsewikQgz9v6/Px8+vfvT0REBDabjXHjxnHPPfdw+PBhEhISaNGiBRaLBZPJhFIKq9VKREQEhmFgs9mIjIwkKioKk8lEZGQkVVVVmEwmTCYTFoulxrYWiwXDMFzbKqWIiIiosa3ZbMZqtda6LYDVaq1zW/fYvW0LeMTuvq3NZnPFXtu21d9n9W19vU9vTqKjo13L6+OkrveplMJsNnt9n/46qet91lb2/jrRXfYxMTEA9Xaiq+xD7f+hLidRUVFER0f75cRqtVJZWcnhw4ex2WykpaWRm5tLTEwMcXFx5Ofnuz5jSkpKaNu2LXl5eVitVte2UVFRJCUlkZeXR1JSEjabjeLiYtq0aUN+fj4Wi8W1bWRkJMnJyRw9epTExEQATpw4QWpqKkVFRVRWVrq2NZvNpKSkcOTIEeLj4zGbzRQWFtKqVSuKi4upqKhwbWsymWjTpg2HDh0iLi6OqKgoCgoKSElJobS0lPLyctq3b8+hQ4cAaNeuHQcPHiQ2NpaYmBiOHz9OcnIylZWVlJaW0q5dO44cOYLNZiM2Npa9e/cSHR1NQkICx44dq+Hk2LFjHu+zupMwIKjtnbPc2rZty6FDhzzKrWXLlpSXl1NWVka7du0oLy9n7969Upe91GXnOepTl6uqqiguLq7hpHXr1hQUFFBVVeXVSUJCAiaTiaKiIp9OWrVqxeHDhxvsxL3svTlxL/uGOvFW9u71xL3svTlxL/uEhASUUuzdu9erk4iICFJTU2s4cS97b05q+384fPiwX068/T+4O/Hn/6FFixY1nKSmplJYWFirE2c98Vb27k68/T+UlZVRUVFRr8+I6mUvnxGnrr1ThmHU1UhUAPcahrHAbVk8UAxcZxjG0loPcHKfR4B7gd6GYeRUX79u3TojKyurxn6FhYW0aNHCn1NgsVgwm82UVdX+a3ZsZK3D5wUHTp+CHsSnXhrisz6fJ82NY8eO0apVq0YdY/PmzdtGjBhR84M8RAh2e1cfdJSXcBLxqRfxqRfxqZdAt3cN/aZnVPtbK0qp6cAs4EpvDY0unL9qX/n3/9W63Se39gtUCGGF06egB/GpF/GplxMnTkjj7Z0m2d5JeelFfOpFfOpFfOol0D79mS3wINC+2rL2butqRSl1P/AscJVhGJ/UL7z6IV+09CI+9SI+9SI+9dK6detgh9AUCJn2TspLL+JTL+JTL+JTL4H26c+3k3XApcAf3ZZdBpQBm2rbUSn1OHAPcJlhGJ83LET/cV5j8f6NfQAor7IxfskWAJZf15uYyLC6rVfAcfoU9CA+9SI+9VJQUEB8fHywwwg2IdPeSXnpRXzqRXzqRXzqJdA+/ck2ngP6KKWeU0r1VEqNBf4EvOCcOUkpNUAptV0pNcC5k1LqL9hnXroe2K6Uaud4pATgfQDgvH4sNjKC2MgIj2RKqZPLBU/OPPNM1qxZU2O5YRjcd999PPnkk34f64knnuD0008nMzNTY4SeDBo0iM8//xyAefPmccstt/jcdty4cSxebL8+fcmSJYwcOdK1Lj09nV9++UVbXIsXL2bmzJk+19d1fWN9WLBgAY899pi244UiOn0KuCYGaeaETHsn5aUX8akX8akX8amXQPusM7kyDGMzMBr7/Tt+BF4EFgCz3TaLA7o7/jq5C4gB/ol9OIXz8Z6GuL0SGRnpem61Gbz+3ck7ME9a9jOvfJuD1ab/C9n777/PxRdfTHp6Ol26dGHYsGG88MILfk0f35SJjIxk/vz5PPDAA35tn5OTw1//+lfWrVvH3r17AxbXhg0bGDZsmF/brlixguuvv97ruuzsbE4//XQA7rjjjkYlKxaLhaeffpp77rnHtSwlJYWdO3e6XrvXz/pQPSkEuOmmm1i+fDlHjx5t0DH95dVXX6VXr15kZGRw8803U1RU5HPbzZs3c/HFF5ORkcGAAQP46KOPXOv2799PSkoK6enprsd9993n81jz5s2jTZs2pKenk5GRwbBhw1i9erXHsWRae72kpaUFO4SgE0rtnZSXXsSnXsSnXsSnXgLt069xcoZhfGwYxlmGYUQbhpFmGMZswzCsbus/NwxDuQ+FcLz29hjW0GANw6CsyurzcaK0wvX85W9y+M+OY659q6wGH2zN4+Vvcmo9hvPh76/iCxcu5P7772fq1Kn8/PPP7N69m5dffpns7GwOHDjQ0LfaJKhvZp+Tk0NycjJt27YNUERNl48++ojMzEzS09N9bqPzl5LY2Fguuugili71a/KyBrF27Vqefvppli1bxpYtW6isrGTWrFlety0sLGTChAncfPPN/Prrr8ydO5dbbrmFPXv2eGy3a9cusrOzyc7OZv78+bWef/To0WRnZ/Prr79y7bXXctNNN1FQUOBaL7/k6SU3N7fujZoBTaW9qwspL72IT72IT72IT70E2mdIXYRUbrFx5d//5/Mxbtl21/P3t+ZRafVMkCqtBu9vzav1GM5HucVWZzxFRUXMnTuXZ555hquvvto1732PHj145pln6NKlC1CzV2Tnzp2kpKR4HGfGjBlkZWXRq1cvHnvsMSwWCwB79uxh9OjRnHbaaZx++unccMMNgD3RnD17Nt27dycjI4MhQ4awZYv9+rKKigoeeeQR+vTpQ7du3bj77rspKSlxnW/hwoX06tWLrl278uc//9nn+1NKecT+9ddf06NHD1555RV69uxJ165dXV+Q16xZw9ixYzl69Cjp6elMmTIFgO+++45LLrmETp06MXToUNauXes6/tKlSxk4cCAZGRn069eP1157zbXu2LFjTJgwgU6dOtG5c2cuvfRS15fp6sMYKysrue2228jIyOD888/nhx9+cK0bPXo0b7zxhtf35+xZWrRoEStWrGDBggWkp6dz5ZVXsnLlSoYOHeqx/eLFi7n00ur3CrXzySefMGTIENfryy+/HIDhw4eTnp7O4sWLUUrV6aNfv35kZGRw5plnsmzZMnbs2MHMmTP54YcfXD0+FRUVAJx33nl88on3a+aXL1/OmDFjXK+HDx9e47Wvfd3jue666zjjjDNISkriwQcf5J///CelpaU1tv32229p0aIF48ePJyIigosuuoj+/fuzfPnyWs/hDxEREdx4442Ulpby66+/upYrpQBYvXq1qx716tWLv/zlL40+Z3NEJggJLaS89CI+9SI+9SI+9RJonyGVXDU1Nm7cSHl5ueuLdEOZPn06Sik2btzIF198wbp16/jb3/4GwNy5cxk2bBi//vorW7ZsYfr06QB89tlnrF+/no0bN7Jv3z7eeOMN17SSjz76KNu3b2ft2rVs3ryZY8eO8cQTTwD23ohnn32WJUuWsGXLFg4dOuQzg/c2WcCxY8c4cOAAP/74IytWrOCpp55i27ZtXHTRRbz99tu0bt2a7OxsFi1aRG5uLuPHj+fee+/ll19+Ye7cudx8880cPGifdCs1NZWlS5eyb98+XnzxRR5++GE2b94MwEsvvURaWho7d+5kx44dPPLII5hM3qvrRx99xCWXXMLu3buZMGECkydPdiUg/jBlyhTGjRvH9OnTyc7O5v333+fSSy/l4MGDroQV4O2332b8+PFej7F161a6du3qer1q1SrAXk7Z2dlcf/31HDlyxKePkpIS/vCHP/D222+zf/9+Vq9eTZ8+fejevTvPPfcc/fr1c/X4REdHA9CtWzd++uknr/EMGTKE7777jqqqKoqKijh8+DBbt251vd66dSsDBw6s1cu2bdvo3bu363XPnj2xWq01eqMAbDZbjd5ewzDYunWrx7L+/fuTlZXFLbfcQk6Of7NUV1VVsWjRIhISElw/WMDJ+nnXXXfx5z//mf3797Nu3TouuOACv44reOL+g4/Q9JHy0ov41Iv41Iv41EugfYZUKhxjNrlmAvRGZUUlUdFRVFhsTFr2M1XWmkP7oiIUb07oRbS59rwypo71APn5+bRq1crjWpoJEybwzTffUFVVxfz5831+GXdy5MgRPvnkE/bs2UNCQgIJCQlMnz6dBQsWMG3aNCIjI8nOzubgwYN06NDB9YU4MjKS4uJiduzYwdlnn023bt0A+xfaxYsXs3btWleydd999zFp0iTmzp3Lu+++y8SJEznzzDMBePjhh3327Dh7z9wxmUzMnj2bqKgo+vbtS+/evfnpp5/o2bNnjW1XrFjB8OHDufjiiwE4//zzGTBgAB9//DFTpkzxuI5oyJAhXHjhhaxfv56zzjoLs9nM4cOH2b9/P126dKk1ETjjjDO4+uqrAbj99tt56aWXWL9+PRdeeGGt7msjOjqasWPHsnz5cnr37k1OTg6bN2/mzTff9Lp9QUEBCQkJtR5z+fLlPn2MGzcOk8nEtm3b6NixI23atKFNmza1Hi8hIYETJ05gs9lqJJ4dO3akdevWbN68mcLCQgYOHMjx48ddr3v27FnnHcZLSko8tlFKkZiYSHFxcY1tBwwYQH5+PkuWLGHcuHF89tlnfPvtt5x33nmA/YPs008/5YwzzqCoqIhHHnmE6667jrVr1/qc8e/DDz8kMzOTyMhIunTpwptvvklSUpJraKCzfprNZnbs2EGvXr1ITk6mb9++tb4vwTtHjhwJ6EQ0gl6kvPQiPvUiPvUiPvUSaJ8h1XOllHLN+OftERdtJjYyguTYSK7MSiUqQnnsHx2h+E1WKsmxkbUeJzYywjXkqDZSUlI4duyYx7Ufy5YtY+/evWRlZfl1wX12djZWq5XevXuTmZlJZmYmd911F3l5eYC9Fwpg1KhRDBw4kEWLFgH2L+a33nors2bNolu3bsyYMYOCggLy8vIoLS1l5MiRruONGTOGwsJCqqqqOHToEB07dnSdPz4+3mcG762nqGXLlkRFRblex8XFeQw5rP7eVq1a5YojMzOTdevWcejQIcA+nGvkyJF07tyZzMxMVq9eTX5+PgAzZsygU6dOjBs3jr59+/Lss8/6vA6uQ4cOrudKKTp06OA6R2OYOHEi7777LlarleXLlzNq1ChatGjhddvk5GSvSYc7Bw4c8OkjPj6e119/nUWLFtGzZ0/GjRtXo9enOsXFxSQmJvrs0RsyZAjr1q3j66+/ZujQoQwdOtT12n0Ioy/i4+M5ceKEx7ITJ054TSJTUlJYsmQJb7zxBt27d+f1119nzJgxrotGExIS6NevH2azmZSUFJ599ll27NjB7t27fZ7/iiuuYO/evezatYuPPvqI888/32O9833//e9/Z82aNZx55plcdtllbNiwoc73JtSkrh8HhKaFlJdexKdexKdexKdeAu0zpHqu6sI9IbrlnA5UOa6xAnuP1eisVG45p4Ov3evNgAEDiI6OZtWqVVx11VU+t4uPj6esrMz1+siRI67nHTt2JDIykl27dnmdTa5t27Y8//zzgH0Y4lVXXcWAAQPIysrit7/9Lb/97W/Jy8vj1ltvZf78+Tz66KPExsby1VdfeZ1coV27dh7DsUpKSlwJTXX8STBro2PHjowdO5aXXnqpxrqKigqmTJnCiy++yBVXXEFkZCSTJ092JVCJiYnMmTOHOXPmsH37dq666ir69OnDqFGjahzLfeIQwzA4cOAA7dq1a1TsAGeddRbJycl8/vnnvP322zz++OM+t83KyqpzWvfafID9Oqjhw4dTXl7OvHnzmDFjBp9++qnPcti5cydnnHGGz/MNGTKEFStWUFhYyOTJk8nPz+fpp5+msLCQ+++/v9ZYwT4McMuWLVxzzTWAfehjREQEnTt39rr9wIEDXTP6gf0HgUmTJnndVimFUqpR06k7vZx11lm8+eabWCwWXn/9dW644QZ27drV4OM2V3