accuracy_vs_fairness.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": [
        "bBwjGdwhXNTv",
        "RDqdZeOX9R6q"
      ]
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "-XTG0H1-Dix1"
      },
      "outputs": [],
      "source": [
        "!pip install aif360"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pip install fairlearn"
      ],
      "metadata": {
        "id": "tCwqXqF9Dv7Q"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# pip install --upgrade seaborn"
      ],
      "metadata": {
        "id": "D3tADW7mUwzi"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        "os.chdir(r\"/usr/local/lib/python3.10/dist-packages/aif360/data/raw/adult\")\n",
        "!wget https://archive.ics.uci.edu/ml/machine-learning-databases/adult/adult.data\n",
        "!wget https://archive.ics.uci.edu/ml/machine-learning-databases/adult/adult.test\n",
        "!wget https://archive.ics.uci.edu/ml/machine-learning-databases/adult/adult.names"
      ],
      "metadata": {
        "id": "zoj-v-t4FSBQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# from aif360.datasets import AdultDataset\n",
        "from aif360.algorithms.preprocessing.optim_preproc_helpers.data_preproc_functions import load_preproc_data_adult\n",
        "from aif360.metrics import ClassificationMetric\n",
        "from aif360.algorithms.preprocessing.reweighing import Reweighing\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.svm import SVC\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from collections import Counter\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "from sklearn.metrics import log_loss, accuracy_score, f1_score, precision_score, recall_score, make_scorer, classification_report, confusion_matrix\n",
        "import itertools\n",
        "from sklearn.model_selection import GridSearchCV, StratifiedKFold\n",
        "\n",
        "import seaborn as sns\n",
        "import pandas as pd\n",
        "import aif360.sklearn.metrics as aim"
      ],
      "metadata": {
        "id": "fZlvY187D6PV",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e0b536fc-f08b-4835-8e8d-b43c9949d705"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:root:No module named 'tempeh': LawSchoolGPADataset will be unavailable. To install, run:\n",
            "pip install 'aif360[LawSchoolGPA]'\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Get data"
      ],
      "metadata": {
        "id": "2Wnpu_wGXGWC"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "df = load_preproc_data_adult(['sex'])\n",
        "privileged = [{'sex': 1}]\n",
        "unprivileged = [{'sex': 0}]"
      ],
      "metadata": {
        "id": "yLSkfFpsGDHz"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train, test = df.split([0.70], shuffle=True)\n",
        "print(\"Train size\", train.features.shape)\n",
        "print(\"Test size\", test.features.shape)\n",
        "weights = np.array(train.instance_weights)\n",
        "\n",
        "X_train = train.features\n",
        "y_train = train.labels.ravel()\n",
        "\n",
        "X_test = test.features\n",
        "y_test = test.labels.ravel()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3N7nw6MbLD8P",
        "outputId": "662d5508-c984-4d78-afba-1b73868b4148"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train size (34189, 18)\n",
            "Test size (14653, 18)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "label_df = pd.DataFrame({'Labels': [1]*len(y_train), 'Class': y_train})\n",
        "sns.countplot(data=label_df, x='Labels', hue='Class', width=0.5, stat='percent')\n",
        "plt.title('Label distribution')\n",
        "plt.xticks([])\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 448
        },
        "id": "Dm9FqTq6T67b",
        "outputId": "40545d8f-8840-4d50-c347-05eb375db37c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Bias - variance"
      ],
      "metadata": {
        "id": "Hon8zTBpXJqj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def rf_results(params):\n",
        "    model = LogisticRegression(\n",
        "        solver=params[0],\n",
        "        C=params[1]\n",
        "    )\n",
        "\n",
        "    model.fit(X_train, y_train)\n",
        "\n",
        "    tn_p = model.predict_proba(X_train)\n",
        "    ts_p = model.predict_proba(X_test)\n",
        "\n",
        "    tn_pr = model.predict(X_train)\n",
        "    ts_pr = model.predict(X_test)\n",
        "\n",
        "    tn_l = log_loss(y_train, tn_p)\n",
        "    ts_l = log_loss(y_test, ts_p)\n",
        "\n",
        "    tn_f = f1_score(y_train, tn_pr, average='macro', zero_division=0)\n",
        "    ts_f = f1_score(y_test, ts_pr, average='macro', zero_division=0)\n",
        "\n",
        "    tn_a = accuracy_score(y_train, tn_pr)\n",
        "    ts_a = accuracy_score(y_test, ts_pr)\n",
        "\n",
        "    return (tn_l, ts_l), (tn_f, ts_f), (tn_a, ts_a)"
      ],
      "metadata": {
        "id": "MGvaCiCeGFGV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "solver = ['lbfgs', 'liblinear', 'newton-cg', 'newton-cholesky', 'sag', 'saga']\n",
        "C = np.linspace(0.001, 0.05, 50)\n",
        "\n",
        "all_combinations = list(itertools.product(solver, C))\n",
        "\n",
        "log_losses = []\n",
        "f_scores = []\n",
        "accuracies = []\n",
        "\n",
        "\n",
        "for i, combination in enumerate(all_combinations):\n",
        "    ll, fsc, acc = rf_results(combination)\n",
        "\n",
        "    log_losses.append(ll)\n",
        "    f_scores.append(fsc)\n",
        "    accuracies.append(acc)\n",
        "\n",
        "    if i%20==0:\n",
        "        print(f\"{i+1}/{len(all_combinations)}\")"
      ],
      "metadata": {
        "id": "fp20Ol6UEDY0"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "sets = {'Accuracy': accuracies, 'F-score': f_scores, 'Log-loss': log_losses}\n",
        "\n",
        "plt.figure(figsize=(20, 4.5))\n",
        "for i, tset in enumerate(sets.items()):\n",
        "    plt.subplot(1, 3, i+1)\n",
        "    if i==2:\n",
        "        souts = sorted(tset[1], key=lambda x: x[0], reverse=True)\n",
        "    else:\n",
        "        souts = sorted(tset[1], key=lambda x: x[0])\n",
        "\n",
        "    tnouts, tsouts = zip(*souts)\n",
        "    plt.plot(tnouts, color='green', label='Train')\n",
        "    plt.plot(tsouts, color='blue', label='Test')\n",
        "    plt.legend()\n",
        "    plt.title(tset[0])\n",
        "    plt.xlabel(\"parameter combinations\")\n",
        "    plt.grid(True, linewidth=0.5, color='gray')\n",
        "\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "-0Jh0rr9Q9vh",
        "outputId": "206e1b8b-d5d8-456f-d0cd-0fb52c86914c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 2000x450 with 3 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Classifiers"
      ],
      "metadata": {
        "id": "vfR0_4O7bZzL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def accurate_model(X_train, y_train, weights):\n",
        "\n",
        "    n_estimators = [10, 20, 50]\n",
        "    max_depth = np.arange(1, 12, 1)\n",
        "    max_features = ['sqrt', 'log2', None]\n",
        "\n",
        "    parameter_grid = {\n",
        "                    'n_estimators': n_estimators,\n",
        "                    'max_depth': max_depth,\n",
        "                    'max_features': max_features,\n",
        "                    }\n",
        "\n",
        "    estimator = RandomForestClassifier(n_jobs=-1, random_state=42, class_weight='balanced')\n",
        "\n",
        "    gs = GridSearchCV(estimator, param_grid=parameter_grid, scoring='neg_log_loss', cv=3, verbose=4)\n",
        "    gs.fit(X_train, y_train, sample_weight=weights)\n",
        "\n",
        "\n",
        "    return gs.best_estimator_"
      ],
      "metadata": {
        "id": "Uq-A0xWfhJuR"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def fair_model(X_train, y_train, weights):\n",
        "\n",
        "    n_estimators = [10, 20, 50]\n",
        "    max_depth = np.arange(1, 12, 1)\n",
        "    max_features = ['sqrt', 'log2', None]\n",
        "\n",
        "    kfold = StratifiedKFold(n_splits=3, shuffle=True, random_state=9)\n",
        "    prms = ['n_estimators', 'max_depth', 'max_features', 'n_jobs', 'random_state', 'class_weight']\n",
        "    fairness_results = []\n",
        "    for n in n_estimators:\n",
        "        for d in max_depth:\n",
        "            for f in max_features:\n",
        "                fair_model = RandomForestClassifier(n_estimators=n, max_depth=d, max_features=f, n_jobs=-1, random_state=42, class_weight='balanced')\n",
        "                cv = 1\n",
        "                for train_index, test_index in kfold.split(X_train, y_train):\n",
        "                    X, x = X_train[train_index], X_train[test_index]\n",
        "                    Y, y = y_train[train_index], y_train[test_index]\n",
        "                    wt = [weights[i] for i in train_index]\n",
        "                    fair_model.fit(X, Y, sample_weight=wt)\n",
        "                    parr = {k: v for k, v in fair_model.get_params().items() if k in prms}\n",
        "                    yp = fair_model.predict(x)\n",
        "                    di = aim.disparate_impact_ratio(pd.Series(y), yp, prot_attr=pd.Series([k[1] for k in x]), zero_division=0)\n",
        "                    di = round(di, 3)\n",
        "\n",
        "                    print(f\"CV: {cv}\", end=\"; \")\n",
        "                    cv+=1\n",
        "                    print(f\"Params: {parr}\", end=\"; \")\n",
        "                    print(f\"DI score: {di}\")\n",
        "                    fairness_results.append((parr, di))\n",
        "\n",
        "    fairness_results = sorted(fairness_results, key=lambda x: x[1], reverse=True)\n",
        "    fair_params = fairness_results[0][0]\n",
        "\n",
        "    fairness = RandomForestClassifier(**fair_params)\n",
        "    for train_index, test_index in kfold.split(X_train, y_train):\n",
        "        X, x = X_train[train_index], X_train[test_index]\n",
        "        Y, y = y_train[train_index], y_train[test_index]\n",
        "        wt = [weights[i] for i in train_index]\n",
        "        fairness.fit(X, Y, sample_weight=wt)\n",
        "\n",
        "    return fairness"
      ],
      "metadata": {
        "id": "uhB8q6HYk6Mp"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def view_proportions(ground_truth, predictions):\n",
        "    verdicts = []\n",
        "\n",
        "    for a, b in zip(ground_truth, predictions):\n",
        "        if a==1 and b==1:\n",
        "            verdicts.append('True positive')\n",
        "        if a==1 and b==0:\n",
        "            verdicts.append('False negative')\n",
        "        if a==0 and b==1:\n",
        "            verdicts.append('False positive')\n",
        "        if a==0 and b==0:\n",
        "            verdicts.append('True negative')\n",
        "\n",
        "    return verdicts"
      ],
      "metadata": {
        "id": "Xq6_ohMRIKF2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Most accurate"
      ],
      "metadata": {
        "id": "bBwjGdwhXNTv"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "acc_uw_model = accurate_model(X_train, y_train, weights)"
      ],
      "metadata": {
        "id": "OV_hsY6ab8wO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "acc_pred_y = acc_uw_model.predict(X_test)\n",
        "print(classification_report(y_test, acc_pred_y))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8V8DHA-phvlv",
        "outputId": "812a17c1-1a4e-478f-f072-23d6dcbe2c5e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "         0.0       0.91      0.70      0.80     11132\n",
            "         1.0       0.46      0.79      0.58      3521\n",
            "\n",
            "    accuracy                           0.73     14653\n",
            "   macro avg       0.69      0.75      0.69     14653\n",
            "weighted avg       0.81      0.73      0.74     14653\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "conf_matrix = confusion_matrix(y_test, acc_pred_y)\n",
        "sns.heatmap(conf_matrix/np.sum(conf_matrix), annot=True, fmt='.2%')\n",
        "plt.xlabel(\"Predicted\")\n",
        "plt.ylabel(\"Actual\")\n",
        "plt.title(\"Visualising accurate model performance with uniform sample weights\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "iyIzBnqTph22",
        "outputId": "b20c5cb6-0ac2-4436-fd67-657be58aa47c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_pred_acc = test.copy()\n",
        "test_pred_acc.labels = acc_pred_y\n",
        "f_metric = ClassificationMetric(test, test_pred_acc, unprivileged_groups=unprivileged, privileged_groups=privileged)\n",
        "\n",
        "print(\"Fairness metrics of accurate model on unweighted samples:-\")\n",
        "print(\"Equal opportunity difference: \", round(f_metric.equal_opportunity_difference(), 4))\n",
        "print(\"Disparate Impact: \", round(f_metric.disparate_impact(), 4))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Tprw73m_psUG",
        "outputId": "af851b3a-9960-49ce-c062-3268dce1c6b3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fairness metrics of accurate model on unweighted samples:-\n",
            "Equal opportunity difference:  -0.3934\n",
            "Disparate Impact:  0.2722\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Most fair"
      ],
      "metadata": {
        "id": "RDqdZeOX9R6q"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "far_uw_model = fair_model(X_train, y_train, weights)"
      ],
      "metadata": {
        "id": "x6ZPRHx5cAc8"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "fpred = far_uw_model.predict(X_test)\n",
        "print(classification_report(y_test, fpred))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PTO3pJrlmDEh",
        "outputId": "386d3842-7293-429b-cf9e-01e7a4c13fc3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "         0.0       0.84      0.83      0.83     11132\n",
            "         1.0       0.48      0.51      0.50      3521\n",
            "\n",
            "    accuracy                           0.75     14653\n",
            "   macro avg       0.66      0.67      0.66     14653\n",
            "weighted avg       0.76      0.75      0.75     14653\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "conf_matrix = confusion_matrix(y_test, fpred)\n",
        "sns.heatmap(conf_matrix/np.sum(conf_matrix), annot=True, fmt='.2%')\n",
        "plt.xlabel(\"Predicted\")\n",
        "plt.ylabel(\"Actual\")\n",
        "plt.title(\"Visualising fair model performance with uniform sample weights\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "XeXoiZC2H070",
