BiasCorrection.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from collections import defaultdict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set plots style\n",
    "plt.style.use('seaborn-v0_8-paper')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Ground Truth Simulation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Confusion matrix values\n",
    "TP = 273\n",
    "FP = 37\n",
    "FN = 52\n",
    "TN = 1190\n",
    "\n",
    "# Total number of samples\n",
    "total_samples = TP + FP + FN + TN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1552"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total_samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sensitivity (Recall): 0.84\n",
      "Specificity: 0.97\n",
      "Precision: 0.88\n",
      "Negative Predictive Value: 0.96\n"
     ]
    }
   ],
   "source": [
    "# Conditional probabilities\n",
    "sensitivity = TP / (TP + FN)\n",
    "specificity = TN / (TN + FP)\n",
    "precision = TP / (TP + FP)\n",
    "negative_predictive_value = TN / (TN + FN)\n",
    "\n",
    "print(f\"Sensitivity (Recall): {sensitivity:.2f}\")\n",
    "print(f\"Specificity: {specificity:.2f}\")\n",
    "print(f\"Precision: {precision:.2f}\")\n",
    "print(f\"Negative Predictive Value: {negative_predictive_value:.2f}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True War Proportion: 0.21\n",
      "True Not War Proportion: 0.79\n"
     ]
    }
   ],
   "source": [
    "# Number of simulations\n",
    "num_simulations = 10000\n",
    "size = 1000\n",
    "\n",
    "# Proportion of true war and not war labels\n",
    "true_war_proportion = (TP + FN) / total_samples\n",
    "true_not_war_proportion = (TN + FP) / total_samples\n",
    "\n",
    "print(f\"True War Proportion: {true_war_proportion:.2f}\")\n",
    "print(f\"True Not War Proportion: {true_not_war_proportion:.2f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# original_samples = []\n",
    "# for sample in range (num_simulations):\n",
    "ground_truth = np.random.choice(\n",
    "    [1, 0], size=size, p=[true_war_proportion, true_not_war_proportion]\n",
    "    )\n",
    "# original_samples.append(ground_truth)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "predicted_samples = []\n",
    "for simulation in range(num_simulations):\n",
    "    # Simulate the classifier predictions with errors\n",
    "    predicted_labels = []\n",
    "    for label in ground_truth:\n",
    "        if label == 1:  # True war\n",
    "            predicted_label = np.random.choice(\n",
    "                [1, 0], p=[(sensitivity), 1 - sensitivity]\n",
    "            )\n",
    "        else:  # True not war\n",
    "            predicted_label = np.random.choice(\n",
    "                [0, 1], p=[specificity, 1 - specificity]\n",
    "            )\n",
    "        predicted_labels.append(predicted_label)\n",
    "\n",
    "    predicted_labels = np.array(predicted_labels)\n",
    "    predicted_samples.append(predicted_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "187"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "np.sum(predicted_samples[3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True War Proportion: 0.2094\n",
      "Predicted War Proportion: 0.1988\n",
      "True Not War Proportion: 0.7906\n",
      "Predicted Not War Proportion: 0.8012\n"
     ]
    }
   ],
   "source": [
    "# Calculate proportions from the predicted labels\n",
    "predicted_war_proportion = np.sum(predicted_samples) / (num_simulations*size)\n",
    "predicted_not_war_proportion = 1 - predicted_war_proportion\n",
    "\n",
    "# Print the proportions\n",
    "print(f\"True War Proportion: {true_war_proportion:.4f}\")\n",
    "print(f\"Predicted War Proportion: {predicted_war_proportion:.4f}\")\n",
    "print(f\"True Not War Proportion: {true_not_war_proportion:.4f}\")\n",
    "print(f\"Predicted Not War Proportion: {predicted_not_war_proportion:.4f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Set plots style\n",
    "plt.style.use('seaborn-v0_8-paper')\n",
    "\n",
    "# Set the size of the chart\n",
    "plt.figure(figsize=(6.4, 4))\n",
    "\n",
    "# Count the number of 1's in each sample for predicted samples\n",
    "modified_count_of_ones = [sum(sample) for sample in predicted_samples]\n",
    "\n",
    "default_colours = plt.rcParams['axes.prop_cycle'].by_key()['color']\n",
    "\n",
    "# Calculate means for both distributions\n",
    "mean_original = np.sum(ground_truth)\n",
    "mean_predicted = np.mean(modified_count_of_ones)\n",
    "\n",
    "# Plot histograms of the counts of 1's\n",
    "# plt.hist(original_count_of_ones, bins=range(0, 51, 1), alpha=0.5, label='Original Sampling', edgecolor='black')\n",
    "plt.hist(modified_count_of_ones, alpha=0.5, edgecolor='black', color=default_colours[1])\n",
    "\n",
    "# Add vertical lines for the means\n",
    "plt.axvline(mean_original, label='Ground Truth', color='black', linestyle='dashed', linewidth=1)\n",
    "plt.axvline(mean_predicted, label='Mean Predicted', color=default_colours[1], linestyle='dashed', linewidth=1)\n",
    "\n",
    "# Add text for the means\n",
    "plt.text(mean_original+0.5, plt.ylim()[1] * 0.98, f'{mean_original:.0f}', color='black')\n",
    "plt.text(mean_predicted+0.5, plt.ylim()[1] * 0.98, f'{mean_predicted:.0f}', color=default_colours[1])\n",
    "\n",
    "# Add titles and labels\n",
    "#plt.title('Predictions distribution')\n",
    "plt.xlabel('Number of 1s (war)')\n",
    "plt.ylabel('Frequency')\n",
    "# plt.xticks(range(0, 51, 5))\n",
    "plt.yticks(range(0, 3600, 500))\n",
    "\n",
    "# Add legend\n",