wl6ULTRMpLL+JTL+JTL+JTL4H2GVal5d5TFGFS3HzOyakW35zQi6nndiTC1LiEwZ2kpCRmzZrF73//e9577z2KioowDINdu3Zx7NjJmQrPOOMMVq9eTV5eHsePH3clS2BPnkaMGMHs2bNd++/du5evvvoKgJUrV7qSh8TERJRSREREsHnzZr7//nuqqqqIjY0lKiqKiIgITCYT119/PbNnz3Ylcbm5ua4vvWPGjGHZsmX89NNPVFRU8Pjjj/v8gtvYqa6vueYa1qxZw+rVq7FarVRUVLB+/Xqys7OprKykoqKClJQUzGYzq1ev9pjc4eOPP2bPnj0YhkFiYqLrvXljy5Yt/POf/8RisfDKK68A9nth1Yc2bdp4nT5+woQJPProoxQUFDB8+HCf+48cOZL169fXOOa+fftcr6+66iqfPo4cOcJ//vMfSkpKiIqKIjY21vV+W7duzcGDB2tcR/b11197TTadDB06lI0bN3Lo0CG6detG//79+emnn9iyZYtffiZOnMiSJUv4+eefOXHiBE8++SRjxowhLi7O6/b/+9//qKyspLi4mPnz53Ps2DEmTpwIwPfff8+OHTuw2WwUFRXx+9//nk6dOrmmwm8IVquVyspKVqxYQVFREWazmfj4eLmxcAOpbZp9oekh5aUX8akX8akX8amXQPsMq+TKOfuHczr18qqTM/4ZxsnlOrnzzjt58sknWbhwIVlZWXTt2pWpU6dy8803c+WVVwJw7bXX0r9/f/r3788ll1zCFVdc4XGMhQsXYrVaGTx4MJmZmVx//fWuSSZ+/PFHRo0aRXp6OhMnTuSxxx6je/funDhxgnvvvZfOnTvTp08fEhMTuffeewF47LHH6Natm+ueQ2PGjGHHjh0AjBgxgnvuuYcJEybQu3dv2rRp43O+/8bOptKxY0feeustXnjhBbp27Urv3r15/vnnsVqtJCYmMm/ePKZOnUqnTp149913ueSSS1z77t69m7Fjx5KRkeHqAbnooou8nufiiy9m1apVdO7cmTfffJN//OMfxMTE1CvWyZMns2fPHjp16uQxq9748ePZvn07V199da0+Lr30Un799VePXsFZs2Zx1113kZmZyVtvvcVpp53m04fNZuOll14iKyuLzp0788UXX7iS8PPPP5/evXvTs2dPMjMzqaiooKysjNWrVzNhwgSfMaWnp9OqVSuP6/R69+5Njx49SE5Odm3jaxjdhRdeyP3338+4cePIysrCbDYzb9481/r77rvP415VCxYsoGvXrmRlZfH999/z3nvvuYaQ7tu3jwkTJrjugXX8+HGWLVvWqETIWR7Lly+nb9++ZGRk8Le//c01GYxQP1JTU4MdglAPpLz0Ij71Ij71Ij71EmifqjHDcnSybt06Iysrq8bywsJCn9e5VKeyspKoqChG/e2HWrf75NZ+DYqxueH02Zypqqqie/furFy5kj59fE+mAvDmm2/yww8/8Nxzz3ldr9PnwoULOXToUKNufBzqNMRnfT5Pmhu5ubmNvrHi5s2bt40YMaLmB7ngga/2rj7oKC/hJOJTL+JTL+JTL4Fu78LqmqumkiiGC+LTfm8rZ+9gXUyePJnJkyf7XK/T5+23367tWKGK1E+9VFZWBjsEoR5IeelFfOpFfOpFfOol0D7DKrlyTghR23Ttgv94m2CjOdG/f38qKipcMzQ2lubuUzfiUy/yq2hoIeWlF/GpF/GpF/Gpl0D7DKtrrpxTotc1zbrgH+5TzDdHNm3axJYtWzj77LO1HK+5+9SN+NSLr5uJC00TKS+9iE+9iE+9iE+9BNpnWCVXjZ06XPBEfOpFfOpFfOpFZlkMLaS89CI+9SI+9SI+9RJon2E1LNA5e5ilpKz27eJjT0U4IU9jZwsUPBGfehGfepHZqEILKS+9iE+9iE+9iE+9BNpnWH07qaqqIjo6mjVdRtS63SWH1te6XrDj9CnoQXzqRXzq5fDhw2RmZgY7DMFPpLz0Ij71Ij71Ij71EmifYTUsUO5grRfxqRfxqRfxqZf4+PhghyDUAykvvYhPvYhPvYhPvQTaZ1j1XDmvwbho96cAWEvLWHuG/Ya9F/70IRFxMhywPsg1LXoRn3oRn3qRYZahhZSXXsSnXsSnXsSnXgLtM6x++rVarYD9mipzfKxHMhURF+taHo7s37+flJQUysvLva4fNGgQn3/+uV/HKisr47rrrqNLly5MnDhRY5QnycnJIT09nYqKCgBGjx7NG2+84XP79PR0fvnlFwDuuOMO181zN2zYwFlnnaU1tmnTpvHee+9pPSacrJ/uTJo0iTVr1mg/V3PAm0+h4RQWFgY7BKEeSHnpRXzqRXzqRXzqJdA+wyq5qi0TtZZ5Tzp0cfXVV9OxY0dOnDjh1/Zff/01PXr0CGhM7mzYsIFhw4b5te0HH3xAbm4uO3bsYOnSpQGJp2PHjmRnZ/t9zUx2djann356jeWDBg1i8+bNrtdnnnlmo5KV7du3s3nzZq666ipAbzndc889rqTQyd13380TTzyh5fi+qKys5N577yUzM5OuXbsyd+7cWrdfvnw5AwYMID09ncsvv5xdu3a51i1ZsoTU1FTS09NdjxUrVvg81ujRo2nfvj3p6el06dKFa665xnW8JUuWMHLkyAa/L/klTy8pKSnBDkGoB1JeehGfehGfehGfegm0z7BKrmw2m+u5YbWyc+7Lrtdf9B/D9sf+ihGAX7tzc3P54osviI6OZuXKldqPf6rJzs6mS5cuzXLqz0WLFjF27NhTdj3PgAEDOHHihEeCqJtnnnmGn3/+mU2bNvHpp5/yzjvv8NZbb3nd9ptvvuHBBx/ktddeY8+ePZx33nlMmjQJi8Xi2qZfv35kZ2e7HuPGjav1/HPnziU7O5v//ve/tGjRgjvuuEPL+3L/fxcaT2lpabBDEOqBlJdexKdexKdexKdeAu0zpJIrwzCwlJT5fFQVl7ieb3v4BXLefN+1r62ikv1vvMe2h1+o9RjOh2EYfse1fPlysrKyuPXWW2v09OTm5jJlyhS6detG586dmT59OiUlJVx77bUcPXrU9ev/L7/8wrx587jllltc+5aXl5OSksL+/fsBWL16NRdccAEZGRn07t2buXPn+h2ne4/OvHnzuOGGG7j77rs57bTTOOuss1zr5syZwzPPPMO//vUvOnXqxP/93/8B8NZbb3HuuefSqVMnxowZw969e13Hnj17Nr179yYjI4Nhw4axbt0617pNmzYxYsQIMjIy6N69O7NmzQK8D2Pct28fI0eOJCMjg/Hjx5OXl+dal5KSws6dO2u8L/eepWnTppGTk8MNN9xAeno6Tz75JL/73e9c53Qybdo0n71Fq1evZujQoQA+y6k2H4ZhMHv2bLp3705GRgZDhgxhy5YtLFq0iHfffZcFCxaQnp7OlVde6TrnkCFD+Pjjj73GM2PGDJ577jkAioqKaN26tcfrdu3aUVRU5HVfJ0uWLOF3v/sdrVq1IiMjg2nTpvlMrv79738zevRozjjjDCIjI7n//vvZt28fGzZsqPUc/pCQkMD48ePZtm2b1/V/+ctf6NWrFxkZGZxzzjl89tlntR5Pkiu9+BpSLDRNpLz0Ij71Ij71Ij71EmifITWuxlpaXuc067VhK69g/2sr2P+a72FMTi7a/anf12ctW7aMSZMmcdlll/Hss8+yZ88eOnfujNVqZeLEiQwcOJBNmzYRFRXFpk2biI+P5+233+bWW29l+/btfscfHx/PwoUL6dGjB9u2bWPs2LH06tXL44u6v3zyySe8+uqrzJ8/n9dee40ZM2bw888/89BDD2E2m9m1axevvvoqJpOJVatW8eyzz7JkyRK6du3KX//6V2666SY+++wzlFL07duXmTNnkpSUxMsvv8xNN93Ejz/+SFxcHA888ABTp05l/PjxFBcXs3XrVp8xLV++nHfeeYdOnTpxxx13MHPmTP7+97/7/Z5efvllNmzYwHPPPcdFF10EwObNm5kwYQJz5szBbDZTUlLCqlWrWLt2bY39S0pK+PXXX13DD32VU20+PvvsM9avX8/GjRtJSkpi165dJCYmMmXKFDZu3Ejbtm155JFHPM7brVs31q/3fnuAoUOHsmzZMmbOnMmGDRtIT0/n66+/dr3OysoiKSnJp5OCggIOHjxIr169XMvOOOMMnnzySa/b22w2j4Td+fznn3/mvPPOA2Dr1q2cfvrptGjRgtGjR/OHP/yB2Ni6/1eKiopYvnw5ffr0qbHOWd/WrFlD+/btyc7OpqqqqtbjRUZG1nlOwX/S0tKCHYJQD6S89CI+9SI+9SI+9RJonyHVc9UU2bhxI7t37+aaa67h9NNPp1+/fq7eq82bN5OTk8OcOXNITEwkOjqawYMHN/hcgwcPJisrC5PJRK9evRg7dqzPL+V1cc455zB69GgiIiKYNGkShw8f5uDBgx7bOL/cLlq0iLvvvpuePXtiNpu5++672b17t6sXZ9y4caSkpGA2m7nzzjupqqpy9TJFRkayZ88e8vLySEhIYMCAAT5juuaaa+jVqxdxcXHMnj2bVatWuSa8aChnnXUWqampfPqpfQbJDz/8kB49eni9fst5gWNCQkKtx6zNR2RkJMXFxezYsQPDMOjWrRvt27cHfPe0JCQkUFBQ4HXdkCFD+O6776iqqmLdunVMmzaNrVu3ul7XVZ+Ki4sBPBKwFi1auJZXZ9SoUXzwwQf88MMPVFRU8PTTT2OxWCgrs9+Ye/Dgwaxbt46dO3eyfPlyvv76ax599NFaY3jooYfo1KkT5557LpWVlSxYsKDGNhEREVRWVrJ9+3aqqqpIT0+nc+fOtR63ruRLqB+5ubnBDkGoB1JeehGfehGfehGfegm0z5BKriLiYrho96c+H+dv+zcX7f6UC7eswhQd5fUYpugoLtyyqtbjXLT7UyLiYvyKaenSpQwZMsSVBV9zzTUsX74cwzDIycmhQ4cO2n5h//777/nNb35D165dOe2001i0aBHHjh1r0LHatGnjeu6c77+kpMTrttnZ2fzxj38kMzOTzMxMOnfujMVicSVjL774Iueeey6nnXYamZmZFBUVueJ64YUX2LNnD4MGDeLCCy/kww8/9BlThw4dXM/T09Ox2WwcOXKkQe/PnQkTJrBs2TLA3ss4YcIEr9u1aNECwGfi4aQ2H+effz633nors2bNolu3bsyYMcNn4uSkuLiY5ORkr+s6duxI69at2bx5M+vWrWPYsGH06tXL9do5hNEXzkTRfaKVoqIinwnk+eefzyOPPMK0adPo1asXZWVldO/e3VW/ne/ZZDJx+umn88gjj/D+++97PZaTOXPm8Ouvv7Jt2zYWL17MaaedVmObzp07M3fuXJ555hm6devGlClTyMnJqfW4gl7kvmGhhZSXXsSnXsSnXsSnXgLtM6RKSynlmk7d2yOmRRLm+FiiU1uScfPVmGI8Z6IzxUSTcfPVRKe2rPU45vhYv+6hU15ezsqVK9m0aRM9evSgR48ePPPMM+Tk5PDll1/SoUMHDhw44PUXdm/Hj4+Pd/UQADUSi6lTpzJq1Ch++ukn9u3bx5QpU+p1bVh9cSaFHTt25Omnn2bv3r2uR25uLueffz4bNmzg+eef54033uDXX39l7969JCUlueLq0qULr776Kjt27OD+++/nlltu8ZlsHDhwwPU8JycHpZRHEugP3ryOHz+eNWvWsH37djZu3MjYsWO97hsfH0+nTp3YvXt3rcerzQfAb3/7Wz777DO++eYbsrOzmT9/PoDPCUJ27txJ7969fb6nwYMH89FHH3Ho0CG