        "outputId": "42a62210-bfc1-4b49-980d-0cc8ae59cb96"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjEAAAHHCAYAAABOTAltAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABVH0lEQVR4nO3dd1hT1/8H8HeIEBCQIUsRRXHgRlERFXFgceGetQVRW+tWqlXar3tQRytu1LpqncVZVKilzopVcVulWkXrAEFUFCsIOb8//JE2JiBoNOv98snzmJN7zz33cnPzyeeccyMRQggQERER6RkTbTeAiIiI6E0wiCEiIiK9xCCGiIiI9BKDGCIiItJLDGKIiIhILzGIISIiIr3EIIaIiIj0EoMYIiIi0ksMYoiIiEgvvbMgZu3atZBIJEhOTn5Xm3jjdrRo0QItWrR463rep9jYWHh5ecHc3BwSiQSPHj0q8rpTpkyBRCJ5d417D95mH/r37w93d3fNNkgDUlNT0aNHD5QuXRoSiQSRkZHabpLRKc77On/ZU6dOvfuGFZG690Vubi6++OILuLm5wcTEBF26dNFO4wxYcnIyJBIJ1q5dq+2mFMnbtDd/3Xnz5mm+YRpQoqgLdurUCb/88gtSU1NhbW2tdpl+/frhxx9/xL179zTWQAIePHiAXr16oWbNmliyZAlkMhksLS213Sx6S2PGjEFcXBwmT54MFxcXNGjQQNtNIgBLly5FyZIl0b9/f2035Y2sXr0ac+fOxejRo1G/fn2UL19e200iI7F3716cOHECU6ZMeX8bFUW0efNmAUCsW7dO7etZWVnC0tJSBAUFCSGEyM3NFf/884+Qy+VF3cQ7sWbNGgFA3LhxQ1GWnZ0tsrOzi1WPNvdn3759AoDYv3//G63/4sUL8c8//2i4Ve/X5MmTRTFOVyUhISGiQoUKmm2QBjg7O4t+/fppuxlGTd37umbNmsLf319l2fxrycmTJ99jCwun7r3du3dv4erqqqUWGYcbN24IAGLNmjXabkqRyOVy8c8//4jc3Nxir5u/r3Pnzn3tssOGDXvj6/SbKnJ3UqdOnWBtbY2NGzeqfX3Xrl3IyspCv379AABSqVTR9aFrzMzMYGZmVqx1tLk/9+/fBwDY2tq+0folSpSAubl5ocvI5XI8f/78jeqnosvNzUVOTg6Al3/XN/2bqvP8+XPI5XKN1WcMdPk6VRTq3tuaPq94bdB/EokE5ubmkEql2m6KxhU5iLGwsEC3bt0QHx+v+FD9r40bN8La2hqdOnUCoL6v+dSpUwgMDISDgwMsLCxQsWJFDBgwQPH6wYMHIZFIcPDgQaW61fXnnT9/Hv3790elSpVgbm4OFxcXDBgwAA8ePHjtvqgbE7No0SLUrFkTJUuWhJ2dHRo0aKAUsKnbH3d3d3Ts2BFHjx5Fo0aNYG5ujkqVKuH7779X2eb58+fh7+8PCwsLlCtXDjNmzMCaNWte2x/fokULhISEAAAaNmwIiUSiSHMfOXIEPXv2RPny5SGTyeDm5oYxY8bgn3/+UapDXb+5RCLB8OHDsWHDBtSsWRMymQyxsbEFtiN/Xw8ePIgGDRrAwsICtWvXVvyttm/fjtq1a8Pc3Bze3t44c+aMSh2//vor/Pz8YGlpCVtbW3Tu3BmXL19WWe7o0aNo2LAhzM3N4eHhgeXLlxfYrh9++AHe3t6wsLCAvb09+vTpg7///rvA5QuTv48///yzYvxRjRo1sH37dpVlHz16hNGjR8PNzQ0ymQyVK1fG7NmzlYKI//YlR0ZGwsPDAzKZDEuXLoVEIoEQAkuWLIFEIlH6+1y/fh09e/aEvb09SpYsicaNG2PPnj1K289/r2zevBn/+9//4OrqipIlSyIzMxP9+/eHlZUVbt26hY4dO8LKygqurq5YsmQJAODChQto1aoVLC0tUaFCBZUvJhkZGRg7dixq164NKysrlCpVCu3atcO5c+fUtmHr1q2YOXMmypUrB3Nzc7Ru3RrXrl1TOWa///472rdvDzs7O1haWqJOnTpYsGCB0jJXrlxBjx49YG9vD3NzczRo0AC7d+9+7d+ufv366Natm1JZ7dq1IZFIcP78eUXZli1bIJFIFOfdq+9rd3d3XLp0CYcOHVL8XV69VmRnZyMsLAyOjo6wtLRE165dkZaW9to2FjQW79UxW/89b1asWKE4bxo2bIiTJ08qrfvf93b+egcOHMClS5cU7c9/j2ZlZeHzzz9XnLPVqlXDvHnzIIRQqrOga0P+sTp69ChGjhwJR0dH2NraYvDgwcjJycGjR48QHBwMOzs72NnZ4YsvvlCpW53XfS4AwLx589CkSROULl0aFhYW8Pb2RnR0tEpd+W3/8ccfUaNGDVhYWMDX1xcXLlwAACxfvhyVK1eGubk5WrRooXLtbdGiBWrVqoXExEQ0adJE0Z6oqKjX7gfw/s9fALhz5w4GDBgAZ2dnyGQy1KxZE6tXr1aqq6AxMfnHydzcHLVq1cKOHTsKHUNY2PnYv39/xTUm/9z773Vt8+bN8Pb2hrW1NUqVKoXatWurvP/fRJHHxAAvx7ysW7cOW7duxfDhwxXlGRkZiIuLQ9++fWFhYaF23fv37+ODDz6Ao6MjJkyYAFtbWyQnJ6v9gCiK/fv34/r16wgNDYWLiwsuXbqEFStW4NKlSzh+/HixvlmtXLkSI0eORI8ePTBq1Cg8f/4c58+fx++//44PP/yw0HWvXbuGHj16YODAgQgJCcHq1avRv39/eHt7o2bNmgBenmQtW7aERCJBeHg4LC0t8d1330Emk722bV999RWqVauGFStWYNq0aahYsSI8PDwAvDwBnz17hiFDhqB06dI4ceIEFi1ahNu3b+PHH398bd2//vqr4m/p4ODw2sGv165dw4cffojBgwfjo48+wrx58xAUFISoqCh8+eWXGDp0KAAgIiICvXr1QlJSEkxMXsbJv/zyC9q1a4dKlSphypQp+Oeff7Bo0SI0bdoUp0+fVmz7woULivNkypQpyM3NxeTJk+Hs7KzSnpkzZ2LixIno1asXBg0ahLS0NCxatAjNmzfHmTNn3ujb6NWrV9G7d2989tlnCAkJwZo1a9CzZ0/ExsaiTZs2AIBnz57B398fd+7cweDBg1G+fHkcO3YM4eHhuHfvnsoA3TVr1uD58+f49NNPIZPJUL9+faxfvx4ff/wx2rRpg+DgYMWyqampaNKkCZ49e4aRI0eidOnSWLduHTp16oTo6Gh07dpVqe7p06fDzMwMY8eORXZ2tiLDmJeXh3bt2qF58+aYM2cONmzYgOHDh8PS0hJfffUV+vXrh27duiEqKgrBwcHw9fVFxYoVAbwMonbu3ImePXuiYsWKSE1NxfLly+Hv748//vgDZcuWVWrD119/DRMTE4wdOxaPHz/GnDlz0K9fP/z++++KZfbv34+OHTuiTJkyGDVqFFxcXHD58mXExMRg1KhRAIBLly6hadOmcHV1xYQJE2BpaYmtW7eiS5cu2LZtm8q+/5efnx82bdqkeJ6RkYFLly7BxMQER44cQZ06dQC8DPwdHR1RvXp1tfVERkZixIgRsLKywldffQUAKufeiBEjYGdnh8mTJyM5ORmRkZEYPnw4tmzZUmD73sTGjRvx5MkTDB48GBKJBHPmzEG3bt1w/fp1mJqaqizv6OiI9evXY+bMmXj69CkiIiIAANWrV4cQAp06dcKBAwcwcOBAeHl5IS4uDuPGjcOdO3cwf/58pbrUXRvOnj2r2H8XFxdMnToVx48fx4oVK2Bra4tjx46hfPnymDVrFvbu3Yu5c+eiVq1aSuf3q4r6ubBgwQJ06tQJ/fr1Q05ODjZv3oyePXsiJiYGHTp0UFr2yJEj2L17N4YNGwbg5fWoY8eO+OKLL7B06VIMHToUDx8+xJw5czBgwAD8+uuvSus/fPgQ7du3R69evdC3b19s3boVQ4YMgZmZmUpw9V/aOH9TU1PRuHFjRfDm6OiIffv2YeDAgcjMzMTo0aML3OaePXvQu3dv1K5dGxEREXj48CEGDhwIV1dXtcu/7nwcPHgw7t69i/3792P9+vVK6+7fvx99+/ZF69atMXv2bADA5cuX8dtvvyne/2+sOH1Pubm5okyZMsLX11epPCoqSgAQcXFxirJXx6Ls2LHjtf3JBw4cEADEgQMHlMrV9T8+e/ZMZf1NmzYJAOLw4cMFtkMIIfz9/ZX6vDt37ixq1qxZyJ6rr6dChQoq27t//76QyWTi888/V5SNGDFCSCQScebMGUXZgwcPhL29vUqdhW371WOn7hhEREQIiUQibt68qShTN54EgDAxMRGXLl0qdNv58vf12LFjirK4uDgBQFhYWChtb/ny5Sp/Ry8vL+Hk5CQePHigKDt37pwwMTERwcHBirIuXboIc3Nzpfr++OMPIZVKlfYhOTlZSKVSMXPmTKV2XrhwQZQoUUKpvKhjYvL3cdu2bYqyx48fizJlyoh69eopyqZPny4sLS3Fn3/+qbT+hAkThFQqFbdu3RJC/HvelipVSty/f19lewDEsGHDlMpGjx4tAIgjR44oyp48eSIqVqwo3N3dRV5enhDi3/dKpUqVVM6DkJAQAUDMmjVLUfbw4UNhYWEhJBKJ2Lx5s6L8ypUrAoCYPHmyouz58+eK7eS7ceOGkMlkYtq0aYqy/DZUr15daYzZggULBABx4cIFIcTL60bFihVFhQoVxMOHD5Xq/e9YlNatW4vatWuL58+fK73epEkTUaVKFZXj918//vijACD++OMPIYQQu3fvFjKZTHTq1En07t1bsVydOnVE165dFc/Vva9fNyYmICBAqd1jxowRUqlUPHr0qNA2vnrdyffq+Zl/3pQuXVpkZGQoynft2iUAiJ9++klRpu697e/vr3I927lzpwAgZsyYoVTeo0cPIZFIxLVr1xRlBV0b8vc/MDBQaf99fX2FRCIRn332maIsNzdXlCtXTu3+/ldRPheEUL3W5eTkiFq1aolWrVoplQMQMplM6e+Zfz1ycXERmZmZivLw8HC1nw0AxDfffKMoy87OVly/cnJyhBDqP5O0cf4OHDhQlClTRqSnpyvV16dPH2FjY6M4buraW7t2bVGuXDnx5MkTRdnBgwcFgDc+HwsaEzNq1ChRqlSpNxqT8zrFmmItlUrRp08fJCQkKKXhNm7cCGdnZ7Ru3brAdfO/FcfExODFixfF2axa/834PH/+HOnp6WjcuDEA4PTp08Wqy9bWFrdv31ZJ1RZFjRo14Ofnp3ju6OiIatWq4fr164qy2NhY+Pr6wsvLS1Fmb2+vGD/0pv57DLKyspCeno4mTZpACKG2O+dV/v7+qFGjRpG3V6NGDfj6+iqe+/j4AABatWqlNAMivzz/GNy7dw9nz55F//79YW9vr1iuTp06aNOmDfbu3QvgZfYgLi4OXbp0UaqvevXqCAwMVGrL9u3bIZfL0atXL6SnpyseLi4uqFKlCg4cOFDk/fqvsmXLKn1jKlWqFIKDg3HmzBmkpKQAeJkB8/Pzg52dndK2AwICkJeXh8OHDyvV2b17dzg6OhZp+3v37kWjRo3QrFkzRZmVlRU+/fRTJCcn448//lBaPiQkpMDs56BBgxT/t7W1RbVq1WBpaYlevXopyqtVqwZbW1ul81UmkykyaHl5eXjw4AGsrKxQrVo1te+t0NBQpTFm+e+H/DrPnDmDGzduYPTo0SrZsfyMaUZGBn799Vf06tULT548URzTBw8eIDAwEFevXsWdO3cKPG7528w/9keOHEHDhg3Rpk0bHDlyBMDLLsCLFy8qvV/fxKeffqqU6fXz80NeXh5u3rz5VvW+qnfv3rCzs1PaDgClv1VR7d27F1KpFCNHjlQq//zzzyGEwL59+5TKC7s2DBw4UGn/fXx8IITAwIEDFWVSqRQNGjR4bVuL+rnw33P84cOHePz4Mfz8/NSej61bt1bKKudfj7p37640s/bV61S+EiVKYPDgwYrnZmZmGDx4MO7fv4/ExES17dPG+SuEwLZt2xAUFAQhhNK1KDAwEI8fPy7ws/Du3bu4cOECgoODYWVlpSj39/dH7dq11a7zNuejra0tsrKysH///tcuW1zFvk9M/gdvfj/67du3ceTIEfTp06fQQUP+/v7o3r07pk6dCgcHB3Tu3Blr1qxBdnb2GzU8IyMDo0aNgrOzMywsLODo6KhIhz9+/LhYdY0fPx5WVlZo1KgRqlSpgmHDhuG3334r0rrqpi/a2dnh4cOHiuc3b95E5cqVVZZTV1Yct27dUgQGVlZWcHR0hL+/P4CiHYP841VUr+6rjY0NAMDNzU1tef4xyL+4V6tWTaXO6tWrIz09HVlZWUhLS8M///yDKlWqqCz36rpXr16FEAJVqlSBo6Oj0uPy5ctqx20VReXKlVW6IqtWrQoAisD96tWriI2NVdluQEAAAKhsuzjH+ebNmwUep/zXi1K3ubm5SuBkY2ODcuXKqeyfjY2N0vkql8sxf/58VKlSBTKZDA4ODnB0dMT58+fVnlevnhf5F7r8Ov/66y8AQK1atdS2FXjZVSmEwMSJE1WO6+TJkwGoHtf/cnZ2RpUqVRQX/CNHjsDPzw/NmzfH3bt3cf36dfz222+Qy+VvHcS8bn81RZPbuXnzJsqWLatye4zinlfq2lXYdeB1bS3q50JMTAwaN24Mc3Nz2Nvbw9HREcuWLSvS+VjU61S+smXLqtzC4tVrwKu0cf6mpaXh0aNHWLFihco2Q0NDC91m/t+7OJ9Lb3M+Dh06FFWrVkW7du1Qrlw5DBgwoNAxmMVRrDExAODt7Q1PT09s2rQJX375JTZt2gQhxGuzChKJBNHR0Th+/Dh++uknxMXFYcCAAfjmm29w/PhxWFlZFTiOJS8vT6WsV69eOHbsGMaNGwcvLy9YWVlBLpejbdu2xZ6hUb16dSQlJSEmJgaxsbHYtm0bli5dikmTJmHq1KmFrltQ4CaKMKDtbeTl5aFNmzbIyMjA+PHj4enpCUtLS9y5cwf9+/cv0jEo6Bt8QQraV20cA7lcDolEgn379qnd/n+/XbyLbbdp0wZffPGF2tfzL3j5inuci6Ogut/mbzVr1ixMnDgRAwYMwPTp02Fvbw8TExOMHj1a7Xmlib9/fr1jx45Vybrle13Q36xZM8THx+Off/5BYmIiJk2ahFq1asHW1hZHjhzB5cuXYWVlhXr16hW5Xeq86f7mD+Z+lbrr29tsRxMKO2eLc24V5Zi87nPhyJEj6NSpE5o3b46lS5eiTJkyMDU1xZo1a9TOltXGdUob52/+Nj/66CPF5I9X5Y+l0YS3OX5OTk44e/Ys4uLisG/fPuzbtw9r1qxBcHAw1q1b91btKnYQA7zMxkycOBHnz5/Hxo0bUaVKFTRs2LBI6zZu3BiNGzfGzJkzsXHjRvTr1w+bN2/GoEGDFJHdq3ejffVbwsOHDxEfH4+pU6di0qRJivKrV6++ye4AACwtLdG7d2/07t0bOTk56NatG2bOnInw8PDXTk9+nQoVKqidraGurKguXLiAP//8E+vWrVMaOPcu0nVvq0KFCgCApKQkldeuXLkCBwcHWFpawtzcHBYWFmr/jq+u6+HhASEEKlasqBI0vI38b1T/Daj//PNPAFCkqD08PPD06VNF5kWTKlSoUOBxyn/9XYuOjkbLli2xatUqpfJHjx7BwcGh2PXlD0S/ePFigcesUqVKAABTU9M3Pq5+fn5Ys2YNNm/ejLy8PDRp0gQmJiZo1qyZ4kOgSZMmr51m+q6mW9vZ2alNvWu6G0qdChUq4JdffsGTJ0+UsjHv87wqTGGfC9u2bYO5uTni4uKUJkOsWbPmnbTl7t27yMrKUsrGvHoNeJU2zl9HR0dYW1sjLy+v2NvM/3tr+nOpsPeOmZkZgoKCEBQUBLlcjqFDh2L58uWYOHHiW/VKvNHPDuRnXSZNmoSzZ88WaWzHw4cPVSK2/DEi+anDChUqQCqVqowpWLp0qdLz/D/iq/W96W3bX52WbWZmhho1akAIoZHxO4GBgUhISFCM7gdedodt2LDhjetUdwyEEBqZsqZpZcqUgZeXF9atW6cUoF68eBE///wz2rdvD+DlPgUGBmLnzp24deuWYrnLly8jLi5Oqc5u3bpBKpVi6tSpKueBEKJIU+3VuXv3Lnbs2KF4npmZie+//x5eXl5wcXEB8DILmJCQoNIm4OUHfW5u7httGwDat2+PEydOICEhQVGWlZWFFStWwN3dvVhjmN6UVCpVOaY//vhjoX36halfvz4qVqyIyMhIlS8o+dtxcnJCixYtsHz5crV3/C7KFOb8NPvs2bNRp04dRXeBn58f4uPjcerUqSJ1JVlaWhbrZz2KysPDA1euXFHal3PnzhW56/pttG/fHnl5eVi8eLFS+fz58yGRSNCuXbt33gZ1ivK5IJVKIZFIlDJWycnJ2Llz5ztpU25urtJtHXJycrB8+XI4OjrC29tb7TraOH+lUim6d++Obdu24eLFi8XaZtmyZVGrVi18//33ePr0qaL80KFDiunobyI/8Hv1/fPq9djExESRJXrTISX53igTU7FiRTRp0gS7du0CgCIFMevWrcPSpUvRtWtXeHh44MmTJ1i5ciVKlSql+BCzsbFBz549sWjRIkgkEnh4eCAmJkalX69UqVKKqaMvXryAq6srfv75Z9y4ceNNdgcffPABXFxc0LRpUzg7O+Py5ctYvHgxOnToUOBPLBTHF198gR9++AFt2rTBiBEjFFOsy5cvj4yMjDf65ufp6QkPDw+MHTsWd+7cQalSpbBt2zaN98tryty5c9GuXTv4+vpi4MCBiinWNjY2Sreonjp1KmJjY+Hn54ehQ4ciNzdXcQ+f/94vwcPDAzNmzEB4eDiSk5PRpUsXWFtb48aNG9ixYwc+/fRTjB07ttjtrFq1KgYOHIiTJ0/C2dkZq1evRmpqqtK3vnHjxmH37t3o2LGjYjp9VlYWLly4gOjoaCQnJ79RxgIAJkyYgE2bNqFdu3YYOXIk7O3tsW7dOty4cQPbtm1TDLh9lzp27Ihp06YhNDQUTZo0wYULF7BhwwbFt83iMjExwbJlyxAUFAQvLy+EhoaiTJkyuHLlCi5duqQIBpcsWYJmzZqhdu3a+OSTT1CpUiWkpqYiISEBt2/fVrlPzasqV64MFxcXJCUlYcSIEYry5s2bY/z48QBQpCDG29sby5Ytw4wZM1C5cmU4OTmhVatWb7Tv/zVgwAB8++23CAwMxMCBA3H//n1ERUWhZs2ayMzMfOv6CxMUFISWLVviq6++QnJyMurWrYuff/4Zu3btwujRoxXZsvetKJ8LHTp0wLfffou2bdviww8/xP3797FkyRJUrlxZ6ZqgKWXLlsXs2bORnJyMqlWrYsuWLTh79ixWrFihdmp7Pm2cv19//TUOHDgAHx8ffPLJJ6hRowYyMjJw+vRp/PLLL8jIyChwe7NmzULnzp3RtGlThIaG4uHDh1i8eDFq1aqlFNgUR36QN3LkSAQGBiomAw0aNAgZGRlo1aoVypUrh5s3b2LRokXw8vIq8HYHRfam05qWLFkiAIhGjRqpff3VqYunT58Wffv2FeXLlxcymUw4OTmJjh07ilOnTimtl5aWJrp37y5Kliwp7OzsxODBg8XFixdVpofdvn1bdO3aVdja2gobGxvRs2dPcffuXZXpokWZYr18+XLRvHlzUbp0aSGTyYSHh4cYN26cePz4caH1VKhQQXTo0EFl39VNpTxz5ozw8/MTMplMlCtXTkRERIiFCxcKACIlJUX9QX5l269OQ/zjjz9EQECAsLKyEg4ODuKTTz4R586dUzlWBU2xfnV6b2EK2ld19RR0m+pffvlFNG3aVFhYWIhSpUqJoKAgxZTC/zp06JDw9vYWZmZmolKlSiIqKqrAnx3Ytm2baNasmbC0tBSWlpbC09NTDBs2TCQlJSmWKc4U6w4dOoi4uDhRp04dIZPJhKenp/jxxx9Vln3y5IkIDw8XlStXFmZmZsLBwUE0adJEzJs3T2UaZkG36y7ob/DXX3+JHj16CFtbW2Fubi4aNWokYmJilJbJn96srm0hISHC0tJSpVzd1Nv/7ne+58+fi88//1yUKVNGWFhYiKZNm4qEhASV87qgNhR0S/ajR4+KNm3aCGtra2FpaSnq1KkjFi1apLLvwcHBwsXFRZiamgpXV1fRsWNHER0drdJudXr27CkAiC1btijKcnJyRMmSJYWZmZnKLfrVva9TUlJEhw4dhLW1tQCg2OeC3ocF3RpCnR9++EFUqlRJmJmZCS8vLxEXF1fgFGt1582r17eiTrEW4uU5O2bMGFG2bFlhamoqqlSpIubOnavyUyoFnZcF7X9+G9LS0pTKCzoP/6uonwurVq0SVapUUbwn16xZU+TrWkHHU935m3/sTp06JXx9fYW5ubmoUKGCWLx4sdo6Xz3H3/f5K4QQqampYtiwYcLNzU2YmpoKFxcX0bp1a7FixYrXtnfz5s3C09NTyGQyUatWLbF7927RvXt34enp+drjJ4Tq+ZibmytGjBghHB0dhUQiUfx9oqOjxQcffCCcnJyEmZmZKF++vBg8eLC4d+9ekY5LYST/3xDSgtGjR2P58uV4+vSpQd4OWt+4u7ujVq1aiImJ0XZTiEgLWrRogfT0dLXdM8bCy8sLjo6OOjm+Up13n5smAFD5KYAHDx5g/fr1aNasGQMYIiJ6r168eKEyfu/gwYM4d+6c2p/H0FVvNCaGis/X1xctWrRA9erVkZqailWrViEzMxMTJ07UdtOIiMjI3LlzBwEBAfjoo49QtmxZXLlyBVFRUXBxccFnn32m7eYVGYOY96R9+/aIjo7GihUrIJFIUL9+faxatQrNmzfXdtOIiMjI2NnZwdvbG9999x3S0tJgaWmJDh064Ouvv0bp0qW13bwi45gYIiIi0kscE0NERER6iUEMERER6SUGMURERKSXOLD3Lb1If/3PkBMZG7vyrbXdBCKd8/TZm91Vvjg0+Zlk6vBmd+l+n5iJISIiIr3ETAwREZGhkOe9fhkDwiCGiIjIUAi5tlvwXjGIISIiMhRy4wpiOCaGiIiI9BIzMURERAZCsDuJiIiI9BK7k4iIiIh0HzMxREREhoLdSURERKSXjOw+MexOIiIiIr3ETAwREZGhYHcSERER6SXOTiIiIiLSfczEEBERGQhju9kdMzFERESGQi7X3KOYlixZAnd3d5ibm8PHxwcnTpwodPlHjx5h2LBhKFOmDGQyGapWrYq9e/cWa5vMxBARERkKLWVitmzZgrCwMERFRcHHxweRkZEIDAxEUlISnJycVJbPyclBmzZt4OTkhOjoaLi6uuLmzZuwtbUt1nYlQgihoX0wSi/Sr2u7CUQ6x658a203gUjnPH12451vI/vPoxqrS1a1WZGX9fHxQcOGDbF48WIAgFwuh5ubG0aMGIEJEyaoLB8VFYW5c+fiypUrMDU1feM2sjuJiIjIUMjzNPbIzs5GZmam0iM7O1tlkzk5OUhMTERAQICizMTEBAEBAUhISFDbzN27d8PX1xfDhg2Ds7MzatWqhVmzZiEvr3g362MQQ0REZCiEXGOPiIgI2NjYKD0iIiJUNpmeno68vDw4OzsrlTs7OyMlJUVtM69fv47o6Gjk5eVh7969mDhxIr755hvMmDGjWLvLMTFERESkIjw8HGFhYUplMplMI3XL5XI4OTlhxYoVkEql8Pb2xp07dzB37lxMnjy5yPUwiCEiIjIUGrzZnUwmK1LQ4uDgAKlUitTUVKXy1NRUuLi4qF2nTJkyMDU1hVQqVZRVr14dKSkpyMnJgZmZWZHayO4kIiIiQ6HB7qSiMjMzg7e3N+Lj4xVlcrkc8fHx8PX1VbtO06ZNce3aNcj/E3T9+eefKFOmTJEDGIBBDBEREb2lsLAwrFy5EuvWrcPly5cxZMgQZGVlITQ0FAAQHByM8PBwxfJDhgxBRkYGRo0ahT///BN79uzBrFmzMGzYsGJtl91JREREhkJLv53Uu3dvpKWlYdKkSUhJSYGXlxdiY2MVg31v3boFE5N/8yZubm6Ii4vDmDFjUKdOHbi6umLUqFEYP358sbbL+8S8Jd4nhkgV7xNDpOp93Cfm+bni3fG2MOZ122usrneF3UlERESkl9idREREZCiM7AcgGcQQEREZCi2NidEWBjFERESGwsgyMRwTQ0RERHqJmRgiIiJDIS/eDyjqOwYxREREhoLdSURERES6j5kYIiIiQ8HZSURERKSX2J1EREREpPuYiSEiIjIU7E4iIiIivWRkQQy7k4iIiEgvMRNDRERkIITgze6IiIhIHxlZdxKDGCIiIkPBKdZEREREuo+ZGCIiIkPB7iQiIiLSS+xOIiIiItJ9zMQQEREZCnYnERERkV5idxIRERGR7mMmhoiIyFCwO4mIiIj0kpEFMexOIiIiIr3ETAwREZGhMLKBvQxiiIiIDIWRdScxiCEiIjIURpaJ4ZgYIiIi0kvMxBARERkKdicRERGRXmJ3EhEREZHuYyaGiIjIULA7iYiIiPSSkQUx7E4iIiIivcRMDBERkaEQQtsteK8YxBARERkKdicRERER6T5mYoiIiAyFkWViGMQQEREZCiO72R2DGCIiIkNhZJkYjokhIiIivcRMDBERkaHgFGsiIiLSS+xOIiIiItJ9zMQQEREZCiPLxDCIISIiMhRGNsWa3UlERESkl5iJISIiMhBCztlJREREpI+MbEwMu5OIiIhILzETQ0REZCiMbGAvgxgiIiJDwTExREREpJc4JoaIiIhI9zETQ0REZCiMLBPDIIaIiMhQGNmvWLM7iYiIiPQSgxh6b1LT0jF+6hw0bdcL3i07o+vHQ3Dx8p8AgBe5ufh26Sp0/XgIGrbugpad+iF8+jzcT3tQaJ0fdA9BrabtVB4zvlkCALhzL1Xt67WatkPcr0cAAI8zn2DYF5PRMKArevQfhst/XlPaxoxvlmDtpm3v4IgQAU2bNsLW6O9w9a/jePrsBjoGtVF6/cuvRuH0mV+QmnYJf985i59i1qNBQ69C6/x87BAcOrIT91Iv4EbySWzashxVqlRSWkYmM8O386fh5t+nkXL/IjZsXAonJwfF63Z2Ntga/R1S7l/EbwkxqFO3htL6386fhhEjB73dzpPmyeWae+gBBjH0XjzOfIKPP/scpiVKIOqb6di1YTnGDh+EUtZWAIDnz7PxR9JfGNy/L7auXozIWf9D8q3bGD5+aqH1bv5uAQ7u3qB4rIycBQD4oKUfAMDFyUHp9YO7N2DYwI9Q0sICfo0bAABWrNuMrGf/4MfVi9CwXh1M/nqhov5zFy/j/KUkfNyryzs4KkRASUsLXLxwGWFjJql9/erVGwgLmwyfhm3xQUBP3Lx1B7t2r4ODg32BdTbz88GK5evRqkU3BAUFw9S0BHb99D1KlrRQLDN7zkS0a98KwR8NQ9vAPnAp44wNm5YpXh/3xXBYW1miWZMgHDl8HIuXRChea9jQCw0a1sWSxas1cARIo+RCc49iWrJkCdzd3WFubg4fHx+cOHGiwGXXrl0LiUSi9DA3Ny/2Njkmht6L1Rt+hIuTI2Z8FaYoK1fWRfF/aytLfLdgltI6X4YNQd9Bo3Ev5T7KuDiprdfezlbp+Xfrt8LNtQwa1qsNAJBKpXAorXyxjz98DIGt/RQX9OvJt9CutT/cy5dDj87tEL17H4CX2aFpcxdj6oRRkEqlb7bjRK+x/+dD2P/zoQJf/3HrbqXn4eNnoH//3qhVyxMHDx5Tu07Xzv2Vnn/26Tgk30pEvXq18dtvJ1CqlDWCQ3phQP/ROHQoAQAwZPA4nD4bj4YNvXDy5FlU8/RAdPRPuHbtBtas3oTQAX0BACVKlMCCRTMxbOgEyPXk2zq9e1u2bEFYWBiioqLg4+ODyMhIBAYGIikpCU5O6q/fpUqVQlJSkuK5RCIp9naNJhOTnp6OOXPmoGvXrvD19YWvry+6du2KuXPnIi0tTdvNM3gHjh5HTc8qCPvfTDTv0Ac9+g9TBAsFefr0GSQSCaytLYu0jRcvXiDm5wPo2uGDAt8Ml65cxZWr19GtY6CirGrlSjhx+hxyc/Nw7PdEVPWoCOBl4NWwXm3Uql61iHtJ9G6ZmpoidEBfPHqUiQsXLhd5vVKlrAEADx8+AgDUq1cLZmZmOHDgqGKZP/+8jlu37qCRT30AwIXzl+Hv3wRSqRStA5rj4sUrAIAxYYNx5PBxnDl9QUN7RRol5Jp7FMO3336LTz75BKGhoahRowaioqJQsmRJrF5dcLZOIpHAxcVF8XB2di727hpFEHPy5ElUrVoVCxcuhI2NDZo3b47mzZvDxsYGCxcuhKenJ06dOqXtZhq023dTsGXnHpQv54rl82egd9cOiJgfhV1796tdPjs7B/OXrUb7AH9YWRYtiIk/nIAnT5+iS/s2BS6zPSYOldzdUK/2v/37gz7uBalUina9BiD+8DFMCx+Nm3/fwe598fgs9ENMnbMIbXuG4vOJs/DkaVbxdpxIA9q2a4WU+xfx4OEVDB8xAJ2CPsaDBw+LtK5EIsHsuRNx7NhJ/PHHyzFoTs6OyM7OxuPHT5SWvX8/Hc7OjgCAb7+JQm5uLi5cOoROnQIxbMh4eHi448N+3TH760VYsHAGLlw6hO/XL1YESaQDtNCdlJOTg8TERAQEBCjKTExMEBAQgISEhALXe/r0KSpUqAA3Nzd07twZly5dKvbuGkV30ogRI9CzZ09ERUWpfEMXQuCzzz7DiBEjCj3YAJCdnY3s7GylMpPsbMhkMo232dDI5QI1Patg9Gf9AQDVq1bG1es3sXXnXnR+Jeh4kZuLzyfOghACE8cNL/I2tsfEoVnjBnByLK329efZ2di7/yAG9++rVG5tZYk5U8YrlQ0YMQGfDxuImJ8P4Pbde/hp00pM+XoBotZsxLgRnxS5TUSacPhQApo07oDSpe3Qf0AffL9+MVr6d0Xaawa+A8D8yGmoUaMa2gT0LNY2MzOfYEDoaKWyPXs34H9fRaB37y5wdy+PenVbY/GSCEwIH4kvw