    "plt.legend(loc='upper right')\n",
    "\n",
    "plt.savefig(\n",
    "    'report_charts/bias_correction_hist1.pdf',\n",
    "    bbox_inches='tight',\n",
    ")\n",
    "\n",
    "# Show the plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Correcting for the bias"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Using conditional probabilities to correct for bias\n",
    "corrected_samples = []\n",
    "\n",
    "for predicted_labels in predicted_samples:\n",
    "    corrected_labels = []\n",
    "    for label in predicted_labels:\n",
    "        if label == 1:  # Predicted war\n",
    "            corrected_label = np.random.choice(\n",
    "                [1, 0], p=[precision, 1 - precision]\n",
    "            )\n",
    "        else:  # Predicted not war\n",
    "            corrected_label = np.random.choice(\n",
    "                [0, 1], p=[negative_predictive_value, 1 - negative_predictive_value]\n",
    "            )\n",
    "        corrected_labels.append(corrected_label)\n",
    "    \n",
    "    corrected_labels = np.array(corrected_labels)\n",
    "    corrected_samples.append(corrected_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(6.4, 4))\n",
    "\n",
    "# Count the number of 1's in each sample for predicted samples\n",
    "modified_count_of_ones = [sum(sample) for sample in predicted_samples]\n",
    "corrected_count_of_ones = [np.sum(sample) for sample in corrected_samples]\n",
    "\n",
    "# Calculate mean\n",
    "mean_corrected = np.mean(corrected_count_of_ones)\n",
    "mean_original = np.sum(ground_truth)\n",
    "mean_predicted = np.mean(modified_count_of_ones)\n",
    "\n",
    "default_colours = plt.rcParams['axes.prop_cycle'].by_key()['color']\n",
    "\n",
    "\n",
    "# Plot histograms of the counts of 1's\n",
    "plt.hist(corrected_count_of_ones, alpha=0.5, edgecolor='black',color=default_colours[0] )\n",
    "plt.hist(modified_count_of_ones, alpha=0.5, edgecolor='black', color=default_colours[1] )\n",
    "\n",
    "# Add vertical lines for the means\n",
    "plt.axvline(mean_original, label='Ground Truth', color='black', linestyle='dashed', linewidth=1)\n",
    "plt.axvline(mean_predicted, label='Mean Predicted', color=default_colours[1], linestyle='dashed', linewidth=1)\n",
    "plt.axvline(mean_corrected, label='Mean Corrected', color=default_colours[0], linestyle='dashed', linewidth=1)\n",
    "\n",
    "# Add text for the means\n",
    "plt.text(mean_original -5, plt.ylim()[1] * 0.98, f'{mean_original:.0f}')\n",
    "plt.text(mean_predicted -5, plt.ylim()[1] * 0.98, f'{mean_predicted:.0f}', color=default_colours[1])\n",
    "plt.text(mean_corrected +1, plt.ylim()[1] * 0.98, f'{mean_corrected:.0f}', color=default_colours[0])\n",
    "\n",
    "# Add titles and labels\n",
    "#plt.title('Predictions distribution')\n",
    "plt.xlabel('Number of 1s (war)')\n",
    "plt.ylabel('Frequency')\n",
    "# plt.xticks(range(0, 51, 5))\n",
    "plt.yticks(range(0, 3600, 500))\n",
    "\n",
    "# Add legend\n",
    "plt.legend(loc='upper right')\n",
    "\n",
    "plt.savefig(\n",
    "    'report_charts/bias_correction_hist2.pdf',\n",
    "    bbox_inches='tight',\n",
    ")\n",
    "\n",
    "# Show the plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Correcting real predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load predictions\n",
    "a3_preds = pd.read_csv('predictions/atres_predictions.csv')\n",
    "la6_preds = pd.read_csv('predictions/la6_predictions.csv')\n",
    "t5_preds = pd.read_csv('predictions/telecinco_predictions.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create list with the different list of predictions for channels\n",
    "predicted_real = [a3_preds['predicted_class'].tolist(), la6_preds['predicted_class'].tolist(), t5_preds['predicted_class'].tolist()]\n",
    "\n",
    "# Using conditional probabilities to correct for bias\n",
    "corrected_real = []\n",
    "\n",
    "for predicted_labels in predicted_real:\n",
    "    corrected_labels = []\n",
    "    for label in predicted_labels:\n",
    "        if label == 1:  # Predicted war\n",
    "            corrected_label = np.random.choice(\n",
    "                [1, 0], p=[precision, 1 - precision]\n",
    "            )\n",
    "        else:  # Predicted not war\n",
    "            corrected_label = np.random.choice(\n",
    "                [0, 1], p=[negative_predictive_value, 1 - negative_predictive_value]\n",
    "            )\n",
    "        corrected_labels.append(corrected_label)\n",
    "    \n",
    "    corrected_labels = np.array(corrected_labels)\n",
    "    corrected_real.append(corrected_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Add new predictions to dataframe\n",
    "a3_preds['corrected_preds'] = corrected_real[0]\n",
    "la6_preds['corrected_preds'] = corrected_real[1]\n",
    "t5_preds['corrected_preds'] = corrected_real[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save predictions with bias corrections\n",
    "a3_preds.to_csv('predictions/a3_bias_corrected.csv', index=False)\n",
    "la6_preds.to_csv('predictions/la6_bias_corrected.csv', index=False)\n",
    "t5_preds.to_csv('predictions/t5_bias_corrected.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Error Bars Plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load predictions\n",
    "a3_preds = pd.read_csv('predictions/atres_predictions.csv')\n",
    "la6_preds = pd.read_csv('predictions/la6_predictions.csv')\n",
    "t5_preds = pd.read_csv('predictions/telecinco_predictions.csv')\n",
    "\n",
    "a3_preds_corrected = pd.read_csv('predictions/a3_bias_corrected.csv')\n",