6devGkCFD+Oijj9iyZQuDBg3yLQNITk6mffv2bNmyxbXsp59+omfPnj73mTJlCt9++y2//PIL999/P9nZ2T6nuzeZTNrq4DXXXMO///1v/vvf/xIdHc1DDz1U6/YyLFAv9f1fE4KLlJdexKdexKdexKdeAu0zpJKrunBPYro/NJ2Ok37jem2KjiLjprF0f2i6tvOtWrUKq9XK+vXr+eKLL/jiiy9Yv349l19+OUuXLqV///6kpaXx8MMPU1xcTEVFhWsYX+vWrSkoKPBINPr06cOGDRvYu3cvJSUlPPXUUx7nKy4upkWLFsTExLBp0ybeffddbe/FG06fU6ZM4fnnn3ddL1VUVMT777+PzWajuLiYiIgIWrZsicVi4emnn/boJXn77bfJy8vDZDKRmJiIYRg+k4x3332XrVu3UlpaypNPPsnll1/u91TtTlq3bs2+ffs8lrVt25bBgwdz2223MWLECFq2bOlz/5EjR3pMyOGtnGrzsXnzZr7//nuqqqqIjY0lKirK9X5btWrlMRGIk/Xr1zNq1CifMQ0dOpTXXnuNgQMHAvahgq+99ho9e/Z09bbVxsSJE3nuuec4duwY+/fv5+WXX2bSpElet62oqODnn3/GZrNx+PBh7rrrLq644gq6du0K2Cf8OHToEAB79+7lscce4/LLL68zhrrYtWsXX375JRUVFcTExBATE1PnbJUyLFAvznIVQgMpL72IT72IT72IT70E2mdYJVfObj5LSRnW8kq63HOja93Qr5Zw+v23Yi2v1Ha+pUuXct1115GRkUHbtm1dj2nTpvHhhx9SUlLC0qVLycnJoW/fvmRlZblmaevWrRvjxo2jf//+ZGZmsnv3bi644ALGjx/PhRdeyODBg109IU6eeeYZnnnmGTIyMnjmmWdc92IKFE6fV1xxBffddx+33XYbGRkZDB48mP/85z8opRg+fDijRo3i3HPP5cwzzyQmJsZjeN9nn33G4MGDSU9P53e/+x2vvPIKiYmJXs937bXXcvfdd9OjRw+Kiopcs+LVh3vvvZe//OUvdOrUySM5nThxIj///LPPIYFOpkyZwrvvvuvqjfFWTrX5OHHiBPfeey+dO3emT58+JCYmcu+99wL2Gwbv2bPHNcMg2Id6xsXF0b9/f58xDRkyhOLiYtcQwLPOOgubzea63so50YUvfv/739OzZ0/OOusshg8fztixYz2Sq0GDBrnuVVVZWcntt9/OaaedxtChQ2nfvr2r5w3gq6++YtiwYXTs2JHf/OY3DBgwgMcff7xWp/5QWVnJn/70J7p27Ur37t05ePBgjXuCVUeGSejFn0lJhKaDlJdexKdexKdexKdeAu1TBXJYWX1Yt26dkZWVVWN5YWGhX7/OA1gsFsxmMx+1q/0i/0sONWwSiOaG02c48MMPP3DttdeydevWOoeT3X777YwcOdLn8MGG4s3n5MmTufHGGxt1M93mSkPqZ30+T5obOtxs3rx524gRI2p+kAse+Grv6oPUZb2IT72IT72IT70Eur0Lj2/ODqxWa9gkA02BcPFpsVhYuHAh1113nV/X6SxcuDAgcXjz+eabbwbkXM2BcKmfTYXjx49L4x1CSHnpRXzqRXzqRXzqJdA+w+qbifMajYt2fxrkSMKDuq55CQWcwy179OhR4xq2U004+GxKiE+91HYtotD0kPLSi/jUi/jUi/jUS6B9hlVy5Rzi6O/Nf4XaaSpDRhtDly5dmsyU3uHgsykhPvVSXl4uv4yGEFJeehGfehGfehGfegm0z7C6ItzXTVqFhiE+9SI+9SI+9eJ+Gwih6SPlpRfxqRfxqRfxqZdA+wyJ5MrfX6jlvjd6EZ96EZ96qa9P6emqnXbt2gU7BKEeSHnpRXzqRXzqRXzqJdA+m3xyFR8fT2FhoV9fjOS+N3oRn3oRn3qpj0/DMCgsLCQ+Pj6AEYU2hw8fDnYIQj2Q8tKL+NSL+NSL+NRLoH02+WuuzGYzCQkJFBUV1bltSUmJfHnSiPjUi/jUS319JiQkyOyCgiAIgiAElJD4pmE2m/268CwmJobo6OhTEFHzQHzqRXzqRXzqRYadhBZSXnoRn3oRn3oRn3ppEsMClVKjlFKblFLlSqkDSqk5Sqk650FWSk1SSm1VSlUopfYope5pdMS18Pzzzwfy8M0O8akX8akX8akX8WlH2rvmifjUi/jUi/jUS6B91plcKaX6AR8CXwL9gLuAO4En6thvNLAY+AdwJjAXeEopNa2RMftkxYoVgTp0s0R86kV86kV86kV8SnvXnBGfehGfehGfegm0T3+GBd4P/GQYxr2O19uUUh2AeUqpOYZhFPvYbxbwT8Mw5jleb1dKZQEPAC83KmofmExNfn6OkEJ86kV86kV86kV8AtLeNVvEp17Ep17Ep14C7dOfow8B/l1t2SogFujvbQelVCRwjo/9MpRSp9UzTr+wWq2BOGyzRXzqRXzqRXzqRXwC0t41W8SnXsSnXsSnXgLtU9U1xblSqgK41zCMBW7L4oFi4DrDMJZ62ac9kAtcbhjGv92W9wK2AIMNw9jgvs9HH31UduTIEVeyl5CQkN+iRYvj9XkzhYWFLeu7j+Ab8akX8akX8akXTT47jhgxIklLQEFA2rvmi/jUi/jUi/jUS6Dbu4bOFmhU+1sd5ef+Li655JLYBsYiCIIgCIFC2jtBEATBb/wZFngQaF9tWXu3dd44Cli87Oec+9DXfoIgCIIQLKS9EwRBEBqFP8nVOuDSassuA8qATd52MAyjCtjoY7/9hmHsq2ecgiAIghBopL0TBEEQGoU/ydVzQB+l1HNKqZ5KqbHAn4AXnDMnKaUGKKW2K6UGuO03DxijlPq9Uqq7Uupm4A7gSd1vQhAEQRA0IO2dIAiC0CjqTK4Mw9gMjAYuBH4EXgQWALPdNosDujv+Ovf7F3ADcBPwE/BHYJZhGA2aljZUbuwYKjTEp1JqkVLKqP44VTE3VZRS5yul3ldK7XM4edTP/aRueqEhPqVu+kYpdZNSaq1S6qhS6oTj/36SH/s1u/rZFNo7aev0Im2dXqS904u0d3ppKu2dXxNaGIbxMfBxLes/x8tFvYZhvAW81dDgnKiTN3Z8CZgMZAGvYY9/Vi37OW/s+CCwEhgKvKSUKm9okhcONNSngw3A2IAGGHokAFuBJcDz/uwgdbNW6u3TgdRN74wAPgB+DxwHxgCLlVIWwzCWe9uhOdfPYLZ30tbpRdq6gCDtnV6kvdNLk2jv6pyKvSmglHoL6GEYRn+3ZXdhH4rRxteNHZVS64BDhmFc7bZsPnC1YRgBufdIKNAIn4uATMMwhp2KOEMRpdReYJFhGI/WsZ3UTT+oh89FSN30G6XUv4Ey9/pXbb3UzyAgbZ1epK0LLNLe6UXau8AQjPYuVG75HDI3dgwR6u3TjbOUUgeVUr8qpd5V9nu5CPVA6mbAkLrpP8lAnrcVUj+DirR1epG2LshI/QwYUj/9J5lT3N6FSnLVnprT2R5y/E3zsU8qENmA/ZoDDfEJ8BH2oRUjgVuxX3PwnVKqt/YIwxupm/qRuuknSqkbgbOB//OxidTP4CFtnV6krQs+Uj/1I/XTT4LV3jX0JsJNAe03dmzm1OUTwzCWub3copT6CtgG3AVMDWBs4YbUTc1I3fQPpdSV2BuZ3zomb/C6WR2Hkfp5apG2Ti/S1p1apH5qRuqnfwSzvQuVniu5saNeGuKzBoZhVAI/AJl6wmo2SN0MMFI3a6KUmgAsB6YZhvFGLZtK/Qwe0tbpRdq64CP1M8BI/axJsNu7UEmu5MaOeqm3T28opczAGUC2vtDCH6mbgUfqpidKqduARcAUwzAW1bat1M+gIm2dXqStCzJSPwOP1E9PmkR7ZxhGk38AZwGV2G/w2BP79JPHgXlu2wwAtgMD3JaNBqzYp2TsDtwMlGPPZIP+vkLJJ/bpQp/DfoFwJnAu8K7D51nBfk9B9pkA9HU8coGXHc+zvLl0LJO6qcmn1M06fd6H/Ze532L/Nc75aO3Np2OZ1M/glJW0dUH2KZ8ndTqV9i6IPqV+1umzSbR3QRdRD2EXA5uBCkcFfAKIcFs/DPvYyGHV9puEfSxqJfArcE+w30tTeNTXJ/bZlT7CfpFfpWOfD+Sf2cNV9cdeby7d9pO6qcGn1M06fe6V+hk6D2nrgutTPk/q9CntXRB9Sv2s02eTaO9C4j5XgiAIgiAIgiAITZ1QueZKEARBEARBEAShSSPJlSAIgiAIgiAIggYkuRIEQRAEQRAEQdCAJFeCIAiCIAiCIAgakORKEARBEARBEARBA5JcCYIgCIIgCIIgaECSK6FZopR6VCllKKUygx1LbSilkpRSryqlch3xrgx2TKcCpdQUx/sdFuxYBEEQQhlp75o20t6FH5JcCQ1CKTXM8WEwzW1ZpuNDvG8QQ3OhlOrriCcz2LE0goeBW4FFwPXAn4MajSAIQjND2rtThrR3QlggyZWgk0zgEaBvcMNw0Rd7PJle1s3BfqfzfacwnoZwEfBfwzAeNAzjTcMwvgh2QIIgCIK0dwFA2jshLJDkSggJlFIxSimzruMZhmExDKPcMAxD1zEDRBugQOcBlVIRSqk4ncdsQAyJwTy/IAhCU0XaO31IeycEA0muBC0opaYAax0v33AMoTCUUovctolTSj2mlNqhlKpQSh1VSr2plEqvdizn+PAzlFIvKqUOAqVAR6WUSSn1kFJqnWP/CqXUTqXUI0qpSPdjAG84Xq51i+fRaufIrHbus5VS/1ZKFSilypRS3yulJnl5v3uVUp8rpXorpdYopUoc8fxVKRVVbds+SqmVSqmDjngPKaVWKaXOqM2nUsoA2gMXuMU/zLFeKaVmKKV+UkqVK6WOKaVWKKW6VzuOczjLTUqpe5RSO4Fy4Fql1FuO/Uxu29/m2P5f1Y7zd4eTCMfrc5VS/1BK/eLwdNzh7exaXJ2jlPpMKVUM/Mtt/TVK/X979xujR1XFcfz76ypVUwltEbQkSCGQQoIRsCS8qduIAlHTJmgTCmpNAFFEqAgGUCyJNTY1RBFNEIFAixFCRH2j1BW2xQqVRipiqSm1YChQiq3U0to/2+OLcx92Oju7Prs8LDT7+yST6d6ZuXNnppmzc/fMHT1RjmODpPmABjs3ZmZvJsc7xzvHOxtKx3pGbMxbAXwHuBb4CfBwKd8AUG7AvwNOAW4D/gocDVwKdEs6NSJeqtW5lOzF+i4wHtgBHALMB+4p037gY8ACMh3i82XbX5A36otLu54q5U8MdgCSzgAeBLaTud7bgbnAUklTImJxbZOjyjH9HLgPOAf4MvCv0h4kHQ78HngF+AHwEvBeoBuYVs5DkxVkzvnNwAvAwlLeOo7vA18BeoFbS52XAmdKOj0i1tfquwJ4N3nutwJ/B95eju8DwJqyXjd5TmdI6oqIvlL+YeDhys+zgfcDdwHPk9fyC8CKci3X1fZ/NLAM+Fk5X/vL+fk0eR3XAd+k//q+PMh5MTN7szneOd453tngIsKTp2FP5E0pgEsayuY1rH8VsA84o1b+wVK+uFK2oNSzDBhXW1/AOxvqvwPoA46qlM0r9XQ3rN/axzGVslXALuC4Stl4YDXZ+3VEpfyZsv2sWr1rgM2Vn2eV9aaP8Dy/CPTWyk4qdf4W6KqUTydv4vc3XJMtwKRaPceXZVdUyjYBd1fbDEwtP19ZWe9dDW09AdgN3FIrb52rz9TKu4DngH8Ch1bKp5CBvvHaefLkydNoTo53jncNbXW88zTo5LRAGy3nAY8D6yUd3prIm8168kXWupsjYn+1INIuAElvkzSx0ls2