2cWq37Sfeo+82QymcpnXnp6OvLy8lQyKc7Ozrhy5YrauqtVq4bVq1ejTp06ePz4MebNm4cmTZrg0qVLKFeuXJHbaBSZmHPnzmHMmDFquxgkEgnGjBmDs2fPvraeiIgI2NjYKD1mL4h6By02PI6l7eHhXl6prJK7G+6lKnfl5Qcwd1PvY2XkrCJnYe6mpOL4qbPoHtS2wGV+PnAU/zzPRqe2rQuta8een2FtZYlWfr44efo8WjdvAtMSJfBBKz+cPHO+SO0h0qRnz/7B9es3cfLkWQwbMgG5ubkIDun12vW++XYq2rZrhfZt++LunRRF+f3UNMhkMtjYKGdQnJwckJqqvnv9o4974PHjTOyJ2Q+/5j6IifkZubm52LFjL/ya+7zdDpLGCLlcYw91n3kRERGvb0QR+Pr6Ijg4GF5eXvD398f27dvh6OiI5cuXF6seo8jEuLi44MSJE/D09FT7+okTJ4rUFxceHo6wsDClMpMndzTSRkNXr04NJN+6rVR289YdpQG7+QHMrb/vYvWir2FrU6rI9e/Ysx/2djZo7tuowGW2x8ShZTMflcHA/5Xx8BGi1mzE98vmAQDkcjle5OYCAHJzc5GXx4GMpH0mJiaQycwKXeabb6ciqNMHaBfYFzdvKr/3zpy5iJycHLRo0RS7dsUCAKpUqYTy5V1x4vfTKnU5ONhjQvhIRTbHRCqFaYmXHx+mpiU48F2XaPAHINV95qnreXBwcIBUKkVqaqpSeWpqKlxcXFSWV8fU1BT16tXDtWvXXr/wfxhFEDN27Fh8+umnSExMROvWrRUBS2pqKuLj47Fy5UrMmzfvtfWoS6O9yEl/J202NB/37oKPB3+OFes2o23r5rjwRxKid+/D5C9GAngZwIR9NRN//HkNS+ZMhVwuR/qDDACATSlrmJqaAgAGjpyA1s2b4MMenRR1y+Vy7NyzH53bBaBECfUX01u37yLx7EUsmzet0HbOXrAcIX26wdnx5f0y6tWpgZ9if0WTRvURvWsf6tWpUej6RMVlaVkSlTwqKJ5XqOCG2nWq42HGY2RkPMS48cOwN+YXpKSkobSDHT4d/DHKlnXBju17FevE7PkBP/30M5ZHfQ/gZRdSz16d0afXp3jy9CmcnF+ez5mPn+D582xkZj7B9+u2ImL2//Dw4SNkPnmKed9MwfHjiTh58qxKG2fPmYhFC7/DvbsvP6SOJySiz4ddER9/BKED+uJ4QuI7PEJULMUckFsYdZ956piZmcHb2xvx8fHo0qULgJfX5fj4eAwfXrQhAXl5ebhw4QLat29frDYaRRAzbNgwODg4YP78+Vi6dCny8vIAvJx+6+3tjbVr16JXr9enZunN1a5eDZERE7Egai2i1m6EaxkXjB81GB0DWwEA7qc9wIGjxwEAPfoPU1p39aLZaFS/DgDg7zv38PBxptLrCSfP4F7qfXTt8EGB298e8zOcnRzQpFH9Apf57fdE3LpzDxGTxinK+nYPwqUrV/HhJ6NRu0Y1DBnQr3g7TvQa9evXxr64zYrns+dMBAD8sD4ao0Z+hWpVPdBvU3eULm2HjIxHSEw8jw/a9MLly1cV61SsVAGlS9spnn/y6ccAgNif/60XAAZ/OhYbfnh548bxX0yHXC7ww8ZlkMnMEP/LYYwePVGlfa0DmqOShzsGDfz3G/nyqHWoX782DhzagcRT5xAxa4EGjgTps7CwMISEhKBBgwZo1KgRIiMjkZWVhdDQUABAcHAwXF1dFd1R06ZNQ+PGjVG5cmU8evQIc+fOxc2bNzFoUPFuoCgRwrh+aOHFixdIT3+ZPXFwcFB8w3/j+tKva6JZRAbFrnzh446IjNHTZzfe+Taypmnui5blpA3FWn7x4sWYO3cuUlJS4OXlhYULF8LH5+V4qRYtWsDd3R1r164FAIwZMwbbt29HSkoK7Ozs4O3tjRkzZqBevXrF2qbRBTGaxiCGSBWDGCJV7yWImdL39QsVkeWUTRqr610xitlJREREZHiMYkwMERGRUdDg7CR9wCCGiIjIUGhwdpI+YHcSERER6SVmYoiIiAwFu5OIiIhIHwk5u5OIiIiIdB4zMURERIaC3UlERESklxjEEBERkV7iFGsiIiIi3cdMDBERkaFgdxIRERHpI2FkQQy7k4iIiEgvMRNDRERkKIwsE8MghoiIyFDwjr1EREREuo+ZGCIiIkPB7iQiIiLSS0YWxLA7iYiIiPQSMzFEREQGQgjjysQwiCEiIjIURtadxCCGiIjIUBhZEMMxMURERKSXmIkhIiIyEMb220kMYoiIiAyFkQUx7E4iIiIivcRMDBERkaEwrp9OYhBDRERkKIxtTAy7k4iIiEgvMRNDRERkKIwsE8MghoiIyFAY2ZgYdicRERGRXmImhoiIyEAY28BeBjFERESGwsi6kxjEEBERGQhjy8RwTAwRERHpJWZiiIiIDAW7k4iIiEgfCSMLYtidRERERHqJmRgiIiJDYWSZGAYxREREBoLdSURERER6gJkYIiIiQ2FkmRgGMURERAbC2LqTGMQQEREZCGMLYjgmhoiIiPQSMzFEREQGwtgyMQxiiIiIDIWQaLsF7xW7k4iIiEgvMRNDRERkINidRERERHpJyNmdRERERKTzmIkhIiIyEOxOIiIiIr0kODuJiIiISPcxE0NERGQg2J1EREREesnYZicxiCEiIjIQQmi7Be8Xx8QQERGRXmImhoiIyECwO4mIiIj0krEFMexOIiIiore2ZMkSuLu7w9zcHD4+Pjhx4kSR1tu8eTMkEgm6dOlS7G0yiCEiIjIQQmjuURxbtmxBWFgYJk+ejNOnT6Nu3boIDAzE/fv3C10vOTkZY8eOhZ+f3xvtL4MYIiIiAyHkEo09iuPbb7/FJ598gtDQUNSoUQNRUVEoWbIkVq9eXeA6eXl56NevH6ZOnYpKlSq90f4yiCEiIiIV2dnZyMzMVHpkZ2erLJeTk4PExEQEBAQoykxMTBAQEICEhIQC6582bRqcnJwwcODAN24jgxgiIiIDIYREY4+IiAjY2NgoPSIiIlS2mZ6ejry8PDg7OyuVOzs7IyUlRW07jx49ilWrVmHlypVvtb86NTtp9+7dRV62U6dO77AlRERE+keTPzsQHh6OsLAwpTKZTPbW9T558gQff/wxVq5cCQcHh7eqS6eCmKKOTJZIJMjLy3u3jSEiIjJiMpmsSEGLg4MDpFIpUlNTlcpTU1Ph4uKisvxff/2F5ORkBAUFKcrk8pfRV4kSJZCUlAQPD48itVGngpj8nSAiIqLik4v3f58YMzMzeHt7Iz4+XpGMkMvliI+Px/Dhw1WW9/T0xIULF5TK/ve//+HJkydYsGAB3NzcirxtnQpiiIiI6M0JLQQxABAWFoaQkBA0aNAAjRo1QmRkJLKyshAaGgoACA4OhqurKyIiImBubo5atWoprW9rawsAKuWvo9NBTFZWFg4dOoRbt24hJydH6bWRI0dqqVVERES6SVt37O3duzfS0tIwadIkpKSkwMvLC7GxsYrBvrdu3YKJiebnEkmE0M3fvDxz5gzat2+PZ8+eISsrC/b29khPT0fJkiXh5OSE69eva7uJAIAX6brRDiJdYle+tbabQKRznj678c63caVqe43V5fnnXo3V9a7o7BTrMWPGICgoCA8fPoSFhQWOHz+OmzdvwtvbG/PmzdN284iIiHSOtu7Yqy06G8ScPXsWn3/+OUxMTCCVSpGdnQ03NzfMmTMHX375pbabR0REpHO0dcdebdHZIMbU1FTRf+bk5IRbt24BAGxsbPD3339rs2lERESkA3R2YG+9evVw8uRJVKlSBf7+/pg0aRLS09Oxfv36Yo9eJiIiMgbamGKtTTqbiZk1axbKlCkDAJg5cybs7OwwZMgQpKWlYcWKFVpuHRERke7R5M8O6AOdzcQ0aNBA8X8nJyfExsZqsTVERESka3Q2iCEiIqLi0ZdZRZqis0FMxYoVIZEUnM7SlfvEEBER6QpjGxOjs0HM6NGjlZ6/ePECZ86cQWxsLMaNG6edRhEREZHO0NkgZtSoUWrLlyxZglOnTr3n1hAREek+fRmQqyk6OzupIO3atcO2bdu03QwiIiKdY2x37NXZTExBoqOjYW9vr+1mEBER6RyOidER9erVUxrYK4RASkoK0tLSsHTpUi22jIiIiHSBzgYxnTt3VgpiTExM4OjoiBYtWsDT01OLLVNmUdZP200g0jkmhcwsJKJ3x9jGxOhsEDNlyhRtN4GIiEivGFt3ks4O7JVKpbh//75K+YMHDyCVSrXQIiIiItIlOpuJEQUMjc7OzoaZmdl7bg0REZHu05NJRRqjc0HMwoULAQASiQTfffcdrKysFK/l5eXh8OHDOjUmhoiISFcYW3eSzgUx8+fPB/AyExMVFaXUdWRmZgZ3d3dERUVpq3lERESkI3QuiLlx4wYAoGXLlti+fTvs7Oy03CIiIiL9wNlJOuLAgQPabgIREZFekWu7Ae+Zzs5O6t69O2bPnq1SPmfOHPTs2VMLLSIiIiJdorNBzOHDh9G+fXuV8nbt2uHw4cNaaBEREZFuE5Bo7KEPdLY76enTp2qnUpuamiIzM1MLLSIiItJtciObY62zmZjatWtjy5YtKuWbN29GjRo1tNAiIiIi3SaHRGMPfaCzmZiJEyeiW7du+Ouvv9CqVSsAQHx8PDZu3Ijo6Ggtt46IiIi0TWeDmKCgIOzcuROzZs1CdHQ0LCwsULduXfz666+wt7fXdvOIiIh0jr6MZdEUiSjo/v46JjMzE5s2bcKqVauQmJiIvLw8bTcJAFDCzFXbTSDSOfwVayJVOdm33/k29jv31lhdbVJVh3ToGp0dE5Pv8OHDCAkJQdmyZfHNN9+gVatWOH78uLabRURERFqmk91JKSkpWLt2LVatWoXMzEz06tUL2dnZ2LlzJwf1EhERFcDYupN0LhMTFBSEatWq4fz584iMjMTdu3exaNEibTeLiIhI58k1+NAHOpeJ2bdvH0aOHIkhQ4agSpUq2m4OERER6Sidy8QcPXoUT548gbe3N3x8fLB48WKkp6dru1lEREQ6z9gyMToXxDRu3BgrV67EvXv3MHjwYGzevBlly5aFXC7H/v378eTJE203kYiISCcZ288O6MUU66SkJKxatQrr16/Ho0eP0KZNG+zevVvbzQLAKdZE6nCKNZGq9zHFeo9zX43V1SF1k8bqeld0LhOjTrVq1TBnzhzcvn0bmzbp/kElIiLSBrlEcw99oBeZGF3GTAyRKmZiiFS9j0zMLpcPNVZX55SNGqvrXdG52UlERET0ZowtK6EX3UlEREREr2ImhoiIyEDoy9RoTWEQQ0REZCDkRjYejd1JREREpJeYiSEiIjIQxjawl0EMERGRgTC2MTHsTiIiIiK9xEwMERGRgdCXO+1qCoMYIiIiAyHXkx9u1BR2JxEREZFeYiaGiIjIQHB2EhEREekljokhIiIivcQp1kRERER6gJkYIiIiA8ExMURERKSXjG1MDLuTiIiISC8xE0NERGQgjG1gL4MYIiIiA2FsQQy7k4iIiEgvMRNDRERkIISRDexlEENERGQg2J1EREREpAcYxBARERkIuQYfxbVkyRK4u7vD3NwcPj4+OHHiRIHLbt++HQ0aNICtrS0sLS3h5eWF9evXF3ubDGKIiIgMhNDgozi2bNmCsLAwTJ48GadPn0bdunURGBiI+/fvq13e3t4eX331FRISEnD+/HmEhoYiNDQUcXFxxdquRAhhbHcp1qgSZq7abgKRzjGRGNnoQqIiyMm+/c63saD8Rxqra9StH4q8rI+PDxo2bIjFixcDAORyOdzc3DBixAhMmDChSHXUr18fHTp0wPTp04u8XWZiiIiI6I3l5OQgMTERAQEBijITExMEBAQgISHhtesLIRAfH4+kpCQ0b968WNvm7CQiIiIDocnZSdnZ2cjOzlYqk8lkkMlkSmXp6enIy8uDs7OzUrmzszOuXLlSYP2PHz+Gq6srsrOzIZVKsXTpUrRp06ZYbWQmhoiIyEBocmBvREQEbGxslB4REREaa6u1tTXOnj2LkydPYubMmQgLC8PBgweLVQczMURERKQiPDwcYWFhSmWvZmEAwMHBAVKpFKmpqUrlqampcHFxKbB+ExMTVK5cGQDg5eWFy5cvIyIiAi1atChyG5mJISIiMhCanJ0kk8lQqlQppYe6IMbMzAze3t6Ij49XlMnlcsTHx8PX17fIbZfL5SrdV6/DTAwREZGBkGtpYmBYWBhCQkLQoEEDNGrUCJGRkcjKykJoaCgAIDg4GK6uroruqIiICDRo0AAeHh7Izs7G3r17sX79eixbtqxY22UQQ0RERG+ld+/eSEtLw6RJk5CSkgIvLy/ExsYqBvveunULJib/dv5kZWVh6NChuH37NiwsLODp6YkffvgBvXv3LtZ2eZ+Yt8T7xBCp4n1iiFS9j/vEfF1Bc/eJmXCz6PeJ0RZmYoiIiAyEsWUlOLCXiIiI9BIzMURERAZCbmS5GAYxREREBkKTd+zVBwxiiIiIDIRx5WE4JoaIiIj0FDMxREREBoLdSURERKSXtHXHXm1hdxIRERHpJWZiiIiIDASnWBMREZFeMq4Qht1JREREpKeYiSEiIjIQnJ1EREREesnYxsSwO4mIiIj0EjMxREREBsK48jAMYoiIiAwGx8QQERGRXuKYGCIiIiI9wEwMERGRgTCuPAyDGCIiIoNhbGNi2J1EREREeomZGCIiIgMhjKxDiUEMERGRgWB3EhEREZEeYCaGiIjIQBjbfWIYxBARERkI4wph2J1EREREeopBDL0Xfs18sHPHWtxKTkRuzh106hSo9HqXLu2wb89GpN67iNycO6hbt+Zr64zf/yNyc+6oPHbv/F5pOU/PytixfQ0epF3G44dXkXBsD9zcyipenzdnMu6nXMSNv06ib9+uSut2794RO3esffMdJ3qNZs18sGP7GiTfOIWc7NtK740SJUpg1swvcTrxFzzM+BPJN05h9apIlCnjXGidX4wbhmO/xeBB+hXc/vsson/8DlWrVlK8XqFCOeRk31b76N6tAwDAzs4WO7avQcaDJJz4PRZer7wnFyyYgdGjP9XgkSBNkENo7KEPGMTQe2FpWRLnz/+BEaO+KvD1346dQPiXM4tcZ49en8DVzUvxqOPVErm5uYjeFqNYplKlCjh0YCeSkq6hdZseqOcdgJmzIvH8eTYAoGOHNujTpwvatf8QE76cgRVRc1G6tB0AoFQpa0yfNh4jRn75FntOVLj898aoUf9Tea1kSQt41auFWbMi4dO4LXr1/hRVq3pg+7bVhdbp19wXy6LWwc+vE9q374sSpqbYE7MRJUtaAAD+/vsu3MrXU3pMnToPT548RWzcAQDAhAkjYGVlBZ/GbXHocAKWLZujqL9Ro/po1LAeFi78ToNHgjRBrsGHPuCYGHovYuMOKC6O6mzYsA3Ay2+IRfXw4SOl5717dcazZ/8gettPirLp08ZjX+yvmBD+b3B0/fpNxf89PSvj0OEEJJ4+j8TT5/HtvKmo6F4eDx48xNcR/8Py5d/j77/vFrlNRMUVF3cAcQW8NzIzn6B9+w+VykaN/p8im1jQuRkU9JHS80GDxuDunfOoX78Ojh79HXK5HKmpaUrLdO7cFtHRMcjKegYA8PSsgq0/7sLVqzewatUGDBrYD8DL7NCSxREY/Nk4yOX68lFnPIztPjHMxJDBCA3tgy1bd+HZs38AABKJBO3btcbVq9exN2YD7t4+h2NHf1JK158//we869eBra0N6terDQsLc1z7KxlNmzREvXq1sGjxKm3tDpFaNjbWkMvlePQosxjrlAIAPMx4pPb1evVqw8urFtas3aQou3D+D7Rs0RRSqRRt2vjjwoXLAICxnw/BocMJOH36/JvvBJGGMIj5f3///TcGDBhQ6DLZ2dnIzMxUeghhXFGvrmrYwAu1a1XH6tX/XoSdnBxgbW2FL8YNQ9zPB9Guw4fYuSsW0Vu/Q3O/xgCAn/cfwsZN23H82B6sWjUfoQNHIyvrGRYvjsCwYRPw2eBgXLp4GIcP7kSNGlW1tXtEAACZTIZZM7/Eli278OTJ0yKtI5FIMG/eFPz22wlc+iNJ7TKhoX1w+fKfOH48UVE2Z+4S5Obm4crl39C5UzsMHjwWlStXxMcf98SsWZFYvDgCV678ho0blqFUKWuN7B+9PWPrTmIQ8/8yMjKwbt26QpeJiIiAjY2N0kPIn7ynFlJhQkP74vyFP3Dy1FlFmYnJy9N7909xWLBwJc6du4Q5c5dgz95f8OmnHyuWmzb9W3jWaIZ69QOwa1csJowfjl/jj+JFbi6+DB8F/xZdsGr1RqxZveB97xaRQokSJbBp4zJIJBIMHxFe5PUWLpyJmjWq4aOPh6l93dzcHH16d8GatZuVyjMznyA4ZDiqVG2MgDY9cPnKVSxZ/DUmhM9A377dULFiedSq5Y9nz/7BV1+NfptdIw0SGvynD4xmTMzu3bsLff369euvrSM8PBxhYWFKZXalPd+qXfT2Spa0QO9enTBl6jyl8vT0DLx48QKXL19VKr9y5SqaNmmktq5q1TzwYd/uaNDoA4T274MjR39HenoGfoz+Cau+mw8rK0s8fZr1zvaFSJ2XAUwUypcvhw8CexU5CxMZOQPt2wWgdUB33LlzT+0y3bt1QMmSFvjhh+hC6woO7oXHjzPx008/Y+uWldi9Ow65ubnYtj0GkyeNLfY+EWmC0QQxXbp0gUQiKbT7RyKRFFqHTCaDTCYr1jr07vXoHgSZzAwbNm5XKn/x4gVOnTqHqlU9lMqrVKmEm7duq61r2ZLZGPfFVGRlPYNUKoWp6cu3iKmpKQBAKpW+gz0gKlh+AFO5sjvafNALGQWMa3lVZOQMdO7UFm0+6Ink5L8LXK5//z6IidmP9PSMApdxcLDHV1+NRsuW3QAAUqmJ4j1hamrK94UO0ZduIE0xmu6kMmXKYPv27ZDL5Wofp0+f1nYTDZqlZUnUrVtTcf+Xiu7lUbduTcX9WuzsbFG3bk3UqP5y3EnVqh6oW7cmnJ0dFXWsWb0AM2dMUKl7QGgf7Nodh4yMhyqvzft2GXr1DMLAAR/Cw8MdQ4f0R8cObRAVpdp1OHDAh0hLz0DMnv0AgGPHTqJli6bwaVQfo0d9gkt/JOHx46IPpiQqCkvLkqhbpwbq1qkBAHB3d0PdOjXg5lYWJUqUwJbNy1G/fh2EhIyAVCqFs7MjnJ0dFUEEAMTGbsaQIf0VzxcunIkP+3ZFcMhwPHnyVLGOubm50rY9PNzh5+eD1Ws2oTDfzJuKyMgVuHs3BQBwLOEU+n3YDZ6elTFwYD8cSzipoaNBb0suhMYe+sBoMjHe3t5ITExE586d1b7+uiwNvZ0G3nUR/8u/6epv5k0BAKz7fisGDhqDoI4fYPWq+YrXN21YBgCYNv0bTJv+LQCgvFtZlSmdVat6oFkzH7Rt10ftdnftisXQYRMw/osRiJw/DUl/XkfP3p/gt2PKF10nJweETxgJP/9/z4+Tp85ifuRy7N71Pe6npWPAgNFvvP9EBfH2rotf9v+oeD5v7hQAwPffb8X0Gd8iKOjlbLpTp/YrrRfQpicOH04AAFSqWAEOpe0Vr302OAQAlN5zADBw0BisX//vtvqH9Mbt2/ewf/+hAtvXpo0/PDzc0T90pKJs6dI18K5fB0eP/ISTp85ixoz5Ba5P9C5JhJF8ch85cgRZWVlo27at2tezsrJw6tQp+Pv7F6veEmaummgekUExYTcrkYqcbPXd2Jr0UYVuGqvrh5vbX7+QlhlNJsbPz6/Q1y0tLYsdwBAREekSffm5AE0xmjExREREZFiMJhNDRERk6PTl/i6awiCGiIjIQBjbFGsGMURERAaCY2KIiIiI9AAzMURERAaCY2KIiIhILxnbmBh2JxEREZFeYiaGiIjIQBjJTfgVGMQQEREZCM5OIiIiItIDzMQQEREZCGMb2MsghoiIyEAY2xRrdicRERGRXmImhoiIyEAY28BeBjFEREQGglOsiYiISC8Z28BejokhIiIivcRMDBERkYEwttlJDGKIiIgMhLEN7GV3EhEREeklZmKIiIgMhLHNTmImhoiIyEDIITT2KK4lS5bA3d0d5ubm8PHxwYkTJwpcduXKlfDz84OdnR3s7OwQEBBQ6PIFYRBDREREb2XLli0ICwvD5MmTcfr0adStWxeBgYG4f/++2uUPHjyIvn374sCBA0hISICbmxs++OAD3Llzp1jblQhjyz1pWAkzV203gUjnmEgk2m4Ckc7Jyb79zrfRolyAxuo6ePuXIi/r4+ODhg0bYvHixQAAuVwONzc3jBgxAhMmTHjt+nl5ebCzs8PixYsRHBxc5O1yTAwREZGBkGswL5GdnY3s7GylMplMBplMplSWk5ODxMREhIeHK8pMTEwQEBCAhISEIm3r2bNnePHiBezt7YvVRnYnERERkYqIiAjY2NgoPSIiIlSWS09PR15eHpydnZXKnZ2dkZKSUqRtjR8/HmXLlkVAQPEySczEEBERGQhNjg8JDw9HWFiYUtmrWRhN+Prrr7F582YcPHgQ5ubmxVqXQQwREZGB0OTN7tR1Hanj4OAAqVSK1NRUpfLU1FS4uLgUuu68efPw9ddf45dffkGdOnWK3UZ2JxERERkIbUyxNjMzg7e3N+Lj4/9th1yO+Ph4+Pr6FrjenDlzMH36dMTGxqJBgwZvtL/MxBAREdFbCQsLQ0hICBo0aIBGjRohMjISWVlZCA0NBQAEBwfD1dVVMaZm9uzZmDRpEjZu3Ah3d3fF2BkrKytYWVkVebsMYoiIiAyEtu6a0rt3b6SlpWHSpElISUmBl5cXYmNjFYN9b926BROTfzt/li1bhpycHPTo0UOpnsmTJ2PKlClF3i7vE/OWeJ8YIlW8TwyRqvdxn5hGZf01VteJu4c0Vte7wjExREREpJfYnURERGQghEYnWes+BjFEREQGwthGiLA7iYiIiPQSMzFEREQGQpM3u9MHDGKIiIgMBLuTiIiIiPQAMzFEREQGgt1JREREpJc4xZqIiIj0kpxjYoiIiIh0HzMxREREBoLdSURERKSX2J1EREREpAeYiSEiIjIQ7E4iIiIivcTuJCIiIiI9wEwMERGRgWB3EhEREekldicRERER6QFmYoiIiAwEu5OIiIhILwkh13YT3isGMURERAZCbmSZGI6JISIiIr3ETAwREZGBEEY2O4lBDBERkYFgdxIRERGRHmAmhoiIyECwO4mIiIj0Eu/YS0RERKQHmIkhIiIyELxjLxEREeklYxsTw+4kIiIi0kvMxBARERkIY7tPDIMYIiIiA2Fs3UkMYoiIiAwEp1gTERER6QFmYoiIiAwEu5OIiIhILxnbwF52JxEREZFeYiaGiIjIQLA7iYiIiPQSZycRERER6QFmYoiIiAwEfwCSiIiI9BK7k4iIiIj0ADMxREREBoKzk4iIiEgvcUwMERER6SVjy8RwTAwRERHpJWZiiIiIDISxZWIYxBARERkI4wph2J1EREREekoijC33RAYpOzsbERERCA8Ph0wm03ZziHQC3xdk6BjEkEHIzMyEjY0NHj9+jFKlSmm7OUQ6ge8LMnTsTiIiIiK9xCCGiIiI9BKDGCIiItJLDGLIIMhkMkyePJmDF4n+g+8LMnQc2EtERER6iZkYIiIi0ksMYoiIiEgvMYghIiIivcQghoiIiPQSgxjSe0uWLIG7uzvMzc3h4+ODEydOaLtJRFp1+PBhBAUFoWzZspBIJNi5c6e2m0T0TjCIIb22ZcsWhIWFYfLkyTh9+jTq1q2LwMBA3L9/X9tNI9KarKws1K1bF0uWLNF2U4jeKU6xJr3m4+ODhg0bYvHixQAAuVwONzc3jBgxAhMmTNBy64i0TyKRYMeOHejSpYu2m0KkcczEkN7KyclBYmIiAgICFGUmJiYICAhAQkKCFltGRETvA4MY0lvp6enIy8uDs7OzUrmzszNSUlK01CoiInpfGMQQERGRXmIQQ3rLwcEBUqkUqampSuWpqalwcXHRUquIiOh9YRBDesvMzAze3t6Ij49XlMnlcsTHx8PX11eLLSMiovehhLYbQPQ2wsLCEBISggYNGqBRo0aIjIxEVlYWQkNDtd00Iq15+vQprl27pnh+48YNnD17Fvb29ihfvrwWW0akWZxiTXpv8eLFmDt3LlJSUuDl5YWFCxfCx8dH280i0pqDBw+iZcuWKuUhISFYu3bt+28Q0TvCIIaIiIj0EsfEEBERkV5iEENERER6iUEMERER6SUGMURERKSXGMQQERGRXmIQQ0RERHqJQQwRERHpJQYxRPRW+vfvjy5duiiet2jRAqNHj37v7Th48CAkEgkePXr03rdNRNrBIIbIQPXv3x8SiQQSiQRmZmaoXLkypk2bhtzc3He63e3bt2P69OlFWpaBBxG9Df52EpEBa9u2LdasWYPs7Gzs3bsXw4YNg6mpKcLDw5WWy8nJgZmZmUa2aW9vr5F6iIheh5kYIgMmk8ng4uKCChUqYMiQIQgICMDu3bsVXUAzZ85E2bJlUa1aNQDA33//jV69esHW1hb29vbo3LkzkpOTFfXl5eUhLCwMtra2KF26NL744gu8+sslr3YnZWdnY/z48XBzc4NMJkPlypWxatUqJCcnK37fx87ODhKJBP379wfw8tfIIyIiULFiRVhYWKBu3bqIjo5W2s7evXtRtWpVWFhYoGXLlkrtJCLjwCCGyIhYWFggJycHABAfH4+kpCTs378fMTExePHiBQIDA2FtbY0jR47gt99+g5WVFdq2batY55tvvsHatWuxevVqHD16FBkZGdixY0eh2wwODsamTZuwcOFCXL58GcuXL4eVlRXc3Nywbds2AEBSUhLu3buHBQsWAAAiIiLw/fffIyoqCpcuXcKYMWPw0Ucf4dChQwBeBlvdunVDUFAQzp49i0GDBmHChAnv6rARka4SRGSQQkJCROfOnYUQQsjlcrF//34hk8nE2LFjRUhIiHB2dhbZ2dmK5devXy+qVasm5HK5oiw7O1tYWFiIuLg4IYQQZcqUEXPmzFG8/uLFC1GuXDnFdoQQwt/fX4waNUoIIURSUpIAIPbv36+2jQcOHBAAxMOHDxVlz58/FyVLlhTHjh1TWnbgwIGib9++QgghwsPDRY0aNZReHz9+vEpdRGTYOCaGyIDFxMTAysoKL168gFwux4cffogpU6Zg2LBhqF27ttI4mHPnzuHatWuwtrZWquP58+f466+/8PjxY9y7dw8+Pj6K10qUKIEGDRqodCnlO3v2LKRSKfz9/Yvc5mvXruHZs2do06aNUnlOTg7q1asHALh8+bJSOwDA19e3yNsgIsPAIIbIgLVs2RLLli2DmZkZypYtixIl/n3LW1paKi379OlTeHt7Y8OGDSr1ODo6vtH2LSwsir3O06dPAQB79uyBq6ur0msymeyN2kFEholBDJEBs7S0ROXKlYu0bP369bFlyxY4OTmhVKlSapcpU6YMfv/9dzRv3hwAkJubi8TERNSvX1/t8rVr14ZcLsehQ4cQEBCg8np+JigvL09RVqNGDchkMty6davADE716tWxe/dupbLjx4+/fieJyKBwYC8RAQD69esHBwcHdO7cGUeOHMGNGzdw8OBBjBw5Erdv3wYAjBo1Cl9//TV27tyJK1euYOjQoYXe48Xd3R0hISEYMGAAdu7cqahz69atAIAKFSpAIpEgJiYGaWlpePr0KaytrTF27FiMGTMG69atw19//YXTp09j0aJFWLduHQDgs88+w9WrVzFu3DgkJSVh48aNWLt27bs+RESkYxjEEBEAoGTJkjh8+DDKly+Pbt26oXr16hg4cCCeP3+uyMx8/vnn+PjjjxESEgJfX19YW1uja9euhda7bNky9OjRA0OHDoWnpyc++eQTZGVlAQBcXV0xdepUTJgwAc7Ozhg+fDgAYPr06Zg4cSIiIiJQvXp1tG3bFnv27EHFihUBAOXLl8e2bduwc+dO1K1bF1FRUZg1a9Y7PDpEpIskoqAReUREREQ6jJkYIiIi0ksMYoiIiEgvMYghIiIivcQghoiIiPQSgxgiIiLSSwxiiIiISC8xiCEiIiK9xCCGiIiI9BKDGCIiItJLDGKIiIhILzGIISIiIr3EIIaIiIj00v8B+t+POkAbDfgAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_pred_fair = test.copy()\n",
        "test_pred_fair.labels = fpred\n",
        "f_metric = ClassificationMetric(test, test_pred_fair, unprivileged_groups=unprivileged, privileged_groups=privileged)\n",
        "\n",
        "print(\"Fairness metrics of fair model on unweighted samples:-\")\n",
        "print(\"Equal opportunity difference: \", round(f_metric.equal_opportunity_difference(), 4))\n",
        "print(\"Disparate Impact: \", round(f_metric.disparate_impact(), 4))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gUJDTkYmH73n",
        "outputId": "93823796-4add-408e-d168-da32c12f2abf"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fairness metrics of fair model on unweighted samples:-\n",
            "Equal opportunity difference:  0.0155\n",