    "la6_preds_corrected = pd.read_csv('predictions/la6_bias_corrected.csv')\n",
    "t5_preds_corrected = pd.read_csv('predictions/t5_bias_corrected.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_errorbars_df(predictions, corrected):\n",
    "    # Group data by year_month\n",
    "    grouped = predictions.groupby('year_month')\n",
    "\n",
    "    # Dictionary to hold the counts of class 1 for each month\n",
    "    monthly_corrected_counts = defaultdict(list)\n",
    "\n",
    "    # Iterate over each month\n",
    "    for month, group in grouped:\n",
    "        predicted_labels = group['predicted_class'].values\n",
    "        total_count = len(predicted_labels)\n",
    "        total_class_one_count = group['predicted_class'].sum()\n",
    "        for _ in range(1000):\n",
    "            corrected_labels = []\n",
    "            for label in predicted_labels:\n",
    "                if label == 1:  # Predicted war\n",
    "                    corrected_label = np.random.choice([1, 0], p=[precision, 1 - precision])\n",
    "                else:  # Predicted not war\n",
    "                    corrected_label = np.random.choice([0, 1], p=[negative_predictive_value, 1 - negative_predictive_value])\n",
    "                corrected_labels.append(corrected_label)\n",
    "            corrected_count = sum(corrected_labels)\n",
    "            monthly_corrected_counts[month].append(corrected_count)\n",
    "\n",
    "    # Calculate the 10th and 90th quantiles of the corrected counts for each month\n",
    "    monthly_quantiles = {month: (np.percentile(counts, 10), np.percentile(counts, 90)) for month, counts in monthly_corrected_counts.items()}\n",
    "    \n",
    "    \n",
    "    grouped_corrected = corrected.groupby('year_month')\n",
    "    # Group the original dataframe by month and calculate the required statistics\n",
    "    results = []\n",
    "    for month, group in grouped_corrected:\n",
    "        total_count = len(group)\n",
    "        total_class_one_count = group['corrected_preds'].sum()\n",
    "        share_class_one = total_class_one_count / total_count\n",
    "        quantile_10th, quantile_90th = monthly_quantiles[month]\n",
    "        share_class_one_10th = quantile_10th / total_count\n",
    "        share_class_one_90th = quantile_90th / total_count\n",
    "        min_class_one_count = min(monthly_corrected_counts[month])\n",
    "        max_class_one_count = max(monthly_corrected_counts[month])\n",
    "        share_class_one_min = min_class_one_count / total_count\n",
    "        share_class_one_max = max_class_one_count / total_count\n",
    "\n",
    "        results.append({\n",
    "            'year_month': month,\n",
    "            'total_count': total_count,\n",
    "            'total_class_one_count': total_class_one_count,\n",
    "            'share_class_one': share_class_one,\n",
    "            'count_class_one_10th': quantile_10th,\n",
    "            'share_class_one_10th': share_class_one_10th,\n",
    "            'count_class_one_90th': quantile_90th,\n",
    "            'share_class_one_90th': share_class_one_90th,\n",
    "            'min_class_one_count': min_class_one_count,\n",
    "            'share_class_one_min': share_class_one_min,\n",
    "            'max_class_one_count': max_class_one_count,\n",
    "            'share_class_one_max': share_class_one_max\n",
    "        })\n",
    "\n",
    "    # Convert to DataFrame for easy plotting or further analysis\n",
    "    results_df = pd.DataFrame(results)\n",
    "    return results_df\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "a3_results = generate_errorbars_df(a3_preds, a3_preds_corrected)\n",
    "la6_results = generate_errorbars_df(la6_preds, la6_preds_corrected)\n",
    "t5_results = generate_errorbars_df(t5_preds, t5_preds_corrected)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjEAAAFxCAYAAACC3/kFAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/H5lhTAAAACXBIWXMAAA9hAAAPYQGoP6dpAADtTUlEQVR4nOy9d3hj93Wn/9570UEA7HU45HAKZ0aaXjSqVrctuch2ZMvaJI6zsdZJvElsx7Gd3WTtZOPsJv65ZBPbSZTiHlmWbVnNUR2V0VRN75W9F/R6y++PS4AACdZhAUff93n4gAQuwe8FwHvPPedzPkcyDMNAIBAIBAKBYIkhL/YCBAKBQCAQCGaDCGIEAoFAIBAsSUQQIxAIBAKBYEkighiBQCAQCARLEhHECAQCgUAgWJKIIEYgEAgEAsGSRAQxAoFAIBAIliQiiBEIBAKBQLAksSz2AuYLXdfp6urC4/EgSdJiL0cgEAgEAsE0MAyDUChEbW0tsjx5ruWaDWK6urqor69f7GUIBAKBQCCYBe3t7SxbtmzSba7ZIMbj8QDmi+D1ehd5NVdPKpXi+eef595778VqtS72cuaMa3W/srnW9/Fa37801/p+Xuv7B9f2Pl5L+xYMBqmvr8+cxyfjmg1i0iUkr9d7zQQxLpcLr9e75D+g2Vyr+5XNtb6P1/r+pbnW9/Na3z+4tvfxWty36UhBhLBXIBAIBALBkkQEMQKBQCAQCJYkIogRCAQCgUCwJBFBjEAgEAgEgiWJCGIEAoFAIBAsSUQQIxAIBAKBYEkighiBQCAQCARLEhHECAQCgUAgWJKIIEYgEAgEAsGSRAQxAoFAIBAIliQiiBEIBAKBQLAkEUGMQCAQCASCmaOlFnsFIogRCAQCgUAwC9TEYq9ABDECgUAgEAhmgQhiBAKBQCAQLEk0EcQIBAKBQCBYiohMjEAgEAgEgiWHroOuLvYqRBAjEAgEAoFghmgJMIzFXoUIYgQCgUAgEMyQAiglgQhiBAKBQCAQzBQRxAgEAoFAIFiSFEBnEoggRiAQCAQCwUxRk4u9AkAEMQKBQCAQCGaKyMQIBAKBQCBYcug6aCITIxAIBAKBYKlRIFkYEEGMYJFJqvpiL0EgEAgEM0FNQDwIL30ZDj66qEsRQcxSYLgV5QcPUBK5uNgrmXPiqrbYSxAIBALBTNCS4G+Ds0/DqV8s6lJEELMUOPpD5NY3aOp/frFXMufEU2YmJimCGYFAIFgaqAkItJvfl61a1KWIIGYp0PkWwLWZiUmawUskKYIYgUAgWBKoCQh0mN+LIEYwKYYBnYcBcCcHINy7yAuaW9LlpGhCBDECgUCwJNBEJkYwXYZbIDaU+VHqOLR4a5ljVE1H1cwBYuHk4k9DFQgEAsEUGIapiRGZGMG06DKzMIZkvlVS58HFXM2cEs/qTNI0g3hKZGMEAoGgoFETYOhmECMpUNKwqMsRQUyhM1JKMhpvBUDquHaCmNgYHUwkIbIxAoFAUNBoCYj0m9kYXx0o1kVdjghiCp2RIEZf+150SUHqPlowMyuulrGZl4jQxQgEAkFhoybAP6KHKW1a3LUggpjCRlOh+yhICkbldQScDUhaAnpOLPbK5oTYmCAmLDIxAoFAUNhkdyaVrlzctSCCmMJm4BykolC6Aix2htwjAqqOA4u7rjlA141xbr2aLnQxAoFAUNBoydHOpPLVi7sWRBBT2Iz4w1CxDmA0iGnfv0gLmjviqoZhjL9fZGMEAoGggMnOxIycmxYTEcQUMiN6GCqbQbIy5B6JetuXvrg37dQ7FiHuFQgEggIl014tMjGC6ZCdiXGXEbeVYbjKIdgBwa7FXdtVMlYPk0ZkYgQCgaBAURNmEBPqAYsDPDWLvSIRxBQsqRj0nQbFbvbhO0sAMCrXm4+3L21dzNj26jS6PvFjAoFAIFhEtASEukyfmOIGkBc/hFj8FQjy03MCdBUq1oDVlenFNyqvMx9f4n4xkwl4RTZGIBAIChA1Cf50Z9Lit1eDCGIKl7QepmKtGcSMMJqJWbri3ngqv6g3jdDFCAQCQQGSMzNp8durQQQxhUtGD7MWbO7M3Ub5apCt0H3MrE8uQaZqow4nVIzJohyBQCAQLDzZnUnlzYu7lhFEEFOodKU7k3IzMVicZolJS5qBzBJkos6kNIYxsfBXIBAIBIuEmpWJqVy7uGsZQQQxhUjMD4MXwe4FT11uEGN1QVoXs0TFvdMJUIQuRiAQCAqIAptenUYEMYVI1xHztqIZbK5cBbjNDVUjupgl6tw7ne4jMUdJIBAICggtCckIRAfBUQyu0sVeEbCIQcy9997Lxo0b2bx5M7feeitHjpgn7gsXLnDTTTexZs0aduzYwalTpxZriYtHWg9TuS43CwMjQUw6E7OfSRWyBUhK09H0qdccEboYgUAgKBzUhOlRBlDSuKhLyWbRgpif/OQnHD9+nKNHj/KZz3yG3/qt3wLgv/23/8YjjzzC+fPn+fznP5+5/21FJhOTK+oFzKDGXQlFVabhUDq1t0SYrtbFMCAq/GIEAoGgMCiw6dVpLIv1h4uLizPfBwIBJEmir6+PQ4cO8fzzzwPwoQ99iE996lNcvHiRVasmr78lEgkSidFunWAwCEAqlSKVSs39Dswjlo5DSECqdA1gg6x9SKkqyA6UivXI4V7U1r0Y7upFXe9MCEcT6Nqo3iX9ffZ9aQKRODbZvmBrmy8y790S+xxOl2t9/9Jc6/t5re8fXNv7OO/7logi+9tRAK2kCX0eX8OZ7INkLGLO/jd/8zd55ZVXAHj22WdJJpM8/PDDnDt3LrPNzp07+T//5/9w5513TvpcX/rSl/jyl7887v4f/ehHuFyuPL9RmDiSQ7zz1B8RtZbywvXfmHC7pr7/ZEPnD7lUcS8nl/36wi1QIBAIBG9LtrZ8h/rhNznQ+Cm6S3bO29+JRqM8/PDDBAIBvF7vpNsuWiYG4Hvf+x4A3/3ud/n85z/PX/7lX876ub74xS/ymc98JvNzMBikvr6ee++9d8oXoZCQzj0Lp8BRdx337VoHpSsAMzJ94YUXuOeee7DqcaTzIej8ISss/Sy/775FXvX0udAbJqWNtljrmkrbiX0s37ALWcn9OEoSNFd5kGVpoZc5p+S8d1brYi9nzrnW9y/Ntb6f1/r+wbW9j/O+b/3nUNoHAdhy94NsSWsz54F0JWU6LGoQk+ZjH/sYn/zkJ1m2bBnd3d2oqorFYsEwDNra2li+fPmUz2G327Hbx5cerFbr0vqw9preL3LlOmSnB8as3Wq1YlXsZueSYkPuPYGMClbnYqx2Rmi6gYaMrIyXYsmKZVwQA5A0JDxL6f2bhCX3WZwh1/r+pbnW9/Na3z+4tvdxXvbNMMBIZYS91so1485Nc8lM1r8owl6/309X1+gU5l/84heUlZVRWVnJ1q1b+cEPfgDAE088wbJly6bUw1xT5Dj1TlAGkxVweM1ARleh6+iCLe9qmMqpNx+i1VogEAgWGS0J8WGzxbqoeuJz0yKwKJmYQCDAgw8+SCwWQ5ZlKioqePrpp5EkiX/8x3/kt37rt/jKV76C1+vl3/7t3xZjiYuDro90Jkkjgx/dE29rc5umdz0nTL+YhhsXbJmzZTYuvML0TiAQCBaZbKfe0sZFXcpYFiWIaWho4MCB/EZtzc3N7N27d4FXVCAMXYZ4AIqXg7MM8pRXMtiKsvxilobp3XRM7sYST2nourHkdTECgUCwZNGyp1cXVmVEOPYWEul5SRXrpk7X2YpGnXvbDywJ07uEOvMgxjAgnBTZGIFAIFg0sjMxBTJuII0IYgqJjB6mOaeUNDg4yL//+78TDodHt1Us4K0DTw1E+sDfusCLnRmGYUw5+HEiIqKkJBAIBIuHGh81Vq0ojMGPaUQQU0h0pidX52Zijhw5QldXF/39/bnb54wgOLhAi5wd8ZQ+62SRCGIEAoFgEdGSo5mYijWLu5YxiCCmUNBS0HMcZIuZrsuamdTW1gZAJBLJnSdkK4LKdElp/0KudsbMpjMpTSw5vXlLAoFAIJgHklEIdpqdscUNi72aHEQQUyj0nTZTdmUrzQmhkilkTaVSdHZ2AqCqKkNDQ6O/kyPuLewgZjadSdmILiWBQCBYBNQkRHrNC21f/eQNJ4uACGIKhXQpaYw/TFdXF7o+qiVJZ2UAsNhMEbBih95TZg9/gXK1QcxclpRUVcXv98/Z8wkEAsE1S7YepmTF4q4lDyKIKRSyTe7ylJKKiooAaG9vz/09pw8q14KhjU6/LkCuppwEcxfEpFIpent7CYfDRCKFG/QJBAJBQaAlR6dXF1hnEoggpnDIycSMdialg5gNGzZkfh6niylwv5iEqqHPrjEpQzylo2pX9ySxWIze3l4kSUKWZfx+P6oqylQCgUAwIWpiNBNTvnpx15IHEcQUAskI9J8x5x+VNoHFnAGl63om89LY2IjVaiUYDBIIBEZ/11ZkOvcCtO1b6JVPi3jyKiOYEa5mBEEkEmFwcBBZHv3IS5LE4ODgXCxNIBAIrk20LI+YinWLu5Y8iCCmEOg+DoYO5WvA7snc3dfXRyKRoLy8HKfTmSkptbZmecJYHVCz0fy+81BBmt7FZ2Fyl4/Zmt4FAgGGh4dzApg0qqrmBoUCgUAgGEVNikyMYAoyeph1eUtJVVVVAPmDGADfMvDWQnTQHF1QYMxm3EA+ZqOLGRoaIhQK5QQwhmFkxNKSJBEKhUilUnOyRoFAILimSIYg3GtWCjzVi72acYggphBIjxuozBX1pktJFRUVALjdZoAzLoixFUHV9eb3HYVnene1nUlpEimd1DR1MYZh0N/fTzQazQQwhmFw9uxZvv/97/Pcc89lAhlZlhkcHMzVGgkEAsHbHTUJgU6zUlDcmLH+KCREEFMIdBwyb7NEvYZhZIKVdCbGbrfjcDgy2YUM2eLeAtPFqJqOqs1dcDCdbIyu6/T29pJMJjMBzPDwME899RSvv/46iUSCrq4uzp49O7pO0XYtEAgEuWTrYUqbFnctEyCCmMUmOmTOPXIUm06IsgKYOo5QKITH48Hr9QJm6aOmpgYYq4txjmZiCsz0bq6yMGmmMr1TVZWenh50XUeSJFRV5eDBg/zsZz+jt7eXoqIitm3bBsDBgweJxWKAmY2JRCIkEok5Xa9AIBAsWbL1MAXYXg0iiFl8OrNKSXn0MNXVuTXIdFYmJ4iRJKjeABYH9J+FRJhCYa6DmMk6lBKJBD09PZmfOzo6eOKJJzh69CiGYbBp0yYefPBBtm7dysqVK0kmk+zfPxr0ybLMwMCAKCsJBAIBjBjdjWRiygtrZlIaEcQsNl2T+8NUVlbmbJ4OasbpYhw+c3CkoY8KhQuAxCwnV09EUtVJquOfMxaL0d/fjyzLRKNRXn75ZZ577jmCwSCVlZV88IMfZOfOnVgspmX2DTfcgNVq5cKFC3R3d+c8l2i7Xhokk8nFXoJAcG2jZXnEVBbW9Oo0IohZbLL1MHmcetOZl7QItaSkBJvNlhGtZsieaF1AupipMjEnOgO82CmhzyD7MVYXEwqFGBwcRJIkzpw5w+OPP86lS5ew2WzccsstvO9976O0tDTnd9xuN9u3bwdgz549Od1K8XhcuPkWOKlUiu7u7lxtmEAgmFvUbLfelYu7lgkQQcxiYhhZnUnrTW0LEI1G6e/vx263jzv5ZuticuYoWd1ZupjCCGJ03Zg0E2MYBv/vlcs81aZwuM0/7efN1sX4/X4CgQB+v5+nnnqKN954g2QyycqVK3nwwQdZt24d0gSK+vXr11NaWsrw8DAnT57M3C/LMsPDwzkzqwSFRSgUwmq1ZrRjAoFgHogNmV/OEvOrABFBzGIS6IBIv9l7763NtK+lW6urq6szJ2Cr1Zr5tbwlJVmGOjOzQOfhgjC9m8rkrmM4Rm/IFNLuPj8w7eeNJFUMw2BgYIBAIMBbb72VEe56PB7e9a53ceedd+JyuSZ8Dl3XkWWZW265BYC33nqLcHhUS5TWxwgKD8MwcgTZgUCAYDC4yKsSCK4xtBT4Ry6UC3DwYxoRxCwmGT3MurylpLQ/DIwGMbqu5+9QAvDVmaPS434YvDR/654mU5ncHWwZyny//8oQ0Wk68iZTOp1d3Vy8eDEj3AXYvHkzv/Zrv0Z9ff2Ev2sYBpIkUVRUhGEYVFVV0dzcjKqq7NuXm8FKpVLiKr8AicViOeJrWZYJBoMikBEI5pLs6dUF2l4NYFnsBbytyTj1NoNtfBCTzrjouo7TaZaaDMOgrKwMq9VKT08PiUQCu92ctWTqYtabavK2fVC+uC1xU+lh3modBqDIahBOwZsXB7l7fdWkv6NpGgNd7Rw6d5T21hbA1A3dcsst40pvYzEMA5vNRnl5OZIkkUgk0DSNnTt30tLSwpUrV2hvb88EQZIkEQgEcDgcOZkwweISDofHjZBIBzJAxpJAIBBcBWoyqzOpMNurQWRiFpeOkSCmcp2pacG8+u/q6kJRlJzOJJvNBoDFYkGWZaqqqjAMI1cXk+3c2/bmguzCZMQn0cNEEiqnuoPYLDLvbzC3e+V836TPl0olOXN4L/tffob21hZsNhu33nor733ve6cMYDRNo6ioiIqKikyJrry8HF3XcTgc7Ny5E4A333wzZ7K1aLsuLFRVndDLJ11aErOwBII5ILszqbx5cdcyCSKIWSx0HbqPgiSbol6LGaR0dnai6zqVlZUoiml8Z7PZxmlj8utiFKgzjdwWe/yAYRjEJ8nEHG33o+kGG+u8bCkzcNsUTnQEGAjnP0EN9Hby6lOPcf74IVQ1RXV9Iw8++CBr166dULibRtd1ysvL8fl8OfcrikJJSQm6rtPc3ExlZSXBYJBjx47lbKdpmjgxFghj52CNRVEUQqGQeL8EgqtFzZ5eXZjt1SCCmMVj8AIkw6ZLr7s8c/dkehgAp9M5uS6m6npTXzNwAeKLpxFIqPqk2uJDraYeZvvyYqwy3LSyDAPYfa4/ZztVTXF07yu8/sxPCQ4P4nQXsenG21m39UYsNseka0jrX6qrqzPluLG43W4cDvN5brnlFiRJ4tixYzknQVmWCYfDws23AIjFYkiSNGlmTJZlEcgIBFdLKg7+DkCCUiHsFYwlo4dZmyklwWhnUrY/TPYJ2G63YxgGFRUVKIpCV1dX7gRmh9csT2EsqundZFkY3TAyephtDWbb3u1rzEDulXN9mRNUT/sVXvjpd7l8+hgg0bDmOnbc8W5KK80AbjLNTVr/Ul1dnTG4m4iysjIkSaKsrIzrrrsOTdN48803x4lHxZDIxSUWi6HrOolEgscee4yf/vSnXLlyJe97kg5kxDwsgWCWhHsgFTG7Z635LwILARHELBY54wZMUa+u65kgpra2NrNpRriLeXC2Wq0ZzYyu63R0dIw+b7YupmXP/O7DJEwWYFzujzAcTbG81EWlx9y39TUeKjx22oainO/oZ//Lz/Dm808Si4TxlVaw4/Z30bRuI4oyGpBM1P2UT/8yGekARtd1tm3bhsvloqOjg5aWlpztDMNgeHh4GnsvmA/Sgt4LFy4QCoUYHh7mxRdf5Mknn6Szs3Pc9ukMmghkBIIZkt1eXcBZGBBBzOKR1qxUjIp6+/r6SCQSlJWVZYS82XqYNOnH8paUFCvUbDK/X8RhkJO1V6dbq7c3jJonyZLE7avLaVb6OPHiT+i8cgGL1Ubz5p1sueUu3F7fuOfJ9zcm0r9Mhd1up6ioCIvFwq5duwDYu3dvjrW9JElEo9GMR4lg4dB1nXg8jmEYnDlzBoBNmzbh8Xjo7+/n2Wef5ZlnnqG/P7ccmQ5kRPApEMyAbD1MaWE69aYRQcxioCag75QZcFSuN43qyD8vKV9rb1oXM+EcpXqz04buI6aAeBGYrDMpXUra3jjaURQYGqCqZy832dqQdZXKugZuuOt+ahtWTphN0QwjM0dpOvqXqfD5fFgsFpqamqitrSUSiXDkyJGcbWRZZmhoSLj5LjBpQW93dzd+vx+fz8eOHTt48MEHufnmm3E6nXR1dfGLX/yCF154ISdoSU8oF4GMQDBNtKzp1eWrF3ctUyCCmMWg96SZritbBc7RjMFk/jDZpIWoVVVVSJJER0cHmpaVlfDWmYLhRMgUEC8wSVVH0/NrRwKxFOd7Q7htCuuqPagjLeW7n3qMyHAfUcnBrxJr0Oo2YbNPLtwFiKa0GelfJkOSJMrLyzEMg5tvvhlZljlx4gRDQ0PjthVuvgtLNBrNzMYCMl1piqKwfv16PvKRj7Bjxw5sNhstLS088cQTvPrqqxkX5vRgUBHICATTIHt6dQF3JoEIYhaHzmynXrOUlO35kg5iDMPI0cOkkSQJq9WKxWKhoqICVVXp6uoa3SB7GGTLG/O3HxMwmR7mrdZhDGDz8hIsiszhPS/R19cHkinctay+lW7dy5st03PKjcSTM9K/TIXFYsHn8+H1etm4cSOGYfDGG2/kiEclSSKZTAo33wUiHo+jqirRaJSWlhYURWHNmjU521itVjZv3sxDDz3Epk2bkGWZ8+fP89hjj7F3795MV5MIZASCaaBme8SsmXzbRUYEMYtB9uTqEVGv3+8nFArh8XgoKioCTJ3GRCfmSXUxFjtUbzS/X4SJ1olJgph0a/WOhhLUVJKetstIksS2W+6mad1GblxRjCzBoY7wpCUpMAdMWpyeGetfpsLj8WCz2di8eTNFRUX09vZy4UJuRittrJZtjCeYH8LhMIqicP78eXRdp6mpKZONHIvdbmfnzp089NBDrFu3DsMwOHnyJI899hhvvfUWqVSKaDTK4ODgAu+FQLCESMUg0AmyBYqXL/ZqJkUEMYtBuvW56rpM69p09TBpnE4nmqZlsjZjO2kyuphFaLOeKBOj6QaH28yr4K0NJfR1taPrOkVFRRT5igHwOixsqHaR1Aze6gjnfZ50VqS4tByLzTnljKbZUF5ejsVi4aabbgJg//79xOPxnG3EkMj5xzCMcYLedevWTfl7LpeLW265hQ9/+MOsXLmSVCrF4cOHeeyxxzh58iThcFgEMgLBRPjbQE+ZAYysLPZqJkUEMQtNIgSDF80yUlaabrp6mDTpMlN60nV7e3uu2LR6g9luPXQJYv65349JmCiIOdsTJJLQWF1ZRInLRm9HCzB+1s3NjebPe/KUlHTdQLFYKS6rRBnRv4QTc58NkWWZ0tJS6uvraWhoIB6Pc+jQoXHbqaoqTNXmkVAolNF9hcNhSktLcwL9qfB6vdx555188IMfpL6+nng8zr59+3j88cc5duzYuG4mgeBtj6aCfySzX8DTq9OIIGah6ToKGFDZDPaizN3pICZdHhqrh9HHCGUlScJut2Oz2SgrKyOZTNLT0zO6gd0zYnrHgrZaa7pBSs0v6j3YMtKV1FCCYRj0tLcA5onGZhn9KG6pc+OwyJzqizIcGw1QdE3H4XThKynLKbNF5iGIATPb5XK52LVrF4qicObMGVO/k0V68KBw850fxgp6161bl3nvc8TsU1BWVsa73vUu3vve91JVVUUkEuH111/nhz/8Ifv37xcmhgJBmuzp1WWFO/gxjQhiFpqutKh3rTkeAPNAPTAwgMPhoLi4GBivh+kNjT9JTjpHyeo0szGwoOLeyUW9I/4wjaWE/EPEIiE8Hi92ux2vY7R0ZrfI7Kh3Yxiwr9XMxui6gcdXjNszfkJxOKHO20mopKQEn8/H1q1bAdizZ8+49mpFURgaGhInwjkmmUySSqUIh8O0tbVhsVhYtco8qGqahs/nm/FrXl1dzXvf+17e+c53UlpaSiAQ4Fe/+hXf+c53uHz58nzshkCwtNCyp1cXtqgXRBCz8LQfMG8r1pnlHnL1MGMHPYKZlfFHk4zF7XZP7hdTt928XcBhkBPpU/pCcVoGoxQ7rayqLKJnpJS0bFkdAB6HBTlLwzxaUjLnPxWXlmNz5C+vGcbkwdPVkD2OoLi4mIGBgUxWIBtd10XXyxwTCoVQFIVz585hGAarVq3KCNqtViterzcziXwmSJLE8uXL+eAHP8gdd9yB1+ulr6+P73//+3zve9/L6/4rELxtyO5Mqijc6dVpRBCz0HSNmKdVb4ARC/10EDPRvCQz0zD+qfJlYnKuTJfvAiToOb5gpncTzUxKG9xtbShBliR6R0pJ6fEKkiThdY4GbmsrnZQ6LbT5k/hxZ/QvEzEfupg0NpuN4uLijMj30KFDRKPRnG0kSSISiQg33znCMIzMrKSzZ88Co4JewzAy/x92u52KiopZmQ9KksSqVatyDPOuXLnCo48+yk9+8hOhlxG8PdESIhMjmIBwPwQ7wVVmmtGNMJUeJvsEnXaoBfMgbLPZcDqdFBcXE4/Hcw+8nmooaYRkxHQIXgAmCmIOjehhdjSWkkomGejtRFEU6urqMtv4soIYdIObV5pjCcZOts5HJDE/mZg0Pp+P5cuXs2rVKpLJJPv3j9cZpctKws336olEIoAZmEejUSorKykvN4eEGoaBx+PJbHs1gQyYuqb169fz0EMPsX37dux2O2fOnOHb3/42Tz75pBBuC95exIMQ6jWbT4qmL6JfLEQQs5Bk62FGRL2pVIru7m4URckcpMfqYULx0SDGH8uaWM0UfjFWV9YwyPnXxRiGQUIdfyJJqjrHOvzIEmyuL6a/qw1D16mprc1x2LUqMm6bktG/vHPDMgBePd8/oQNwmsg86mLSlJWVsXPnTqxWKxcvXsw1GMxCtO5ePdFoFFmW87ZVOxwOZDn30HW1gQyYRodbtmzhwQcfZOvWrSiKwtGjR/l//+//8atf/SoTWAkE1zRDVwADShpgDgxE5xsRxCwkmaGPzaarLtDZ2Ymu61RUVKAoZj9+th4moWokskzfhqPJnJO1y+VC07RMEJPjFyNJUGsKUheiQyme0vOWvU52BkioOutrvBTZLaN6mKwsDJhBkM9lzehfGsrcNJW7GYwkOdk5+dWwYUB0HvxisrFYLNTW1rJ9u6k12rNnz7gOGUmSSCQS4oR3FaRSKRKJBMFgkM7OTmw2G01NTcDohPJ8pAOZqw1mnU4nW7du5eGHH2b79u0YhsH+/fv59re/LVyaBdc22e3VBT74MY0IYhaStFNv5Xqw5JrcTaiHiedqPTTNIJh1n81mQ5bliXUxDeZE5syog3lkInHtwayuJMMwMv4w9fX1mW3S849WNtTjctoy99/RbKYzXz6X29qcj/nUxaRxu91s2rSJsrIy/H4/J06cGLeNLMv4/X7h5jtL0oLedBZmzZo1mYyd1Wqd0K0XzEAmPf/qakiXanfs2MHv/d7vcd1115FMJuno6BDlQsG1S7YepkwEMYJsDGNU1Fuzadzk6on0MKH4+BPhcCS3U8lqteJ2u/F6vUQikdyBhVXXg91rRteR+S1z5AtiDMPI6GG2N5QQHB4kFgnj9foy7eRgBgfp+Uel7tEg5rY1FcgS7L00OKHeJs1CBDFgdpHdfPPNABw5ciTv1bkkSaKsNAvSgl5N0zh//jxgDntMPzadCeXpQOZqgw1JkkilUmiaxgc/+EEqKipIJBJ5u9MEgmuCJdaZBCKIWTj8rRD3mxOmPWbAous67e3tSJKUycRk62F03ch7Yg4n1ByBr91uxzCM/K3WtiIz8wPzrovJF2R0+mP0BONUeOwsL3VlsjDp1uo02a69JS5bOsaj1G1j07JiYimN/VfGT5POJpbUxpkCzgeSJLF+/Xqam5tRVZW9e/fm3U64+c6cWCyGYRhcuXKFeDxOTU0NJSWmwHusoHcy7HY7ZWVlcxbI9Pf3c+uttwLw+uuvi2yM4NpETSyZ6dVpRBCzUKRnGFWszehhent7SSaTlJSU5PhfpIkk87dWG4apjUnjcrnQdT2/uFeWoXaz+X17/pPtXJHPI+ZQ62gWRpKkjB4mXUrKl/ZXZIkSV1ZJaa1ZUnplipKSYZiv2UJgt9t5xzvegcPhoLW1dbxHD+YJMBQKkUql8jyDIB/hcHhGgt7JcDqdcxbIqKpKSUkJDoeDoaGhvGVEgWDJo2VlYqZZTpqJc/Z8IIKYhSIt6q0cDWKmmpeUr5SUJlvga7Vac3Qx44ZBLtsxsobxs3/mioSq5Q24DrWMTK1uLCWVTDDY04WiWDL+MBOdYLJLSjc2leGwyhxpG84J3vIx363W2VRXV7Nrl6k5evPNN/NqYGRZzi3vCSZEVVUSiQRDQ0P09PTgcDhobGwEJhf0TsZcBjLZFwqvvfaayMYIrj0igxAbNm1AHL4pN9d1nXA4/6DehUIEMQtFRtR7HVhMzcvYIGY6epg0KdUglMgV+Ho8HtxuN8FgMLeMsfxGkGToPWmqz+eBeHL8AT2aVDnVFcSmyGyo89HX1Y5h6NTU1mSEmpYJTOwcVoUihyXz/Y1NZegGvHZ+cs+YhdLFgHliu/HGG6mqqiIcDnP06NG824mT3fRIC3rT5nbNzc05HXuTCXonYy4DGa/XS0VFBUNDQxw/fvyqnk8gKDiGLpm3JY3T2jwejy/68U0EMQuBpkLPSTOQqNkEmAHL2CAmWw8TT2k5upd8DIVHsxLpclReXYynGkpXQCo2Ki6eY/KJeo+1+1F1gw3LfDisSsalN7u1OjtoG0tZ0Wg25vaRLqWpjO/iqYXRxaSx2Wzce++9SJLEsWPH8Pv947aRloDXQiEQi8VIpVJcuHABmLmgdzLmMpDZuHEjYGZjFjuVLhDMGZoKwy3m99Nsr04kEjMq8c4HIohZCAbOgRozAwm3aWg3PDxMOBzOZE8gVw8znYxCOKGS0syDcnqOUl6/GFmBajN4ovXNOdih8eQLYg5m6WEMw8joYZYvXw6YJQKXyzXhc3od1sx0603Liil12bjYH6Z9KDrh7xgGhBdIF5OmqamJjRs3ous6b7755oTtvWJA5MREo1F0Xefy5cskk0mWLVuWEXvPRNA7GelA5mrfh+XLl1NeXs7w8LDIxgiuHbL1MOXTm16dTE5e3l8IRBCzEHSkRb3rMpOrZ6qHcdmUcU9rGKPt1haLBUVR8gcxAHXbRtZy4Kp2ZSLGdiYZhsFbmdbqUoLDA8SjYbw+Hz6fWWtVFAUbI/vpb4Pk+OAkrY1RZInb1lQAUwt8IwtYUkrzzne+E5fLRWdn54TTkOPx+AKvaukQiUTmTNA7GU6nk9LS0qvKyEiSxLZt5v+TyMYIrhmyO5PKp26vNgyjIJoWZnRkSKVSvPrqq3zzm9/