DjhtJI2TdCRwOnBvRGyo7G83cCMZdM6ubfZcRPyqVrYcOEL9Odb/LvPZksaPpG0NPlnmi6K/Z42IeAzoAc6pp2oASyJia7UgsrdvExmQkHQCeaP/Ednj2F1Wbc17K9vubP1b0gRJk+nvIZze0OaXySBW9SGyN/SnEbG9UvfzZI+fmdnByPHO8c7xbgzzw5WNlmnkzWVLwzSNfJG17ummiiRdIOlxsndta6ljSVl82AjbN7XM1zYsW1tbp+WZhnW3lfmkMl9B3jivBbaVfPWvSZoywna209bxZIpIVeO5JIPjjJKHPhN4FXislM8s63STaR5rWhtJel/JS98K/IcMJluAk2m+BhvrvzhUjqOeUgH96SBmZgcbxzvHO8e7MczvXNloGQf8CbhukOW7G8p21gtKzvISYCXwRbInag9wKrCIN7bDoD7SUl/jWkmQPY/A+ZIWAx8nc7kXAt+SNCsiHnxDWjqwrQPOZdFLfx76TGBlROyV1AssUo5YdUD+eQlMD5BpFjcCfyYDzn4yN35Cw36a9j/US7x+wdfMDlaOd453dY53Y4gfrqyThhrmdT0wOSJ6Xuc+5pI9eGdGxH9bhZKOHWZ76jaW+UkNy06srTNsEbGG7AlbqByx6S/kN0lGEmyqbd1cW3YieX5ebLOu3jLvJoPKTZXyCcCnyBd5f1jZ5uQyLYiIG6qVSZoI7G1z3/8o82kNy5rKzMzeKhzvBuF418jxbgxxWqB10o4yn9iw7G7gOEkX1heUYVbf0+Y++sgeo67K9u8ALhtmew4QEZvJF3znSHotHaLkcs8nexofaLONr5E0SVK9V+pZcoSlycOtr/h1mV9VG1b2NOCjwG8iYk87FVXy0C8mR2B6qJSvJQPZ9WXV3spmrR7MA+4fki4gc9jbtZoceelCSYdW6plC/lJhZvZW5XhX43g3JMe7McR/ubJOWkve4L8k6VVyBJyNEbGK/HP6WcCtks4G/kCmN0wlRxi6F/hGG/v4JXAu0CNpCTnc6ufIUY/qVpOB6ZrSw7QTeDIinhyk7vlkz9ojkn5c2n8e+cLq1TFw6Nx2fBa4XNL99OeBzyKP+5oR1EdEPCXpJnJo2p5S95HksLivAF8fZpXLyZv7DvKcVcvnUMs/J3PG1wFXl0C/gXy/4Fyyd66LNkREn6SvkkPVPirpdnJo2kvIc3XKMI/DzGy0ON4N5Hg3+HE43o0hfriyjomIXZLmAt8m/6x+CHAnsCoi9kg6C7gcOJ/Mx95Ljp60jLzhtLOPpSVwXEYGsBfIYWn/SH6Do7rus5IuIm++t5D/328AGoNNRDwiaUZZ50ryRdm/kUOqLm33PNQsJ0d0mk32lO0mRxiaFxF3jrBOyO94PE1+a+N75Iu5PcB1MfCbH//PQ2SwWRkR+2rlczjwex9ExD5JnyC/jXIR+f2QR4GPkDnox7S744i4R/nxyOvJ3PxNpY5twO3DPA4zs1HheNfI8W4Ijndjh/L9QzMzMzMzM3s9/M6VmZmZmZlZB/jhyszMzMzMrAP8cGVmZmZmZtYBfrgyMzMzMzPrAD9cmZmZmZmZdYAfrszMzMzMzDrAD1dmZmZmZmYd4IcrMzMzMzOzDvDDlZmZmZmZWQf44crMzMzMzKwD/gftnGu4wkMIvQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1050x337.5 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC:\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZwAAAEvCAYAAABxH3fiAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8/fFQqAAAACXBIWXMAAAuJAAALiQE3ycutAABNZ0lEQVR4nO2deXxU1dn4v0/2jSQkIQnI6saiFRVXsO5aN2q14luXWq3tW6utonWrVqu2WmvVWrW+tr9WbUFBfX0VFZfWBaxI1YrUBZAqImAIIWTfM5nz++PeGSaT2XJmhmGS5/v5zGeSe8+588z33rnPXc49R4wxKIqiKEqyyUh1AIqiKMrwQBOOoiiKskPQhKMoiqLsEDThKIqiKDsETTiKoijKDkETjqIoirJD0ISTxojIXiJi3NcdqY5nKCAiE12fS1Idi7IdEbnWXS8zIpQZKSLdbrknoizvEbfcTVHK+X5fE8PMLxSRuSLymojUikiPiDSJyDsicpuI7BZU/ip3eQdF+tyhiiac9Oa8gL/PFhFdn8qQQ0QqgJ8Czxpj3otQ9FtAjvv3bBEpSXJchwL/AX4LHAKsAv4XeBMYhxPzahE5LaDa74E64M5kxrazojuoNMVNLme7/24GdgGOTl1EipI0fgoUA7+KUu7b7nsNkAfMSVZAIrIv8BowGrgDqDTGHG2MOdsYcwowBjgZJyFN8NUzxnQA9wJfFZETkxXfzoomnPTlaGAs8B5wtzvt2+GLK0r6ISL5wHeBtcaYf0YotztwKLAF+Ik7OSm/BxERYB5OUrvJGHONMaYtsIxxeAE4EFgStIj5gAF+mIz4dmY04aQvvh/TfOAxwAucLiKFoQqLyBLftWgROVtE/ikirSLSFFBGROQs93p0o4h0ichqEblJRApCLHN3Efm5iLwlIpvd69c1IvKEiOw/mC8jIudHuqYecM39yKDpRkTWi0iGiFwpImvcuGtE5PciUhxmeeNEZJ6IbBWRDhF5T0TOjSHOE0RksVuvW0TWicjdIlIeKWYR+ZqIvO5e3zciUhqjl4NFZKGIfOl+3mYReVVEvh/KQ5hlHOnOf2QQ8VWKSK+IbBGRrDDLPdot+2rQ9EFtR1E4EygFHo1Szvd7WAg8DTThnEVMHOTnxcIJwN7AJuDWSAWNMR3GmJVB074AlgEniciYJMS306IJJw1xk8rpQB+wwBhTA7wKFAGnRaqLc3liHtADPA985C4zg+3J60BgJfACUAj8HHg9xM7ie+68EThnWs/gXJ+eA7wlIsfE8TUHy3zgZuAz4CUgG7gYWOQekfoRkUnAO8C5QAuwCGgH/sr2o+MBiMjtwIvAscAnwLOAB7gceFtEqsJUPdutV+i+v4tzhBsREbkMeAv4L6AW+D+c9bU38Jto9QdBqPi6gVeASpzvG64eBCQDy+0oEqe4769HKec7WJhvjOkGngQkYHoiOdl9f9IY47FcxutAJk7yGj4YY/SVZi+cH5EBXgox7eUwdZa48zuBI0LMv8qdvxQYHTA9B/iTO+/2oDqHALuGWNZJOAltLSAxfqfz3c+4Kcz8R9z5RwZNN+5rLTAuYHolsC5MnRfd6X8GsgKmn4KTQAywJKjOHHf6x8CeAdMFuMmdtzBMzAb4r0Gu48NxzlpbgOOD5mUBJ4XwsD7Mso505z8ymPiAc9x580LMywUa3e2pJJ7tKIqHze46KYhQ5jB3uWtCTPskyvYUcnsLsX1NDJj2pjvt3MGs06DlnuIu4y+2y0jHV8oD0JfFSoO/BW/wOEeQbThnPaND1Fni1rk/xLwsYCvOUX6ouvnuD78ByIgxxkfdz/tKjOXPj7QDIHrCOSFEnSvdeT8PmLarO605cEcZMH8hoRPOSnf69BB1BHjf3TFWhIj5eYt1/IJb9+IYy8eTcELGBxQAre52VRg073S37hPJ2o5wDhoM8HmUcn90y/0saJ34DjgOirA9hdzeQmxfEwOmrXanfW2w6zVgGbu7y/i37TLS8aWX1NIMERkNHIPzo37aN90Y045zySUD58g0HM+GmLY/UAG8ZYzZHDzTGNOJc8lsJLBHUDyFInKmiNwuIv/PvS/wCM5lH4LLJ4lenMs/wXzivgdeJz/MfX/JGNMcos6C4AkiUglMBz41xvw7eL5x9iDLcC6RhHpOJJTzsLj3TI7EOXiYN5i6loSMzzgtqp7BOZg5NWi2bxsLvLdivR2FodJ9bwhXQERy2X726Y/FXSfz3X/PC1E1HiR6kaj4vtOoBCwrbQh5M1DZqTkXJ6k87SaZQObh3Dz9NuHb+W8IMW2i+36siJgon1+BuyN3b+AvBMLduwDn/k6yqTWhr6W3uu+5AdN8yeeLMMtaH2LaRPd99xj9BBPKeSTKcc4GvjTGtEYrnAAixTcfZ5s7B+e+DOI833Iyzk7zxYCyE933QW1HEShx3yM5+DpOo4I3jTGfB82bB9wAfEtELjfG9AbMixYfQff+AsvXA5OJL1m0uO+lcSwj7dCEk374WuMcKiJvBs3znbHuIyL7GGM+CFG/K8Q0X71PcY7UI7EN/A0XnsTZcdyKc2bwBdBujDEichtOA4VEHA0GxhgK7yCWYxOP77NrgZejlA2VyEI5j4WoO8UYiXYlI1J8r+B87+NFZJQxZitwBk4Sf9gY0xPic2LejqLQ5L6HbGno4vs9TArxewDnMmc5cCL9z+Q63feQrTpDzAs8uFsJzMI5o5uPHb5k2mRZPy3RhJNGiPOw2Vfcf3dzX+H4Ns4N3FjY5L6vMcacH2Odw3GSzVPGmJ+FmL97jMvx4dtxFYWZP26QywtHjfs+Icz8UNN9fuoH4Sce6nF2iKNFZESMZzm9JMGdMaZPRBYCc3Fay91P6MtpYLcdRaLOfS8LNVOcHgh8rbx2cV/h+Db9E85G933XCHV88zpwGkj4WAxcAswRkavDnF1HY6T7vtWibtqi93DSC9/R3H3GGAn1AvZzy5wjIpkxLvddnJvoR4hIyB93CHw/mI3BM9wdwXExLseH75r/5BDLK8c5mkwEvqPgE8I8o/Ot4AnGmE3AGmCaiOyZoDjCYozpw2nkkUnk+3GBbAbKXffBHB9nSL6j+HNEZBfgCJxLj8FnMTbbUVjcs6kaYKyEfr7sLJzm789F+D2MwjnLmR307NNS9/3oMMuG7fet3nTXiY+XcForjgWuj/QdRKTAPVAMZqr7vjJS/aGGJpw0wU0evuceHgtXzjgPma3G6XIjpudgjPPcwh0491v+T0QGHPWJyB4i8t2ASb7r798MfP7E/fH+icFfm34X57LFiSJycMDyRrjLi3RZJWaMMZ/htPIrBu4KTMoichLhu0O5Bef38lSoHYiI7CIiP0pEjC6/xrmk9uvg55lEJMuNNZAl7vuNQWW/jbNjtsY4/ZetxmkG/zMcD4+6N+YDy9lsR9F4g/CNMXyNASL9HupxLgvm4jxE6pv+Fs42Vwo8KE6PBoFxfpXtVwjuCVqmwbmv1QXc5DaYKQqqLyJyPPA2TgOQYHyddy4NMW/okupmcvqK7YVz6cAAn8VQ9kaCnp9ge7PoiWHqZOA8+GhwHvp7G3gc557FJ+70lUF1XmJ7E+NFwFM4lwi2AA+7884fxHf0xd2FkxSex7ms8glOa6lwzaLXh1nekYRuDrwrzn0Jg3O/YQHOD9+Lc8loQLNot96t7rw+nNZWT+I0X/7IndYUVP6RUDEPwseVbkwG+BfOjvVvrt/gz5qKk7AN8IEb2wc4R/d3hfEQc3w4R/Im4DU1UdtRlM/1PQv086DpU9zpbUR4Rscte55b9h9B03fHuedm3O32GZzLhG8HeP9lhOXOxDmzNK77V936zwFfBjj4Roi6b7rrZpd49w3p9Ep5APqKcUVtf64l7A8goOweAT/GInfaEiIknIC6p+Ls6Lfg3FfZ4u7sfg3sH1Q2D+fp/jU4SaLG3YmNY/vDkOcP8nte5i6v2/0x/wHnGn7InSMWCcedN8F1ug3nfsn7wHdwWlqFTDhuvcOBJ9wdSg/O/ZZ/A/cBhweVDRnzIH18Fae5u2991OAcsV8