            "Disparate Impact:  0.8343\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(figsize=(13, 5))\n",
        "\n",
        "plt.subplot(1, 2, 1)\n",
        "acc_verd = view_proportions(y_test, acc_pred_y)\n",
        "fdf = pd.DataFrame({'sex': [k[1] for k in X_test], 'classification': acc_verd})\n",
        "sns.countplot(data=fdf, x='classification', hue='sex', stat='percent')\n",
        "plt.title(\"Accurate model\")\n",
        "\n",
        "plt.subplot(1, 2, 2)\n",
        "fair_verd = view_proportions(y_test, fpred)\n",
        "fdf = pd.DataFrame({'sex': [k[1] for k in X_test], 'classification': fair_verd})\n",
        "sns.countplot(data=fdf, x='classification', hue='sex', stat='percent')\n",
        "plt.title(\"Fair model\")\n",
        "\n",
        "\n",
        "plt.suptitle(\"Counts of errors w.r.t. protected class\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 515
        },
        "id": "JbrePNN4fYrD",
        "outputId": "4c2c8d96-6385-4605-ae13-aba13d9c7606"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1300x500 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Reweighing"
      ],
      "metadata": {
        "id": "wQffiLFVmX1j"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "rw = Reweighing(unprivileged_groups=unprivileged, privileged_groups=privileged)\n",
        "\n",
        "train_rw = rw.fit_transform(train)\n",
        "\n",
        "X_train = train_rw.features\n",
        "y_train = train_rw.labels.ravel()\n",
        "weights = np.array(train_rw.instance_weights)"
      ],
      "metadata": {
        "id": "qMkWdIHlSnN_"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "acc_rw_model = accurate_model(X_train, y_train, weights)"
      ],
      "metadata": {
        "id": "E--M2TJInj4H"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "acc_pred_y = acc_rw_model.predict(X_test)\n",
        "print(classification_report(y_test, acc_pred_y))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bZKoVifInqwU",
        "outputId": "d489f019-dace-445e-acb8-5fa12b442b5d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "         0.0       0.89      0.71      0.79     11132\n",
            "         1.0       0.44      0.72      0.55      3521\n",
            "\n",
            "    accuracy                           0.71     14653\n",
            "   macro avg       0.67      0.72      0.67     14653\n",
            "weighted avg       0.78      0.71      0.73     14653\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "conf_matrix = confusion_matrix(y_test, acc_pred_y)\n",
        "sns.heatmap(conf_matrix/np.sum(conf_matrix), annot=True, fmt='.2%')\n",
        "plt.xlabel(\"Predicted\")\n",
        "plt.ylabel(\"Actual\")\n",
        "plt.title(\"Visualising accurate model performance with reweighed sample weights\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "07M5gM6kntX2",
        "outputId": "87b96fb5-9332-4f19-c114-5cbd3a1f09f8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAm8AAAHHCAYAAADpkb+WAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABbKklEQVR4nO3dd1gUV9sG8HtZZelNqg0QFcUYVBDEiJWI0dh7xV5jwxJNXjXGKJaYECtqbLHXaKKJjVgwsYK9xd5pNgQVlD3fH35sXHeBRWeE1fvntdflnp05c2Z2Z/bhOWfOKoQQAkRERERkFEzyuwFEREREZDgGb0RERERGhMEbERERkRFh8EZERERkRBi8ERERERkRBm9ERERERoTBGxEREZERYfBGREREZEQYvBEREREZEdmCtyVLlkChUODatWtybeKN21G7dm3Url37reshysnbfGa++eYbKBQK6Rv1llJTU9GzZ0+4urpCoVBgyJAh+d2kD86ePXugUCiwZ88eg5ddv369/A3LJ29yPX913Y8++kjaBmVDoVDgm2++eSfbetcK6vUqO2/T3qx1k5OTJW5V3hgcvDVp0gQWFhZ4/Phxtst07NgRpqamuHfvniSNow/DypUrERkZmd/NIANMmjQJS5YsQb9+/bBs2TJ07tw5v5tE4DlEVBBNmjQJmzZtkqVug4O3jh074unTp/j111/1vv7kyRNs3rwZDRo0QJEiRdC5c2c8ffoU7u7ukjVWKjt27MCOHTvytE5B3h9jxy8e4/HXX3+hWrVqGDduHDp16gQ/P7/8btIHp2bNmnj69Clq1qypKfuQz6E3uZ7Th+1///sfnj59Kvt2CkTw1qRJE1hbW2PlypV6X9+8eTPS0tLQsWNHAIBSqYSZmVmBTKWamprC1NQ0T+sU5P0paJ48eZLfTSAJqdVqPHv2DACQmJgIOzs7yep+8eIFMjIyJKvvQ2BiYgIzMzOYmOTPkOW0tLR82W523uR6Th+2QoUKwczMLL+b8VYMPvvNzc3RokULREdHIzExUef1lStXwtraGk2aNAGgf7zP0aNHERoaCkdHR5ibm8PT0xPdu3fXvJ7dWI5r165BoVBgyZIlmrKTJ0+ia9euKFWqFMzMzODq6oru3bsb1GWrb4zEzJkzUaFCBVhYWMDe3h7+/v5agaq+/fHw8MDnn3+O/fv3IyAgAGZmZihVqhR++eUXnW2ePHkStWrVgrm5OYoXL47vvvsOixcvNmhMVF729fbt2+jRoweKFi0KlUoFT09P9OvXT+sL8uHDhxg6dCg8PDygUqlQvHhxdOnSRdOHn91YLX3vT9aYkdjYWNSsWRMWFhb46quvALwM6Bs1aqRpi5eXFyZMmIDMzEyt9bdu3Yrr169DoVBAoVDAw8ND83p6ejrGjRuH0qVLQ6VSoUSJEhg5ciTS09NzPGavti3r2FtYWKB06dKa8T979+5FYGAgzM3N4e3tjV27dunUcezYMXz22WewsbGBlZUV6tWrh4MHD+osd+bMGdStW1fr/VWr1Xrb9eeffyI4OBiWlpawtrZGo0aNcObMmVz3J6d9jI2NRfXq1TXnVVRUlM6yhh5LhUKBL774AitWrECFChWgUqmwbds2KBQKXL16FVu3btW8V1mfkcTERPTo0QMuLi4wMzODr68vli5dqlVv1nn8/fffIzIyEl5eXlCpVDh79qxmHMm///6LTp06wdbWFk5OThgzZgyEELh58yaaNm0KGxsbuLq6Yvr06Vp1Z2RkYOzYsfDz84OtrS0sLS0RHByM3bt3Z9uG+fPna9pQtWpVHDlyROeYnT9/Hm3atIGTk5Pmc/L1119rLXP79m10794dLi4uUKlUqFChAhYtWpTre9eiRQtUqVJFq6xx48ZQKBT47bffNGWHDh2CQqHAn3/+CUD3PMztHAJeBuATJ05E8eLFYWZmhnr16uHSpUu5tjHrfTl79iw6dOgAe3t71KhRQ/P68uXL4efnB3Nzczg4OKBdu3a4efOm5vUZM2ZAqVTi4cOHmrLp06dDoVAgPDxcU5aZmQlra2t8+eWXWm2OjIxEhQoVYGZmBhcXF/Tp0wcPHjzQaqO+6/n169fRpEkTWFpawtnZGUOHDsX27duzHSt49uxZ1KlTBxYWFihWrBimTp2qs4yh5096ejqGDh0KJycnzXfirVu3cjzOr8rtu+j69evo378/vL29YW5ujiJFiqB169Y61+us6/j+/fsxaNAgODk5wc7ODn369EFGRgYePnyILl26wN7eHvb29hg5ciSEEJr1Xz1XfvzxR7i7u8Pc3By1atXC6dOnDdqX3D4f+pw8eVLnHIiNjYVCodA5Xz777DMEBgZqlRlyfdU35u3p06cYNGgQHB0dNe/b7du3sx2r+PDhQ3Tt2hV2dnawtbVFt27dtBIXCoUCaWlpWLp0qea87Nq1KwDg8ePHGDJkiOY72NnZGZ9++ini4uJyPDavKmTwknjZdbp06VKsXbsWX3zxhab8/v372L59O9q3bw9zc3O96yYmJqJ+/fpwcnLCqFGjYGdnh2vXrmHjxo15aYLGzp07ceXKFXTr1g2urq44c+YM5s+fjzNnzuDgwYN5ypAtWLAAgwYNQqtWrTB48GA8e/YMJ0+exKFDh9ChQ4cc17106RJatWqFHj16ICwsDIsWLULXrl3h5+eHChUqAHh5ca9Tpw4UCgVGjx4NS0tL/Pzzz1CpVJLu6507dxAQEICHDx+id+/eKFeuHG7fvo3169fjyZMnMDU1RWpqKoKDg3Hu3Dl0794dVapUQXJyMn777TfcunULjo6OBh+3LPfu3cNnn32Gdu3aoVOnTnBxcQHw8uJhZWWF8PBwWFlZ4a+//sLYsWORkpKCadOmAQC+/vprPHr0CLdu3cKPP/4IALCysgLw8uLdpEkT7N+/H71790b58uVx6tQp/Pjjj/j3338NSkc/ePAAn3/+Odq1a4fWrVtj7ty5aNeuHVasWIEhQ4agb9++6NChA6ZNm4ZWrVrh5s2bsLa2BvAyIAsODoaNjQ1GjhyJwoULY968eahdu7Ym8AOA+Ph41KlTBy9evMCoUaNgaWmJ+fPn6z0Xli1bhrCwMISGhmLKlCl48uQJ5s6dixo1auDYsWM6X7qGePDgARo2bIg2bdqgffv2WLt2Lfr16wdTU1PNH0d5PZZ//fWX5jx3dHSEm5sbli1bhqFDh6J48eIYNmwYAMDJyQlPnz5F7dq1cenSJXzxxRfw9PTEunXr0LVrVzx8+BCDBw/Wqnvx4sV49uwZevfuDZVKBQcHB81rbdu2Rfny5TF58mRs3boV3333HRwcHDBv3jzUrVsXU6ZMwYoVKzB8+HBUrVpV03WYkpKCn3/+Ge3bt0evXr3w+PFjLFy4EKGhoTh8+DAqVaqk1YaVK1fi8ePH6NOnDxQKBaZOnYoWLVrgypUrKFy4MICXXyLBwcEoXLgwevfuDQ8PD1y+fBm///47Jk6cCABISEhAtWrVNAGvk5MT/vzzT/To0QMpKSk53tARHByMzZs3IyUlBTY2NhBC4O+//4aJiQliYmI0fwjHxMTAxMQEn3zyid56cjqHskyePBkmJiYYPnw4Hj16hKlTp6Jjx444dOhQtu17VevWrVGmTBlMmjRJ8wU/ceJEjBkzBm3atEHPnj2RlJSEmTNnombNmjh27Bjs7OwQHBwMtVqN/fv34/PPP9fan5iYGE39x44dQ2pqqlZXcJ8+fbBkyRJ069YNgwYNwtWrVzFr1iwcO3YMf//9t+Z9el1aWhrq1q2Lu3fvYvDgwXB1dcXKlSt1AvksDx48QIMGDdCiRQu0adMG69evx5dffomKFSvis88+A5C386dnz55Yvnw5OnTogOrVq+Ovv/5Co0aNDDrOhnwXHTlyBP/88w/atWuH4sWL49q1a5g7dy5q166Ns2fPwsLCQqvOgQMHwtXVFePHj8fBgwcxf/582NnZ4Z9//kHJkiUxadIk/PHHH5g2bRo++ugjdOnSRWv9X375BY8fP8aAAQPw7Nkz/PTTT6hbty5OnTqludbrY8jnQ5+PPvoIdnZ22Ldvn845cOLECc35olar8c8//6B3796add/m+tq1a1esXbsWnTt3RrVq1bB3794c37c2bdrA09MTERERiIuLw88//wxnZ2dMmTJF05aePXsiICBA00YvLy8AQN++fbF+/Xp88cUX8PHxwb1797B//36cO3dOJ0DNlsiDFy9eCDc3NxEUFKRVHhUVJQCI7du3a8oWL14sAIirV68KIYT49ddfBQBx5MiRbOvfvXu3ACB2796tVX716lUBQCxevFhT9uTJE531V61aJQCIffv2ZdsOIYSoVauWqFWrluZ506ZNRYUKFXLYc/31uLu762wvMTFRqFQqMWzYME3ZwIEDhUKhEMeOHdOU3bt3Tzg4OOjUqY+h+9qlSxdhYmKi9xir1WohhBBjx44VAMTGjRuzXUbfvgqh//2pVauWACCioqIManefPn2EhYWFePbsmaasUaNGwt3dXWfZZcuWCRMTExETE6NVnvV5+/vvv3XWeVVW21auXKkpO3/+vAAgTExMxMGDBzXl27dv1/mMNWvWTJiamorLly9ryu7cuSOsra1FzZo1NWVDhgwRAMShQ4c0ZYmJicLW1lbrOD5+/FjY2dmJXr16abUzPj5e2NraapWPGzdOGHJ6Zu3j9OnTNWXp6emiUqVKwtnZWWRkZAgh8nYss47PmTNndLbn7u4uGjVqpFUWGRkpAIjly5dryjIyMkRQUJCwsrISKSkpQoj/zmMbGxuRmJioVUfW/vbu3VtT9uLFC1G8eHGhUCjE5MmTNeUPHjwQ5ubmIiwsTGvZ9PR0rTofPHggXFxcRPfu3TVlWW0oUqSIuH//vqZ88+bNAoD4/fffNWU1a9YU1tbW4vr161r1Zp0nQgjRo0cP4ebmJpKTk7WWadeunbC1tdV7DmQ5cuSIACD++OMPIYQQJ0+eFABE69atRWBgoGa5Jk2aiMqVK2ue6zsPszuHspYtX7681vH56aefBABx6tSpbNsnxH/vS/v27bXKr127JpRKpZg4caJW+alTp0ShQoU05ZmZmcLGxkaMHDlSCPHy2BUpUkS0bt1aKJVK8fjxYyGEED/88IMwMTERDx48EEIIERMTIwCIFStWaNW/bds2nfLXr+fTp08XAMSmTZs0ZU+fPhXlypXL9vr1yy+/aMrS09OFq6uraNmypabM0PPn+PHjAoDo37+/1nIdOnQQAMS4ceNETgz5LtL3mTpw4IDOfmRdx0NDQ7U+s0FBQUKhUIi+fftqyrLOtVePY9a5Ym5uLm7duqUpP3TokAAghg4dqil7/Xpl6OcjO40aNRIBAQGa5y1atBAtWrQQSqVS/Pnnn0IIIeLi4gQAsXnzZiHE211fY2NjBQAxZMgQrXW7du2q875lrfvqdUUIIZo3by6KFCmiVWZpaal1ncpia2srBgwYkOMxyE2eBk0olUq0a9cOBw4c0ErRrly5Ei4uLqhXr16262ZF2Vu2bMHz58/zslm9Xs1qPHv2DMnJyahWrRoA5Cn1mNW2W7du6e02yY2Pjw+Cg4M1z52cnODt7Y0rV65oyrZt24agoCCtv/4dHBw04wNzY8i+qtVqbNq0CY0bN4a/v79OHVnZuQ0bNsDX1xfNmzfPdpm8UqlU6NatW47tfvz4MZKTkxEcHIwnT57g/Pnzuda7bt06lC9fHuXKlUNycrLmUbduXQDI9i/pV1lZWaFdu3aa597e3rCzs0P58uW10u1Z/8963zIzM7Fjxw40a9YMpUqV0izn5uaGDh06YP/+/UhJSQEA/PHHH6hWrRoCAgI0yzk5Oem8vzt37sTDhw/Rvn17rf1RKpUIDAw0aH/0KVSoEPr06aN5bmpqij59+iAxMRGxsbEA8n4sa9WqBR8fH4O2/8cff8DV1RXt27fXlBUuXBiDBg1Camoq9u7dq7V8y5Yt4eTkpLeunj17av6vVCrh7+8PIQR69OihKbezs9M5x5RKpWbck1qtxv379/HixQv4+/vrvR60bdsW9vb2mudZ53BWnUlJSdi3bx+6d++OkiVLaq2bdZ4IIbBhwwY0btwYQgit4xoaGopHjx7leC2qXLkyrKyssG/fPgAvswtZQxji4uLw5MkTCCGwf/9+rWvMm+jWrZvWuLDX9zc3ffv21Xq+ceNGqNVqtGnTRmu/XV1dUaZMGc3nycTEBNWrV9fs47lz53Dv3j2MGjUKQggcOHBAs+9ZGRfg5efV1tYWn376qVb9fn5+sLKyyvFc2bZtG4oVK6bJ2gCAmZkZevXqpXd5KysrdOrUSfPc1NQUAQEBWsfG0PPnjz/+AAAMGjRIaxuGTqljyHfRq9fV58+f4969eyhdujTs7Oz0ft569OihdW0PDAzUOaeyzjV9n4dmzZqhWLFimucBAQEIDAzU7Ks+hn4+shMcHIy4uDjN+Mr9+/ejYcOGqFSpkiZjGxMTA4VCoenGf5vr67Zt2wAA/fv31yofOHBgtuu8fk4EBwfj3r17mu+FnNjZ2eHQoUO4c+dOrstmJ88jXrO+kLL64G/duoWYmBi0a9cOSqUy2/Vq1aqFli1bYvz48XB0dETTpk2xePFig8Yu6XP//n0MHjwYLi4uMDc3h5OTEzw9PQEAjx49ylNdX375JaysrBAQEIAyZcpgwIAB+Pvvvw1a9/ULOwDY29trjcu4fv06SpcurbOcvjJ9DNnXpKQkpKSk5Dpn0eXLlyWf16hYsWJ6BwyfOXMGzZs3h62tLWxsbODk5KS5SBryHl28eBFnzpyBk5OT1qNs2bIAoHfs5euKFy+uE5Ta2tqiRIkSOmUANO9bUlISnjx5Am9vb506y5cvD7VarRm7cf36dZQpU0ZnudfXvXjxIgCgbt26Ovu0Y8cOg/ZHn6JFi8LS0lKrLOsYZf2RlddjmfX5MkTW/r8+gL58+fKa1w2t+/XzydbWFmZmZjrd+ba2tjpjn5YuXYqPP/4YZmZmKFKkCJycnLB161a9n7XXt5MVyGXVmfUlltO5kpSUhIcPH2L+/Pk6xzXrj5mc3lOlUomgoCCtL6Pg4GDUqFEDmZmZOHjwIM6ePYv79++/dfCW2/7m5vX37OLFixBCoEyZMjr7fu7cOa39Dg4ORmxsLJ4+fYqYmBi4ubmhSpUq8PX11ez76wHqxYsX8ejRIzg7O+vUn5qamuNxvX79Ory8vHTO++yut/quEa9fww09f65fvw4TExNN91gWfdcRfQz5Lnr69CnGjh2LEiVKQKVSwdHREU5OTnj48KFBn/Wsa52+a6C+z4O+a1vZsmVzHKudl8+HPsHBwXjx4gUOHDiACxcuIDExEcHBwahZs6bW+eLj46MZdvE219es9+31z3lO39Fvc05NnToVp0+fRokSJRAQEIBvvvnG4D+ksuRpzBsA+Pn5oVy5cli1ahW++uorrFq1CkKIXLNIWRNFHjx4EL///ju2b9+O7t27Y/r06Th48CCsrKyyzfy8OsA9S5s2bfDPP/9gxIgRqFSpEqysrKBWq9GgQYNsB4pnp3z58rhw4QK2bNmCbdu2YcOGDZgzZw7Gjh2L8ePH57hudgGreGXg59uScl8NkZf3AYDesV0PHz5ErVq1YGNjg2+//RZeXl4wMzNDXFwcvvzyS4ParVarUbFiRfzwww96X3/94qNPdu/Pu3jfXpe1z8uWLYOrq6vO64UK5fl0zNO283Issxu7KoWc6tb3vhjyXi1fvhxdu3ZFs2bNMGLECDg7O0OpVCIiIgKXL19+ozpzk/V+durUCWFhYXqX+fjjj3Oso0aNGpg4cSKePXuGmJgYfP3117Czs8NHH32EmJgYzZiitw3e3nZ/X3/P1Gq15iYKfXW/OuauRo0aeP78OQ4cOKAJUIGX+xQTE4Pz588jKSlJax/VajWcnZ2xYsUKve3JLnP7Jgw5NlJciwxhyHfRwIEDsXjxYgwZMgRBQUGwtbWFQqFAu3bt9F5X83INlOr6l5fPhz7+/v4wMzPDvn37ULJkSTg7O6Ns2bIIDg7GnDlzkJ6ejpiYGK0epHd9fX2bc6pNmzYIDg7Gr7/+ih07dmDatGmYMmUKNm7cqBlnmZs32puOHTtizJgxOHnyJFauXIkyZcqgatWqBq1brVo1VKtWDRMnTsTKlSvRsWNHrF69Gj179tRErq/emQTo/uX+4MEDREdHY/z48Rg7dqymPCvyfhOWlpZo27Yt2rZti4yMDLRo0QITJ07E6NGj3/qWYnd3d713dhlyt5eh++rk5AQbG5tc7wLy8vLKdRlD34ec7NmzB/fu3cPGjRu1BiFfvXpVZ9nsgkUvLy+cOHEC9erVe+dTtDg5OcHCwgIXLlzQee38+fMwMTHRXLDd3d31fvZeXzfrr3FnZ2eEhIRI1tY7d+4gLS1NK/v277//AoBmgK6cx9Ld3R0nT56EWq3Wyr5ldY2/i7kR169fj1KlSmHjxo1a+zdu3Lg3qi+rqzyncyXrbsLMzMw3fj+Dg4ORkZGBVatW4fbt25oAJivD4OLigrJly+Y4MBx48yEPb8rLywtCCHh6emqyT9kJCAiAqakpYmJiEBMTgxEjRgB4uY8LFixAdHS05vmr9e/atQuffPJJnv+QcHd3x9mzZyGE0Douhlxvs2Po+ePu7g61Wo3Lly9rZdv0XUeyk9t30fr16xEWFqZ1x/WzZ890rtdS0Xdt+/fff3Mc/J+Xz4c+WV3XMTExKFmypFbAn56ejhUrViAhIUHnMwO82fU16327evWqVqbxbT4zQM7npZubG/r374/+/fsjMTERVapUwcSJEw0O3t5ooqCsLNvYsWNx/Phxg8ZuPXjwQCcizRoDltV16u7uDqVSqRkfkWXOnDlaz7Mi3tfre9NJKl+fcsPU1BQ+Pj4QQkgyPi80NBQHDhzA8ePHNWX379/P9q/KVxm6ryYmJmjWrBl+//13HD16VKeerPVbtmyJEydO6J1sOWuZrJPg1fchMzMT8+fPz7W9ObU7IyND570EXl6s9KX727Rpg9u3b2PBggU6rz19+lTW+aaUSiXq16+PzZs3a3UPJCQkYOXKlahRowZsbGwAAA0bNsTBgwdx+PBhzXJJSUk6729oaChsbGwwadIkvZ+rpKSkN2rrixcvMG/ePM3zjIwMzJs3D05OTppJdOU8lg0bNkR8fDzWrFmj1aaZM2fCysoKtWrVeuO6DaXv83bo0CHNmKq8cnJyQs2aNbFo0SLcuHFD67WsbSiVSrRs2RIbNmzQG+QZ8n4GBgaicOHCmDJlChwcHDR3qAcHB+PgwYPYu3evQVm37M4hubRo0QJKpRLjx4/XuTYJIbSuqWZmZqhatSpWrVqFGzduaH0RP336FDNmzICXlxfc3Nw067Rp0waZmZmYMGGCzrZfvHiRY6ASGhqK27dva0018ezZM72ffUMZev5kffHOmDFDaxlDv5sM+S5SKpU6x3zmzJnZ9oy8rU2bNuH27dua54cPH8ahQ4dyDDLy8vnITnBwMA4dOoTdu3drPjOOjo4oX7685o7OV8+Nt7m+hoaGAtCNNWbOnJlrO3NiaWmp81nNzMzUOVednZ1RtGjRPA0je6PMm6enJ6pXr47NmzcDgEHB29KlSzFnzhw0b94cXl5eePz4MRYsWAAbGxs0bNgQwMs+99atW2PmzJlQKBTw8vLCli1bdPqqbWxsULNmTUydOhXPnz9HsWLFsGPHDr1ZHUPUr18frq6u+OSTT+Di4oJz585h1qxZaNSokWbaiLcxcuRILF++HJ9++ikGDhyomSqkZMmSuH//fo7ReV72ddKkSdixYwdq1aqluZ397t27WLduHfbv3w87OzuMGDEC69evR+vWrdG9e3f4+fnh/v37+O233xAVFQVfX19UqFAB1apVw+jRo3H//n04ODhg9erVePHihcH7XL16ddjb2yMsLAyDBg2CQqHAsmXL9KaU/fz8sGbNGoSHh6Nq1aqwsrJC48aN0blzZ6xduxZ9+/bF7t278cknnyAzMxPnz5/H2rVrsX37dr03Z0jlu+++w86dO1GjRg30798fhQoVwrx585Cenq41D9TIkSOxbNkyNGjQAIMHD9ZMFZKVkcpiY2ODuXPnonPnzqhSpQratWsHJycn3LhxA1u3bsUnn3yCWbNm5bmdRYsWxZQpU3Dt2jWULVsWa9aswfHjxzF//nzNdApyHsvevXtj3rx56Nq1K2JjY+Hh4YH169fj77//RmRkpCTnUG4+//xzbNy4Ec2bN0ejRo1w9epVREVFwcfHB6mpqW9U54wZM1CjRg1UqVIFvXv3hqenJ65du4atW7dq/hCbPHkydu/ejcDAQPTq1Qs+Pj64f/8+4uLisGvXLty/fz/HbVhYWMDPzw8HDx7UzPEGvMxCpaWlIS0tzaDgLbtzSC5eXl747rvvMHr0aFy7dg3NmjWDtbU1rl69il9//RW9e/fG8OHDNcsHBwdj8uTJsLW1RcWKFQG8/MLy9vbGhQsXNPNfZalVqxb69OmDiIgIHD9+HPXr10fhwoVx8eJFrFu3Dj/99BNatWqlt219+vTBrFmz0L59ewwePBhubm5YsWKFpgflTbKUhp4/lSpVQvv27TFnzhw8evQI1atXR3R0tMEZHEO+iz7//HMsW7YMtra28PHxwYEDB7Br1y4UKVIkz/tliNKlS6NGjRro168f0tPTERkZiSJFimDkyJHZrpPXz4c+wcHBmDhxIm7evKl1DtSsWRPz5s2Dh4cHihcvril/m+urn58fWrZsicjISNy7d08zVUhWD8abZrb9/Pywa9cu/PDDDyhatCg8PT3h7e2N4sWLo1WrVvD19YWVlRV27dqFI0eO6MxfmaM3vU119uzZAoDW7byven26ibi4ONG+fXtRsmRJoVKphLOzs/j888/F0aNHtdZLSkoSLVu2FBYWFsLe3l706dNHnD59Wmcah1u3bonmzZsLOzs7YWtrK1q3bi3u3Lmjc1uvIVOFzJs3T9SsWVMUKVJEqFQq4eXlJUaMGCEePXqUYz36pk3QV78QQhw7dkwEBwcLlUolihcvLiIiIsSMGTMEABEfH6//IOdxX4UQ4vr166JLly7CyclJqFQqUapUKTFgwACtaQLu3bsnvvjiC1GsWDFhamoqihcvLsLCwrSmO7h8+bIICQkRKpVKuLi4iK+++krs3LlT76322d3a/vfff4tq1aoJc3NzUbRoUTFy5EjNlByv1pGamio6dOgg7OzsBACtKQ8yMjLElClTRIUKFYRKpRL29vbCz89PjB8/Xuv90Se7tmX3vgHQuX07Li5OhIaGCisrK2FhYSHq1Kkj/vnnH511T548KWrVqiXMzMxEsWLFxIQJE8TChQuznXIlNDRU2NraCjMzM+Hl5SW6du2qdS7kZaqQChUqiKNHj4qgoCBhZmYm3N3dxaxZs3SWNfRY6jsOWbI7dgkJCaJbt27C0dFRmJqaiooVK2qdr0L8N/XAtGnTdNbP2t+kpCSt8rCwMGFpaZntfmdRq9Vi0qRJwt3dXahUKlG5cmWxZcsWERYWpvV5yqkN+s6n06dPa849MzMz4e3tLcaMGaOz7wMGDBAlSpQQhQsXFq6urqJevXpi/vz5OtvQZ8SIEQKAmDJlilZ56dKlBQCtqWqE0D9VSHbnUNay69at06pD3/RL+mT3vmTZsGGDqFGjhrC0tBSWlpaiXLlyYsCAAeLChQtay23dulUAEJ999plWec+ePQUAsXDhQr31z58/X/j5+Qlzc3NhbW0tKlasKEaOHCnu3LmjWUbf9fbKlSuiUaNGwtzcXDg5OYlhw4aJDRs2CABaUwRld414/XMjhOHnz9OnT8WgQYNEkSJFhKWlpWjcuLG4efOmQVOFGPJd9ODBA825ZmVlJUJDQ8X58+eFu7u71rQUWd9Zr08dZei59uq5Mn36dFGiRAmhUqlEcHCwOHHihN46X2fo50OflJQUoVQqhbW1tXjx4oWmfPny5QKA6Ny5s9713vT6mpaWJgYMGCAcHByElZWVaNasmbhw4YIAoDVVUXbHT1+McP78eVGzZk1hbm4uAIiwsDCRnp4uRowYIXx9fYW1tbWwtLQUvr6+Ys6cObkek1cphJBxhDblaMiQIZg3bx5SU1NzvFOXKDu1a9dGcnKywTOeE32oIiMjMXToUNy6dUtr6gvS79q1a/D09MS0adNyzZK9r44fP47KlStj+fLlBk/t9a7kz4/jfYBe/xHce/fuYdmyZahRowYDNyIiCb1+vX327BnmzZuHMmXKMHAjvfT9UH1kZCRMTEy0bowoKOSbm4C0BAUFoXbt2ihfvjwSEhKwcOFCpKSkYMyYMfndNCKi90qLFi1QsmRJVKpUCY8ePcLy5ctx/vx5g24Sow/T1KlTERsbizp16qBQoUL4888/8eeff6J3796STQUjJQZv70jDhg2xfv16zJ8/X/MDuwsXLiyQET0RkTELDQ3Fzz//jBUrViAzMxM+Pj5YvXo12rZtm99NowKqevXq2LlzJyZMmIDU1FSULFkS33zzDb7++uv8bppeHPNGREREZEQ45o2IiIjIiDB4IyIiIjIiDN6IiIiIjAhvWHgPPE++kt9NICqQJvjzbm6i1317Tf67bqX8XirsWEqyut4XzLwRERERGRFm3oiIiEha6sz8bsF7jcEbERERSUuo87sF7zUGb0RERCQtNYM3OXHMGxEREZERYeaNiIiIJCXYbSorBm9EREQkLXabyordpkRERERGhJk3IiIikha7TWXF4I2IiIikxXneZMVuUyIiIiIjwswbERERSYvdprJi8EZERETS4t2msmK3KREREZERYeaNiIiIJMVJeuXF4I2IiIikxW5TWTF4IyIiImkx8yYrjnkjIiIiMiLMvBEREZG0OEmvrBi8ERERkbTYbSordpsSERERGRFm3oiIiEhavNtUVgzeiIiISFrsNpUVu02JiIiIjAgzb0RERCQtdpvKisEbERERSUoIThUiJ3abEhERERkRBm9EREQkLaGW7pFHs2fPhoeHB8zMzBAYGIjDhw9nu+ySJUugUCi0HmZmZm+z5+8Eu02JiIhIWvk05m3NmjUIDw9HVFQUAgMDERkZidDQUFy4cAHOzs5617GxscGFCxc0zxUKxbtq7htj5o2IiIiklU+Ztx9++AG9evVCt27d4OPjg6ioKFhYWGDRokXZrqNQKODq6qp5uLi4vO3ey47BGxERERVY6enpSElJ0Xqkp6frLJeRkYHY2FiEhIRoykxMTBASEoIDBw5kW39qairc3d1RokQJNG3aFGfOnJFlP6TE4I2IiIikpc6U7BEREQFbW1utR0REhM4mk5OTkZmZqZM5c3FxQXx8vN5ment7Y9GiRdi8eTOWL18OtVqN6tWr49atW7IcFqlwzBsRERFJS8JfWBg9ejTCw8O1ylQqlSR1BwUFISgoSPO8evXqKF++PObNm4cJEyZIsg05MHgjIiKiAkulUhkUrDk6OkKpVCIhIUGrPCEhAa6urgZtq3DhwqhcuTIuXbr0Rm19V9htSkRERNJSq6V7GMjU1BR+fn6Ijo5+pRlqREdHa2XXcpKZmYlTp07Bzc0tz7v8LjHzRkRERNLKpx+mDw8PR1hYGPz9/REQEIDIyEikpaWhW7duAIAuXbqgWLFimjFz3377LapVq4bSpUvj4cOHmDZtGq5fv46ePXvmS/sNxeCNiIiI3gtt27ZFUlISxo4di/j4eFSqVAnbtm3T3MRw48YNmJj81+n44MED9OrVC/Hx8bC3t4efnx/++ecf+Pj45NcuGEQhhBD53Qh6O8+Tr+R3E4gKpAn+Y/K7CUQFzrfXVsi+jWd/S7cNs086SlbX+4KZNyIiIpJWPv3CwoeCNywQERERGRFm3oiIiEhSQmTmdxPeawzeiIiISFrsNpUVgzciIiKSVj5NFfKh4Jg3IiIiIiPCzBsRERFJi92msmLwRkRERNJit6ms2G1KREREZESYeSMiIiJpsdtUVgzeiIiISFrsNpUVu02JiIiIjAgzb0RERCQtdpvKisEbERERSYvBm6zYbUpERERkRJh5IyIiImnxhgVZMXgjIiIiabHbVFYM3oiIiEhazLzJimPeiIiIiIwIM29EREQkLXabyorBGxEREUmL3aayYrcpERERkRFh5o2IiIikxW5TWTF4IyIiImkxeJMVu02JiIiIjAgzb0RERCQtIfK7Be81Bm9EREQkLXabyordpkRERERGhJk3IiIikhYzb7Ji8EZERETS4iS9smLwRkRERNJi5k1WHPNGREREZESYeSMiIiJpcaoQWTF4IyIiImmx21RW7DYlIiIiMiLMvBEREZG0mHmTFYM3IiIikhanCpEVu02JiIiIjAgzb0RERCQpoebdpnJi8EZERETS4pg3WbHblIiIiMiIMPNGRERE0uINC7Ji8EZERETS4pg3WTF4IyIiImlxzJusOOaNiIiIyIgw80ZERETSYuZNVgzeiIiISFqCY97kxG5TIiIiIiPCzBt9MGYvXI65i1ZolXmWLI7fVy3QKhNCoN/wsdh/8Ch+ihiDejWrZ1vnzj1/Y+2mrTh74RIepTzG+sWzUK6sl9Yy46fOwIEjx5CUfB8WFmao9JEPhvbvjlLuJQAAj1Ie46vvvsfhuJNwL14UE74aivJlS2vW/276bBQv6oqu7Vu+7SEg0iu4fxP4hPrD0asonj/LwM24i9gxeTXuXbmrWcavfR183LQ63Cp4wszaHJM+7oVnKU9yrFdhokCdIS3h2/wTWDnZ4XHCAxxbvw97Z27SLNP8+z6o3Kqm1noX957AsrCpAAClaSE0ndwL5T71Q2rSQ2wZsxhX/j6jWfaT3o1gW7QI/vjmFwmOBEmG3aayYvBGH5TSnu74+adJmudKpVJnmWVrNkFhYH1Pnz1DlY8rILRuTXwz5Se9y/h4l0aj+nXg5uKMRymPMWfhcvQe+jW2r1sMpVKJ+UtXI+3JU6xbNBNrft2KcZNnYO2iGQCAE6fP4eSZCxg9pG+e95XIUB6B5XBo2S7cPnEZJoWU+HREG4T9MgozPx2J50/TAQCm5ipc2nsSl/aexKdftjOo3uC+jVG1Uwh+HRaFxIu3ULRiKTSf1hvPHj/FoSXbNctd3HMCv46Yp3n+Iv255v/+7euiaEVPLGgxDmVr+6LVTwMw1b8/AMCuuBP82tXBvCZjpDgMJCVOFSIrBm/0QVEqlXAs4pDt6+f/vYylqzdgzcIZqN2kY671NWlQDwBw+25Ctsu0btpQ8/9ibi4Y2DsMLcP64/bdBJQsXhRXrt3AZ/VqwaNkcbRq+hnW//YnAOD5ixf4dtosjB81WG+QSSSVrCxXlo3D52FUXBSKVvTE9cPnAQAHFm0DAHhUK29wvSX8yuL8zlj8u/s4AODhrWRUbBKE4r6lcOiV5V5kPEdq0iO9dTiVLooLu2KRdPE2HtxIROjXHWHhYI0n9x+j8cRu2DllNdJTnxq+s0TvAQZv71hycjIWLVqEAwcOID4+HgDg6uqK6tWro2vXrnBycsrnFr7fbty6jTpNOkKlMoVvhXIY0rcb3FydAbzMoo0cPwVfDxuQY4D3Np48fYZNW3egeFFXuLm8fK/Lli6Fw3En0LJxA/xzKBZlvTwBAItWrEPVyhXxUfmysrSFKDtm1hYAgKcPU9+qnpux/8KvQ10U8XTFvavxcClfEu7+3tj23XKt5TyqlcfIo3Pw7FEarhw4i+jv12m2HX/uBnyb10AhVWGUrvUxUhIe4Mn9x/i4aXW8SH+Oc9uPvlUbSSb8hQVZMXh7h44cOYLQ0FBYWFggJCQEZcu+/FJOSEjAjBkzMHnyZGzfvh3+/v753NL308c+3vju62HwKFkcyffuY86iFejSfwQ2LZsLS0sLTJ0xH5U+8kHd4CDJt7164xZMn7MQT58+g2fJ4pj/40QULlwYANCzcxtM+H4WPmvTHcXcnPHt6CG4fvM2fvszGivm/YDxU2fiwJE4VChXBt98ORjWVpaSt48oi0KhwGdjO+P6kQtI/PfWW9UVM/d3qKzNMTB6GkSmGgqlCaK/X4eTm//RLHNx7wmc3XYED24mwcHdGSEj2qLzkpFY0GIchFogbu1euJQriYG7puLJ/cdYO2AGzG0tUTe8FRa1+w71hrXGR42r4cGNRPw6Yj4eJzx420NAUmC3qawYvL1DAwcOROvWrREVFQWFQntUlRACffv2xcCBA3HgwIFs60hPT0d6erpWmUl6OlQqlSxtfp8EB1XV/N+7tCcq+nijfsswbPsrBg52tjgUewLrF8+SZduN6tdBUNXKSLp3H0tWbsDwsRFYNnc6VCpTWFtZYuo3X2ot333gKAwb0ANbduzGrTt38fuqBfhm8k+IWrwSIwb2kqWNRADQaEJXOHsXx8JW3751XRU+D8THTT/B+sGzkfjvbbj5uOOzsZ3wOOEBjm+IAQCc/v2gZvnECzeRcO4GhsZEwrOaD678cwbqF5nYOnYJtr5Sb7NpvXFwyXa4VfBAufp+mPPZV6jR93M0/KYL1vTTP/aU6H3CqULeoRMnTmDo0KE6gRvw8q/doUOH4vjx4znWERERAVtbW63HlJ+iZGrx+83G2gruJYrhxq07OBR7HDdv30VQg1bwrdkIvjUbAQCGfj0RXb8Y+dbbsrayhHuJYvCvVBE/TvwaV6/fRPS+f/Qu++vWHbC2skTd4CAciTuJejWro3ChQqhfNxhHjp1867YQZafR+DB4162Mxe0mIiX+/lvXFzq6A2Lm/o7Tvx9E4oWbOPHrfhxYuA3B/Ztku86Dm0lIu5cCBw8Xva97BvnAuUxxHFq6A57VyuPi7hN4/jQdp7cchGcexuORvIRaLdmDdDHz9g65urri8OHDKFeunN7XDx8+DBcX/ResLKNHj0Z4eLhWmcnj25K18UPy5MlT3Lx9F40b1EODusFo2aSB1uvNO/fDyEG9UfuTQEm3K4SAEEBGxnOd1+4/eIioxSvxy9zvAQBqtRrPX7wAALx48QKZmbyQkTwajQ9D+VB/LGr3HR7eSpKkzsLmphCvjX1Sq9V6/4DNYuPqAHN7KzxOfKjzWiFVYTT6tivWD5kNoRZQKE1g8v91KQsXgkLJfESBwW5TWTF4e4eGDx+O3r17IzY2FvXq1dMEagkJCYiOjsaCBQvw/fff51iHSqXS6SJ9npEsW5vfJ9NmLUDtTwJR1NUFicn3MPvn5VAqTdAwpBYc7O303qTg5uKE4kVdNc8bt++FwX27IqTWJwBeztF2Nz4Ricn3AABXb7wcI+RYxB6ORRxw8/ZdbIveh+oBVeBgZ4v4pGQsXLYWKpUpgqtX1dnelJ/mIaxdC7g4OQIAKn/sg9+3/YXqAVWwfvOfqPyxj+THhejzCV1RsWl1rOr1AzLSnsHKyRYA8CzliWbaDisnW1g52cHB/eV1y8W7BNLTnuHR7WQ8fZQGAOi6YjTObj+Kw7/sBABciD6GmgOa4dHte0i8eAtuFTxQvcdniFu3FwBgaqFC7cEtcHbbEaQmPYRDSRfUH90e968l4NI+3SxzrYHNcHH3ccSfuQ4AuHH0X4SO7oBj6/YisMunuHn0X3kPFBmONyzIisHbOzRgwAA4Ojrixx9/xJw5c5CZmQng5fQVfn5+WLJkCdq0aZPPrXx/JSQmY+S4KXiYkgIHO1tU/rgCVsz7EQ72dgbXcfXGLaSm/jcx6e6Yg/jfpB80z0eMmwwA6Ne9Iwb06ASVqSniTpzGsrWbkPI4FUUc7ODv+xGWR/2AIq9t9+9Dsbhx+y4ixo7QlLVv2Rhnzl9Eh15DUNHHG/265z59CVFeBXT+FADQfY32fGkbh8/D8fX7AABVO9ZDnSH/TRTdY91YnWXs3V1g6WCtWWbruKWoN6wVPp/QDZaONnic8ABHV/6FPTM2AgDUmWq4li+JSi2DYWZjiceJD3B53ylE/7AOmRkvtNriXLY4PmpUDXMafqUpO/vHYXhWK48ea8ci+cpdrB88W6pDQlSgKYTgD5Dlh+fPnyM5+WXGzNHRUXPn4RvVlXxFqmYRvVcm+HPyVqLXfXttRe4LvaW0b6X7Q9NyrPztNTbMvOWTwoULw83NLb+bQUREJD3eaCArju4kIiIiMiLMvBEREZG0eLeprBi8ERERkbR4t6ms2G1KREREZESYeSMiIiJpsdtUVgzeiIiISFL8WSt5sduUiIiIyIgweCMiIiJpqYV0jzyaPXs2PDw8YGZmhsDAQBw+fNig9VavXg2FQoFmzZrleZvvGoM3IiIiklY+BW9r1qxBeHg4xo0bh7i4OPj6+iI0NBSJiYk5rnft2jUMHz4cwcHBb7PX7wyDNyIiIpKWUEv3yIMffvgBvXr1Qrdu3eDj44OoqChYWFhg0aJF2a6TmZmJjh07Yvz48ShVqtTb7vk7weCNiIiICqz09HSkpKRoPdLT03WWy8jIQGxsLEJCQjRlJiYmCAkJwYEDB7Kt/9tvv4WzszN69OghS/vlwOCNiIiIpCVht2lERARsbW21HhERETqbTE5ORmZmJlxcXLTKXVxcEB8fr7eZ+/fvx8KFC7FgwQJZDoNcOFUIERERSUpIOM/b6NGjER4erlWmUqneut7Hjx+jc+fOWLBgARwdHd+6vneJwRsREREVWCqVyqBgzdHREUqlEgkJCVrlCQkJcHV11Vn+8uXLuHbtGho3bqwpU////HSFChXChQsX4OXl9Zatlwe7TYmIiEha+XC3qampKfz8/BAdHf1fM9RqREdHIygoSGf5cuXK4dSpUzh+/Ljm0aRJE9SpUwfHjx9HiRIlJDkUcmDmjYiIiKSVT7+wEB4ejrCwMPj7+yMgIACRkZFIS0tDt27dAABdunRBsWLFEBERATMzM3z00Uda69vZ2QGATnlBw+CNiIiI3gtt27ZFUlISxo4di/j4eFSqVAnbtm3T3MRw48YNmJgYf6ejQgjBX481cs+Tr+R3E4gKpAn+Y/K7CUQFzrfXVsi+jcf9P5OsLus5f0pW1/uCmTciIiKSloR3m5Iu488dEhEREX1AmHkjIiIiSXFElrwYvBEREZG02G0qKwZvREREJC0Gb7LimDciIiIiI8LMGxEREUlKyt82JV0M3oiIiEhaDN5kxW5TIiIiIiPCzBsRERFJK39+2vSDweCNiIiIJMUxb/JitykRERGREWHmjYiIiKTFzJusGLwRERGRtDjmTVbsNiUiIiIyIsy8ERERkaR4w4K8GLwRERGRtNhtKisGb0RERCQpZt7kxTFvREREREaEmTciIiKSFrtNZcXgjYiIiCQlGLzJit2mREREREaEmTciIiKSFjNvsmLwRkRERJJit6m82G1KREREZESYeSMiIiJpMfMmKwZvREREJCl2m8qLwRsRERFJisGbvDjmjYiIiMiIMPNGREREkmLmTV4M3oiIiEhaQpHfLXivsduUiIiIyIgw80ZERESSYrepvBi8ERERkaSEmt2mcmK3KREREZERYeaNiIiIJMVuU3kxeCMiIiJJCd5tKit2mxIREREZEWbeiIiISFLsNpUXgzciIiKSFO82lReDNyIiIpKUEPndgvcbx7wRERERGRFm3oiIiEhS7DaVF4M3IiIikhSDN3mx25SIiIjIiDDzRkRERJLiDQvyYvBGREREkmK3qbzYbUpERERkRJh5IyIiIknxt03lxeBNj99++83gZZs0aSJjS4iIiIwPfx5LXgze9GjWrJlByykUCmRmZsrbGCIiIqJXMHjTQ63mnwxERERvSs1uU1kxeCMiIiJJccybvBi8GSAtLQ179+7FjRs3kJGRofXaoEGD8qlVREREBROnCpEXg7dcHDt2DA0bNsSTJ0+QlpYGBwcHJCcnw8LCAs7OzgzeiIiI6J3iPG+5GDp0KBo3bowHDx7A3NwcBw8exPXr1+Hn54fvv/8+v5tHRERU4Agh3YN0MXjLxfHjxzFs2DCYmJhAqVQiPT0dJUqUwNSpU/HVV1/ld/OIiIgKHKFWSPYgXQzeclG4cGGYmLw8TM7Ozrhx4wYAwNbWFjdv3szPphEREdEHiGPeclG5cmUcOXIEZcqUQa1atTB27FgkJydj2bJl+Oijj/K7eURERAUOpwqRFzNvuZg0aRLc3NwAABMnToS9vT369euHpKQkzJ8/P59bR0REVPAIoZDsQbqYecuFv7+/5v/Ozs7Ytm1bPraGiIiIPnQM3oiIiEhSvEtUXgzecuHp6QmFIvu07ZUrV95ha4iIiAo+jnmTF4O3XAwZMkTr+fPnz3Hs2DFs27YNI0aMyJ9GERER0QeLwVsuBg8erLd89uzZOHr06DtuDRERUcHHGw3kxbtN39Bnn32GDRs25HcziIiIChz+woK8mHl7Q+vXr4eDg0N+N4OIiKjA4Zg3eTHzlovKlSujSpUqmkflypXh5uaGr776ij+PRUREVMDMnj0bHh4eMDMzQ2BgIA4fPpztshs3boS/vz/s7OxgaWmJSpUqYdmyZe+wtW+GmbdcNG3aVOtuUxMTEzg5OaF27dooV65cPrbsP+ZFg/O7CUQFUku3qvndBKIPUn6NeVuzZg3Cw8MRFRWFwMBAREZGIjQ0FBcuXICzs7PO8g4ODvj6669Rrlw5mJqaYsuWLejWrRucnZ0RGhqaD3tgGIUQ7FE2doVMi+V3E4gKJAZvRLrWXN8k+zYOFW0hWV2BdzYavmxgIKpWrYpZs2YBANRqNUqUKIGBAwdi1KhRBtVRpUoVNGrUCBMmTHij9r4L7DbNhVKpRGJiok75vXv3oFQq86FFREREH4709HSkpKRoPdLT03WWy8jIQGxsLEJCQjRlJiYmCAkJwYEDB3LdjhAC0dHRuHDhAmrWrCnpPkiNwVsusktMpqenw9TU9B23hoiIqOATEj4iIiJga2ur9YiIiNDZZnJyMjIzM+Hi4qJV7uLigvj4+Gzb+ujRI1hZWcHU1BSNGjXCzJkz8emnn77dAZAZx7xlY8aMGQAAhUKBn3/+GVZWVprXMjMzsW/fvgIz5o2IiKggkfJu09GjRyM8PFyrTKVSSVa/tbU1jh8/jtTUVERHRyM8PBylSpVC7dq1JduG1Bi8ZePHH38E8DLzFhUVpdVFampqCg8PD0RFReVX84iIiD4IKpXKoGDN0dERSqUSCQkJWuUJCQlwdXXNdj0TExOULl0aAFCpUiWcO3cOERERDN6M0dWrVwEAderUwcaNG2Fvb5/PLSIiIjIO+XG3qampKfz8/BAdHY1mzZoBeHnDQnR0NL744guD61Gr1XrH1BUkDN5ysXv37vxuAhERkVFR59N2w8PDERYWBn9/fwQEBCAyMhJpaWno1q0bAKBLly4oVqyYZsxcREQE/P394eXlhfT0dPzxxx9YtmwZ5s6dm097YBgGb7lo2bIlAgIC8OWXX2qVT506FUeOHMG6devyqWVERET0qrZt2yIpKQljx45FfHw8KlWqhG3btmluYrhx4wZMTP67VzMtLQ39+/fHrVu3YG5ujnLlymH58uVo27Ztfu2CQTjPWy6cnJzw119/oWLFilrlp06dQkhIiE7fen7gPG9E+nGeNyJd72Ket32urSWrq2Y8kySvY+YtF6mpqXqnBClcuDBSUlLyoUVEREQFm5ppIVlxnrdcVKxYEWvWrNEpX716NXx8fPKhRURERAWbGgrJHqSLmbdcjBkzBi1atMDly5dRt25dAEB0dDRWrlyJ9evX53PriIiI6EPD4C0XjRs3xqZNmzBp0iSsX78e5ubm8PX1xV9//QUHB4f8bh4REVGBI5gxkxWDNwM0atQIjRo1AgCkpKRg1apVGD58OGJjY5GZmZnPrSMiIipY8muqkA8Fx7wZaN++fQgLC0PRokUxffp01K1bFwcPHszvZhEREdEHhpm3HMTHx2PJkiVYuHAhUlJS0KZNG6Snp2PTpk28WYGIiCgb7DaVFzNv2WjcuDG8vb1x8uRJREZG4s6dO5g5c2Z+N4uIiKjAU0v4IF3MvGXjzz//xKBBg9CvXz+UKVMmv5tDREREBICZt2zt378fjx8/hp+fHwIDAzFr1iwkJyfnd7OIiIgKPGbe5MXgLRvVqlXDggULcPfuXfTp0werV69G0aJFoVarsXPnTjx+/Di/m0hERFQgCSgke5AuBm+5sLS0RPfu3bF//36cOnUKw4YNw+TJk+Hs7IwmTZrkd/OIiIjoA8PgLQ+8vb0xdepU3Lp1C6tWrcrv5hARERVIaoV0D9LFGxbegFKpRLNmzdCsWbP8bgoREVGBw98klReDNyIiIpKUyO8GvOfYbUpERERkRJh5IyIiIklxig95MXgjIiIiSakVHPMmJ3abEhERERkRZt6IiIhIUrxhQV4M3oiIiEhSHPMmL3abEhERERkRZt6IiIhIUvxlBHkxeCMiIiJJ8RcW5MVuUyIiIiIjwswbERERSYp3m8qLwRsRERFJimPe5MXgjYiIiCTFqULkxTFvREREREaEmTciIiKSFMe8yYvBGxEREUmKY97kxW5TIiIiIiPCzBsRERFJijcsyIvBGxEREUmKwZu82G1KREREZESYeSMiIiJJCd6wICsGb0RERCQpdpvKi92mREREREaEmTciIiKSFDNv8mLwRkRERJLiLyzIi8EbERERSYq/sCAvjnkjIiIiMiLMvBEREZGkOOZNXgzeiIiISFIM3uTFblMiIiIiI8LMGxEREUmKd5vKi8EbERERSYp3m8qL3aZERERERoSZNyIiIpIUb1iQF4M3IiIikhTHvMmL3aZERERERoSZNyIiIpKUmrk3WTF4IyIiIklxzJu8GLwRERGRpJh3kxfHvBEREREZEWbeiIiISFLsNpUXgzciIiKSFH9hQV7sNiUiIiIyIsy8ERERkaQ4VYi8GLwRERGRpBi6yYvdpkRERERGhJk3IiIikhTvNpUXgzciIiKSFMe8yYvdpkRERERGhJk3IiIikhTzbvJi8EZERESS4pg3eTF4IyIiIklxzJu8OOaNiIiI3huzZ8+Gh4cHzMzMEBgYiMOHD2e77IIFCxAcHAx7e3vY29sjJCQkx+ULCgZvREREJCkh4SMv1qxZg/DwcIwbNw5xcXHw9fVFaGgoEhMT9S6/Z88etG/fHrt378aBAwdQokQJ1K9fH7dv387rLr9TCiEEc5tGrpBpsfxuAlGB1NKtan43gajAWXN9k+zbGOzRTrK6frq22uBlAwMDUbVqVcyaNQsAoFarUaJECQwcOBCjRo3Kdf3MzEzY29tj1qxZ6NKlyxu3WW7MvBEREVGBlZ6ejpSUFK1Henq6znIZGRmIjY1FSEiIpszExAQhISE4cOCAQdt68uQJnj9/DgcHB8naLwcGb0RERCQpIeG/iIgI2Nraaj0iIiJ0tpmcnIzMzEy4uLholbu4uCA+Pt6gdn/55ZcoWrSoVgBYEPFuUyIiIpKUlFOFjB49GuHh4VplKpVKwi28NHnyZKxevRp79uyBmZmZ5PVLicEbERERFVgqlcqgYM3R0RFKpRIJCQla5QkJCXB1dc1x3e+//x6TJ0/Grl278PHHH79Ve98FdpsSERGRpNQQkj0MZWpqCj8/P0RHR//XDrUa0dHRCAoKyna9qVOnYsKECdi2bRv8/f3far/fFWbeiIiISFL5NY1FeHg4wsLC4O/vj4CAAERGRiItLQ3dunUDAHTp0gXFihXTjJmbMmUKxo4di5UrV8LDw0MzNs7KygpWVlb5tBe5Y/BGRERE74W2bdsiKSkJY8eORXx8PCpVqoRt27ZpbmK4ceMGTEz+63ScO3cuMjIy0KpVK616xo0bh2+++eZdNj1POM/be4DzvL2dokVdETHpKzQIrQsLCzNcunwNPXuGIzbupN7lF/78I8K6tNEpP3P2Anwr1QUABNcIxLBh/VClckUULeqKFq2647fftmstHz60D4YP6w8AmPb9HPwYOU/zWkDVypg5cxKqf/I5MjMzpdrVDw7neTNM+QAfNO7THJ4VveDg4oBpvSJwdMchzevZzQu2fNIS/D5P/2ufdmqATzs1gFNxZwDArYs3sOGntTi+Jw4AYGlrhTbh7fFxcCU4FnNEyr0UHNlxCGumr8TTx080ywz4YTAqBH2Eu9fuImrETFw7c1Wzje4TeiPxRgK2LNgswVH4cLyLed76eLSWrK5519ZJVtf7gpk3+qDZ2dli355N2LP3H3zeuBOSku+hTGlPPHj4KNt1hoaPxVdfT9I8L1SoEOKO7sSGDVs0ZZaWFjh58iwWL1mNDesW6tRRsWJ5fDNuBJo2C4NCocDmTUuwc9denD59HkqlErNnT0a/fiMZuNE7obIww/VzV7F77S4Mnz9a5/Xe/l21nleuXQV9pn6BQ39kP3fWvbv3sHLKMsRfvQOFQoGarepgxILR+LJhOG5dvAkHFwfYuzhg2cQluH3xJhyLO6HnxL6wd3HAj/2mAgBaDGwNM0tzjGo0DJ92aoDekwfgq8bDAQBlKpdF6UplsXjcz9IdCJIMf5heXgze6IM2ckR/3Lp1Bz17/Xcb+rVrN3NcJyXlMVJSHmueN2kSCnt7WyxZukZTtm37bmzbvjvbOry9S+PUqXPYvedvAMCpU+fg7V0ap0+fx/Bh/RATcxBHY0+86W4R5cnxPXGajJg+j5Ieaj33/zQQZw6cRuLNBP0rAIiLPqL1fM20FajfqQHKVPHGrYs3cfPfG/ih7xTN6wk34rFm2gp8ETkUJkoTqDPVKFa6OP75PQZ3r95B9KodqNehPgBAWUiJnhP7Yd6XsyDUDBMKIsEfppcV7zalD9rnn9dHbOxJrF41D3duncCRw9vRo3uHPNXRvVt7REfH4MYNw38L7/TpcyhTxhMlShRFyZLFUKZMKZw5cx6lSrkjLKwtxo6bmtddIXonbB1tUbmuH3av2WXwOgoTE1RvXAMqczP8G3c+2+UsbCzwNPUJ1JkvA7LrZ6/io+ofw0RpAt+alXHj/DUAQJO+zXH24GlcOXX5rfaFyFgxeCtAbt68ie7du+e4jL6fCeGwxTdXyrMk+vTpjEuXrqLh5x0wb94viPzxW3TubNh4DTc3FzQIrYOFi1flabvnz1/C/8ZMwbY/V+PPP1bh6/9NxvnzlzB39hSMGv0d6tevjePHonHk8HYE1wh8k10jkkWtlnXxLO0pDm/L/eeGSni7Y+nZVVhxcR16TuyH7/tMxu2Lt/Qua21vjRYD22DXqh2ask1zNyIzMxMz9kWhamggokbOgquHG2q2rIMNM9ai58S+mBEThSGzR8Dc2kKyfaS3p5bwQbrYbVqA3L9/H0uXLsWiRYuyXSYiIgLjx4/XKlOYWEGhtJG7ee8lExMTxMaexP/GTAYAHD9+BhUqeKNPr85Ytiz3QbJdOrfGw4cp2Lx5W563PX/BMsxfsEzzvHPn1nicmoqDB2Nx9vQ+VKveCMWLuWHF8jkoXTYIGRkZed4GkdRqt6mH/Zv24Xn681yXvXPlNkZ+NhQW1pao1jAIA6YPwjdtv9YJ4MytzPHl4jG4dekm1v/434+QP338BDMH/aC17JhV32LFpKWo0awmnEu6YmidAeg9eQBaDW6LZd8tlmYn6a2x21ReDN7eod9++y3H169cuZJrHfp+JsS+SLm3ateH7O7dRJw9969W2fnzl9CieUOD1u8a1g4rVmzA8+e5f5HlpEgRe4z5eijq1GuJgIDKuHjxCi5duopLl66icOHCKFu2FE6fzr67iehdKFfVB8VKF8dPX3xv0PKZz18g4frLebOunr4ML98yaNitMRZ8NVezjJmlGUb/Mg7P0p5ieu/JyHyR/U06tVvXxZOUNBzdeRjD5n2JozsOIfNFJg7+8TfahOdtuAORMWPw9g41a9YMCoUix25OhUKRYx36fiYkt3Uoe/8cOALvsl5aZWXLlDJo/FqtmkEoU8YTi5bkrctUn+nfj8dPMxbg9u278Pf3RaHChTWvFSqkhFKpfOttEL2tOm1DcPnkJVw/d+2N1leYKFDI9L/PtrmVOb5aNg7P019gao+JOWbzrB1s0HJwW4xtOfr/6zKBstDL80JZqJDW3F2U/9jdKS9+2t8hNzc3bNy4EWq1Wu8jLi77u71IHj/9tACBgVUw6suB8PLyQLt2zdCzZ0fMiVqiWWbid6OweNFPOut269Yehw7F4cyZCzqvWVpawNe3Anx9KwAAPD1Kwte3AkqUKKqzbEi9YJQt44k5c19u8+jREyjn7YUGoXXQs0dHZGaqceECB2aTfFQWZnD38YS7jycAwLmEM9x9PFGkqKNmGXMrc1RrVB1/rd6pt47/rfwWoWH/Zazbj+yE8gE+cCrujBLe7mg/shN8qn2E/Zv2aur7etk3UJmbYd7IWTC3toCtkx1sneyg0BOIdR3XA1sWbMaDhPsAgH+Pnkdwi9ooVro4QjrUx4XYc5IdD3p7aiEke5AuZt7eIT8/P8TGxqJp06Z6X88tK0fSOxp7Aq1a98R3343C/74egqvXbiJ82DisWvWrZhlXVxeUfC3osrGxRovmDTE0fKzeev39fBG9a73m+fTvvwEALP1lLXr0HKopNzMzw08/TUSHjv007/3t23cxeMgY/LzgB6SnZ6B7jyF49uyZVLtMpMPr49IYt+Y7zfOwsT0AAHvW/YW5w2cAAKo3DoZCocDfv8XorcOlpCus7f8be2vjaIf+PwyBvbM9njxOw43z1zGp83ic2v9yChzPj7xQpoo3AGBGTJRWXV980htJtxI1z31rVoKLuxtmDYnUlG1buhWlPi6N7zZNxeUTF7E+cg2IPhT8hYV3KCYmBmlpaWjQoIHe19PS0nD06FHUqlUrT/XyFxaI9OMvLBDpehe/sNDJvYVkdS2/vlGyut4XzLy9Q8HBwTm+bmlpmefAjYiIqKBR825TWXHMGxEREZERYeaNiIiIJMV53uTF4I2IiIgkxalC5MXgjYiIiCTFMW/y4pg3IiIiIiPCzBsRERFJimPe5MXgjYiIiCTFMW/yYrcpERERkRFh5o2IiIgkxR9vkheDNyIiIpIU7zaVF7tNiYiIiIwIM29EREQkKd6wIC8Gb0RERCQpThUiL3abEhERERkRZt6IiIhIUrxhQV4M3oiIiEhSnCpEXgzeiIiISFK8YUFeHPNGREREZESYeSMiIiJJ8W5TeTF4IyIiIknxhgV5sduUiIiIyIgw80ZERESS4t2m8mLwRkRERJJit6m82G1KREREZESYeSMiIiJJ8W5TeTF4IyIiIkmpOeZNVuw2JSIiIjIizLwRERGRpJh3kxeDNyIiIpIU7zaVF4M3IiIikhSDN3lxzBsRERGREWHmjYiIiCTFX1iQF4M3IiIikhS7TeXFblMiIiIiI8LMGxEREUmKv7AgLwZvREREJCmOeZMXu02JiIiIjAgzb0RERCQp3rAgLwZvREREJCl2m8qL3aZERERERoSZNyIiIpIUu03lxeCNiIiIJMWpQuTF4I2IiIgkpeaYN1lxzBsRERGREWHmjYiIiCTFblN5MXgjIiIiSbHbVF7sNiUiIiIyIsy8ERERkaTYbSovBm9EREQkKXabyovdpkRERERGhJk3IiIikhS7TeXF4I2IiIgkxW5TebHblIiIiMiIMPNGREREkmK3qbwYvBEREZGkhFDndxPeawzeiIiISFJqZt5kxTFvRERE9N6YPXs2PDw8YGZmhsDAQBw+fDjbZc+cOYOWLVvCw8MDCoUCkZGR766hb4HBGxEREUlKCCHZIy/WrFmD8PBwjBs3DnFxcfD19UVoaCgSExP1Lv/kyROUKlUKkydPhqurqxS7/k4weCMiIiJJqSEke+TFDz/8gF69eqFbt27w8fFBVFQULCwssGjRIr3LV61aFdOmTUO7du2gUqmk2PV3gsEbERERFVjp6elISUnReqSnp+ssl5GRgdjYWISEhGjKTExMEBISggMHDrzLJsuOwRsRERFJSspu04iICNja2mo9IiIidLaZnJyMzMxMuLi4aJW7uLggPj7+Xe36O8G7TYmIiEhSUv7CwujRoxEeHq5VZkxdnHJg8EZEREQFlkqlMihYc3R0hFKpREJCglZ5QkKCUd2MYAh2mxIREZGkhIT/DGVqago/Pz9ER0drytRqNaKjoxEUFCTHbuYbZt6IiIhIUnmd4kMq4eHhCAsLg7+/PwICAhAZGYm0tDR069YNANClSxcUK1ZMM2YuIyMDZ8+e1fz/9u3bOH78OKysrFC6dOl82QdDMHgjIiKi90Lbtm2RlJSEsWPHIj4+HpUqVcK2bds0NzHcuHEDJib/dTreuXMHlStX1jz//vvv8f3336NWrVrYs2fPu26+wRQiv8Jjkkwh02L53QSiAqmlW9X8bgJRgbPm+ibZt+Fk6y1ZXUmPLkhW1/uCmTciIiKSFPNC8mLwRkRERJKScqoQ0sW7TYmIiIiMCDNvREREJCl2m8qLwRsRERFJKq8/KE95w25TIiIiIiPCzBsRERFJit2m8mLwRkRERJLi3abyYrcpERERkRFh5o2IiIgklZcflKe8Y/BGREREkmK3qbzYbUpERERkRJh5IyIiIknxblN5MXgjIiIiSXHMm7wYvBEREZGkmHmTF8e8ERERERkRZt6IiIhIUsy8yYvBGxEREUmKoZu82G1KREREZEQUgrlNIsmkp6cjIiICo0ePhkqlyu/mEBUIPC+IpMXgjUhCKSkpsLW1xaNHj2BjY5PfzSEqEHheEEmL3aZERERERoTBGxEREZERYfBGREREZEQYvBFJSKVSYdy4cRyUTfQKnhdE0uINC0RERERGhJk3IiIiIiPC4I2IiIjIiDB4IyIiIjIiDN6IiIiIjAiDNyKJzJ49Gx4eHjAzM0NgYCAOHz6c300iylf79u1D48aNUbRoUSgUCmzatCm/m0T0XmDwRiSBNWvWIDw8HOPGjUNcXBx8fX0RGhqKxMTE/G4aUb5JS0uDr68vZs+end9NIXqvcKoQIgkEBgaiatWqmDVrFgBArVajRIkSGDhwIEaNGpXPrSPKfwqFAr/++iuaNWuW300hMnrMvBG9pYyMDMTGxiIkJERTZmJigpCQEBw4cCAfW0ZERO8jBm9Ebyk5ORmZmZlwcXHRKndxcUF8fHw+tYqIiN5XDN6IiIiIjAiDN6K35OjoCKVSiYSEBK3yhIQEuLq65lOriIjofcXgjegtmZqaws/PD9HR0ZoytVqN6OhoBAUF5WPLiIjofVQovxtA9D4IDw9HWFgY/P39ERAQgMjISKSlpaFbt2753TSifJOamopLly5pnl+9ehXHjx+Hg4MDSpYsmY8tIzJunCqESCKzZs3CtGnTEB8fj0qVKmHGjBkIDAzM72YR5Zs9e/agTp06OuVhYWFYsmTJu28Q0XuCwRsRERGREeGYNyIiIiIjwuCNiIiIyIgweCMiIiIyIgzeiIiIiIwIgzciIiIiI8LgjYiIiMiIMHgjIiIiMiIM3ojovdC1a1c0a9ZM87x27doYMmTIO2/Hnj17oFAo8PDhw3e+bSL6MDB4IyJZde3aFQqFAgqFAqampihdujS+/fZbvHjxQtbtbty4ERMmTDBoWQZcRGRM+NumRCS7Bg0aYPHixUhPT8cff/yBAQMGoHDhwhg9erTWchkZGTA1NZVkmw4ODpLUQ0RU0DDzRkSyU6lUcHV1hbu7O/r164eQkBD89ttvmq7OiRMnomjRovD29gYA3Lx5E23atIGdnR0cHBzQtGlTXLt2TVNfZmYmwsPDYWdnhyJFimDkyJF4/Zf+Xu82TU9Px5dffokSJUpApVKhdOnSWLhwIa5du6b5/U17e3soFAp07doVAKBWqxEREQFPT0+Ym5vD19cX69ev19rOH3/8gbJly8Lc3Bx16tTRaicRkRwYvBHRO2dubo6MjAwAQHR0NC5cuICdO3diy5YteP78OUJDQ2FtbY2YmBj8/fffsLKyQoMGDTTrTJ8+HUuWLMGiRYuwf/9+3L9/H7/++muO2+zSpQtWrVqFGTNm4Ny5c5g3bx6srKxQokQJbNiwAQBw4cIF3L17Fz/99BMAICIiAr/88guioqJw5swZDB06FJ06dcLevXsBvAwyW7RogcaNG+P48ePo2bMnRo0aJddhIyICwG5TInqHhBCIjo7G9u3bMXDgQCQlJcHS0hI///yzprt0+fLlUKvV+Pnnn6FQKAAAixcvhp2dHfbs2YP69esjMjISo0ePRosWLQAAUVFR2L59e7bb/ffff7F27Vrs3LkTISEhAIBSpUppXs/qYnV2doadnR2Al5m6SZMmYdeuXQgKCtKss3//fsybNw+1atXC3Llz4eXlhenTpwMAvL29cerUKUyZMkXCo0ZEpI3BGxHJbsuWLbCyssLz58+hVqvRoUMHfPPNNxgwYAAqVqyoNc7txIkTuHTpEqytrbXqePbsGS5fvoxHjx7h7t27CAwM1LxWqFAh+Pv763SdZjl+/DiUSiVq1aplcJsvXbqEJ0+e4NNPP9Uqz8jIQOXKlQEA586d02oHAE2gR0QkFwZvRCS7OnXqYO7cuTA1NUXRokVRqNB/lx5LS0utZVNTU+Hn54cVK1bo1OPk5PRG2zc3N8/zOqmpqQCArVu3olixYlqvqVSqN2oHEZEUGLwRkewsLS1RunRpg5atUqUK1qxZA2dnZ9jY2Ohdxs3NDYcOHULNmjUBAC9evEBsbCyqVKmid/mKFStCrVZj7969mm7TV2Vl/jIzMzVlPj4+UKlUuHHjRrYZu/Lly+O3337TKjt48GDuO0lE9BZ4wwIRFSgdO3aEo6MjmjZtipiYGFy9ehV79uzBoEGDcOvWLQDA4MGDMXnyZGzatAnnz59H//79c5yjzcPDA2FhYejevTs2bdqkqXPt2rUAAHd3dygUCmzZsgVJSUlITU2FtbU1hg8fjqFDh2Lp0qW4fPky4uLiMHPmTCxduhQA0LdvX1y8eBEjRozAhQsXsHLlSixZskTuQ0REHzgGb0RUoFhYWGDfvn0oWbIkWrRogfLly6NHjx549uyZJhM3bNgwdO7cGWFhYQgKCoK1tTWaN2+eY71z585Fq1at0L9/f5QrVw69evVCWloaAKBYsWIYP348Ro0aBRcXF3zxxRcAgAkTJmDMmDGIiIhA+fLl0aBBA2zduhWenp4AgJIlS2LDhg3YtGkTfH19ERUVhUmTJsl4dIiIAIXIboQvERERERU4zLwRERERGREGb0RERERGhMEbERERkRFh8EZERERkRBi8ERERERkRBm9ERERERoTBGxEREZERYfBGREREZEQYvBEREREZEQZvREREREaEwRsRERGREWHwRkRERGRE/g/8bS47oci1KQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_pred_acc = test.copy()\n",
        "test_pred_acc.labels = acc_pred_y\n",
        "f_metric = ClassificationMetric(test, test_pred_acc, unprivileged_groups=unprivileged, privileged_groups=privileged)\n",
        "\n",
        "print(\"Fairness metrics of accurate model on reweighed samples:-\")\n",
        "print(\"Equal opportunity difference: \", round(f_metric.equal_opportunity_difference(), 4))\n",
        "print(\"Disparate Impact: \", round(f_metric.disparate_impact(), 4))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "C1v9kbzDnwV9",
        "outputId": "ec1cdec9-8365-4905-d5b5-dd8ea5d3c5dd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fairness metrics of accurate model on reweighed samples:-\n",
            "Equal opportunity difference:  0.0329\n",
            "Disparate Impact:  0.9351\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "far_rw_model = fair_model(X_train, y_train, weights)"
      ],
      "metadata": {
        "id": "3UF05PUmn04s"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "fpred = far_rw_model.predict(X_test)\n",
        "print(classification_report(y_test, fpred))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xB_UCBozn5SG",
        "outputId": "ae2ae4ca-ead5-4430-b97a-2dd1fbbe92ba"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "         0.0       0.89      0.72      0.79     11132\n",
            "         1.0       0.44      0.71      0.55      3521\n",
            "\n",
            "    accuracy                           0.72     14653\n",
            "   macro avg       0.67      0.71      0.67     14653\n",
            "weighted avg       0.78      0.72      0.73     14653\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "conf_matrix = confusion_matrix(y_test, fpred)\n",
        "sns.heatmap(conf_matrix/np.sum(conf_matrix), annot=True, fmt='.2%')\n",
        "plt.xlabel(\"Predicted\")\n",
        "plt.ylabel(\"Actual\")\n",
        "plt.title(\"Visualising fair model performance with reweighed sample weights\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "57tzBJIxn9Gf",
        "outputId": "e9b8306f-cd6a-44cb-a371-896d58a7f4bd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_pred_fair = test.copy()\n",
        "test_pred_fair.labels = fpred\n",
        "f_metric = ClassificationMetric(test, test_pred_fair, unprivileged_groups=unprivileged, privileged_groups=privileged)\n",
        "\n",
        "print(\"Fairness metrics of fair model on reweighed samples:-\")\n",
        "print(\"Equal opportunity difference: \", round(f_metric.equal_opportunity_difference(), 4))\n",
        "print(\"Disparate Impact: \", round(f_metric.disparate_impact(), 4))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ETH47-fvoAdc",
        "outputId": "dca77bb0-7e65-44de-973e-9dc0dabe7e43"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fairness metrics of fair model on reweighed samples:-\n",
            "Equal opportunity difference:  0.0461\n",
            "Disparate Impact:  0.961\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(figsize=(13, 5))\n",
        "\n",
        "plt.subplot(1, 2, 1)\n",
        "acc_verd = view_proportions(y_test, acc_pred_y)\n",
        "fdf = pd.DataFrame({'sex': [k[1] for k in X_test], 'classification': acc_verd})\n",
        "sns.countplot(data=fdf, x='classification', hue='sex', stat='percent')\n",
        "plt.title(\"Accurate model\")\n",
        "\n",
        "plt.subplot(1, 2, 2)\n",
        "fair_verd = view_proportions(y_test, fpred)\n",
        "fdf = pd.DataFrame({'sex': [k[1] for k in X_test], 'classification': fair_verd})\n",
        "sns.countplot(data=fdf, x='classification', hue='sex', stat='percent')\n",
        "plt.title(\"Fair model\")\n",
        "\n",
        "\n",
        "plt.suptitle(\"Counts of errors w.r.t. protected class\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 515
        },
        "id": "VHCDXfLXoDl1",
        "outputId": "7fa6cf92-35d6-456d-e23a-69b0c0559ba1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1300x500 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABDIAAAHyCAYAAAAZee4zAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABstUlEQVR4nO3deVwV5fv/8fcBZV9MFMVEILdc0IrU1FxKDSlNczcLtUzLLaVFbXFpo/KjaWaafcrtg7lUtqe5Z2a5pLmkpoRSiZqWoKigcP/+8Mf5emIRETwMvJ6Px3nUzNwzc83MOYfL69xzj80YYwQAAAAAAGABLs4OAAAAAAAAIL8oZAAAAAAAAMugkAEAAAAAACyDQgYAAAAAALAMChkAAAAAAMAyKGQAAAAAAADLoJABAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAABQAEePHlW3bt0UEBAgm82mKVOmODskoNDZbDaNHz++2G0LAFC6UcgAAOQoPj5egwYN0g033CAPDw/5+fmpefPmmjp1qs6ePevs8CRJb7/9tubMmeOUfY8cOVLLly/XmDFjNH/+fLVv394pcZQmCxYssFTB6Kuvvrom/3C32nkBAOBqlXF2AACA4ufLL79U9+7d5e7urujoaNWvX1/p6en67rvv9NRTT2n37t2aNWuWs8PU22+/rQoVKqhfv37XfN+rV69Wp06d9OSTT17zfZdWCxYs0K5duzRixAhnh5IvX331laZPn17kxQyrnRcAAK4WhQwAgIOEhAT16tVLISEhWr16tYKCguzLhgwZogMHDujLL790YoTFw7Fjx1SuXLki3YcxRufOnZOnp2e2ZefOnZObm5tcXAreuTI1NVXe3t5XE2KRKK5xFcY5BwAAV4+/xAAAB6+//rpOnz6t9957z6GIkaVGjRp6/PHH7dMXLlzQiy++qOrVq8vd3V2hoaF65plnlJaW5rBebvfHh4aGOvSomDNnjmw2mzZs2KCYmBhVrFhR3t7euu+++/TXX385rLd7926tW7dONptNNptNrVu3liSdP39eEyZMUM2aNeXh4aGAgADdfvvtWrFixWWP/7ffflP37t1Vvnx5eXl56bbbbnMo3GTFZ4zR9OnT7fvOS2ZmpqZMmaJ69erJw8NDlSpV0qBBg/TPP/9kOxcdOnTQ8uXLdeutt8rT01PvvPOO1q5dK5vNpoULF+q5557T9ddfLy8vL6WkpEiSlixZooiICHl6eqpChQp64IEH9Oeffzpsu1+/fvLx8VF8fLzuvvtu+fr6qk+fPpKk/fv3q2vXrqpcubI8PDxUtWpV9erVS8nJybke05tvvilXV1edPHnSPm/SpEmy2WyKiYmxz8vIyJCvr69GjRqV43ayzue6des0ePBgBQYGqmrVqtnatW7dWl9++aUOHTpkP+ehoaF5nvectG7dWvXr19fWrVvVrFkzeXp6KiwsTDNnznRod7XnvF+/fpo+fbok2eO99H2S3/eEJH399ddq1aqVfH195efnp0aNGmnBggX5Oi9paWkaN26catSoIXd3dwUHB+vpp5/O9vlMS0vTyJEjVbFiRfn6+uree+/VH3/8ke/zeu7cOY0fP161atWSh4eHgoKC1KVLF8XHx+e6zqFDhzR48GDVrl1bnp6eCggIUPfu3XXw4EGHdvn5PB85ckT9+/dX1apV5e7urqCgIHXq1CnbtgAAJQM9MgAADj7//HPdcMMNatasWb7aDxgwQHPnzlW3bt30xBNP6Mcff1RsbKz27NmjpUuXFjiOYcOG6brrrtO4ceN08OBBTZkyRUOHDtWiRYskSVOmTNGwYcPk4+OjZ599VpJUqVIlSdL48eMVGxurAQMGqHHjxkpJSdGWLVv0008/qV27drnu8+jRo2rWrJnOnDmj4cOHKyAgQHPnztW9996rDz/8UPfdd59atmyp+fPn68EHH1S7du0UHR192WMZNGiQ5syZo/79+2v48OFKSEjQW2+9pW3btmnDhg0qW7asve2+ffvUu3dvDRo0SI888ohq165tX/biiy/Kzc1NTz75pNLS0uTm5mbfbqNGjRQbG6ujR49q6tSp2rBhg7Zt2+bQa+TChQuKjIzU7bffrv/85z/y8vJSenq6IiMjlZaWpmHDhqly5cr6888/9cUXX+jkyZPy9/fP8ZhatGihzMxMfffdd+rQoYMkaf369XJxcdH69evt7bZt26bTp0+rZcuWeZ6jwYMHq2LFiho7dqxSU1OzLX/22WeVnJysP/74Q2+88YYkycfH57LnPif//POP7r77bvXo0UO9e/fW4sWL9dhjj8nNzU0PPfSQQ9uCnvNBgwbp8OHDWrFihebPn58thvy+J+bMmaOHHnpI9erV05gxY1SuXDlt27ZNy5Yt0/3335/necnMzNS9996r7777TgMHDlSdOnW0c+dOvfHGG/r111/1ySef2OMZMGCA/ve//+n+++9Xs2bNtHr1at1zzz35Op8ZGRnq0KGDVq1apV69eunxxx/XqVOntGLFCu3atUvVq1fPcb3Nmzfr+++/V69evVS1alUdPHhQM2bMUOvWrfXLL7/Iy8tLUv4+z127dtXu3bs1bNgwhYaG6tixY1qxYoUSExMLVPACABRzBgCA/y85OdlIMp06dcpX++3btxtJZsCAAQ7zn3zySSPJrF692j5Pkhk3bly2bYSEhJi+ffvap2fPnm0kmbZt25rMzEz7/JEjRxpXV1dz8uRJ+7x69eqZVq1aZdtmw4YNzT333JOvY7jUiBEjjCSzfv16+7xTp06ZsLAwExoaajIyMhyOZ8iQIZfd5vr1640kExcX5zB/2bJl2eaHhIQYSWbZsmUObdesWWMkmRtuuMGcOXPGPj89Pd0EBgaa+vXrm7Nnz9rnf/HFF0aSGTt2rH1e3759jSQzevRoh21v27bNSDJLliy57LFcKiMjw/j5+Zmnn37aGGNMZmamCQgIMN27dzeurq7m1KlTxhhjJk+ebFxcXMw///yT43ayrvftt99uLly4kOc+77nnHhMSEnJFcf5bq1atjCQzadIk+7y0tDRz0003mcDAQJOenm6MKZxzPmTIEJNTqpXf98TJkyeNr6+vadKkicO+jDEOn43czsv8+fONi4uLw/vZGGNmzpxpJJkNGzYYY/7vczx48GCHdvfff3+un9tLvf/++0aSmTx5crZll8b5721del6zbNy40Ugy8+bNs8+73Of5n3/+MZLMxIkT84wTAFBycGsJAMAuq9u8r69vvtp/9dVXkuRwK4EkPfHEE5J0VWNpDBw40KErfosWLZSRkaFDhw5ddt1y5cpp9+7d2r9//xXt86uvvlLjxo11++232+f5+Pho4MCBOnjwoH755Zcr2p508RYEf39/tWvXTsePH7e/IiIi5OPjozVr1ji0DwsLU2RkZI7b6tu3r8N4GVu2bNGxY8c0ePBgeXh42Offc889uvHGG3M8/4899pjDdFaPi+XLl+vMmTP5Pi4XFxc1a9ZM3377rSRpz549OnHihEaPHi1jjDZu3CjpYi+N+vXrX3Y8kUceeUSurq753v/VKFOmjAYNGmSfdnNz06BBg3Ts2DFt3brVoW1hnPN/y+97YsWKFTp16pRGjx7tsC9Jl72dKWs/derU0Y033uiwnzvvvFOS7PvJ+hwPHz7cYf38Dh760UcfqUKFCho2bFi2ZXnFeel5PX/+vE6cOKEaNWqoXLly+umnn+zLLvd59vT0lJubm9auXZvjrTkAgJKHQgYAwM7Pz0+SdOrUqXy1P3TokFxcXFSjRg2H+ZUrV1a5cuXyVXTITbVq1Rymr7vuOknK1z9UXnjhBZ08eVK1atVSeHi4nnrqKe3YseOy6x06dMjhVo4sderUsS+/Uvv371dycrICAwNVsWJFh9fp06d17Ngxh/ZhYWG5buvfy7LiySnmG2+8MVu8ZcqUyTb+RFhYmGJiYvTf//5XFSpUUGRkpKZPn57n+BhZWrRooa1bt+rs2bNav369goKCdMstt6hhw4b220u+++47tWjR4rLbyuu4C1uVKlWyDSZaq1YtSco2psLVnvOc5Pc9kTW+RP369fN3YDnsZ/fu3dn2kXWsWfvJ+hz/+xaQnI4xJ/Hx8apdu7bKlLmyO5bPnj2rsWPHKjg4WO7u7qpQoYIqVqyokydPOrz/Lvd5dnd312uvvaavv/5alSpVUsuWLfX666/ryJEjVxQPAMA6GCMDAGDn5+enKlWqaNeuXVe0Xn5+Hc5NRkZGjvNz+3XeGHPZbbZs2VLx8fH69NNP9c033+i///2v3njjDc2cOVMDBgwocKwFkZmZqcDAQMXFxeW4vGLFig7TOT2hJD/L8sPd3T3HJ25MmjRJ/fr1s5+v4cOHKzY2Vj/88EOOA29muf3223X+/Hlt3LhR69evtxcsWrRoofXr12vv3r3666+/8lXIuNpjKypFEdeVvieuZj/h4eGaPHlyjsuDg4MLZT8FNWzYMM2ePVsjRoxQ06ZN5e/vL5vNpl69eikzM9PeLj+f5xEjRqhjx4765JNPtHz5cj3//POKjY3V6tWrdfPNNzvrEAEARYRCBgDAQYcOHTRr1ixt3LhRTZs2zbNtSEiIMjMztX//fnuvBenioJknT55USEiIfd51113n8IQLSUpPT1dSUlKBY82rgFK+fHn1799f/fv3tw82OX78+DwLGSEhIdq3b1+2+Xv37rUvv1LVq1fXypUr1bx580L/R3FWPPv27bPfLpBl3759VxRveHi4wsPD9dxzz+n7779X8+bNNXPmTL300ku5rtO4cWO5ublp/fr1Wr9+vZ566ilJF//h+e6772rVqlX26cJwNQWzSx0+fDjbI15//fVXSbrswJBXcs5zize/74msHhK7du3K1uvpUnnt5+eff1abNm3yPHdZn+OsnhWXHk9+VK9eXT/++KPOnz/vMHDt5Xz44Yfq27evJk2aZJ937ty5bN8TUv4+z9WrV9cTTzyhJ554Qvv379dNN92kSZMm6X//+1++YwIAWAO3lgAAHDz99NPy9vbWgAEDdPTo0WzL4+PjNXXqVEnS3XffLeniE0QulfUL8KVPPahevbp9PIUss2bNyrVHRn54e3vn+I+eEydOOEz7+PioRo0a2R45+W933323Nm3aZB/fQZJSU1M1a9YshYaGqm7dulccY48ePZSRkaEXX3wx27ILFy7kGH9+3XrrrQoMDNTMmTMdju3rr7/Wnj178vXUiZSUFF24cMFhXnh4uFxcXC57vjw8PNSoUSN98MEHSkxMdOiRcfbsWb355puqXr26/TG+ycnJ2rt3b75uW0lMTLQXkLJ4e3vnuu7evXuVmJh42e1KF8/7O++8Y59OT0/XO++8o4oVKyoiIiLPda/knGcVSv59jfP7nrjrrrvk6+ur2NhYnTt3zqHdpT2TcjsvPXr00J9//ql3330327KzZ8/anw4TFRUl6eIjdS/17891brp27arjx4/rrbfeyrYsrx5Urq6u2ZZPmzYt23fC5T7PZ86cyXZ+qlevLl9f38u+hwEA1kSPDACAg+rVq2vBggXq2bOn6tSpo+joaNWvX1/p6en6/vvvtWTJEvXr10+S1LBhQ/Xt21ezZs3SyZMn1apVK23atElz585V586ddccdd9i3O2DAAD366KPq2rWr2rVrp59//lnLly9XhQoVChxrRESEZsyYoZdeekk1atRQYGCg7rzzTtWtW1etW7dWRESEypcvry1btujDDz/U0KFD89ze6NGj9cEHHygqKkrDhw9X+fLlNXfuXCUkJOijjz7K8baMy2nVqpUGDRqk2NhYbd++XXfddZfKli2r/fv3a8mSJZo6daq6detWoOMvW7asXnvtNfXv31+tWrVS79697Y8CDQ0N1ciRIy+7jdWrV2vo0KHq3r27atWqpQsXLmj+/PlydXVV165dL7t+ixYt9Oqrr8rf31/h4eGSpMDAQNWuXVv79u2zv1ckaenSperfv79mz57tMD8n0dHRWrduncM/dCMiIrRo0SLFxMSoUaNG8vHxUceOHSVdHMekVatWWrt27WVjrlKlil577TUdPHhQtWrV0qJFi7R9+3bNmjXrsj0KruScZxVFhg8frsjISLm6uqpXr175fk/4+fnpjTfe0IABA9SoUSPdf//9uu666/Tzzz/rzJkzmjt3bp7n5cEHH9TixYv16KOPas2aNWrevLkyMjK0d+9eLV68WMuXL9ett96qm266Sb1799bbb7+t5ORkNWvWTKtWrdKBAwcuey6zrtW8efMUExOjTZs2qUWLFkpNTdXKlSs1ePBgderUKcf1OnTooPnz58vf319169bVxo0btXLlSgUEBDi0u9zn+ddff1WbNm3Uo0cP1a1bV2XKlNHSpUt19OhR9erVK1/HAACwGGc+MgUAUHz9+uuv5pFHHjGhoaHGzc3N+Pr6mubNm5tp06aZc+fO2dudP3/eTJgwwYSFhZmyZcua4OBgM2bMGIc2xlx8XOeoUaNMhQoVjJeXl4mMjDQHDhzI9fGrmzdvdlg/63GYa9assc87cuSIueeee4yvr6+RZH8U60svvWQaN25sypUrZzw9Pc2NN95oXn75ZfujNfMSHx9vunXrZsqVK2c8PDxM48aNzRdffJGtnfL5+NUss2bNMhEREcbT09P4+vqa8PBw8/TTT5vDhw/b24SEhOT4mMmsY8/tEamLFi0yN998s3F3dzfly5c3ffr0MX/88YdDm759+xpvb+9s6/7222/moYceMtWrVzceHh6mfPny5o477jArV67M13F9+eWXRpKJiopymD9gwAAjybz33nv2eVnXdvbs2dnm/ft6Zz0m9VKnT582999/vylXrpyR5PDI0Uuvf15atWpl6tWrZ7Zs2WKaNm1qPDw8TEhIiHnrrbcc2hXGOb9w4YIZNmyYqVixorHZbNmOJz/vCWOM+eyzz0yzZs2Mp6en8fPzM40bNzYffPBBvs5Lenq6ee2110y9evWMu7u7ue6660xERISZMGGCSU5Otrc7e/asGT58uAkICDDe3t6mY8eO5vfff8/X41eNufgo1Weffdb+PVC5cmXTrVs3Ex8fb2/z7239888/pn///qZChQrGx8fHREZGmr1792b7Trjc5/n48eNmyJAh5sYbbzTe3t7G39/fNGnSxCxevPiycQMArMlmTD5GTQMAACgBWrdurePHj1/xgLYAAKD4YIwMAAAAAABgGRQyAAAAAACAZVDIAAAAAAAAlsEYGQAAAAAAwDLokQEAAAAAACyDQgYAAAAAALAMChkAAAAAAMAyKGQAAAAAAADLoJABAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgBcgbVr18pms2nt2rVXvO6cOXNks9l08ODBQo8LAICS6Gr+7hYX5A5A4aOQATjZ22+/LZvNpiZNmjg7lEL11Vdfafz48c4OAwAAOEHWP8Bzeo0ePdrZ4QGwuDLODgAo7eLi4hQaGqpNmzbpwIEDqlGjhrNDKhRfffWVpk+fTjEDAIBS7IUXXlBYWJjDvPr16+d7/ZYtW+rs2bNyc3Mr7NAAWBiFDMCJEhIS9P333+vjjz/WoEGDFBcXp3Hjxjk7rBylpqbK29vb2WEAAAALiYqK0q233lrg9V1cXOTh4XHZdmfOnJGXl1eB9wPAWri1BHCiuLg4XXfddbrnnnvUrVs3xcXF5dju5MmTGjlypEJDQ+Xu7q6qVasqOjpax48ft7c5d+6cxo8fr1q1asnDw0NBQUHq0qWL4uPjJeV+f+bBgwdls9k0Z84c+7x+/frJx8dH8fHxuvvuu+Xr66s+ffpIktavX6/u3burWrVqcnd3V3BwsEaOHKmzZ886rD99+nRJcuhKmiUzM1NTpkxRvXr15OHhoUqVKmnQoEH6559/LnvOsmJLTExUhw4d5OPjo+uvv96+v507d+rOO++Ut7e3QkJCtGDBgmzb+O2339S9e3eVL19eXl5euu222/Tll19ma/fHH3+oc+fO8vb2VmBgoEaOHKm0tLQc4/rxxx/Vvn17+fv7y8vLS61atdKGDRsuezwAAJRGhw4d0uDBg1W7dm15enoqICBA3bt3zzYWRE75S+vWrVW/fn1t3bpVLVu2lJeXl5555plc90XuAJQ89MgAnCguLk5dunSRm5ubevfurRkzZmjz5s1q1KiRvc3p06fVokUL7dmzRw899JBuueUWHT9+XJ999pn++OMPVahQQRkZGerQoYNWrVqlXr166fHHH9epU6e0YsUK7dq1S9WrV7/i2C5cuKDIyEjdfvvt+s9//mP/lWPJkiU6c+aMHnvsMQUEBGjTpk2aNm2a/vjjDy1ZskSSNGjQIB0+fFgrVqzQ/Pnzs2170KBBmjNnjvr376/hw4crISFBb731lrZt26YNGzaobNmyecaWkZGhqKgotWzZUq+//rri4uI0dOhQeXt769lnn1WfPn3UpUsXzZw5U9HR0WratKm9W+vRo0fVrFkznTlzRsOHD1dAQIDmzp2re++9Vx9++KHuu+8+SdLZs2fVpk0bJSYmavjw4apSpYrmz5+v1atXZ4tn9erVioqKUkREhMaNGycXFxfNnj1bd955p9avX6/GjRtf8fkHAKAkSE5OdvjhRZIqVKigzZs36/vvv1evXr1UtWpVHTx4UDNmzFDr1q31yy+/XLZ3xYkTJxQVFaVevXrpgQceUKVKlfJsT+4AlDAGgFNs2bLFSDIrVqwwxhiTmZlpqlatah5//HGHdmPHjjWSzMcff5xtG5mZmcYYY95//30jyUyePDnXNmvWrDGSzJo1axyWJyQkGElm9uzZ9nl9+/Y1kszo0aOzbe/MmTPZ5sXGxhqbzWYOHTpknzdkyBCT01fM+vXrjSQTFxfnMH/ZsmU5zv+3rNheeeUV+7x//vnHeHp6GpvNZhYuXGifv3fvXiPJjBs3zj5vxIgRRpJZv369fd6pU6dMWFiYCQ0NNRkZGcYYY6ZMmWIkmcWLF9vbpaammho1ajicx8zMTFOzZk0TGRlpP9dZ5yksLMy0a9fOPm/27NlGkklISMjzGAEAsLqsv3k5vYzJOZ/YuHGjkWTmzZtnn5dT/tKqVSsjycycOTNfsZA7ACUPt5YAThIXF6dKlSrpjjvukHTxFoyePXtq4cKFysjIsLf76KOP1LBhQ3u1/1JZt2t89NFHqlChgoYNG5Zrm4J47LHHss3z9PS0/39qaqqOHz+uZs2ayRijbdu2XXabS5Yskb+/v9q1a6fjx4/bXxEREfLx8dGaNWvyFduAAQPs/1+uXDnVrl1b3t7e6tGjh31+7dq1Va5cOf3222/2eV999ZUaN26s22+/3T7Px8dHAwcO1MGDB/XLL7/Y2wUFBalbt272dl5eXho4cKBDHNu3b9f+/ft1//3368SJE/bjSU1NVZs2bfTtt98qMzMzX8cEAEBJM336dK1YscLhJTnmE+fPn9eJEydUo0YNlStXTj/99NNlt+vu7q7+/ftfUSzkDkDJwa0lgBNkZGRo4cKFuuOOO5SQkGCf36RJE02aNEmrVq3SXXfdJUmKj49X165d89xefHy8ateurTJlCu8jXaZMGVWtWjXb/MTERI0dO1afffZZtjEtkpOTL7vd/fv3Kzk5WYGBgTkuP3bs2GW34eHhoYoVKzrM8/f3V9WqVbMVbvz9/R3iPHToUI6Puq1Tp459ef369XXo0CHVqFEj2/Zq166d7XgkqW/fvrnGm5ycrOuuu+6yxwUAQEnTuHHjHAf7PHv2rGJjYzV79mz9+eefMsbYl+Unn7j++uuv6Ekm5A5AyUIhA3CC1atXKykpSQsXLtTChQuzLY+Li7MXMgpLbj0zLu39cSl3d3e5uLhka9uuXTv9/fffGjVqlG688UZ5e3vrzz//VL9+/fL160FmZqYCAwNzHdj030lGTlxdXa9o/qXJUWHLOuaJEyfqpptuyrGNj49Pke0fAAArGjZsmGbPnq0RI0aoadOm8vf3l81mU69evfKVT1zaoyM/yB2AkoVCBuAEcXFxCgwMtI+WfamPP/5YS5cu1cyZM+Xp6anq1atr165deW6vevXq+vHHH3X+/PlcB8rMquqfPHnSYf6hQ4fyHffOnTv166+/au7cuYqOjrbPz+omeqncCifVq1fXypUr1bx58ytOQgpDSEiI9u3bl23+3r177cuz/rtr1y4ZYxyO5d/rZg2k6ufnp7Zt2xZV2AAAlCgffvih+vbtq0mTJtnnnTt3LlueUhyQOwDFD2NkANfY2bNn9fHHH6tDhw7q1q1bttfQoUN16tQpffbZZ5Kkrl276ueff9bSpUuzbSvr14KuXbvq+PHjeuutt3JtExISIldXV3377bcOy99+++18x571q8Wlv1IYYzR16tRsbb29vSVlL5z06NFDGRkZevHFF7Otc+HChSJPYO6++25t2rRJGzdutM9LTU3VrFmzFBoaqrp169rbHT58WB9++KG93ZkzZzRr1iyH7UVERKh69er6z3/+o9OnT2fb319//VVERwIAgHW5urpm6/Uwbdq0XHuKOhO5A1D80CMDuMY+++wznTp1Svfee2+Oy2+77TZVrFhRcXFx6tmzp5566il9+OGH6t69ux566CFFRETo77//1meffaaZM2eqYcOGio6O1rx58xQTE6NNmzapRYsWSk1N1cqVKzV48GB16tRJ/v7+6t69u6ZNmyabzabq1avriy++yNeYFFluvPFGVa9eXU8++aT+/PNP+fn56aOPPso2VoZ08Y+0JA0fPlyRkZFydXVVr1691KpVKw0aNEixsbHavn277rrrLpUtW1b79+/XkiVLNHXqVIdBsgrb6NGj9cEHHygqKkrDhw9X+fLlNXfuXCUkJOijjz6y307zyCOP6K233lJ0dLS2bt2qoKAgzZ8/P9vj4FxcXPTf//5XUVFRqlevnvr376/rr79ef/75p9asWSM/Pz99/vnnRXY8AABYUYcOHTR//nz5+/urbt262rhxo1auXKmAgABnh5YNuQNQ/FDIAK6xuLg4eXh4qF27djkud3Fx0T333KO4uDidOHFCAQEBWr9+vcaNG6elS5dq7ty5CgwMVJs2beyDcbq6uuqrr77Syy+/rAULFuijjz5SQECAbr/9doWHh9u3PW3aNJ0/f14zZ86Uu7u7evTooYkTJ6p+/fr5ir1s2bL6/PPPNXz4cMXGxsrDw0P33Xefhg4dqoYNGzq07dKli4YNG6aFCxfqf//7n4wx6tWrlyRp5syZioiI0DvvvKNnnnlGZcqUUWhoqB544AE1b968IKc13ypVqqTvv/9eo0aN0rRp03Tu3Dk1aNBAn3/+ue655x57Oy8vL61atUrDhg3TtGnT5OXlpT59+igqKkrt27d32Gbr1q21ceNGvfjii3rrrbd0+vRpVa5cWU2aNNGgQYOK9HgAALCiqVOnytXVVXFxcTp37pyaN2+ulStXKjIy0tmhZUPuABQ/NlOUI9kAAAAAAAAUIsbIAAAAAAAAlkEhAwAAAAAAWAaFDAAAAAAAYBkUMgAAAAAAgGVQyAAAAAAAAJZBIQMAAAAAAFhGGWcHUNQyMzN1+PBh+fr6ymazOTscAACKFWOMTp06pSpVqsjFhd83ihI5CQAAubuSnKTEFzIOHz6s4OBgZ4cBAECx9vvvv6tq1arODqNEIycBAODy8pOTlPhChq+vr6SLJ8PPz8/J0QAAULykpKQoODjY/vcSRYecBACA3F1JTlLiCxlZXTf9/PxIGgAAyAW3OhQ9chIAAC4vPzkJN8MCAAAAAADLoJABAAAAAAAsg0IGAAAAAACwDKeOkTFjxgzNmDFDBw8elCTVq1dPY8eOVVRUlCSpdevWWrduncM6gwYN0syZM691qAAACzDG6MKFC8rIyHB2KMWGq6urypQpwxgYAABcQ+QkOStbtqxcXV2vejtOLWRUrVpVr776qmrWrCljjObOnatOnTpp27ZtqlevniTpkUce0QsvvGBfx8vLy1nhAgCKsfT0dCUlJenMmTPODqXY8fLyUlBQkNzc3JwdCgAAJR45Se5sNpuqVq0qHx+fq9qOUwsZHTt2dJh++eWXNWPGDP3www/2QoaXl5cqV67sjPAAABaRmZmphIQEubq6qkqVKnJzc6MHgi7+GpSenq6//vpLCQkJqlmzplxcuKsUAICiQk6SO2OM/vrrL/3xxx+qWbPmVfXMKDaPX83IyNCSJUuUmpqqpk2b2ufHxcXpf//7nypXrqyOHTvq+eefz7NXRlpamtLS0uzTKSkpRRo3AMD50tPTlZmZqeDgYHru/Yunp6fKli2rQ4cOKT09XR4eHs4OCQCAEoucJG8VK1bUwYMHdf78+asqZDj9Z5mdO3fKx8dH7u7uevTRR7V06VLVrVtXknT//ffrf//7n9asWaMxY8Zo/vz5euCBB/LcXmxsrPz9/e2v4ODga3EYAIBigN4GOeO8XN6MGTPUoEED+fn5yc/PT02bNtXXX39tX966dWvZbDaH16OPPurEiAEAxRl/e3NWWL1TnN4jo3bt2tq+fbuSk5P14Ycfqm/fvlq3bp3q1q2rgQMH2tuFh4crKChIbdq0UXx8vKpXr57j9saMGaOYmBj7dEpKCsUMAACQJ8btAgDAOpxeyHBzc1ONGjUkSREREdq8ebOmTp2qd955J1vbJk2aSJIOHDiQayHD3d1d7u7uRRcwAAAocRi3CwAA6yh2/V0yMzMdxri41Pbt2yVJQUFB1zAiAABQmmRkZGjhwoU5jttVoUIF1a9fX2PGjGE0egAAnMSpPTLGjBmjqKgoVatWTadOndKCBQu0du1aLV++XPHx8VqwYIHuvvtuBQQEaMeOHRo5cqRatmypBg0aODNsAABQAu3cuVNNmzbVuXPn5OPjk23crpCQEFWpUkU7duzQqFGjtG/fPn388ce5bo8ByAEAKBpOLWQcO3ZM0dHRSkpKkr+/vxo0aKDly5erXbt2+v3337Vy5UpNmTJFqampCg4OVteuXfXcc885M2QAAFBCFfa4XbGxsZowYcK1Ch8AgFLDqbeWvPfeezp48KDS0tJ07NgxrVy5Uu3atZMkBQcHa926dTpx4oTOnTun/fv36/XXX5efn58zQwYAlHIffvihwsPD5enpqYCAALVt21apqamSpP/+97+qU6eOPDw8dOONN+rtt9+2r/fQQw+pQYMG9l/o09PTdfPNNys6Otopx4HsssbtioiIUGxsrBo2bKipU6fm2PbScbtyM2bMGCUnJ9tfv//+e5HEDQAonUpzTlLsxsgAAKC4SkpKUu/evfXQQw9pz549Wrt2rbp06SJjjOLi4jR27Fi9/PLL2rNnj1555RU9//zzmjt3riTpzTffVGpqqkaPHi1JevbZZ3Xy5Em99dZbzjwk5OFqx+1yd3e3P8416wUAQGEo7TmJ059aAsBR4gvhzg7hilUbu9PZIQDXRFJSki5cuKAuXbooJCRE0sXbDCRp3LhxmjRpkrp06SJJCgsL0y+//KJ33nlHffv2lY+Pj/73v/+pVatW8vX11ZQpU7RmzRr+cVtMMG4XkJ0VcxKJvASlQ2nPSShkAACQTw0bNlSbNm0UHh6uyMhI3XXXXerWrZvc3NwUHx+vhx9+WI888oi9/YULF+Tv72+fbtq0qZ588km9+OKLGjVqlG6//XZnHAZywLhdAAArKe05CYUMAADyydXVVStWrND333+vb775RtOmTdOzzz6rzz//XJL07rvv2sdOuHSdLJmZmdqwYYNcXV3zHFsB1957772X67KscbsAACguSntOwhgZAABcAZvNpubNm2vChAnatm2b3NzctGHDBlWpUkW//fabatSo4fAKCwuzrztx4kTt3btX69at07JlyzR79mwnHgkAALCy0pyT0CMDAIB8+vHHH7Vq1SrdddddCgwM1I8//qi//vpLderU0YQJEzR8+HD5+/urffv2SktL05YtW/TPP/8oJiZG27Zt09ixY/Xhhx+qefPmmjx5sh5//HG1atVKN9xwg7MPDQAAWEhpz0koZAAAkE9+fn769ttvNWXKFKWkpCgkJESTJk1SVFSUJMnLy0sTJ07UU089JW9vb4WHh2vEiBE6d+6cHnjgAfXr108dO3aUJA0cOFBffvmlHnzwQX377bcO3T0BAADyUtpzEpsxxjg7iKKUkpIif39/JScnW2oUVpReVhwhnNHB4Wznzp1TQkKCwsLC5OHh4exwip28zg9/J68dzjWsxoo5iUReAuciJ8lbYeUkjJEBAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACyjjLMDAACgqEQ8Ne+a7m/rxOgCrTd9+nRNnDhRR44cUcOGDTVt2jQ1btw41/ZLlizR888/r4MHD6pmzZp67bXXdPfddxc0bAAAcA2QlxQeemQAAOBEixYtUkxMjMaNG6effvpJDRs2VGRkpI4dO5Zj+++//169e/fWww8/rG3btqlz587q3Lmzdu3adY0jBwAAJY1V8hIKGQAAONHkyZP1yCOPqH///qpbt65mzpwpLy8vvf/++zm2nzp1qtq3b6+nnnpKderU0YsvvqhbbrlFb7311jWOHAAAlDRWyUsoZAAA4CTp6enaunWr2rZta5/n4uKitm3bauPGjTmus3HjRof2khQZGZlrewAAgPywUl5CIQMAACc5fvy4MjIyVKlSJYf5lSpV0pEjR3Jc58iRI1fUHgAAID+slJdQyAAAAAAAAJZBIQMAACepUKGCXF1ddfToUYf5R48eVeXKlXNcp3LlylfUHgAAID+slJdQyAAAwEnc3NwUERGhVatW2edlZmZq1apVatq0aY7rNG3a1KG9JK1YsSLX9gAAAPlhpbykTJFuHQAA5CkmJkZ9+/bVrbfeqsaNG2vKlClKTU1V//79JUnR0dG6/vrrFRsbK0l6/PHH1apVK02aNEn33HOPFi5cqC1btmjWrFnOPAwAAFACWCUvoZABAIAT9ezZU3/99ZfGjh2rI0eO6KabbtKyZcvsA2clJibKxeX/OlA2a9ZMCxYs0HPPPadnnnlGNWvW1CeffKL69es76xAAAEAJYZW8xGaMMUW6BydLSUmRv7+/kpOT5efn5+xwgMtKfCHc2SFcsWpjdzo7BJRy586dU0JCgsLCwuTh4eHscIqdvM4PfyevHc41rMaKOYlEXgLnIifJW2HlJIyRAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACyDQgYAAAAAALAMChkAAAAAAMAyKGQAAAAAAADLoJABAAAAAAAsg0IGAAAAAACwjDLODgAAgKKS+EL4Nd1ftbE7r3idb7/9VhMnTtTWrVuVlJSkpUuXqnPnznmus3btWsXExGj37t0KDg7Wc889p379+hUsaAAAcE0U97zESjkJPTIAAHCi1NRUNWzYUNOnT89X+4SEBN1zzz264447tH37do0YMUIDBgzQ8uXLizhSAABQklkpJ6FHBgAAThQVFaWoqKh8t585c6bCwsI0adIkSVKdOnX03Xff6Y033lBkZGRRhQkAAEo4K+Uk9MgAAMBCNm7cqLZt2zrMi4yM1MaNG50UEQAAKI2cmZM4tZAxY8YMNWjQQH5+fvLz81PTpk319ddf25efO3dOQ4YMUUBAgHx8fNS1a1cdPXrUiREDAOBcR44cUaVKlRzmVapUSSkpKTp79qyTogIAAKWNM3MSpxYyqlatqldffVVbt27Vli1bdOedd6pTp07avXu3JGnkyJH6/PPPtWTJEq1bt06HDx9Wly5dnBkyAAAAAABwIqeOkdGxY0eH6ZdfflkzZszQDz/8oKpVq+q9997TggULdOedd0qSZs+erTp16uiHH37Qbbfd5oyQAQBwqsqVK2frnXj06FH5+fnJ09PTSVEBAIDSxpk5SbEZIyMjI0MLFy5UamqqmjZtqq1bt+r8+fMO99zceOONqlatGvcBAwBKraZNm2rVqlUO81asWKGmTZs6KSIAAFAaOTMncXohY+fOnfLx8ZG7u7seffRRLV26VHXr1tWRI0fk5uamcuXKObSvVKmSjhw5kuv20tLSlJKS4vACAKC4On36tLZv367t27dLuvgos+3btysxMVGSNGbMGEVHR9vbP/roo/rtt9/09NNPa+/evXr77be1ePFijRw50hnhlxiM2wUAKO2slJM4vZBRu3Ztbd++XT/++KMee+wx9e3bV7/88kuBtxcbGyt/f3/7Kzg4uBCjBQCgcG3ZskU333yzbr75ZklSTEyMbr75Zo0dO1aSlJSUZE8gJCksLExffvmlVqxYoYYNG2rSpEn673//y6NXrxLjdgEASjsr5SQ2Y4wp8r1cgbZt26p69erq2bOn2rRpo3/++cehV0ZISIhGjBiRa5UnLS1NaWlp9umUlBQFBwcrOTlZfn5+RR0+cNUSXwh3dghXrNrYnc4OAaXcuXPnlJCQoLCwMHl4eDg7nGInr/OTkpIif39//k7moHz58po4caK6deumihUrasGCBerWrZskae/evapTp442btyY73G7ONewGivmJBJ5CZyLnCRvhZWTOL1Hxr9lZmYqLS1NERERKlu2rMM9N/v27VNiYmKe99y4u7vbu4VmvQAAAPKrsMbt4nZXAACKhlOfWjJmzBhFRUWpWrVqOnXqlBYsWKC1a9dq+fLl8vf318MPP6yYmBiVL19efn5+GjZsmJo2bcoTSwAAQKHbuXOnmjZtqnPnzsnHx8c+btf27dsLNG5XbGysJkyYUMRRAwBQ+ji1kHHs2DFFR0crKSlJ/v7+atCggZYvX6527dpJkt544w25uLioa9euSktLU2RkpN5++21nhgwAAEqorHG7kpOT9eGHH6pv375at25dgbc3ZswYxcTE2KezbncFAABXx6mFjPfeey/P5R4eHpo+fbqmT59+jSICAACllZubm2rUqCFJioiI0ObNmzV16lT17NlT6enpOnnypEOvjKNHj6py5cq5bs/d3V3u7u5FHTYAAKVOsRsjAwAAoDi42nG7AABA0XBqjwwAAApTMXsQV7HBebk8xu0CABQm/vbmrLDOC4UMAIDllS1bVpJ05swZeXp6Ojma4ufMmTOS/u88ITvG7QIAFAZykrylp6dLklxdXa9qOxQyAACW5+rqqnLlyunYsWOSJC8vL9lsNidH5XzGGJ05c0bHjh1TuXLlrjppKMkYtwsAUBjISXKXmZmpv/76S15eXipT5upKERQyAAAlQtagi1mJA/5PuXLl8hyUEgAAFB5ykty5uLioWrVqV13coZABACgRbDabgoKCFBgYqPPnzzs7nGKjbNmy9MQAAOAaIifJnZubm1xcrv6ZIxQyAAAliqurK/9wBwAATkdOUnR4/CoAAAAAALAMChkAAAAAAMAyKGQAAAAAAADLoJABAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACyDQgYAAAAAALAMChkAAAAAAMAyKGQAAAAAAADLoJABAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACyDQgYAAAAAALAMChkAAAAAAMAyKGQAAAAAAADLoJABAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACzDqYWM2NhYNWrUSL6+vgoMDFTnzp21b98+hzatW7eWzWZzeD366KNOihgAAAAAADiTUwsZ69at05AhQ/TDDz9oxYoVOn/+vO666y6lpqY6tHvkkUeUlJRkf73++utOihgAAAAAADhTGWfufNmyZQ7Tc+bMUWBgoLZu3aqWLVva53t5ealy5crXOjwAAAAAAFDMFKsxMpKTkyVJ5cuXd5gfFxenChUqqH79+hozZozOnDmT6zbS0tKUkpLi8AIAAMgLt7sCAGAdTu2RcanMzEyNGDFCzZs3V/369e3z77//foWEhKhKlSrasWOHRo0apX379unjjz/OcTuxsbGaMGHCtQobAACUAFm3uzZq1EgXLlzQM888o7vuuku//PKLvL297e0eeeQRvfDCC/ZpLy8vZ4QLAECpVmwKGUOGDNGuXbv03XffOcwfOHCg/f/Dw8MVFBSkNm3aKD4+XtWrV8+2nTFjxigmJsY+nZKSouDg4KILHAAAWB63uwIAYB3F4taSoUOH6osvvtCaNWtUtWrVPNs2adJEknTgwIEcl7u7u8vPz8/hBQAAcCW43RUAgOLLqT0yjDEaNmyYli5dqrVr1yosLOyy62zfvl2SFBQUVMTRAQCA0ojbXQEAKN6cWsgYMmSIFixYoE8//VS+vr46cuSIJMnf31+enp6Kj4/XggULdPfddysgIEA7duzQyJEj1bJlSzVo0MCZoQMAgBKK210BACjenFrImDFjhqSLo4Bfavbs2erXr5/c3Ny0cuVKTZkyRampqQoODlbXrl313HPPOSFaAABQ0mXd7vrtt99e0e2uORUy3N3d5e7uXiRxAgBQmjn91pK8BAcHa926ddcoGgAAUFpxuysAANZRbJ5aAgAA4Czc7goAgHVQyAAAAKUet7sCAGAdFDIAAECpx+2uAABYh4uzAwAAAAAAAMgvChkAAAAAAMAyKGQAAAAAAADLoJABAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACyDQgYAAAAAALAMChkAAAAAAMAyKGQAAAAAAADLoJABAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACyDQgYAAAAAALAMChkAAAAAAMAyKGQAAAAAAADLoJABAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACyDQgYAAAAAALAMChkAAAAAAMAyKGQAAAAAAADLcGohIzY2Vo0aNZKvr68CAwPVuXNn7du3z6HNuXPnNGTIEAUEBMjHx0ddu3bV0aNHnRQxAAAAAABwJqcWMtatW6chQ4bohx9+0IoVK3T+/HndddddSk1NtbcZOXKkPv/8cy1ZskTr1q3T4cOH1aVLFydGDQAAAAAAnKWMM3e+bNkyh+k5c+YoMDBQW7duVcuWLZWcnKz33ntPCxYs0J133ilJmj17turUqaMffvhBt912mzPCBgAAAAAATlKgHhk33HCDTpw4kW3+yZMndcMNNxQ4mOTkZElS+fLlJUlbt27V+fPn1bZtW3ubG2+8UdWqVdPGjRsLvB8AAFByFEZewu2uAABYR4EKGQcPHlRGRka2+Wlpafrzzz8LFEhmZqZGjBih5s2bq379+pKkI0eOyM3NTeXKlXNoW6lSJR05ciTH7aSlpSklJcXhBQAASq7CyEu43RUAAOu4oltLPvvsM/v/L1++XP7+/vbpjIwMrVq1SqGhoQUKZMiQIdq1a5e+++67Aq2fJTY2VhMmTLiqbQAAgOKvMPMSbncFAMA6rqiQ0blzZ0mSzWZT3759HZaVLVtWoaGhmjRp0hUHMXToUH3xxRf69ttvVbVqVfv8ypUrKz09XSdPnnTolXH06FFVrlw5x22NGTNGMTEx9umUlBQFBwdfcUywvoin5jk7hAJZ6uvsCADAGooqL5Gu/HbXnAoZaWlpSktLs0/TSxQAgMJxRYWMzMxMSVJYWJg2b96sChUqXNXOjTEaNmyYli5dqrVr1yosLMxheUREhMqWLatVq1apa9eukqR9+/YpMTFRTZs2zXGb7u7ucnd3v6q4AABA8VfYecml2y2M213pJQoAQNEo0FNLEhISCmXnQ4YM0YIFC/Tpp5/K19fXngj4+/vL09NT/v7+evjhhxUTE6Py5cvLz89Pw4YNU9OmTenCCQAAJBVeXpKlsG53pZcoAABFo8CPX121apVWrVqlY8eO2X8RyfL+++/naxszZsyQJLVu3dph/uzZs9WvXz9J0htvvCEXFxd17dpVaWlpioyM1Ntvv13QsAEAQAlUGHmJVLi3u9JLFFm43RUACleBChkTJkzQCy+8oFtvvVVBQUGy2WwF2rkx5rJtPDw8NH36dE2fPr1A+wAAACVbYeQlRXG7KwAAKBoFKmTMnDlTc+bM0YMPPljY8QAAAFyRwshLuN0VAADrKFAhIz09Xc2aNSvsWAAAAK5YYeQl3O4KAIB1uBRkpQEDBmjBggWFHQsAAMAVK4y8xBiT4yuriCH93+2uf//9t1JTU/Xxxx/nOj4GAAAoOgXqkXHu3DnNmjVLK1euVIMGDVS2bFmH5ZMnTy6U4AAAAC6HvAQAgNKlQIWMHTt26KabbpIk7dq1y2FZQQf+BAAAKAjyEgAASpcCFTLWrFlT2HEAAAAUCHkJAAClS4HGyMhy4MABLV++XGfPnpWUv8epAgAAFAXyEgAASocCFTJOnDihNm3aqFatWrr77ruVlJQkSXr44Yf1xBNPFGqAAAAAeSEvAQCgdClQIWPkyJEqW7asEhMT5eXlZZ/fs2dPLVu2rNCCAwAAuBzyEgAASpcCjZHxzTffaPny5apatarD/Jo1a+rQoUOFEhgAAEB+kJcAAFC6FKhHRmpqqsMvHln+/vtvubu7X3VQAAAA+UVeAgBA6VKgQkaLFi00b948+7TNZlNmZqZef/113XHHHYUWHAAAwOWQlwAAULoU6NaS119/XW3atNGWLVuUnp6up59+Wrt379bff/+tDRs2FHaMAAAAuSIvAQCgdClQj4z69evr119/1e23365OnTopNTVVXbp00bZt21S9evXCjhEAACBX5CUAAJQuBeqRIUn+/v569tlnCzMWAACAAiEvAQCg9ChQj4zZs2dryZIl2eYvWbJEc+fOveqgAAAA8ou8BACA0qVAhYzY2FhVqFAh2/zAwEC98sorVx0UAABAfpGXAABQuhSokJGYmKiwsLBs80NCQpSYmHjVQQEAAOQXeQkAAKVLgQoZgYGB2rFjR7b5P//8swICAq46KAAAgPwiLwEAoHQpUCGjd+/eGj58uNasWaOMjAxlZGRo9erVevzxx9WrV6/CjhEAACBX5CUAAJQuBXpqyYsvvqiDBw+qTZs2KlPm4iYyMzMVHR3NvagAAOCaIi8BAKB0ueJChjFGR44c0Zw5c/TSSy9p+/bt8vT0VHh4uEJCQooiRgAAgByRlwAAUPoUqJBRo0YN7d69WzVr1lTNmjWLIi4AAIDLIi8BAKD0ueIxMlxcXFSzZk2dOHGiKOIBAADIN/ISAABKnwIN9vnqq6/qqaee0q5duwo7HgAAgCtCXgIAQOlSoME+o6OjdebMGTVs2FBubm7y9PR0WP73338XSnAAAACXQ14CAEDpUqBCxpQpUwo5DAAAgIIhLwEAoHQpUCGjb9++hR0HAABAgZCXAABQuhRojAxJio+P13PPPafevXvr2LFjkqSvv/5au3fvLrTgAAAA8oO8BACA0qNAhYx169YpPDxcP/74oz7++GOdPn1akvTzzz9r3LhxhRogAABAXshLAAAoXQpUyBg9erReeuklrVixQm5ubvb5d955p3744YdCCw4AAOByyEsAAChdClTI2Llzp+67775s8wMDA3X8+PGrDgoAACC/yEsAAChdClTIKFeunJKSkrLN37Ztm66//vqrDgoAACC/yEsAAChdClTI6NWrl0aNGqUjR47IZrMpMzNTGzZs0JNPPqno6OjCjhEAACBX5CUAAJQuBSpkvPLKK7rxxhsVHBys06dPq27dumrRooWaNWum5557rrBjBAAAyBV5CQAApUuZgqzk5uamd999V2PHjtXOnTuVmpqqm2++WTVq1Cjs+AAAAPJEXgIAQOlSoEKGJL333nt64403tH//fklSzZo1NWLECA0YMKDQggMAAMgP8hIAAEqPAhUyxo4dq8mTJ2vYsGFq2rSpJGnjxo0aOXKkEhMT9cILLxRqkAAAALkhLwEAoHQpUCFjxowZevfdd9W7d2/7vHvvvVcNGjTQsGHDSBgAAMA1Q14CAEDpUqDBPs+fP69bb7012/yIiAhduHDhqoMCAADIL/ISAABKlwIVMh588EHNmDEj2/xZs2apT58+Vx0UAABAfpGXAABQulzVYJ/ffPONbrvtNknSjz/+qMTEREVHRysmJsbebvLkyVcfJQAAQB7ISwAAKD0KVMjYtWuXbrnlFklSfHy8JKlChQqqUKGCdu3aZW9ns9kKIUQAAIDckZcAAFC6FKiQsWbNmkLZ+bfffquJEydq69atSkpK0tKlS9W5c2f78n79+mnu3LkO60RGRmrZsmWFsn8AAGB9hZWXAAAAayjQGBmFJTU1VQ0bNtT06dNzbdO+fXslJSXZXx988ME1jBAAAAAAABQnTi1kREVF6aWXXtJ9992Xaxt3d3dVrlzZ/rruuuuuYYQAAKC0+Pbbb9WxY0dVqVJFNptNn3zyicPyfv36yWazObzat2/vnGABACjFnFrIyI+1a9cqMDBQtWvX1mOPPaYTJ07k2T4tLU0pKSkOLwAAgMuhpygAANZQ4KeWXAvt27dXly5dFBYWpvj4eD3zzDOKiorSxo0b5erqmuM6sbGxmjBhwjWOFAAAWF1UVJSioqLybJPVUxQAADhPsS5k9OrVy/7/4eHhatCggapXr661a9eqTZs2Oa4zZswYh8espaSkKDg4uMhjBQAAJV9WT9HrrrtOd955p1566SUFBATk2DYtLU1paWn2aXqJAgBQOIr9rSWXuuGGG1ShQgUdOHAg1zbu7u7y8/NzeAEAAFyt9u3ba968eVq1apVee+01rVu3TlFRUcrIyMixfWxsrPz9/e0vflgBAKBwFOseGf/2xx9/6MSJEwoKCnJ2KACQp8QXwp0dQoFUG7vT2SEAxdaV9hSllygAAEXDqYWM06dPO/SuSEhI0Pbt21W+fHmVL19eEyZMUNeuXVW5cmXFx8fr6aefVo0aNRQZGenEqAEAABx7iuZUyHB3d5e7u7sTIgMAoGRzaiFjy5YtuuOOO+zTWb9a9O3bVzNmzNCOHTs0d+5cnTx5UlWqVNFdd92lF198kaQAAAA4HT1FAQBwDqcWMlq3bi1jTK7Lly9ffg2jAQAApRk9RQEAsAZLjZEBAABQVOgpCqCksuLYXYzbhbxQyAAAABA9RQEAsApLPX4VAAAAAACUbhQyAAAAAACAZVDIAAAAAAAAlkEhAwAAAAAAWAaFDAAAAAAAYBkUMgAAAAAAgGVQyAAAAAAAAJZBIQMAAAAAAFgGhQwAAAAAAGAZFDIAAAAAAIBlUMgAAAAAAACWQSEDAAAAAABYBoUMAAAAAABgGRQyAAAAAACAZVDIAAAAAAAAlkEhAwAAAAAAWAaFDAAAAAAAYBkUMgAAAAAAgGVQyAAAAAAAAJZBIQMAAAAAAFgGhQwAAAAAAGAZFDIAAAAAAIBlUMgAAAAAAACWUcbZAaBwJL4Q7uwQCqTa2J3ODgEAAAAAYCH0yAAAAAAAAJZBIQMAAAAAAFgGhQwAAAAAAGAZFDIAAAAAAIBlUMgAAAAAAACWwVNLAAAAUKis+DQ1nqQGANZBjwwAAAAAAGAZFDIAAAAAAIBlUMgAAAAAAACWQSEDAAAAAABYBoUMAAAAAABgGRQyAAAAAACAZVDIAAAAAAAAlkEhAwAAAAAAWAaFDAAAAAAAYBkUMgAAAAAAgGU4tZDx7bffqmPHjqpSpYpsNps++eQTh+XGGI0dO1ZBQUHy9PRU27ZttX//fucECwAAAAAAnM6phYzU1FQ1bNhQ06dPz3H566+/rjfffFMzZ87Ujz/+KG9vb0VGRurcuXPXOFIAAAAAAFAcOLWQERUVpZdeekn33XdftmXGGE2ZMkXPPfecOnXqpAYNGmjevHk6fPhwtp4bAAAAV4ueogAAWEOxHSMjISFBR44cUdu2be3z/P391aRJE23cuNGJkQEAgJKInqIAAFhDGWcHkJsjR45IkipVquQwv1KlSvZlOUlLS1NaWpp9OiUlpWgCBAAAJUpUVJSioqJyXPbvnqKSNG/ePFWqVEmffPKJevXqdS1DBQCgVCu2PTIKKjY2Vv7+/vZXcHCws0MCAAAWV5CeomlpaUpJSXF4AQCAq1dse2RUrlxZknT06FEFBQXZ5x89elQ33XRTruuNGTNGMTEx9umUlJQrKmZEPDXvyoMtBpb6OjsCAABKroL0FI2NjdWECROKPDYAAEqbYlvICAsLU+XKlbVq1Sp74SIlJUU//vijHnvssVzXc3d3l7u7+zWKEgAAIGdX++OKxA8sAADkxKmFjNOnT+vAgQP26YSEBG3fvl3ly5dXtWrVNGLECL300kuqWbOmwsLC9Pzzz6tKlSrq3Lmz84IGAAClTkF6ivLjCgAARcOphYwtW7bojjvusE9n/WrRt29fzZkzR08//bRSU1M1cOBAnTx5UrfffruWLVsmDw8PZ4UMAABKoYL2FAUAAIXPqYWM1q1byxiT63KbzaYXXnhBL7zwwjWMCgAAlEb0FAUAwBqK7RgZAAAA1xI9RQEAsAYKGQAAAKKnKAAAVuHi7AAAAAAAAADyi0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACyDQgYAAAAAALAMChkAAAAAAMAyKGQAAAAAAADLoJABAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACyDQgYAAAAAALAMChkAAAAAAMAyKGQAAAAAAADLoJABAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACyDQgYAAAAAALCMMs4OAAAAAAAA5C3xhXBnh1Ag1cbuLPRt0iMDAAAAAABYBj0yAADIgxV//SiKXz4AAACKC3pkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDKKdSFj/PjxstlsDq8bb7zR2WEBAAAAAAAnKdaFDEmqV6+ekpKS7K/vvvvO2SEBAIBSiB9YAAAoHor941fLlCmjypUrOzsMAAAA1atXTytXrrRPlylT7FMpAABKnGL/13f//v2qUqWKPDw81LRpU8XGxqpatWq5tk9LS1NaWpp9OiUl5VqECQAASgF+YAEAwPmK9a0lTZo00Zw5c7Rs2TLNmDFDCQkJatGihU6dOpXrOrGxsfL397e/goODr2HEAACgJMv6geWGG25Qnz59lJiYmGvbtLQ0paSkOLwAAMDVK9aFjKioKHXv3l0NGjRQZGSkvvrqK508eVKLFy/OdZ0xY8YoOTnZ/vr999+vYcQAAKCkutIfWPhxBQCAolHsby25VLly5VSrVi0dOHAg1zbu7u5yd3e/hlEBAIDSICoqyv7/DRo0UJMmTRQSEqLFixfr4YcfztZ+zJgxiomJsU+npKRQzAAAoBAU6x4Z/3b69GnFx8crKCjI2aEAAIBS7nI/sLi7u8vPz8/hBQAArl6xLmQ8+eSTWrdunQ4ePKjvv/9e9913n1xdXdW7d29nhwYAAEo5fmABAMA5ivWtJX/88Yd69+6tEydOqGLFirr99tv1ww8/qGLFis4ODQAAlDJPPvmkOnbsqJCQEB0+fFjjxo3jBxYAAJygWBcyFi5c6OwQAAAAJPEDCwAAxUWxLmQAAAAUF/zAAgBA8VCsx8gAAAAAAAC4FIUMAAAAAABgGRQyAAAAAACAZVDIAAAAAAAAlkEhAwAAAAAAWAaFDAAAAAAAYBkUMgAAAAAAgGVQyAAAAAAAAJZBIQMAAAAAAFgGhQwAAAAAAGAZFDIAAAAAAIBlUMgAAAAAAACWQSEDAAAAAABYBoUMAAAAAABgGRQyAAAAAACAZVDIAAAAAAAAlkEhAwAAAAAAWAaFDAAAAAAAYBllnB0AAOQl4ql5zg6hQJb6OjsCAAAAoGSikAEAAAAA+cAPLEDxwK0lAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACyDQgYAAAAAALAMChkAAAAAAMAyKGQAAAAAAADLoJABAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDLKODsAAEDpEPHUPGeHUCBLfZ0dAQAAKEzkJNZHjwwAAAAAAGAZFDIAAAAAAIBlUMgAAAAAAACWYYlCxvTp0xUaGioPDw81adJEmzZtcnZIAACglCIvAQDAuYp9IWPRokWKiYnRuHHj9NNPP6lhw4aKjIzUsWPHnB0aAAAoZchLAABwvmJfyJg8ebIeeeQR9e/fX3Xr1tXMmTPl5eWl999/39mhAQCAUoa8BAAA5yvWhYz09HRt3bpVbdu2tc9zcXFR27ZttXHjRidGBgAAShvyEgAAiocyzg4gL8ePH1dGRoYqVarkML9SpUrau3dvjuukpaUpLS3NPp2cnCxJSklJydc+M9LOFjBa5zpVNsPZIRRIfq9LQXAtrx2uY3ZWvI4S1zInVryWV3Ids9oaY4oqnBLjSvOSq81JJD431xLff9lZ8TpKXMucWPFach2zs+J1lPJ/La8kJynWhYyCiI2N1YQJE7LNDw4OdkI01059ZwdQULH+zo6g2LHkteQ6ZmPJ6yhxLXNgyWtZgOt46tQp+ftz/QtTac1JpNLzuSnpLHkdJa5lDix5LbmO2VjyOkpXfC3zk5MU60JGhQoV5OrqqqNHjzrMP3r0qCpXrpzjOmPGjFFMTIx9OjMzU3///bcCAgJks9mKNF5nSUlJUXBwsH7//Xf5+fk5OxxcBa5lycB1LDlKw7U0xujUqVOqUqWKs0Mp9q40LymNOYlUOj43pQHXseTgWpYMpeE6XklOUqwLGW5uboqIiNCqVavUuXNnSReTgFWrVmno0KE5ruPu7i53d3eHeeXKlSviSIsHPz+/EvumLm24liUD17HkKOnXkp4Y+XOleUlpzkmkkv+5KS24jiUH17JkKOnXMb85SbEuZEhSTEyM+vbtq1tvvVWNGzfWlClTlJqaqv79+zs7NAAAUMqQlwAA4HzFvpDRs2dP/fXXXxo7dqyOHDmim266ScuWLcs20BYAAEBRIy8BAMD5in0hQ5KGDh2a660kuNh1ddy4cdm6r8J6uJYlA9ex5OBaIifkJXnjc1MycB1LDq5lycB1dGQzPG8NAAAAAABYhIuzAwAAAAAAAMgvChkAAAAAAMAyKGQAAAAAAADLoJBRiObMmWP558PbbDZ98sknebbp16+fOnfufE3iQdHiWhZP+f0uyc/n1cpKwnfq1Vi7dq1sNptOnjzp7FBgUSXhM0ReUrpwLYsfcpKLSsL36dUojjkJhYx/6devn2w2W7bXgQMHnB3aNZGUlKSoqChJ0sGDB2Wz2bR9+3aHNlOnTtWcOXOufXD5lNP1u/Q1fvx4Z4d4zVn1Wl6qNF3Xnj176tdff7VPjx8/XjfddFO2dpd+Xour0v6dml+tW7fWiBEjHOY1a9ZMSUlJ8vf3d05QKBZK+2eIvKRksuq1zFKarik5SeljlZzEEo9fvdbat2+v2bNnO8yrWLGik6K5tipXrnzZNsXpDZyTpKQk+/8vWrRIY8eO1b59++zzfHx87P9vjFFGRobKlCmdH4Xifi0vVZquq6enpzw9PS/bLj+f1+KgNH+nXg03NzfLXGMUrdL8GSIvKV2K+7XMUpquKTkJpOKZk9AjIwfu7u6qXLmyw8vV1VWTJ09WeHi4vL29FRwcrMGDB+v06dO5bufnn3/WHXfcIV9fX/n5+SkiIkJbtmyxL//uu+/UokULeXp6Kjg4WMOHD1dqamqu28uqgL7zzjsKDg6Wl5eXevTooeTkZHubzMxMvfDCC6patarc3d110003admyZfbl6enpGjp0qIKCguTh4aGQkBDFxsbal1/aLSwsLEySdPPNN8tms6l169aSHLv9zZo1S1WqVFFmZqZDrJ06ddJDDz1kn/700091yy23yMPDQzfccIMmTJigCxcu5HqsV+PS6+bv7y+bzWaf3rt3r3x9ffX1118rIiJC7u7u+u6773LsyjhixAj7MUsXz21sbKzCwsLk6emphg0b6sMPP8wzltDQUL3yyit66KGH5Ovrq2rVqmnWrFkObX7//Xf16NFD5cqVU/ny5dWpUycdPHjQvvzChQsaPny4ypUrp4CAAI0aNUp9+/Z1iHfZsmW6/fbb7W06dOig+Ph4+3KrXstLFbfr+uKLL6p3797y9vbW9ddfr+nTpzu0SUxMVKdOneTj4yM/Pz/16NFDR48etS/P6/vh0u6Lc+bM0YQJE/Tzzz/bfznI+rXq0s9rs2bNNGrUKIcY/vrrL5UtW1bffvutJCktLU1PPvmkrr/+enl7e6tJkyZau3ZtPs7+1Snu36nz589XaGio/P391atXL506dcreJj/vj88++0w1a9aUh4eH7rjjDs2dO9eh++WJEyfUu3dvXX/99fLy8lJ4eLg++OAD+/r9+vXTunXrNHXqVPs1PnjwoEM3zpSUFHl6eurrr7922PfSpUvl6+urM2fOSLr89wmsqbh/hshL8lbc/n6Rl1y94nZNyUnyr7h/n5KT5B+FjCvg4uKiN998U7t379bcuXO1evVqPf3007m279Onj6pWrarNmzdr69atGj16tMqWLStJio+PV/v27dW1a1ft2LFDixYt0nfffaehQ4fmGcOBAwe0ePFiff7551q2bJm2bdumwYMH25dPnTpVkyZN0n/+8x/t2LFDkZGRuvfee7V//35J0ptvvqnPPvtMixcv1r59+xQXF6fQ0NAc97Vp0yZJ0sqVK5WUlKSPP/44W5vu3bvrxIkTWrNmjX3e33//rWXLlqlPnz6SpPXr1ys6OlqPP/64fvnlF73zzjuaM2eOXn755TyPtSiNHj1ar776qvbs2aMGDRrka53Y2FjNmzdPM2fO1O7duzVy5Eg98MADWrduXZ7rTZo0Sbfeeqv9Wj322GP2qv358+cVGRkpX19frV+/Xhs2bJCPj4/at2+v9PR0SdJrr72muLg4zZ49Wxs2bFBKSkq2exBTU1MVExOjLVu2aNWqVXJxcdF9991n/+Nfkq/lpa7ldZ04caIaNmyobdu2afTo0Xr88ce1YsUKSRf/0HTq1El///231q1bpxUrVui3335Tz5497evn9f1wqZ49e+qJJ55QvXr1lJSUpKSkJIftXLq9hQsXyhhjn7do0SJVqVJFLVq0kCQNHTpUGzdu1MKFC7Vjxw51795d7du3t38/XGvF4Ts1Pj5en3zyib744gt98cUXWrdunV599VX78su9PxISEtStWzd17txZP//8swYNGqRnn33WYR/nzp1TRESEvvzyS+3atUsDBw7Ugw8+aP9cTp06VU2bNtUjjzxiv8bBwcEO2/Dz81OHDh20YMECh/lxcXHq3LmzvLy88vV9gpKlOHyGyEsKB3lJybmWWchJyEnISYo4JzFw0LdvX+Pq6mq8vb3tr27duuXYdsmSJSYgIMA+PXv2bOPv72+f9vX1NXPmzMlx3YcfftgMHDjQYd769euNi4uLOXv2bI7rjBs3zri6upo//vjDPu/rr782Li4uJikpyRhjTJUqVczLL7/ssF6jRo3M4MGDjTHGDBs2zNx5550mMzMzx31IMkuXLjXGGJOQkGAkmW3btjm06du3r+nUqZN9ulOnTuahhx6yT7/zzjumSpUqJiMjwxhjTJs2bcwrr7zisI358+eboKCgHGMoTP++JmvWrDGSzCeffOLQ7t/HZIwxjz/+uGnVqpUxxphz584ZLy8v8/333zu0efjhh03v3r1z3X9ISIh54IEH7NOZmZkmMDDQzJgxwxhz8TzUrl3b4XqkpaUZT09Ps3z5cmOMMZUqVTITJ060L79w4YKpVq1atngv9ddffxlJZufOncaYknEtL1Ucrmv79u0d5vXs2dNERUUZY4z55ptvjKurq0lMTLQv3717t5FkNm3aZIzJ+/vh38c3btw407Bhw2ztLv28Hjt2zJQpU8Z8++239uVNmzY1o0aNMsYYc+jQIePq6mr+/PNPh220adPGjBkzJtdjvVrF/TvVy8vLpKSk2Oc99dRTpkmTJsaY/L0/Ro0aZerXr++w/NlnnzWSzD///JPjfo0x5p577jFPPPGEfbpVq1bm8ccfd2iT9b7O2s7SpUuNj4+PSU1NNcYYk5ycbDw8PMzXX39tjMnf9wmsp7h/hshLrkxx+PtFXlK4isM1JSfJn+L+fUpOcmWsebNWEbvjjjs0Y8YM+7S3t7eki1Xj2NhY7d27VykpKbpw4YLOnTunM2fOyMvLK9t2YmJiNGDAAM2fP19t27ZV9+7dVb16dUkXuyPt2LFDcXFx9vbGGGVmZiohIUF16tTJMbZq1arp+uuvt083bdpUmZmZ2rdvn7y8vHT48GE1b97cYZ3mzZvr559/lnSxu1C7du1Uu3ZttW/fXh06dNBdd91VwDN1UZ8+ffTII4/o7bfflru7u+Li4tSrVy+5uLjYj3XDhg0O1fGMjIw8z11Ru/XWW6+o/YEDB3TmzBm1a9fOYX56erpuvvnmPNe9tAqf1fXw2LFjki6emwMHDsjX19dhnXPnzik+Pl7Jyck6evSoGjdubF/m6uqqiIgIh66W+/fv19ixY/Xjjz/q+PHj9mWJiYmqX79+vo/TitfyUtfyujZt2jTb9JQpUyRJe/bsUXBwsEMFu27duipXrpz27NmjRo0a5fn9UBAVK1bUXXfdpbi4OLVo0UIJCQnauHGj3nnnHUnSzp07lZGRoVq1ajmsl5aWpoCAgALvNz+K83dqaGiow+cvKCjI/vnMz/tj3759atSokcPySz+v0sXPyCuvvKLFixfrzz//VHp6utLS0q7483L33XerbNmy+uyzz9SrVy999NFH8vPzU9u2be3nIK/vE1hXcf4MkZcUDvKSnFnxWmYhJyEnIScp2pyEQkYOvL29VaNGDYd5Bw8eVIcOHfTYY4/p5ZdfVvny5fXdd9/p4YcfVnp6eo4Xf/z48br//vv15Zdf6uuvv9a4ceO0cOFC3XfffTp9+rQGDRqk4cOHZ1uvWrVqRXZst9xyixISEvT1119r5cqV6tGjh9q2bXvZ++/y0rFjRxlj9OWXX6pRo0Zav3693njjDfvy06dPa8KECerSpUu2dT08PAq836uR9aWVxcXFxaH7m3Sxe2WWrHvkvvzyS4eETbp4r11e/t01z2az2f+gnz59WhEREQ5fdFmuZOChjh07KiQkRO+++679ftL69etfcdctK17LS13L63q18vp+KKg+ffpo+PDhmjZtmhYsWKDw8HCFh4dLunisrq6u2rp1q1xdXR3Wu3RQsqJQnL9TL/f5lK7+/TFx4kRNnTpVU6ZMsd9/O2LEiCv+fLq5ualbt25asGCBevXqpQULFqhnz572AeQK6/sExU9x/gxdLfKSi8hLct+O1a5lFnISchJykqLNSShk5NPWrVuVmZmpSZMm2avAixcvvux6tWrVUq1atTRy5Ej17t1bs2fP1n333adbbrlFv/zyS7YP0uUkJibq8OHDqlKliiTphx9+kIuLi2rXri0/Pz9VqVJFGzZsUKtWrezrbNiwwaEa5+fnp549e6pnz57q1q2b2rdvr7///lvly5d32Jebm5uki5W7vHh4eKhLly6Ki4vTgQMHVLt2bd1yyy325bfccov27dt3xcd6LVWsWFG7du1ymLd9+3b7F0rdunXl7u6uxMREh3N7tW655RYtWrRIgYGB8vPzy7FNpUqVtHnzZrVs2VLSxevx008/2R99deLECe3bt0/vvvuu/b7D7777zmEbpelaXqoor+sPP/yQbTqrwl6nTh39/vvv+v333+2/gPzyyy86efKk6tata18nt++Hf3Nzc7vstZMuDoA2cOBALVu2TAsWLFB0dLR92c0336yMjAwdO3bM/j5xpuLynZqX/Lw/ateura+++sph3ubNmx2mN2zYoE6dOumBBx6QdPF+5V9//dXhvZDfa9ynTx+1a9dOu3fv1urVq/XSSy/Zl+Xn+wQlR3H5DJGXFA3ykotKwrXMQk5CTnI1yEmyY7DPfKpRo4bOnz+vadOm6bffftP8+fM1c+bMXNufPXtWQ4cO1dq1a3Xo0CFt2LBBmzdvtn+pjBo1St9//72GDh2q7du3a//+/fr0008vOwiMh4eH+vbtq59//lnr16/X8OHD1aNHD/vjcJ566im99tprWrRokfbt26fRo0dr+/btevzxxyVJkydP1gcffKC9e/fq119/1ZIlS1S5cmX7aMSXCgwMlKenp5YtW6ajR486jEL+b3369NGXX36p999/3z4AU5axY8dq3rx5mjBhgnbv3q09e/Zo4cKFeu655/I81mvpzjvv1JYtWzRv3jzt379f48aNc/hj4+vrqyeffFIjR47U3LlzFR8fr59++knTpk3T3LlzC7zfPn36qEKFCurUqZPWr1+vhIQErV27VsOHD9cff/whSRo2bJhiY2P16aefat++fXr88cf1zz//yGazSZKuu+46BQQEaNasWTpw4IBWr16tmJgYh/2Upmt5qaK8rhs2bNDrr7+uX3/9VdOnT9eSJUvsn7O2bdsqPDxcffr00U8//aRNmzYpOjparVq10q233nrZ74d/Cw0NVUJCgrZv367jx48rLS0tx3be3t7q3Lmznn/+ee3Zs0e9e/e2L6tVq5b69Omj6Ohoffzxx0pISNCmTZsUGxurL7/88kpP7VUrLt+pecnP+2PQoEHau3evRo0apV9//VWLFy92GMFdkmrWrKkVK1bo+++/1549ezRo0CCH0eKli9f4xx9/1MGDBx26Yf9by5YtVblyZfXp00dhYWFq0qSJfVl+vk9QchSXzxB5SdEgL3GMycrXMgs5CTkJOUkh5ySFMtJGCZLTQDxZJk+ebIKCgoynp6eJjIw08+bNcxj05NJBYNLS0kyvXr1McHCwcXNzM1WqVDFDhw51GOBl06ZNpl27dsbHx8d4e3ubBg0aZBsQ61JZg+u8/fbbpkqVKsbDw8N069bN/P333/Y2GRkZZvz48eb66683ZcuWNQ0bNrQPumKMMbNmzTI33XST8fb2Nn5+fqZNmzbmp59+si/XJQP1GGPMu+++a4KDg42Li4t9MKKczlFGRoYJCgoykkx8fHy22JctW2aaNWtmPD09jZ+fn2ncuLGZNWtWrsdaWHIbgCmnAW/Gjh1rKlWqZPz9/c3IkSPN0KFD7cdszMUBsaZMmWJq165typYtaypWrGgiIyPNunXrct1/SEiIeeONNxzmNWzY0IwbN84+nZSUZKKjo02FChWMu7u7ueGGG8wjjzxikpOTjTHGnD9/3gwdOtT4+fmZ6667zowaNcp0797d9OrVy76NFStWmDp16hh3d3fToEEDs3bt2hJ3LS9VHK7rhAkTTPfu3Y2Xl5epXLmymTp1qkObQ4cOmXvvvdd4e3sbX19f0717d3PkyBFjzOW/H/59fOfOnTNdu3Y15cqVM5LM7NmzjTHZP6/GGPPVV18ZSaZly5bZ4k5PTzdjx441oaGhpmzZsiYoKMjcd999ZseOHbke69Wywnfqpd544w0TEhJin87P++PTTz81NWrUMO7u7qZ169ZmxowZRpI9thMnTphOnToZHx8fExgYaJ577jkTHR3tcF727dtnbrvtNuPp6WkkmYSEhFzf108//bSRZMaOHZvtmC73fQLrscJniLwk/4rD3y/yksJVHK4pOUn+WOH79FLkJHmzGfOvm7VQbI0fP16ffPKJtm/f7uxQ4ESZmZmqU6eOevTooRdffNHZ4ZRKoaGhGjFihEaMGOHsUFAMvfzyy5o5c6Z+//13Z4cCFCnyEkjkJc5GToK8lOSchDEygGLu0KFD+uabb9SqVSulpaXprbfeUkJCgu6//35nhwZA0ttvv61GjRopICBAGzZs0MSJE6+q+ygAFGfkJUDxVZpyEgoZQDHn4uKiOXPm6Mknn5QxRvXr19fKlStzvXcRwLW1f/9+vfTSS/r7779VrVo1PfHEExozZoyzwwKAIkFeAhRfpSkn4dYSAAAAAABgGTy1BAAAAAAAWAaFDAAAAAAAYBkUMgAAAAAAgGVQyAAAAAAAAJZBIQMo4Q4ePCibzabt27cX+b7mzJmjcuXKOcybNWuWgoOD5eLioilTpmj8+PG66aabijyW0NBQTZkypcj3AwAA8oecBEBh4aklQAl38OBBhYWFadu2bUX+x/rs2bM6deqUAgMDJUkpKSmqUKGCJk+erK5du8rf31+ZmZlKS0tTQEBAoexzzpw5GjFihE6ePOkw/6+//pK3t7e8vLwKZT8AAODqkJOQkwCFpYyzAwBQcnh6esrT09M+nZiYqPPnz+uee+5RUFCQfb6Pj0+Rx1KxYsUi3wcAACieyEmAko1bS4ASIjMzU6+//rpq1Kghd3d3VatWTS+//HK2dhkZGXr44YcVFhYmT09P1a5dW1OnTnVos3btWjVu3Fje3t4qV66cmjdvrkOHDkmSfv75Z91xxx3y9fWVn5+fIiIitGXLFkmO3TjnzJmj8PBwSdINN9wgm82mgwcP5tiN8/3331e9evXk7u6uoKAgDR061L5s8uTJCg8Pl7e3t4KDgzV48GCdPn3aHmf//v2VnJwsm80mm82m8ePHS8rejTMxMVGdOnWSj4+P/Pz81KNHDx09etS+PCuu+fPnKzQ0VP7+/urVq5dOnTp15RcDAIBSjJyEnAQoahQygBJizJgxevXVV/X888/rl19+0YIFC1SpUqVs7TIzM1W1alUtWbJEv/zyi8aOHatnnnlGixcvliRduHBBnTt3VqtWrbRjxw5t3LhRAwcOlM1mkyT16dNHVatW1ebNm7V161aNHj1aZcuWzbafnj17auXKlZKkTZs2KSkpScHBwdnazZgxQ0OGDNHAgQO1c+dOffbZZ6pRo4Z9uYuLi958803t3r1bc+fO1erVq/X0009Lkpo1a6YpU6bIz89PSUlJSkpK0pNPPpnjMXfq1El///231q1bpxUrVui3335Tz549HdrFx8frk08+0RdffKEvvvhC69at06uvvprfSwAAAEROQk4CXAMGgOWlpKQYd3d38+6772ZblpCQYCSZbdu25br+kCFDTNeuXY0xxpw4ccJIMmvXrs2xra+vr5kzZ06Oy2bPnm38/f3t09u2bTOSTEJCgn3euHHjTMOGDe3TVapUMc8++2zuB/cvS5YsMQEBAbnuM0tISIh54403jDHGfPPNN8bV1dUkJibal+/evdtIMps2bbLH5eXlZVJSUuxtnnrqKdOkSZN8xwYAQGlHTuKfrR05CVD46JEBlAB79uxRWlqa2rRpk6/206dPV0REhCpWrCgfHx/NmjVLiYmJkqTy5curX79+ioyMVMeOHTV16lQlJSXZ142JidGAAQPUtm1bvfrqq4qPjy9w3MeOHdPhw4fzjHvlypVq06aNrr/+evn6+urBBx/UiRMndObMmXzvZ8+ePQoODnb49aVu3boqV66c9uzZY58XGhoqX19f+3RQUJCOHTt2hUcFAEDpRU6SN3ISoHBQyABKgEsHs7qchQsX6sknn9TDDz+sb775Rtu3b1f//v2Vnp5ubzN79mxt3LhRzZo106JFi1SrVi398MMPki7et7l7927dc889Wr16terWraulS5cWSdwHDx5Uhw4d1KBBA3300UfaunWrpk+fLkkO8RaWf3dHtdlsyszMLPT9AABQUpGTFA5yEiBvFDKAEqBmzZry9PTUqlWrLtt2w4YNatasmQYPHqybb75ZNWrUyPEXjJtvvlljxozR999/r/r162vBggX2ZbVq1dLIkSP1zTffqEuXLpo9e3aB4vb19VVoaGiucW/dulWZmZmaNGmSbrvtNtWqVUuHDx92aOPm5qaMjIw891OnTh39/vvv+v333+3zfvnlF508eVJ169YtUOwAACA7chJyEuBaoJABlAAeHh4aNWqUnn76ac2bN0/x8fH64Ycf9N5772VrW7NmTW3ZskXLly/Xr7/+queff16bN2+2L09ISNCYMWO0ceNGHTp0SN98843279+vOnXq6OzZsxo6dKjWrl2rQ4cOacOGDdq8ebPq1KlT4NjHjx+vSZMm6c0339T+/fv1008/adq0aZKkGjVq6Pz585o2bZp+++03zZ8/XzNnznRYPzQ0VKdPn9aqVat0/PjxHLt3tm3bVuHh4erTp49++uknbdq0SdHR0WrVqpVuvfXWAscOAAAckZOQkwDXAoUMoIR4/vnn9cQTT2js2LGqU6eOevbsmeO9lIMGDVKXLl3Us2dPNWnSRCdOnNDgwYPty728vLR371517dpVtWrV0sCBAzVkyBANGjRIrq6uOnHihKKjo1WrVi316NFDUVFRmjBhQoHj7tu3r6ZMmaK3335b9erVU4cOHbR//35JUsOGDTV58mS99tprql+/vuLi4hQbG+uwfrNmzfToo4+qZ8+eqlixol5//fVs+7DZbPr000913XXXqWXLlmrbtq1uuOEGLVq0qMBxAwCAnJGTkJMARc1mjDHODgIAAAAAACA/6JEBAAAAAAAsg0IGAAAAAACwDAoZAAAAAADAMihkAAAAAAAAy6CQAQAAAAAALINCBgAAAAAAsAwKGQAAAAAAwDIoZAAAAAAAAMugkAEAAAAAACyDQgYAAAAAALAMChkAAAAAAMAyKGQAAAAAAADL+H/uYuujZhB44wAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    }
  ]
}