YouyBwN9xWnW14jzNfkQ4D4OJz/Xl2wm/F0P5mLejKMvJc9fRJ0HTfcl/fgzLGOGuYy9BDyq729ZNwArXmc/x/wHHxrDsQpyeJl7HOTjqxWkI8I4b44DfW4DLZ223i3R9iStAURRlp0ScsZ6uAg41ETrwTBdE5GfAL3B6i3gxWvmhhCYcRVF2atxGI58Bbxhjvp7qeOLB7Ufuc5wztsNTHc+ORhsNKIqyU2OM2YYzFs5sETkg1fHEycU4PShcmepAUoGe4SiKoig7BD3DURRFUXYImnAURVGUHcKQ79rm9ddfNzk5Odb1+/r6yMyM9YF9JRj1Z4+6s0fd2ROvu66urtZjjjkm5IPaQz7h5OTkMG3aNOv6PT09xJOwhjvqzx51Z4+6sydedytWrNgUbp5eUotCTU1N9EJKWNSfPerOHnVnTzLdacKJgp6Wx4f6s0fd2aPu7EmmO004UaioCNX5rhIr6s8edWePurMnme404URhy5YtqQ4hrVF/9qg7e9SdPcl0N+QbDcRLQUFBqkNIa9Tf4PF4PLS3t5Ofn09zc3Oqw0lL1J09sbjLysqisDDSYKlh6tkGlQhE5HDgJ8C+wHjgZmPMTTHUOwenu/TdcHrtvdcYc08yYtSWLvGh/gaHx+Ohra2NkpISCgsLycrSY0Ib1J09sbhrb2+nu7ub3NzcQS071ZfUioBVwNU445NERURmA/NwxtyYDtyGM0jVRckIsKmpKRmLHTaov8HR3t5OSUkJIkJfX1/0CkpI1J09sbgrKCigq6tr0MtO6SGAMeYFnAGsEJFfx1jtWuBpY8zt7v9rRGQa8FPgwUTHWFYW90i5wxr1N3hEBCDmI/TO3sg7iPzs4ddiS89u7InFnW8bHfSyrWqlCBHJxhlk6odBsxYDl4vIBGPMF4n8zI6ODoqLEzK68bBE/dnj9XpjaqJ66l8+iDj/b9/bL1EhpQ2xulMGkkx3aZVwgAogG2ckyEB8l+PG4AwZ66ehoYEZM2aQmZmJ1+tlzpw5zJ07l9pap0pVVRW1tbXk5+eTl5dHY2MjI0eOpKuri87OTjweDxs2bMDr9TJmzBhqamrIy8ujoKCAhoYGSkpK/Dd5q6qqqK+vp6+vz182JyeH4uJi6uvrKS4uxuv10tbWRmVlJQ0NDXg8Hn/Z7OxsSktL2bp1KyNGjACgtbWViooKWlpa6Onp8ZfNysqirKyMuro6/zXX5uZmysvLaWtro7u72182IyODyspKamtrKSgoICcnh6amJsrKyujo6KCrq4vRo0f7nVRXV7N58+Z+TkpLS+np6aGjo4Pq6mrq6ur6OcnNzaWoqIht27b1c9Lb24vH4+n3PcM5GTVqFE1NTfT29oZ0UlRUREZGBi0tLWGdlJeXs2XLFmsnges+lJPAdW/rJNS69zlpb2/334zt7e0lIyODvr4+jDFkZ2fT29uLiJCZmYnH4yEjI/pVca/X61+W71JdVlYWXq8Xr9frXy7g/zuwbGZmJsaYiGUzMjLweDwDyno8nrCxB5bNysqK+D19O0Bf7NGceDyeiMvNysoadk4Cv2dw2cDYvV5vTE7a29vp6uoasI+IxE4zPIGIrMcZAvemCGXG4DQSONm9HOebvhfOuPIDRgRctmyZiadrG5sbY8p21N/gaG5upqSkBHASRSwJxXdJravXy3899hEAj5+9N3nZTt3heEktVnfKQGJ1F7itBrJixYrVxxxzTMidbrqtka2ABxgdNL3afQ8+84kb3xGuYof6s8d3NBmN/OxM8rMz/QkGQGT79EQxbtw4/6u8vJwxY8b4/7/77rsHvbxLLrmEm2++Oaay3/zmNxk7diytra39ps+ePZuHH36437S1a9cOeHjx6aef5thjj2Xs2LFMnjyZ0047jddff33QMcfKsmXLmDlzJrvssgvHHXccq1evDlt2y5YtfOc732G33XZjypQp3HXXXf3mT58+vZ/rQw89NOyy3nzzTcrLyxk3bhzjx49nv/3249577+23rFdeeSVi7LFudzakVcIxxvQC7wAnBs06CdiQ6Ps3ipIu9HkND727vQ+scxZ+zB/f3kSfN3FXMDZu3Oh/jR07lr/+9a/+/6+44oqEfU4wNTU1LF26lNzcXJ555plB13/wwQe55ppruPTSS/nkk0/4+OOPueSSS3jxxRcTHyzOZfxzzjmHyy+/nHXr1nHSSSdx7rnn4vF4Qpa/6KKLqKioYNWqVbzwwgv85S9/YeHChf3KBLpevnx5xM8fNWoUGzduZMOGDfzP//wPt99+O6+++mrCvl88pDThiEiRiOwrIvsCOUC1+/80d/5BIrJGRA4KqHY7cJqIXC0ik0Xku8AlOEPQJpyqqqpkLHbYoP7sycrKorO3L6bXg//cxIufbPPX7e0zPLuqngf/uSmm+vFeWn/00Uc5+OCDmTRpEqeddhrr168HwBjD9ddfz+TJkxk/fjyzZs3io48+4pFHHuHJJ5/kgQceYNy4cZx66qlhl/34448zbdo0vve977FgwYJBxdXS0sKtt97KHXfcwde//nX/vb1jjz2WO+64Y0D5L7/8krFjx/qP8m+44YYB/994440RP/O5555j9913Z86cOeTm5nLppZfS1tbGsmXLBpRta2vjjTfe4MorryQ3N5ddd92Vc845h/nz5w/qe4bjkEMOYerUqaxatWrAvHXr1jF79mwmTJjA7rvvznnnnQc492eSRaobDRwABJ7X/sB9fQFMBAqAye47AMaY50TkPOBnwC9x7ulca4xJeJNocC4JTZw4MRmLHhaoP3vaOnuYs3CNdf2ePsOiVfUsWlUfteyi7+xjfflt8eLF3HnnnTz22GPsscce3HfffVxwwQW89tprvPbaa7z11lu88847FBcX85///IcRI0Zw/vnn8+6771JZWcnPf/7ziMtfuHAh55xzDieddBJ33nkn69atY9ddd40ptnfeeYeuri5OOeWUmMrvsssuVFZW8t5773HIIYewbNmyAf9fc801EZexevVq9tprL///mZmZTJ06ldWrV3PEEUf0K2uM8b8Cp3388cf9yl1yySV4vV6mTJnC9ddfzyGHHBL1uxhjWL58OatXrw7p+LbbbuPII49k0aJF9PT0sHLlSsC5pJas+64pPcMxxiwxxkiI18Sg+UuC6j1qjJlqjMkxxkxKVi8D4HTzoNij/uxJl5vejzzyCJdddhlTp04lKyuLyy67jM8++4xPP/2U7Oxs2tra+OSTTzDGsOeeezJ6dPAt2PC88847fPbZZ5xxxhnsvvvu7LfffoM6y2lsbKS8vHxQz+XMnDmTt956i5aWFrZs2cK5557r///jjz+OurNvb28f8ChAcXExbW1tA8qOGDGCmTNncscdd9DZ2cnatWt57LHH6Ozs9Jf5wx/+wMqVK/nggw84/fTTOfPMM9m4cWPYz9+6dSsTJ05kt9124yc/+Qm33HILhx9++IBy2dnZbNy40d/60ve9krndpfoMZ6cnLy8v1SGkNerPnrzsDBZ9Z5+o5bo9Xs5Z+DG9fQMvi+VkCvO/tRe5WZF3InlR5kdi48aN3HDDDdx0003+aR6Ph82bN3P44Yfzve99j2uvvZYNGzZw4okn8otf/ILS0tKYlr1gwQJmzZrFmDFjADjjjDN44IEHuO666/o1bw7E4/H4E8zIkSPZtm1bv2nROOyww3jiiSfYe++9OeSQQ5g1axa//vWv2XvvvZk2bVrIllmBFBYWDmjc0NraSlFRUcjyDz74INdeey3Tp0+nsrKSb33rW/zv//6vf35ggrvgggt46qmneOWVV7jgggtCLm/UqFGsWRP9zPimm27iV7/6FccffzwjRozgoosu4vzzz7d+qDMW0uMQKoU0NjamOoS0Rv3Z4/V6/S3NIr1K87M5dVoFOZn9dxS5mcLXp1VQmp8ddRnx7GTGjh3LHXfcwfr16/2vmpoa/1H1D37wA1577TX++c9/snHjxphbtHV1dfHMM8/w3nvvMWXKFKZMmcJvfvMbNm3axBtvvOH/7A0bNvSrt379esaOHQvAQQcdRF5eHosXL475+8yaNYt3332X119/ncMOO4z999+fjz76iNdff52ZM2dGrT916lQ++ugj//9er5dVq1YxderUkOXHjh3L/PnzWbt2LW+++Sa9vb0ccMABYZefkZER9z03cO6v3nPPPXz88cfce++9XHfddaxatSqp3QJpwonCyJEjUx1CWqP+7BnM094XHrgLJ04u9/+fkynMnlbBhQfukozQ+nH++edzzz33+G9Mt7S0sGjRIrxeLytWrOBf//oXvb295Ofnk5OT4/9elZWV/sYFoVi8eDF9fX289dZbLF26lKVLl/LWW29x8skn+y+rnX766SxcuJB33nkHY4w/oZ1++umAcynr+uuv5+qrr+b555+nvb2dvr4+lixZEvZejK/Z96OPPsqsWbPIzs5mr7324tFHH+Wwww6L6mP27Nl8+umnPPXUU3R3d3PfffdRVFTErFmzQpZfu3YtLS0t9Pb28tJLLzFv3jyuvPJKADZt2sTy5cvp6emhp6eHv/71r7z//vscddRRUeOIxjPPPMOXX34JOJf2fA+K6gBsKcSmgzplO+rPnliPYjt7++jp83L2vtX+aX/65lS+vf9oevq8yQrPzymnnMIVV1zB97//fcaPH8/MmTN58cUXERFaW1u5/PLL2XXXXdlnn30YMWIEl19+OQDnnnsu69at87dsC2bBggWcffbZjB8/nqqqKv/roosu4vnnn6elpYWjjjqKX/ziF8ydO5cJEyYwe/ZsjjjiCObOnetfzkUXXcSvfvUr7rnnHiZPnsy0adP43e9+x0knnRT2O82aNYvCwkL23HNP///t7e3+Z2DmzJkT9kytrKyM+fPnc+eddzJp0iSee+455s+f77+k9+STT/Z7lmbp0qUccMABTJo0id/85jc89NBD/s9ta2vjmmuu8T+js2DBAhYsWMCkSZMGsYZCs3LlSo4//njGjRvHWWedxc0338zkyZMTcvYUjp2mp4FkEW9PA+vXr9dWVnGg/gZH4NPbsfbScPyf3o84fzj2paY9XNgTq7vh0NPADqe6ujp6ISUs6s+eZD4PMdRRd/YM5edwdnrq6uoYP358qsNIW9SfPbE+DxFLS7bhRjKfJRnqJNOdJpwoeL3JvwY+lFF/yWc4ds6ppCd6SS0Kvvb/ih3qzx69LGSPurMnme404UShpqYmeiElLOrPnmT22jvUUXf2JNOdXlKLgj4pHx/qz55YuxjxtHdGnJ9VOPy6F0qXboF2RrRrmxRSUFAQvZASFvVnT6w//Fd2Oybi/BNq30pEOGmFJhx7kulO10oUGhoaUh1CWqP+7Ak3fooSHXVnTzLd6RlOFKJ11KdERv3ZE2sXI8d+5gyu1dfRyetfcbrhP+rD58ks2Hkvpd1999189tln/P73v0/K8pPZPctQR7u2SSF6pBQf6s+eWHsBySrMJ6swv3+CEfFPTxSJHGL6iiuuiDvZPPXUU5SVlTFv3rx+0x977DFOOOGEAeWPO+44HnvsMf//69at48ILL2SPPfZgwoQJHHroofz2t7+lp6cnrrjC0dzczAUXXMD48ePZa6+9+POf/xyx/H333cf06dMZP348Z511Flu2bPHPu/3226msrOy3TiKNBBo4TPWee+7J+eef71/e7bffzoUXXugvm8zeZzThRKG9vT3VIaQ16s+ewTzDZPr6WHvb9jEIl844jTU334dJYM+/sQ4xvaMOMh577DFGjhw56FFAAb744guOO+44qqqqeOONN/jiiy947LHHWLduXb8deyK5+uqr6evr4+OPP2bBggXcdttt/OMf/whZ9sknn+Thhx/m2WefZe3atVRUVPDf//3f/crMnj273zoJ7J8tFL71tXz5crZt28b1118fslwyn53ThBMFHSI5PtSfPVlZWXjaO2N6rb7xXjbNX+Sv6+3uYcPD/8fqG++NqX48R7UbNmzwd1g5ffp0jjvuOACuv/569t57b8aPH8+RRx7Zb4jlwKNqX/3HH3+cfffdl0mTJnH11VdHjKmmpoalS5dyzz338Pbbb7Nu3bp+86MNt3D77bdzwAEHcNttt/kHhJs0aRL33Xcf48aNG1D+G9/4Bk8++SQAq1atoqysrN//e+65Z8R429vbWbRoEddddx0jRoxgn3324ayzzuLRRx8NWX7x4sWce+65TJgwgby8PK6++mr+8Y9/ROxdO1bKy8s59dRTQw47bYzxd+IZOCR4otB7OFGor68PuQEqsaH+7OlpbWPplBOt63u7utnw5yfZ8Ocno5Y99rNX4778tnTpUpYtW+a/B7Dvvvvyk5/8hOLiYh588EEuuOACVq5cGbbl4pIlS1i2bBn19fUcffTRHHfccf7kFczjjz/OHnvswezZsznggANYsGBBvyP2aAn09ddf54Ybboj5u82aNYs333yTOXPmsGzZMiZOnNjv/0MPPTRikvvss88wxjBlyhT/tK985Ss88MADIct7vd4Bw06Dk9x8neG+8sor7LrrrowaNYqzzz6bH//4xzG1MNu6dSuLFi1i+vTpA+aFGxI8UegZThSSORjRcED92ZNuPblfc801FBUV+YcVnzNnDmVlZWRlZfGjH/2I3t5e1q5dG7b+T3/6UwoLC5kwYQJf/epX+fe//x227MKFCznzzDMBOPPMM3n88ccH5auhoWFQZ9+zZs3irbec5uXLli3jiiuu6Pd/uLFufLS3tw/YcZeUlIQcdhrga1/7GvPnz+ezzz6jo6ODX//614iIf+jpU089leXLl/Ppp5/yxz/+kb/+9a9hk5ePCy64gEmTJnHUUUcxbtw4br311gFl4h0SPBp6hhMF7ZolPtSfPbnFI/wt0CLR19nF0hmn4e0eeLM7IzeHI957msz8yA/gZhbE/4DuLrv0H+zt/vvvZ968edTW1vrHxtm2bVvY+pWVlf6/CwsLw97/e+edd/j0008544wzAOdy13XXXccbb7zBEUccQXZ2dsj7SB6Px99tS1lZ2aDu1cyYMYOamhpqamp49913uf/++/nd735HTU0Ny5cv5+qrr45YP9Sw0y0tLWGHnT777LOpqanhjDPOoLOzk4svvpiioiL/7ylw9NDp06dzxRVX8Mgjj/CjH/0obAwPP/wwxx57bMQ44x0SPBp6hhMF7ZolPtSfPR6Px9/SLNIrt2Ik47/7TTLy+vfwm5GXy/jvfpPcipFRl5GIcewDl7F8+XLuueceHn74YT7//HPWr19PcXFxQs7afI0EjjvuOKZMmcKsWbPwer3+6WPHjuXLL7/sd/O7r6+PjRs3+i/vHnXUUTz33HMxf2Zubi4zZszgoYceYsyYMRQVFTFz5kweeugh+vr6wg4f7WO33XZDRPjkk0/80z788MOw9USEq666ivfff581a9Zwwgkn4PV6w5ZP1LDT4JwJ2QwJHguacKKQk5OT6hDSGvVnz2CSwOSfXczYc77u/z8jN4fxF5zO5J9dnIzQotLW1kZmZiYjR47E4/Fwxx13DDjCt6Grq4tnnnmGO+64wz/s9NKlS3nwwQd5/vnnaW1tZf/996eoqIg777yTzs5OOjs7ueOOOygtLWW//ZzB6K655hreffddbrjhBmprawGn5drcuXPZuHFjyM+eNWsWf/zjH/3DTB922GH88Y9/ZObMmVHXVWFhIV//+te57bbbaG1t5aOPPvKPaBqKpqYm/32f9evXM3fuXH74wx/6zzQWL15MY2MjAB9//DF33303p5xyyqB9BrNixQref//9kEOCJwJNOFEoLi5OdQhpjfqzJ9Yfuqe9k76uHnab+x3/tMP+8Ri7X/k9+rqS80xJNI4++miOP/54Dj74YKZPn05eXt6AS242LF68mOzsbM4999x+w06fdtppVFVV8cwzz5Cbm8uCBQv417/+xT777MM+++zDypUrWbBggX+cl4kTJ/L3v/+dL7/8ksMOO4wJEyZw1llnMXHixLD3dmbNmkVbW5v/fs3MmTNpa2tj5syZwMCho4P5zW9+g4gwbdo0zjzzTH76059y+OGH++cHPkvT2NjI2WefzdixYzn55JP56le/yrXXXusv+8wzz3DggQcybtw4zjvvPL71rW/x4x//OD65QGtrK1dddVXIIcETgQ4xHQUdIjk+1N/gsBli+qXqmRHnD8e+1HSIaXt0iOkUokfo8aH+7NHuWexRd/Yk0522UouCjlgZH+rPnlivPsTSkm24MdSv3CSTZLrThBOFtrY2KioqUh1G2qL+7Ik1WQ/H8W6ioQc69mjXNikk8NkAZfCoP3uysvR40BZ1Z08y3WnCiYKO5xIf6s8e7aXBHnVnTzLdacKJgnavHx/qb3BkZWXR3t6OMUbvQ8SBurMnFnc9PT1WI4PqeWcUtGuW+FB/g6OwsJDu7m5aWlrwer10dXWlOqS0RN3ZE4u7jIyMsN3yREITThRqamr0OZI4UH+DJzc3l9zcXH2GKQ7UnT3JdJfyS2oicryIvCciXSLypYj8UkQiNgQXkXwRuVNE1rv1vhCR34pI6H7P48DX2Z9ih/qzR93Zo+7sSaa7lJ7hiMh+wPPA74FzgWnAn924ro1Q9S7gNOC7wCq33kNAIfDfEeoNmkT1kjpcUX/2qDt71J09yXSX6jOcK4EPjTGXG2NWG2OeAm4ELhWRSBcIvwo8Zox50RjzhTHmRWABEHmMVQu2bt2a6EUOK9SfPerOHnVnTzLdpTrhzAJeCJq2GMgHZkSo9w/gZBHZA8B9P8mtm1ASOdrdcET92aPu7FF39iTTXaobDYwGNgdNq3XfIzVvugy4G1grIh6c7/EH4KfBBRsaGpgxYwaZmZl4vV7mzJnD3Llz/V2SV1VVUVtbS35+Pnl5eTQ2NjJy5Ei6urro7OwkLy+PDRs24PV6GTNmDDU1NeTl5VFQUEBDQwMlJSV4PB7a29upqqqivr6evr4+f9mcnByKi4upr6+nuLgYr9dLW1sblZWVNDQ04PF4/GWzs7MpLS1l69at/pXe2tpKRUUFLS0t9PT0+MtmZWVRVlZGXV0dhYWFZGVl0dzcTHl5OW1tbXR3d/vLZmRkUFlZSW1tLQUFBeTk5NDU1ERZWRkdHR10dXUxevRov5Pq6mo2b97cz0lpaSk9PT10dHRQXV1NXV1dPye5ubkUFRWxbdu2fk5ycnLYtGlTv+8ZzsmoUaNoamqit7c3pJOioiIyMjJoaWkJ66S8vJwtW7ZYOwlc96GcBK57Wyeh1n2gE9+6r6urIy8vb4CTwHUfzUlmZiYVFRUDnASu+1BOwv0eqqur/QOXRXMS6vcQ6CSW30NJSUk/J77fQ3Nzc0Qn7e3ttLa2hlz3gU5i+T1E20eEWvfpvI/wer1kZmZa7yMikdLeokWkG7jcGPNAwLQCoB04yxizMEy9HwM/Aa4GVgNTgd8Cjxhjrg8sq71Fpxb1Z4+6s0fd2ROvu0i9Raf6DGczzllOIKMD5g1ARPKAO4ELjTFPuJM/FJFc4GERudUY05GoALUfsPhQf/aoO3vUnT3JdJfqezjLgBODpp0EdALvhamT7b6C+1/oA8R9JYyWlpZELm7Yof7sUXf2qDt7kuku1QnnLmAfEblLRKaKyOnALcC9xpg2ABE5SETWiMhBAMaYVuBV4FYROVFEJorIScBtwEvGmPZEBhjtmqQSGfVnj7qzR93Zk0x3KU04xpgVwGzgKGAlcD/wABB4H6YAmOy++zgLeAl4EFgD/A/O8zzfTnSM2jVLfKg/e9SdPerOnmS6S/UZDsaYl40x+xtjco0xY4wx1xtj+gLmLzHGiDFmScC0emPMxcaYCcaYPPf9R8aYhHdNXFNTk+hFDivUnz3qzh51Z08y3aU84ezs6Lga8aH+7FF39qg7e3Q8nBRSVlaW6hDSGvVnj7qzR93Zk0x3mnCiUFdXl+oQ0hr1Z4+6s0fd2ZNMd5pwolBYWJjqENIa9WePurNH3dmTTHeacKKg14LjQ/3Zo+7sUXf26D2cFNLc3JzqENIa9WePurNH3dmTTHeacKJQXl6e6hDSGvVnj7qzR93Zk0x3mnCi0NbWluoQ0hr1Z4+6s0fd2ZNMd5pwotDd3Z3qENIa9WePurNH3dmTTHeacKKgXWTEh/qzR93Zo+7sGdJd2+zsaBcZ8aH+7FF39qg7e7RrmxSSkaGK4kH92aPu7FF39iTTna6VKFRWVqY6hLRG/dmj7uxRd/Yk050mnCj4xvBW7FB/9qg7e9SdPcl0pwknCgUFBdELKWFRf/aoO3vUnT3JdKcJJwo5OTmpDiGtUX/2qDt71J09yXSnCScKTU1NqQ4hrVF/9qg7e9SdPcl0pwknCjquRnyoP3vUnT3qzh4dDyeFdHR0pDqEtEb92aPu7FF39iTTnSacKHR1daU6hLRG/dmj7uxRd/Yk050mnCiMHj061SGkNerPHnVnj7qzJ5nuNOFEQdvzx4f6s0fd2aPu7NHncBRFUZS0RxNOFKqrq1MdQlqj/uxRd/aoO3uS6U4TThQ2b96c6hDSGvVnj7qzR93Zk0x3mnCikJ+fn+oQ0hr1Z4+6s0fd2ZNMd5pwopCXl5fqENIa9WePurNH3dmTTHeacKLQ2NiY6hDSGvVnj7qzR93Zk0x3mnCiUFpamuoQ0hr1Z4+6s0fd2ZNMd5pwotDT05PqENIa9WePurNH3dmTTHeacKKgfTLFh/qzR93Zo+7s0b7UUoi2548P9WePurNH3dkzpJ/DEZHjReQ9EekSkS9F5JcikhlDvQki8qiIbHXrrhWRbyQ6vrq6ukQvclih/uxRd/aoO3uS6S4raUuOARHZD3ge+D1wLjAN+LMb17UR6u0C/BNYAswGaoEJQFuiY/R6vYle5LBC/dmj7uxRd/Yk011KEw5wJfChMeZy9//VbjK5XUR+aYwJl0BuA9YbY84KmLY+GQGOGTMmGYsdNqg/e9SdPerOnmS6S/UltVnAC0HTFgP5wIxQFUQkA/gGsFxEFohInYh8KCI/E5GEJ9CamppEL3JYof7sUXf2qDt7kuku1Wc4o4Hgjnt8fWOHS7OjgGLgYuBe4Gs4l+LuA0YA1wQWbmhoYMaMGWRmZuL1epkzZw5z5871d8FdVVVFbW0t+fn55OXl0djYyMiRI+nq6qKzs5OMjAw2bNiA1+tlzJgx1NTUkJeXR0FBAQ0NDZSUlODxeGhvb6eqqor6+nr6+vr8ZXNyciguLqa+vp7i4mK8Xi9tbW1UVlbS0NCAx+Pxl83Ozqa0tJStW7cyYsQIAFpbW6moqKClpYWenh5/2aysLMrKyqirq6OwsJCsrCyam5spLy+nra2N7u5uf9mMjAwqKyupra2loKCAnJwcmpqaKCsro6Ojg66uLkaPHu13Ul1dzebNm/s5KS0tpaenh46ODqqrq6mrq+vnJDc3l6KiIrZt29bPiYiwadOmft8znJNRo0bR1NREb29vSCdFRUVkZGTQ0tIS1kl5eTlbtmyxdhK47kM5CVz3tk5CrftAJ75139bWRnt7+wAnges+mpPMzEwqKioGOAlc96GchPs9VFdXs2XLlpichPo9BDqJ5fdQUlLSz4nv99Dc3BzRicfjYf369SHXfaCTWH4P0fYRodZ9Ou8jenp6aGpqst5HREKMMRELAIhIoVs27D0SESkCjDGmPeoCt9fpBi43xjwQMK0AaAfOMsYsDFFnDPAl8K4x5qCA6VcCNwEjTMCXWrZsmZk2bVqsIQ2gtbXVv2KVwaP+7FF39qg7e+J1t2LFitXHHHNMyJ1u1EtqIjIFaCTCTXyXa4AGEdl9ELFtxjnLCWR0wLxQbAV6gVVB0z8GCoGKQXx+VLZt25bIxQ071J896s4edWdPMt3Fcg/nYpzLXDdHKfcLt9zFg/j8ZcCJQdNOAjqB90JVMMb0Am8Dk4NmTcY5M6ofxOdHpaSkJJGLG3aoP3vUnT3qzp5kuosl4RwLPOHu6MNijOkBngCOH8Tn3wXsIyJ3ichUETkduAW413f5TkQOEpE1InJQQL1fAQeLyI0isruIfB24HvidieUa4SDweDyJXNywQ/3Zo+7sUXf2JNNdLAlnIrA6xuWtASbF+uHGmBU4z9EcBawE7gcewEkePgpwzl4KAuq9AJwFnAl8hJO4fgv8PNbPjpX29phvSSkhUH/2qDt71J09yXQXSys1b4zlfMsb1BmGMeZl4OUI85cAEmL648Djg/ksG6qqqpL9EUMa9WePurNH3dmTTHexnOFsAA6IcXkzgI324ex86M3H+FB/9qg7e9SdPaluNPAKcJaITIxUyJ1/NvC3+MPaedBrwfGh/uxRd/aoO3tSfQ/nLpxLWq+KyJGhCrjTX3H/vTsBce00aBcZ8aH+7FF39qg7e1LatY0x5gucjjXH4CSddSLytIj8RUT+T0TWAa8CuwDnuuWHDNpFRnyoP3vUnT3qzp6Ud21jjHlaRA7GabJ8Ak7LNR89wHPAjcaYDxIeYYrJyclJdQhpjfqzR93Zo+7sSaa7mPtSc5PJN0QkF9gDpz+zFuBTY0xXkuJLOcXFxakOIa1Rf/aoO3vUnT3JdDfo3qKNMd3GmI+MMW+570M22QDU1ye044Jhh/qzR93Zo+7sSaa7qGc4InJ4hNkGpxua9caYIbmG9UgpPtSfPerOHnVnTzLdxXJJbQkxPMwpIv8ELjPG/CveoHYmdOTA+FB/9qg7e9SdPake8fOHUeYX4oxH803gdRE5yBgTa1c4Oz1tbW1UVCS0A+phhfqzR93Zo+7sSaa7qAnHGPOHWBYkIrcA7wPXAd+OM66dhlGjRqU6hLRG/dmj7uxRd7HT2dvX7/8RI8v7TcvPzkzYZyVsxE9jzAYR+RNwTqKWuTPQ1NREYWFhqsNIW9SfPerOHnUXO6f+JfLTLH/73n4J+6xBt1KLwqc4Q0APGXp7I47KoERB/dmj7uxRdzsnCTvDcakGwg5DnY5oFxnxof7sUXf2qLvYWfSdfQDo6vXyX499BMDjZ+9NXnaiz0cSeIYjIjnAfwFDqpWadpERH+rPHnVnj7qLnfzsTPKzM/slmLzsDP/0RBLLczh7RilSAEzBac02Fbg6AXHtNGRnZ6c6hLRG/cVO8M1br2Ql7ebtUEe3u/jo9niTsr3FckltDdGfwxGcB0Avc0fjHDKUlpamOoS0Rv3FTuibt1v9fyXy5u1QR7e7wdHnNTz07vazwnMWfsyp0yq48MBdyMwYMP6lNbEknFuInHC6gM+BvxtjGhMSVQoJPsrcVFvH+PHj/f/rUebg2Lp1q7YWUnY46bzdeY3BGOhz373G4DXRpwf+33/69mnGGPrc98DpL6yp5+2NLf4YevsMz61yOo/574PHJuy7xfIczk2xLkxEjgbON8acF09QqST0UWaT/y89yhwcRUVFqQ4hbVj0nX146N0aXvxkGz1924/xcjKFEyeXJ/SzTMw7MfBi8Hq3v5sQ/4faiYXbudntNAe3023v6CW3dlPYeO0+Y+B3Cfyu/f7H0Bfys6N/x52F7j7Ds6vqOW/GGPKyEnO7P+5WaiKyG/Ad4DxgPI7TtE04SmLJyEhsSxffD7vPGPq823+wvr/7jPsDN8b/d58x7v+45dyy7t99XgZOC5je59sZeLf/Hf6z3c+JEEf/z8GtY+j1GD6obRtwOaHH/eH/e3MrBrE+6g0stxPt15LIkGow248Mce5jZGQIGe67AJnue4YIGeK8iwz8PzNgep/Xy4am7pDbhAEaO3oZXZybkLitEo6IjADOBM4HZgJe4C3gHuDphESWInbkUSYMPDIK3hH5d64Rd3ghdnzudP9OLuQOL/SOb/tnh/g7xM52e92B89s7OsnJ3RZ6Z+uPKUwiCLGj35mOAHckBljf2L1DP1NgwM4p1p1Y/+kgAdOcnaWQkQEZ7rsgZLrv/cqJuPUjf37w9NaWZkpLSyMvixA76qAdeKTPiO27+6aHimPgdw3+rFDTnfWSuPsqXR4vZ8z7oN/+LnAbGFmQuAYYg0o4InIsTpL5Bk7rtP+4s75tjFmYsKhSiIgMSDaw/ShzfUMniPQ7ag11tDvwyDbUDnq47EB3zJGmb8fh+3H6/s4Ud+cmQqY4Ozbf3xniq9P/b1+Z7fXduoL7d//PCPf39rpE/bw+r+HeZZvwhNgosjKEG4+dSG5WZtidVWaYHXng/xn4ykffaSZyp7ajaWsr0cu5MZKXlcHXp1Xw7Kr6fvu93Exh9rSKhF1Og9iaRe+Bk2S+DYwFNgL3AvOAXmAtzqifQ4LGjt6wlxsM8O/a9h0ZzoCdVqQdaaidma9MyPlByw3+jNCf1/+zneUG7TwD/m5pbqS8rKzfTtwXy8CdeOBygj47KO5QMWWk8Q4SnAYrn23rDHt2fcj40tQFl2a0tLRowhkEFx64C719hkVuQ4EcN9lceOAuCf2cWM5wPgGagaeAecaYpb4Z7v2bIUV+hKdrMzPgslnjnKPMCEfMkXaeA3auEXb66X6UCbB+fQsTJ45MdRhpQbg+rXrcHcElM8ft4IjSl56eIXMMnHR8LXPP3rfan3D+9M2plORn0dPnJT9jx3fe2Qt0ADv28D4FnPnoR2Hn9XnhhMna5flg0C5GlFSg213shDrQOe+JVf6/E9kyN5aEsy9wAXA2cImIfAr8FZifsCiUIUtNTQ0TJ05MdRhpga9PKx8bNmzo9wyYEju63e2cxPIczgfA5SJyFXAyzv2cG3EeCP0Q59bGkOlHIvBH3+3xsubzL5kyaRdyE3jjbDiRlZXo/mGHLsEPFRfmZuuDxpbodhc7wQc6NV/WMGaX5JwhxrxWjDEeYBGwSETKcca9+Q5O46B5IvIDnPs8zxhjvkxGsDuCwB94fnYm0ydVkZ8/ZPLpDqe8PPFNyYcL6s4edRc7wQc1Y6oqknagY3XYbozZZoy51xgzA9gHuA+n4877gC8SGF/K2bJlS6pDSGvUnz3qzh51Z08y3cV9ncgY85Ex5ic4TaZPJc0f/AwmXftj2llQf/aoO3vUnT3JdJfIIab7gOfc15BBrwXHh/qzR93Zo+7sSaY7vRMehebm5lSHkNaoP3vUnT3qzp5kukt5whGR40XkPRHpEpEvReSXIhLzHSsRuUBEjIgsSUZ8evMxPtSfPerOHnVnTzLdpTThiMh+wPPAG8B+wKXAj4BbY6z/FeA2t35SaGsbuj3O7gjUnz3qzh51Z08y3aX6DOdK4ENjzOXGmNXGmKdwnvG5VEQidoTk9lj9JE6S+jxZAXZ379geeoca6s8edWePurMnme5SnXBmAcFDUi8G8oEZUer+P2CJMebJZATmQ7vIiA/1Z4+6s0fd2ZNMd6luyjEa2Bw0rdZ9D/utReQSYBpwULQPaGhoYMaMGWRmZuL1epkzZw5z586lttb5mKqqKmpra8nPzycvL4/GxkZGjhxJV1cXnZ2deDwecnJy8Hq9jBkzhpqaGvLy8igoKKChoYGSkhI8Hg/t7e1UVVVRX19PX1+fv2xOTg7FxcXU19dTXFyM1+ulra2NyspKGhoa8Hg8/rLZ2dmUlpaydetWRowYAUBraysVFRW0tLTQ09PjL5uVlUVZWRl1dXUUFhaSlZVFc3Mz5eXltLW10d3d7S+bkZFBZWUltbW1FBQUkJOTQ1NTE2VlZXR0dNDV1cXo0aP9Tqqrq9m8eXM/J6WlpfT09NDR0UF1dTV1dXX9nOTm5lJUVMS2bdv6Oent7SU/P7/f9wznZNSoUTQ1NdHb2xvSSVFRERkZGbS0tIR1Ul5ezpYtW6ydBK77UE4C172tk1DrPtCJb91v2rSJ3XfffYCTwHUfzUlmZiYVFRUDnASu+1BOwv0eqqur/c9pRHMS6vcQ6CSW30NJSUk/J77fQ3Nzc0QnW7duJT8/P+S6D3QSy+8h2j4i1LpP531ER0cHY8aMsd5HREKMSd2ALCLSDVxujHkgYFoBTiehZ4UaY8e97/MKMMsYs8ad9ggw0RhzZHD5ZcuWmWnTplnHqP1ZxYf6s0fd2aPu7InX3YoVK1Yfc8wxIXe6qT7D2YxzlhPI6IB5oTgCGAl8FNB1fwYgIuIBjjfGvJaoACsrKxO1qGGJ+rNH3dmj7uxJprtU38NZBpwYNO0koBN4L0ydeTjd6ewb8HoW+Jf799uJDNB3CqnYof7sUXf2qDt7kuku1Wc4dwH/FJG7gD/h9Md2C3CvMaYNQEQOwhkO4TxjzDvGmG3AtsCFiEgTUGqMCT+YjSUFBQWJXuSwQv3Zo+7sUXf2JNNdSs9wjDErgNnAUcBK4H7gAeD6gGIFwGT3fYeTk5OTio8dMqg/e9SdPerOnmS6S/UZDsaYl4GXI8xfgjMEQqRlnJ/YqLbT1NREaWlpshY/5FF/9qg7e9SdPcl0l+p7ODs9I0eOTHUIaY36s0fd2aPu7EmmO004Uejq6kp1CGmN+rNH3dmj7uxJpjtNOFHo7OxMdQhpjfqzR93Zo+7sSaY7TThRGD06+DEhZTCoP3vUnT3qzp5kutOEEwVtzx8f6s8edWePurMnme404SiKoig7BE04Uaiurk51CGmN+rNH3dmj7uxJpjtNOFHYvDlcl25KLKg/e9SdPerOnmS604QThby8vFSHkNaoP3vUnT3qzp5kutOEEwXtkyk+1J896s4edWfPkO1LLR1oaGhIdQhpjfqzR93Zo+7sSaY7TThR0P6Y4kP92aPu7FF39iTTnSacKEQbMlWJjPqzR93Zo+7sSaY7TThR6OjoSHUIaY36s0fd2aPu7EmmO004UdD2/PGh/uxRd/aoO3v0OZwUUldXl+oQ0hr1Z4+6s0fd2ZNMd5pwouD1elMdQlqj/uxRd/aoO3uS6U4TThTGjBmT6hDSGvVnj7qzR93Zk0x3mnCiUFNTk+oQ0hr1Z4+6s0fd2ZNMd5pwopCbm5vqENIa9WePurNH3dmTTHeacKJQVFSU6hDSGvVnj7qzR93Zk0x3mnCisG3btlSHkNaoP3vUnT3qzp5kutOEE4WSkpJUh5DWqD971J096s6eZLrThBMFj8eT6hDSGvVnj7qzR93Zk0x3mnCi0N7enuoQ0hr1Z4+6s0fd2ZNMd5pwolBZWZnqENIa9WePurNH3dmTTHeacKKg42rEh/qzR93Zo+7s0fFwUoheC44P9WePurNH3dmj93BSiHaRER/qzx51Z4+6s0e7tkkh2kVGfKg/e9SdPerOHu3aJoXk5OSkOoS0Rv3Zo+7sUXf2JNOdJpwoFBcXpzqEtEb92aPu7FF39iTTnSacKNTX16c6hLRG/dmj7uxRd/Yk013KE46IHC8i74lIl4h8KSK/FJHMKHWuEpHlItIoIk0i8qaInJCM+EaMGJGMxQ4b1J896s4edWdPMt2lNOGIyH7A88AbwH7ApcCPgFujVD0aeAg4CjgYeBt4XkRmJS9aRVEUJR5SfYZzJfChMeZyY8xqY8xTwI3ApSISto9sY8yJxpj/Z4xZaYz5xBjzE2AtcHqiA2xtbU30IocV6s8edWePurMnme5SnXBmAS8ETVsM5AMzYl2IewluBJDwi4+jRo1K9CKHFerPHnVnj7qzJ5nuspK25NgYDWwOmlbrvg/m6aOf4SScecEzGhoamDFjBpmZmXi9XubMmcPcuXOprXU+pqqqitraWvLz88nLy6OxsZGRI0fS1dVFZ2cn4IwP4fV6GTNmDDU1NeTl5VFQUEBDQwMlJSV4PB7a29upqqqivr6evr4+f9mcnByKi4upr6+nuLgYr9dLW1sblZWVNDQ04PF4/GWzs7MpLS1l69at/uuora2tVFRU0NLSQk9Pj79sVlYWZWVl1NXVUVhYSFZWFs3NzZSXl9PW1kZ3d7e/bEZGBpWVldTW1lJQUEBOTg5NTU2UlZXR0dFBV1cXo0eP9juprq5m8+bN/ZyUlpbS09NDR0cH1dXV1NXV9XOSm5tLUVER27Zt6+cEoLGxsd/3DOdk1KhRNDU10dvbG9JJUVERGRkZtLS0hHVSXl7Oli1brJ0ErvtQTgLXva2TUOs+0Ilv3dfU1LDrrrsOcBK47qM5yczMpKKiYoCTwHUfykm430N1dTVbtmyJyUmo30Ogk1h+DyUlJf2c+H4Pzc3NEZ00NjaSnZ0dct0HOonl9xBtHxFq3afzPqK7u5uqqirrfUQkxBgziP16YhGRbuByY8wDAdMKgHbgLGPMwhiWcTFwF3CqMeZvwfOXLVtmpk2bZh3j+vXrmThxonX94Y76s0fd2aPu7InX3YoVK1Yfc8wxIXe6qb6kthnnLCeQ0QHzIiIiVwJ3At8IlWwSgXaRER/qzx51Z4+6s2cod22zDDgxaNpJQCfwXqSKIvIL4OfAScaYl5MTnnaRES/qzx51Z4+6s2cod21zF7CPiNwlIlNF5HTgFuBeY0wbgIgcJCJrROQgXyUR+R1OC7dvA2tEpNp9lSU6wKysVN/mSm/Unz3qzh51Z08y3aV0rRhjVojIbOBXOM/fbAMewGka7aMAmOy++7jUfX86aJFLgSMTGWNZWcJz2LBC/dmj7uxRd/Yk012qz3AwxrxsjNnfGJNrjBljjLneGNMXMH+JMUaMMUsCpkmY15GJjq+uri7RixxWqD971J096s6eZLpLecLZ2SkqCvv8qRID6s8edWePurMnme404UQhI0MVxYP6s0fd2aPu7EmmO10rUWhpaUl1CGmN+rNH3dmj7uxJpjtNOFGoqKhIdQhpjfqzR93Zo+7sSaY7TThR0COl+FB/9qg7e9SdPXqGk0Ki9Q2kREb92aPu7FF39iTTnSacKGgXGfGh/uxRd/aoO3uGctc2Oz3aRUZ8qD971J096s6eody1zU5PZmbE0a6VKKg/e9SdPerOnmS604QTBW3tEh/qzx51Z4+6s0dbqaUQ32BTih3qzx51Z4+6syeZ7jThRKGwsDDVIaQ16s8edWePurMnme404URBuzmPD/Vnj7qzR93Zk0x3mnCi0NzcnOoQ0hr1Z4+6s0fd2ZNMd5pwoqDjasSH+rNH3dmj7uwZ0uPh7Ox0dHSkOoS0Rv3Zo+7sUXf2JNOdJpwodHV1pTqEtEb92aPu7FF39iTTnSacKGgXGfGh/uxRd/aoO3u0a5sUol1kxIf6s0fd2aPu7NGubVKIjhwYH+rPHnVnj7qzR0f8TCGVlZWpDiGtUX/2qDt71J09yXSnCScKtbW1qQ4hrVF/9qg7e9SdPcl0pwknCvn5+akOIa1Rf/aoO3vUnT3JdKcJJwp5eXmpDiGtUX/2qDt71J09yXSnCScKjY2NqQ4hrVF/9qg7e9SdPcl0pwknCiNHjkx1CGmN+rNH3dmj7uxJpjtNOFHQJ5bjQ/3Zo+7sUXf2aE8DKaSzszPVIaQ16s8edWePurMnme404UShuro61SGkNerPHnVnj7qzJ5nuNOFEQYeqjQ/1Z4+6s0fd2aNDTCuKoihpjyacKOipeXyoP3vUnT3qzp4hfUlNRI4XkfdEpEtEvhSRX4pIZgz1zhGRVSLSLSLrRGRuomPr6+jigZtupa9DW7zYcs8996Q6hLREt7340O3OjmRvdylNOCKyH/A88AawH3Ap8CPg1ij1ZgPzgL8C04HbgF+LyEWJiMv09bHm5vt4deoJ7P7w33l16gmsufk+TF9fIhY/rHjyySdTHUJaodteYtDtbnDsqO0uK6FLGzxXAh8aYy53/18tIrsAt4vIL40xbWHqXQs8bYy53f1/jYhMA34KPBhPQN1bG/j0zj/z5cLFeLt7yEbwdvew4aGn8DS3MeUXc+NZ/LAjlww87dpENRZ66htZ97u/UPPUy7rtxYlud7ETdrt7+Cn6OrrY69dXJeyzxBiTsIUN+sNF1gPzjDE3BEzbDfgUONIYszREnWygHfihMebPAdOPAV4BJhpjvvBNX7ZsmZk2bVrMMb1UPdPimyiKogxNjlv3GpkFsfevtmLFitXHHHNMyJ1uqs9wRgObg6b5+sYON85pBZAdpZ4/4bS2tnY999xz/kuHRUVFDSUlJWE7Cyp78PqpkQLOHD3qU8nO6o1URtlOc3PzyEi+le14NmzWbS9B6HYXO9G2u5UffjDY7W5suBmpTjihMEHvwUiM9QE44YQTtJ9yRVGUnYBUt1LbjHOWE8jogHmh2Ap4QtTzteULV09RFEVJIalOOMuAE4OmnQR0Au+FqmCM6QXeCVNvQ+D9G0VRFGXnIdUJ5y5gHxG5S0SmisjpwC3Avb4WaiJykIisEZGDAurdDpwmIleLyGQR+S5wCfCrHf4NFEVRlJhIacIxxqwAZgNHASuB+4EHgOsDihUAk913X73ngPOAC4APgRuAa40xg2oSvTM/dLqzY+NORB4RERP82lEx7yyIyOEiskhEvnAd3BRjPd3uLNzpducgIheIyOsislVEWt3f7zkx1EvYdpfyRgPGmJeBlyPMX0KIhgLGmEeBR20/N+Ch098D5wLTgD/jOLk2Qj3fQ6fXAc8AhwG/F5GuwSa8dMXWncty4PSkBrjzUwSsAh4D7omlgm53fgbtzkW3OzgGeBa4GmgETgPmiYjHGPN4qAqJ3u5S+hxOKhGRR4EpxpgZAdMuxblcVxnuoVMRWQbUGmO+GTDtbuCbxpgJSQ57pyAOd4/gPCd15I6IMx1wn0V7xBhzU5Ryw367C2YQ7h5Bt7uQiMgLQGfgdhU0P6HbXarv4aSSWcALQdMWA/nAjIHF/Q+dHhim3ngRGS4//EG7C2B/EdksIp+LyFMisldSIhxC6HaXEHS7C00pUB9qRjK2u+GccJL10OlwwMYdwEs4l+COA76Hc1/uXRHZO+ERDi10u4sP3e5CICLfAQ4A/hCmSMK3u5Tfw9nJSOhDp8OMaO4wxiwM+PcjEfkHsBqn09b/TmJs6Y5ud3Gg291ARORUnETzA7fxVshiURYz6O1uOJ/h6EOn9ti4G4Axpgd4H5iYmLCGLLrdJZDhvt2JyLeAx4GLjDEPRyia8O1uOCccfejUnkG7C4WIZAFfATYmLrShh253iWU4b3ci8n3gEeB8Y8wjkcomZbszxgzLF7A/0IPz8OlUnCaTjcDtAWUOAtYABwVMmw304TQtnAx8F+jCOVpI+ffaWd3hNGe9C6fBwUTgYOAp193+qf5OO9hfEbCv+6rBGVJjX2CabneJdafbXT93V+CcsfwA5yzF9xq1o7a7lEtI8Qr4GrAC6HY33luBzID5R+JcpzwyqN45ONeAe4DPgbmp/i47uzucFmwv4dxw7HHrPDvcfvRBboJf63W7S6w73e76uVuf6u1u2D6HoyiKouxYhvM9HEVRFGUHoglHURRF2SFowlEURVF2CJpwFEVRlB2CJhxFURRlh6AJR1EURdkhaMJRlDTDN6BYquNQlMGiCUdRoiAiR7qjRF4UMG2iiNwkIvumMDRFSSs04SiKHROBn+N0q6IoSgxowlGUnQwRyXM7mFSUIYUmHEUZJCJyPvC6++/D7uU24w5l7CtTICI3i8gnItItIltFZL6IjAta1k1u3a+IyP0ishnoAMYOMqaviciTIvJFwOc9ISJ7hCibJSK3iMgmEekUkfdE5CS9N6QkGz2KUpTB8wZwG3Ad8EfgH+70zwBEJAf4O7Af8GfgQ2A8cAlwpIjsb4ypC1rmfKAJuB3IBdoGGdN5OB1V/gFnHJPJOIOLHSEiextjtgaUfQD4Pk6nlovd2BbidO6oKElDE46iDBJjzDoR+TtOwllujJkfVOQynG7wv2qMWe6bKCL/C/wLuMp9BbIFOMEY47UM6/vGmI7ACSLyHLAEZ1jlX7nT9sZJNv8HnGHc3ntF5FWcBKQoSUMTjqIknrNwRpT8j4hUBEzfBPwHODZEnfvjSDb4ko2ICDACyAE+xjlrOjCg6Cnu+z0moKt4Y8zLIrIKmGYbg6JEQxOOoiSeKTiXt7aGmV8TYtqn8XygiOyJcznuOJxBxwIpDfh7ovv+SYjFrEUTjpJENOEoSuLJwBma9/ow87tDTOsIMS0mRGQEsBTnrOZ2YBXOPSCDc28msHGQ2H6OosSLJhxFsSNSa67/AOXGmFd2UCxH4wwVfL4x5i++iSKSD4wMKvu5+z4ZCG64sGfSIlQUtFm0otjia0UWvEMHeBTYTUS+FzxDHEYlOJY+9z349/yTENMWu+9z3fs9vri+hl5OU5KMnuEoih2+y1YXi0g70AJ8box5G7gb+Brw/0TkBOBNnPHgJwGnAk8AP0tgLMtw7hfdJSLjcVq8HQHMBLYFFjTGfCgiDwHfBV4QkedxmkVfhNN8+ysJjEtR+qEJR1EsMMZ0isjZwC+B+3Dun/wFeNsY0+OeMVwGnAOcDPTitFL7G859lUTG0ugmtjtxzmr6cB5MPZLtD6gGchGwGbgAOAon0czBeW5n90TGpiiBSEDLSEVRhjEi8gGQa4yZnOpYlKGJ3sNRlGGG25ggeNrXcC6n/X3HR6QMF/QMR1GGGSJyGfANnJ4FmoB9cHojaAP2M8ZsSFlwypBGE46iDDNEZBZwC06iKQUace713GCMWZvC0JQhjiYcRVEUZYeg93AURVGUHYImHEVRFGWHoAlHURRF2SFowlEURVF2CJpwFEVRlB2CJhxFURRlh/D/AdSK4yI7WLxpAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 450x300 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "mt.plot_graphs()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After hyperparameter tuning targeting reduction of overfitting and improved overall performance, we conclude: \n",
    "\n",
    "With **95%** confidene that the gradient boosting machine algorithm was **not** able to achieve performance any different than classifying the samples by flipping a coin."
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "9386f0cef3f7e847ba4ef8cad0c29e3ff984e6eb5154ed6dc75c3de9b359a7f7"
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}