kS1/6Et/85jd59dVXC2JHCpqOEbfcirUTBjHZehhdN8adiGt8+fUAQ9HckpLP58PhcGQyPRkazOGFdB292r0ZR0rTUbXcq9vLAxGGokmWlTip9jnoyZSSluWsl+CIdX/cb2asBi5AzJ/ZptRtyzhf37nWDGJ2n+9Hn+RqejGCGKfTyd133w3Avn378l6hiOGQ+dF1PTP3q7+/H7fbnZOtm42gdzLmorRUX19PRUUFfr9/Qi2UQLCk0JIz8ohJpVIFkV2eVhAzNDTE5z73OWpra3nkkUd4+eWXuXjxIi+//DKf+MQnqK2t5XOf+5ww95qIzOTqjeMmV+fTw4THtFbHUxr/5z/P05fnHJhSDYLxVOY5YNQ4L/03ANP0zuGDYAcEOuds19LrG8toa3UpQMYfZvlys7Va13VcJEAdk31JhmH4CvSehnA/CjolI9mYxjI3DaUu+kMJTnUFJ1xPLKlPOWtpPti8eTP19fVEo1HeeuutcY8XQuq1EAmFQjlZmObm5kzm5WoEvZNxtRkZkY0RXHOk4iOZGAlKVky5eSwWywjvF5NpBTE33ngjZWVlHDlyhHPnzvHkk0/ygx/8gCeffJLz589z9OhRysrKMi6mgixSceg/B4odqq8DzNT54OAgdrs9U1rJ1sOMLSU9c6KbH+xv50eXlLyRb7qklJ6jlLfV2mKDqg3m9y175mrvgPx6mNHW6hKztbq3C8Uy2lqNruNUh5FP/Yx3nP1zlP/8Arz6N3DwX+DUL+Dii3Dxebj4MmXJLiQtiSRJ0/aMWcgupTSSJPG+970PWZY5derUuKBe0zSRtcxDNBollUpx6dIlJEmaU0HvZLhcLkpKSmYdyCxbtozKykqCwSCHD8//WA+BYF4JdpnNH94asE594ZBMJunu7l708Sr5+1vHcPTo0UkPJnV1dXzhC1/gD//wD+dsYdcMPSdAV6FqLTiKAWhvN+uO1dXVGf+JXD1M7olu7yXzZHglJHGyK8im5WU5j4fipsDXqihYrdZJdDFbofUNaN8Hmz48Z7s4tr06EEtxrieE06qwrsZLX9slDMOgpqYmE7nb1BBStA35wHco1lXoaMnzzCYOYJ3dR8pRxh/Yq7jeamXoUim+ik0YRdWoripSrkpUZyWGxfzniyRUfE7rhM85X5SXl3PjjTeyZ88e9uzZw3vf+97MY4qiEIlEcsYtvN2Jx+OoqsqFCxdQVZXGxsaM0H2uBL2Tkf5bw8PDM9bdpLMxzz33HK+++ipbtmyZ0DJAICh4hi6at9PIwgD09/fz9NNPU1ZWxqc+9al5XNjkTOs/bqIAxu/3c/HiRZYvX05lZeW0r5ri8TgPPfQQp0+fxul0UllZybe//W1WrVrF7bffTmtrayZD8bGPfYxPf/rT09ydAiQtpM3Sw6Tbn/PpYeIpjZQ6mm3pD8c51zs6m+f7+9vZUF+KnNW2mxb4VnodWK3WjANwf38/sVhs9H2p3wX8HXTN7VXj2EzMkbZhDGDL8mKsijzq0rvM1MMYahKHHoRD/4qkq7SWvYO6Gx7AEh+C6GDW1xBEByA6iJIIoCQCOLjMr6UzmPt/MW4tqt2H6qxEd1dB2TLwVEFRtdlm7qmGoirzdsRwcD54xzvewcmTJ+nt7eX8+fOsWjWq9BclpVzm0qF3tlxNIFNXV0dVVRW9vb28+eab3HbbbfOxRIFgftE1GB6x5ZiGU6+u63R2mrKE9HlssZj1ZcNPf/pTPvOZz7Bs2TIuXrzI//gf/2NGmZhHHnmEd7/73UiSxN///d/zO7/zO+zevRuAr3/96zzwwAOzXVphkXbqrVoPVjOYyM7EQK4eZmwp6Xi7aVr3wKYa9pzr4lxvmL2XB7l5ZXnOdkNRM4hxOBwkEgmqq6tpa2ujra2N5uYRkVbDiOld3xkzbWi9+lS9rhvj/GwOtuS2Vqf1MA0NDebvhHtxp9rh4osYVienax6ktmY9KBOfQIxUnI6uDogOcrK1nwOXetnkCXJ7jYqSCGCN9WGJ9mFJ+LEkAuC/AJNJf9Y/AB/+7tXs+oRYrVbuu+8+fvzjH7N//36WLRsVMycSCQzDEN4xjAp6+/r6GB4exuPxUDfiIaRp2rxnYbKZbSCTzsY8++yz7N+/n23btmWeSyBYMmR3JpWtnnLzaDRKX59Z0q+rq5ti6/ll2kFMZ2dnzmL/4R/+gVOnTuHxeOjt7WXDhg3TDmIcDgf33Xdf5uddu3bx1a9+dQbLXkJkRL2bQZIydURFUSgrM8tCuXqY3FLS/hFtyT3rK/FFO/juBYUfH2hj14oyFHn0RJhSDULxFG6Xi+Hh4UwQ09raOhrEuErND+jAOdNBeMWtV717Y7Mwmm5wuM0MYrY1lBIY6icejeDzFZsnpVQMqxpDOfQvgIF+/YMk8eZ55lwkqwN3xXIGw9XUeg1+dCnE9wLwH7cXUeywoDlKUB3m62mJ9WON9lGtBHAn+iHcA6HekdseswPq9C+gbR8s33XVr0E+1qxZQ3NzM+fOnePQoUMZPxxJkohGo+JEhynolSQpJwuTDu6sVuukRojzwWwDmdraWqqrq+np6WHPnj3cfffdi24AJhDMCDWe1Zk09eDHZDJJT08PsISCmA996EN89KMf5Q/+4A+QJAmfz8czzzzD9u3befnll6moqJj1Ir75zW/y/ve/P/PzF77wBf7sz/6M9evX89d//dc0NTVN+RyJRIJEIpH5ORg0u1dSqdTiiSnjAazDLRg2D2rJKkilaG1tRdf1TBYmlUrh9XpJpVJoukE4lsh0JkWTGodahrBZZHY2eEleMdgXdHOuN8LLZ7q5a0TkmqY/EMVR6kKSJCorzccuX76cs/9y9SaUgXNoLW+iL7v6E3g4lkTPGmVwtjtEOKGystxNsUPm/HnTM6W2ttbs4Aj24ho+De37MBzFJNb/Gpz2Z0z7JsNpVdB1A7sCN9ZZ2N2m8kprivetliA6gCU6gG71kHKUkiivYdBtw5anNV3e9w8oL/0v9Ff/Fu2h/7jq12Ai7rnnHi5fvsz58+dZvXo1mqahKAqhUCjjsHwtkP58zfT/LBQKEYvFuHz5MrIss3LlSjRNwzAM3G73ovzf2mw2ioqK8Pv94zov0h1I+TqRtmzZwnPPPcfRo0dpbm4eFbAvIWb7Pi4lruV9vKp9S0Sx+NuRgFRJE0zxHP39/YTDYdxuNy6Xa85fz5k8n2RMs9FbVVX+9m//lmeffZZvfOMbVFZW8sd//MecOHGCpqYmvvKVr2TsuGfCV77yFZ566ileeuklXC4X7e3t1NfXYxgG//AP/8C3vvUtTp8+PeXzfOlLX+LLX/7yuPt/9KMfTeoKO59UBE9y06W/oc9zPXtX/QkAPT099PT0UFVVlRHgTsSxQYl/Pa+wrljnk+vMk/zJIYl/PqdQajf4H5s1LBNc8BmGwYkTJ9B1nQ0bNmQOyMuG3mRb63doL7mJw42fnLudHeHpNpkXOmXeWadz33KdCxcuEIlEWLlypZmJMQxuPf8XlEYvcaLu17lcee+s/s7pYYl/PKvQUGTwmQ0za29VtDj3nPoMdi3M7uYvE3BNT8g2G3p7e+nu7sbr9U4rGH+70dfXR1dXF8XFxZlhj0uVixcvEg6Hqauru6qLOoFgoZEMjfcc+x0MJJ7e9KgpO5iEoaEh2tra5u3/NhqN8vDDDxMIBDLO3RMx7UyMxWLhi1/8Ig899BB/+Id/SENDA48++uhV1a2/+tWv8rOf/YwXX3wxE2jU15s+IpIk8alPfYo//uM/ZnBwMFN6mYgvfvGLfOYzn8n8HAwGqa+v5957753yRZgv5NdPwyUoW7mZ++69Gyw2fvzjHwOmp0hdXR2KomQOeF3+GP7oaAT6n69fAXr5yK3Xcc+Wal544QX+6MN3cfDfj3C0PcA52yruvz5XVFXhteOSVAKBAP39/XR1dXHdddeNnkD7V8I/fYdl6hWq3/VOcyTBVXC5P0w8NZpFuXj+OBDlrhs2UFts49ixY1gsFm65+WaUYDvOjjcojV7CKKpm7T2/zkpXLS/sPco9d92JNTEi7GXiuDqp6rQPx1hVZ/DYlQitYXD5fNR78//TLS/3YC2qMKeyKqMfd7nkCrz619zGAbT7fv+qXoPJiMfjfOMb3yAUCrF+/XocDgeaplFRUXHNZGNSqRQvvPAC99xzT05pdDKGhoZIJBI88cQTANx0002Z7KTVap3y/30hiEQiORkZTdM4duwYmzZtyuuPUVtby7PPPsvQ0BC33347VVVV8+JxM1/M5n1calzL+3hV+3b5VeSjGkZJE/fd/55JN00kEjz11FMArF+/nq1bt875OTZdSZkOMxL2plsef/nLX/LEE09w11138bnPfY4HH3xwxov82te+xo9//GNefPHFTMupqqoMDg5SVVUFwBNPPEFVVdW0Dmh2uz1vDd1qtS7eh7XLNDxTqq9Hcbozim5JkqiurkZRFJxOZ2Z9UTWGPHKi1XSDvZdNPcy919VgtZr322w2/uSda3n40f08/lYn96yvwWEdPaCGEgbVlV6CwSA1NTV0dXXR0dExqoupuQ6cpUihbqzhDiib3oyMfBiGQcqQkUcEuYPhBFcGo3gdFtZUF9PdegHDMKiuqcWmhkBPUnzuRwBIO34bq9MLvpETl92BtajB/DnYZTr45sGqyLjtFuIpjTsarPz8fJLd7Sl+a0P+k4Wmqrji/RAfAGcJFFWaguZdn4R9/4B87hnkoQum8HoesFqtNDQ0cOXKFXp7e2lqakJRFJLJ5DWni5nu/1rarry3t5dgMEhxcTG1tbVIkoSmaZSUlBTECaa4uBhFUQgEAjkaF0VR8gYxdXV11NbW0tXVxcWLF3E4HLhcroIwBJsJi3rMXCCu5X2c1b4FzLK/VNY05e9mi3pramqwWCxz/lrO5PmmrT577LHHKC8vZ+PGjdTX11NaWsrLL7/Mnj17eM973jPek2QSOjo6+OxnP4vf7+eOO+5g8+bN3HDDDSQSCe6//342bNjApk2b+Na3vsUvf/nLaT9vwdFzzLyt22r+2NNDMpmkrKwMm82GpmmZ9ud4Ssux7r/UH2Y4muK6Wi+1xbldRDetKufmVWUMR1M8c6I757GkqhNJathstky56sqVK6MbSFJmkvbVDoNMqHqOs3DapXdbQwmKLGVGDTTU1UJsGEfry1iC7eaE1FV3g6cGxgogLXZzUGb5GrDlt5svHvF/ubvRvH2pZWL762hmGKQBsSHoPwsDF812952PmA+98bWZ7/wMWLNmDZA7Wfzt3GqdHodRKILeyfB4PPh8vmkb4qVdfI8fP46mafT39xeENbtAMCEzbK8OBAL4/X7sdjslJSXzvLipmXYQ85nPfIY33niDrq4unnvuOf70T/+UoqIivvGNb/AXf/EXfPzjH5/2H122bBmGYXDp0iWOHj3K0aNH2b9/P263m0OHDnHixAmOHTvGSy+9xKZNm2a1Y4tOsBvCfeCugOJGYHQMQDrTBKOjAoJjupLSHT53rasiH398r5lZeeKtjnGzgoYjKWw2GxUVFciyTHd3d65Qqm67eZtu/54lsWSuFuXgSCfV9obSnNbq5eUuUGN4z46IaHd+Auxes1tqImxuKF9tGi9ZcrMsRXYLFkVidYnMcq9MT8Tg1EB+XUwslefkkwzB4EXY8TtmMHPyCRi8NL2dngWrV5sti+3t7ZmTYTKZvKrZPUuZaDRKLBajpaUFRVEyr898O/TOlpkEMtXV1dTV1RGLxTh9+jSqquL3++d/kQLBbFGzp1dPPTNprFnrYjPtIMZms2Ushjs7O3OulrZu3crLL788LwtcsnQeMm8r1oIt/9DHyfxh9l42XXrvXpfbgZRmy/IS7l5XRSih8oujuYYowXgKxWpDlmUqKytzjIkA0y8GzGGQV3GVGFdHA4eUpnOsw48swdblJfgH+0nEovh8PoosKu5Lz2KJD5pZoPobwDvNtjxnsfka+upBNjMvkgQ+hxVJkjLZmBdb8qvZU5pOSsuzj1rCzMxs/U0w9HnNxni9XpxOJ4lEgt7eXgBkWSYaHT+1+1onlUqRTCY5d+4chmGwcuXKzLFkIRx6Z4vH45l23X9sNiYcDovhn4LCRZv+9GpN0+juNrP/i21yl2baQcyjjz7Kf//v/x2v18uf//mf83d/93c5jxdCRFZQtO0zbytNp17DMMYFMem6n6YbOVmNnmCMy/0Rqrx2rq/1TfgnPnvvGiTgyaNdBGKjJ3HDgLihZKz+YcwIguU3moLe/rMQG571Lmav+WRngHhKZ221lyKHJZOFWVbhQ0qG8Zw3BZzsfATc5ZnAblpIkvk7letN911Jxuu0IgF3Npiv4attKZL5ghXGZ4wyaElY8y5zrtWx/wB/+/TXNEPSDtTpkpIkSW/LE1vaG+bs2bOAKQxMs1AOvbMlPU17qvJQVVUV9fX1xONxTp06haIoDA0NLfqMGYEgL2r29Oo1k24ai8UyF2JTddcuFNM+Ytx1112cOnWKaDTK/v37Z9VO/bYiY3K3CWSF4eFhIpEIXq83M6gxnToPx3OnVh9t9wNw59oqZHni4HBdjZf7N9YQS2n89K2OnMf8URWbzZYJmHJ0MTYXlK81T+Kd46ctT5dso7vM1OpGs0ba027+vYZKH0UXfo6cCmM03GxO8vbM0kNDls3hZJXrsRSVU+SwUuWW2VihEE7B/q78J4loapKTh7MY1t5vzrfa883ZrWsapIOYlpaWzEnw7aaLMQyDWCxGe3s7kUiE8vJyystN5+mFdui9GqYTaG3daurgjh8/TjJpDi8V+hhBQRL3Q7jX1CC6J7cGCIfDDA4OFkwHIUwziElnEOZqu2sew4Ce4+b3I6LemehhDlwxtSUTlZKy+ey9zcgSPHuim8HwqNlfUtVJGhJVVVVIkkRXV1euSdfIumjfP7N9y3r+bIlAZmp1QynJRJyh/h4sikKtW8N96SkMZKSdj5hzi5SrHJKnWKF4Ob7lG9Ct3hyBbz7yTdnOYeOHzczU4e+Zjr7zgMPhwOfzEQqFGB42Az7DMHIMGq91otEohmEsCUHvZJSWlk6pj6msrKS+vp5EIsGpU6cAM1AT+hhBwTFodiZR0mBmvSehvb0dwzCorKwsmKzptFZx//338/GPf5yXX3553NVjKpVi9+7dfPzjH+f++++fl0UuOYYuQyJk6jiKzEzI2KGP2XqYcJYwN5JQOdLmx2GVuXlVOVOxotzN+zfXkdR0HjuUWw6JaxYURaG8vBxVVTO1TMDUpQB0zm4YZHZg0OWP0RWIU15ko6HMRV9nKxgGNeU+is//BFlLkmy6Cyqap4z0Z4LT5UapWMmu61ZjlWF/t0owMf7komoGyckcgT3VsOpeszb8xtfnbH3ZSJI0rkvp7aaLiUQiRCIR2tvbsVqtrFxpdkIUqqB3IqxW67SEvmltzIkTJ0gmk8iynHkNBIKCIT29unTyziTDMOjoMDP+haKHgWkGMUeOHOGmm27is5/9LF6vl7Vr17Jz507Wrl2Lx+Ph05/+NDfeeCNHjhyZ7/UuDdLZjcp1E4p603qYWDK3tfp4px9VN7hlVUWO/8tkfPbeNVhkiedP99ITiGfujxsyqqbn18U0jsxN6j0JyZmfSOM5paTRriRJkuhpMyP7FcUyrtYX0WUb0raPg7d2ykh/ppQX2XAW+djZVI6qw6sdE+lipugs2fyw6VL51nfBPz8ZxXyt1m+XTIyqqiQSiYwWZvXq1Zn/gUIW9E6Ex+PBbrdPWh6qqKigoaGBRCLByZMnATNwHR4eFvoYQWGg69Nur85uTCgUPQxMM4ixWCx84hOf4MiRI1y6dImvf/3r/NEf/RFf//rXuXTpEkeOHOGRRx7BYrnKMsG1QvsB87ZyHVgcRCIRhoaGMiWFbD3M2IGPh0YmQE+nlJRmWYmLD26tQ9MNfnSgNesRiZhGfl1MSQO4ys1a6NDlGe9idhCTmVrdmDu1+rrh55EMnWjTfdiq14JjYpHybPE5rVgUiTubzQzPC+0yqqsGpNwAMDaZLgaguB6a7gA1Bq99FRLhOV9rfX09TqeT/v7+zNV4KpV6W7Rah0IhAM6dOweYpaQ0hS7onYi0nmcy0tqYEydOZAJWWZaFPkZQGMygMykcDtPf348sywU1VmPGR466ujre/e538/DDD/Pud7970SdYFiRpUW/NFpCknCxMuoQ0qocZPblqusG+K2Zr9Z1rpx/EAPzBXauxW2R2n+undXA0XR1VpUwQ09HRMXrClCSoHdHFtO2d2f4xWk6KJTVOdgawyBKblhUT6GklkUhQ7FKo7N2NbnWTvP4j02+pniGSJFHqtrF1eQleh4UzPSHaU0XES9aiOisAKbPOKdnyX8zbEz+F7qNmSXAOkWU544mSXVJ6O5QXYrEYra2txGIxqqqqKC01PYKWkqB3LJIkUV5ePmkQWl5eTmNjI8lkMpONAdB1naGhoYVYpkAwMTOYXt3Z2Ymu61RUVBRUwmLpXf4UOloK+s6YmYDazcDoCSst6k3rYVRNzzm5nu0JEoypbKovptI7s5kry0pcfGhrHQbww/2j5RBJsZMyJEpLS0kmkxm7aADqd5i3M+xQUjWdlGpeRR7rMMtfG+p8OKwK/VfMYZ1rdLPOGlr1ARzVa8A6fzNkSt02rBaZ21abVwe7z/WDrKC6a0iUNKPZi1F1g4Q6RcajtAkab4FUBE7+zMxQxac/w2M6pMc/vJ1araPRKLquZwa5ZmdhlpKgNx92ux2PxzNpViVfNkaSJKLR6NsigBUUMGpyNBMzRTkpbXJXSKUkEEHM3NN72kzRla00Z/UwsR4mPMZp90ibH4C7Z5iFSfOJW5tw2xT2Xh7kQq+ZRbDYbARjav4RBA23mLc9J0zXxmkSzwoGRlurS1Hiw3T3mN0962IH0RylhFa9D2dF46z2Z7pYFRmf08odI6/bK+f6MicVQ7GR8iwnUbyaKNMQj275dfP2xOOQjIwEMoE5W+uqVatQFIWurq6MSD6ZTF7TpYVIJEIwGKS7uxu73c6KFebU8KUm6J0In8836ZVpWVkZK1asIJVKceLEicz9iqLg9/tz3bQFgoUk0mce31zl4JjYzDGVStEzcmzPFvXqur7oM6hEEDPXpEszFWvB5iaZTNLT04PFYqG8vHyMHia/S+9EowamYnmZmw9tXQbAD/aPXukndInKqjy6mLqtIFtg4BxEp5/aTmePDMPgrbSod7kPfbiNgeEgNlLU00Vo7Uewla9AutqW6mlQVmRjdWURtT4H3YE453pzS0GGxUnI3WhmWyyTnDgr1kL9TkgE4cxTgAFDVyDmn5N12mw2mpqa0HU9o/SHa1fgq+s68Xg801bd3NycOeEvRUHvRFRUVEwrG3Py5Eni8VHxvSRJDAwMXNNBrKCAGRjpTCppnHSz9NBHSZJybEIK4ULkqoOYWCz2tjPtmpR0aab6elCsdHR0jOurT6fPs4OYTn+UtqEodcVO1tXM7sCuyBL/Zddyip1WDrf5OdlpZhAUixV3sSlCTPf5A2aJp3K9WQJLj0mYBmlRb8tghIFwkrpiJ8ttIbpHlOsrjRaMomoiTffhKFmYVjyXzYLLrmRlY/rHbRNOqKa4uKIZHMUTP1k6G3PsP0YyVAYMt1yVu3E2Y0tK13KrdSgUQtd1Lly4AMDataN196sS9Ib7oYAE0bIsU1pamuvFlEVpaSlNTU3jsjFgBnqDg4MLsUyBIJehkZlxU5SSOjs7SaVSmeHFaWw226K79c/4CPL5z3+egwfNwYHPPPMMpaWllJSU8Oyzz8754pYk3UfN29pck7ux/jDRpIqmj159pUtJd62rvKoPRV2Jiwe3m9mY7+9rxTAMbA4ncV2iuLiYeDzOwMDA6C8sG9HFpDuqpkE6iEl3Um1f7sUS66erx3zeVVwhuP7X0d2VGav2haDMbef2ZjOIef18P6kx3jCabphrlyTzysM+Qfq0eqM54yk2BOd/NXKnYbYiziBjNRHpVuu2traMKPRazcREo1GuXLlCIpGgrq4u41x8VYJeNQmhLkjOfQfZ1eB0OikqKppQ6JvOxpw6dWpcNiYej2c6uASCBUHXzYszgLJVk26aPfQxm8UuJcEsgpjvf//7XHfddQD8xV/8Bf/6r//K448/zhe/+MU5X9ySIxkxpyFbHFB1PTDqzZJOwaXf9LGlpP0jLr2zLSWlKbJbeN+mOsqLbJzuDnK4zY/FYkXVDMorq3LWBJhzlMB0GNamrs3rWQLZgyN6mBsqdQxDp6fHzMQ0+CTiK+7F6i5e0NbZYpeVumIn62u8hBIqb7WOz5xkdEiSZE7Itk1wIt3yG+bt0R+ZIwkAMMB/9YGMx+Ohrq6OZDKZMSBUVXXCq/ilSjweR9O0HIfeNFcl6A12mkM757h7bC4oLi6eUB9TUlLCypUrSaVSHD9+POcxWZYJBAIiqy1YOLSs6dUVE7dX67pOV1cXMF4P43DMX8PGdJnxGSYcDuNyuRgcHOTChQs89NBD3HfffbknxrcrnYfB0MwPhN2Dpml0dXUhSRKVlZUT+sMEYymOd/hx2xR2NZVe9TKqvHYe2rEcgO/vawHMkpK3xCwpXb6c5QvTcJN523tqWle2cVXDMMz1n+sJ4rTKbCqJMzw4QFyTqDL6MTZ8GIoqF7zrJN1ufXtzukupb9w2kWwxtSybGhmre/yT1W0zfX7CvXDhhdzH/K0QGRj/OzMgXVbJLimFw4WVWbhawuEww8PD9PX14XQ6aWhoAK6yjp4ImbNe4gGIjC8ZLjZTtV1v3boVSZI4derUuK40WZaFPkawcKhZHjGTtFdnD33MDmIKQQ8DswhiGhsb+dGPfsTf//3fc8cddyBJEn6/vyDSSotO2qm3Yh1YXfT29o6rI9rtdlKanuMge7TDj27AbWsqsFum59I7GcUuG3etraTG5+BSf4Q3Lw1isdpw+MyBXTm6GF8dFFWZJ4TBi1M+dzxlrvtwm7nmLVVWbIrE4DlT0NzoipBsuB0NGbc7T3Awz5S6bdy2ugKLLHGgZWhcB9jYn5Flsx48VuwrSVnZmB+CPiZLEmg3dRmzJFsXYxiGKcC+hkpKYwW9a9euzWTlZi3oNQzzyjHUA//xMDz7x9PKHi40FouFkpKSvIFMcXExK1euRFXVcdkYMF+bnHKvQDBfqCOZGEmeVNjb09Njen8VF+cELYWgh4FZBDF/8zd/wx//8R/z7//+7/zZn/0ZAE8//TQ7d+6c88UtOTKi3g0gyxPOSwqPKSWlyx5XW0pKY7PIlLht/JcbzCvfH+5vxWJzYLU5cLmLiEQiuYPo6swZL7RNPQwybXKXHjVwQ42EkgjQNWD6XVSv3QWuMhRFyRGALRQ2i0xtsZMdjaWkNIM9F3NPCLqex/hOVsyasGVManT5jWaAE+iAy7vH/7FgB4THZ3umQ3l5OaWlpYTD4YzpWSKRuGauwkOhEKlUiosXLyJJ0twIeiP9pjnX4e+ZpduOQ6OW6QWG2+3G6XTmfT+3bNmCJEmcPn16XDYmHcwKfYxg3gm0mf9P3lqwTJw1H6vrTFMoiYsZH0ne9a530dXVxZUrV9i8eTMAH/nIR3jyySfnem1Lj8zk6u3AqPZkrD9Mth4mpensvzyIJMEdzXNn5VzisnHr6nIay1y0D8fYc2UYSZYpLjeFrzmt1st3mbddR8dnHMYQS2poupEJvHbWWrCf+SmdRiV2SaV8/a0gSYtqYFZaZMvxjBnLuGwMmJO1y1aBkrXu7GzMkR+YOoyxBDsh1DvjNUqSlMnGpD8n15LxXSwW4/Lly6RSKerr6zMC71kLerWUmYEJduaKrS++MOmvLSalpaV5g7Xi4mJWrVqFqqocO3Zs3OOyLOP3+6+pzJygAMm0V6+YdLPOzk5gvB6mEEpJMMsW63PnzvG///f/5lOf+hQAly5dyqSN37ZEBs0rdocPyldjGEZG0V1VVZXRwxiGQSgxmgI/3RUkktTYtryEsqK5O/F7nRasisyv70pnY9owJMsEupibzdu+k+YV7iTEUxoXekOE4iori2WqjQEGWs+DJFFdXoLsKl70D3iR3cLNK8sosls41RWkNxjPeTySL4gBUKxm5kXOusJovBWKl8PwFWjdk//3Ql0Q7M7/2CTk08VcC0FMMpkkmUzOrUNvsNPUm731PTOYLBrJWl55bS6WPC9Mpo/Jzsbka69XFEXoYwTzS3pm3iTt1WmfMxjv1FsIol6YRRDz85//nJtuuomWlha+973vARAMBvnMZz4z54tbUrTvM28r1oHNzdDQELFYDK/Xi8tlTrK22+1Ek1qOvcXhtrktJaWRJIkSt5WdjaWsqSqiL5Tg9ZYQ3lIz25MOsACzpVixwcCFSTtv4ilT1HuoxdSC3FBroezMD7hkmCLi5avNrrVCEHzVFDu5dbUZsO0+n6tdCSdUeoNxVC1PZsViNzMy8kiHiazA5pGZSoe/b+oy8hHugWDXjNa4bNmyjEg+Leq9FrpTQqEQQ0NDDA0NUVRUxLJlZsv/rD8XibDp0eNvNTMvFgfc/SXzsfYDkChc636r1YrP5xsXyPh8PlavXo2maXmzMWmCwbkdeyEQAGZd3d9ifl+2esLNenp6iEQiFBUV5dhlWK3WgtDDwCyCmP/5P/8nv/rVr3j00UdRFFOEunnz5rwitbcV2ZOrrY4J9TDZpSTDMNg34tI7k6nV06XEZQqvfnNXIwA/O96HZHFgczgJBoOjB0iLzWwJ19VJTe/S/jBvXTF1Jnf7OnC37+aiZD5/fWMTMLqvi0mx08qdI6/pK2f7cq5oDQP6ggnO9oTo8sdIjp2pZHVA6crRSdir7gJPDQych46DE//RcC8EOqe9RlmWM54x6c+LqqpL2obeMAxisdjcCXrTYl6At75rZmGu+4D5f+arh+jAjGd/LTQejwe73T4uq5LOxpw5cybvDCVJkgiHwyIbI5h7tCT4px78WMj+MGlmHMR0d3ezfbup+UifqBRFmXSS69uC9OTq2i3AqOZkvB5m9ATVNhSlKxBneamLVZVzbwrnsCo4bQqb6ovZWOdjOJpid0uY4rKKnDUCptU+mEHMBAfNeEpneHiYi0MpPDaJnV0/pJtKYjgpLi7B7XZjGEZBpBllWeLmlWVUex10+mNc6BvfvmwYMBhOcr43RPtQNBOkAWBzmWlWSTazMpsfNu8//L2JszFgziLxt0/8+BjGuvcqirKk3XvD4TCJRIJLly7l6H5gloLeyACoMTP1fekVsDph00fMx9JGjRdfnKPVzx/l5eXj7vN6vaxZs0ZkYwQLjxrPaq9eM+Fm6dEohaqHgVkEMRs2bODpp5/Oue+5557LiHzflhiGOUQRMqLefHqYlKZnWpQBDrebpaS711XNW+ai1G12CP3GiDbm6TN+XCMjCHKHQY74xfScmlAXE0tpHLlo7tfDZRco6j3IecWM4pcvrwfMD/hCuvRORlmRfVLPmDSGAf5oigu9YVoHI0STI9kym9v0kZFkWPNOcJdD70nonviEA5jZAX/b5NuMsHLlSiwWC11dXRkh51IWdEajUS5evIimaTQ2NmZKqbMS9GoqhEa0Rm/9O2DA9b82OjIiPYW99Y2CGkGQj4n0MelszNmzZ0U2RrBwpKJm+VuxgXdZ3k10Xc+YcRaqHgZmEcR89atf5eMf/zgf/ehHicVi/PZv/zaf+MQn+Nu//dv5WN/SwN9mWtR7qqF4GeFwmEAggNPpzNis2+32cS69B0ZceuejlJTG57QiSbC2xsuOxhLCSY1jAfMD2NqWdaKtv8G87TtlahDykAz2c7AzChh8PPkjAM7ZNwFkjMxsNtuCuvROht2i8N5N5j/faxcG8mtgxhCMqVzqi3C5P2xmzewe00NBscPGh8yNjnx/6j8eHZxW+6/VaqWpqSlHCL5UW61TqRSJRGLuHHrTYt6BC6aA1+aGjR8efbxmsynC7jluZsAKHLvdjsfjyXlvPR4Pzc3NaJrG0aNHJ/xdkY0RzClDV8z/reLlpldWHgYHBwkEAjgcjsx5DApLDwOzCGJ27NjBiRMn2Lp1K7/9279Nc3Mzhw4dYsuWLfOxvqVBa9bkaqs7UxqoqjIzLKN6mNFSkj+a5FRXEI/Dwo4VV+/SOxGKLOFzmqWsdDbm2StJLFYb/uHh0as/TzV468yT7+CFcc+TSqXQ/F0c7lG5RzlMdeQsfkc9fTEFq9VGZaUZiC2GN8xkbFxWTHOVh0AsxZF2/7R/L5LQaBmIcrEvREB3QUkDrHuvmQXofAv6Tk/9JLEhczbJFAHJ2C4lSZKWZEkpHA7T19dHIBDA5/NRW1sLzFLQm4yYrx/AoX8zbzd8mITiZjiaoiNqQVecULPBbL++9Moc7sn84fP5xo0l2LJlC7Isc/bs2byuzZIkEQqFlmRgKyhQ0sampU0TbpIticgOWgrtGD+rS+aamho+97nP8Q//8A98/vOfp66ubq7XtbRIi2ErrwPFklcPYxhGjj/JW23DGAbc3lyJVZnfzEW6pLSivIhbV5cTVyFsMSPry5ezSkojpbB84tX4cBcn+hLEVZ3/YX8MgGO1HwWgtrYWWZbRNG1RXHonw+Owcs91ZudXPs+YqYglddqGopwLWBh2NZojFQAO/2CaTzA8ZWlpzZo1SJJEe3s7mqYt2VbrbIfedevWZQ58hmHg9U4wbHMiRsS8Rt9ZaHsT3eahte5+2oaiDIRT+K3VRAw7LBvRcl16ec72Y76pqKjICUiKiopobm5G1/UJszGSJBEIBBZohYJrnsH09OqJBz+mM8PZpaTs0TmFQv5JZZPw8Y9/PG8qyW63U19fzwc+8IGcNPLbgu4RUW/d+MnV6Tc9Mqa1Om0WN5+lpDRuuwW7VSaR0vkvOxvYc3GAkyEH2y1w5vwFNmwwh1XScBOcedLU96RipogSQE2QCvRyoFvlg8rrNOodJLwNXDIagEs0NpoZnsVy6Z2KD2yu5du7L7H/8hCRhIrbPuOPPUlVp0N1MrjyN2g69mPktjfNq5kppr8CkJj85ON2u6mrq6Ojo4Ouri7q6+uXZKt1NBqlpaUFRVFYvXq0bdPhcMwo/ZwM9BELBokmVbxv/jNuYGjVB0nKZhlUdVViKDb8qh3Psh2w/zumxYGaNDvtChxZliktLWVgYCCnw/PcuXOcO3eOTZs2jdMPpbUxPp+voFL5giWIYYy2V5fnb682DCPv0Md0ZaGQmHEKwOFw8PjjjxMKhSguLiYUCvH444+jqioHDhxg69at/Md//Md8rLUw0XXoPQ1IULeNRCLBwMAAFouFsjJzVpGphxktJSVVnYMtQyiyxO1r5j+IAXPCM0BdiZN3rCqlSzXFt52dHaMbpTuU+saIe4OdJFSNo51RPm35KQC9G3+f3m6znTjtA1JoH+40K8qL2N5QQlLT2Xtp8KqeK1bUwOAaUxuTOPh9VH0GKf5J2q/TgX+6pKRp2pILZC5evIiu6zQ1NWWEf9MR9BqGQSSh0hOIc6F7mNaWS/SF4qhdJ3D3vYVq8+Fveo+5rWJHdZpi7RBO1OIV4Cozs119S8dw0+l0UlRUlBH6FhUVsXbtWpGNmWeCsaVrXzBn5Ax+zJ9wSHs9Wa1WSktH5Q6FpoeBWQQxnZ2d/PKXv+Txxx/na1/7Go8//jhPPfUUPT09/OIXv+Cxxx7jL//yL+djrYVJ70lT6V3SAEWVmeGKlZWVyLKc1x/mRKefeEpne0MJPtfC9NubnjHm9w/vaiQouUgaMsHhIYYCI3X46g2mkdjARbPDBkamBge4PJzi9th/UicNEitdz2XfzSQTcYpLSnG5XAXXdpeNLEu8d5Opz5hNSWks/Zv/O5rFha3tNTqvnKUvlCClTSOYiQ1OGMikW5Hb2towDANFUfJ2qxQi8XgcwzA4d+4cMD1Br6rp+KNJ2oeinOkOcbk/Qn8ogebvAsP8Xyk7Ywqoh1d/CGNkQGfKXUv6g6wpLkIpA5aNlEEvPj9v+zgfFBcX5+hjNm/ejKIonDt3Lq+QV3QqXR3xlEb3GPfutyVaYtQjZoJMTNrRvaqqKqdRoxAz7TMOYl599VVuu+22nPtuvfVWXn31VQDe8573ZMopbwsyk6vXgtXFpUtmrTFbD5NUdRLZrdVtfgDuWT+3Lr2TYVVkikbKKNU+F7evLKFPN7Mxp8+P1EcVq+nea2jmcD3DgEAnmg6HWv18ymLOx+rZ9jl6usxIPt1aXQguvZPxvk01uG0KJzoD9IeuroVZcxQzuP5jSBiUnH+cQCxF62CEnmCcxFjjvLFE+swZQGMoKyujtLSUSCSSmWK8VDIxgUCAUChEOBymtLQ0I/Ie+5mIJTX6QnEu9Yc52xOifSiGP5pCG8lmSWoMJWGKeZ39x3ENHEe1lxBYcR8Ams2HbsvK6kgSAS1LF3Pl9SlF1IVEuu06HZS43W7Wrl2LYRgiGzMP9IcSaNO52LjWiQXM45DdY2Yx87BU9DAwiyCmubmZr3/96zn3feMb38hcSfb09MxuwNtSJe0WWrMJJCnHHCj9pmeXkrJdeud61MBUlLhHo+gPbKyk3zDFlsfPXMicSDLDILuPmR4daoykqrGs9WeUSGHaincSXn4nvR0tADSOtFYXgkvvZHid5lBIA9h9/uqzMQMbPoGuOPB0vIol0oOBOdizbShKlz8+flJ2NqHuvNOvx3YpJZPJgjeRjEajaJqWCbyyBb2apmNYnXQMRznTHeRiX5jeQIJoQssba9jCI1kqw6Ds7A8BGFrzYQyLA5BR3TXjfieKm0T1VkAyBfbxpXWCt1gsFBcXZ97ndDbm/Pnz13w2JpVKLdh+JFQNf1SUkgDTeRxM64gJjtn5hj5CYUoGZhzEPProo3zrW9+ipqaGbdu2UVNTw9///d/zL//yL4A5DPKv/uqv5nyhBUv3yLiF2q1omkZPTw+SJGWuRsf6w1weiNAXStBU4WZF+cJ28ngdFiyK+aGt8LlpWGZ+QFvb2xmOjlz1p4OY3lOmjT4QGOrj3bFn0Q2J4I1fIB6P4h/ow2qzZTotCjFCH8sHt5pddK+c67/qg6fmLGdo7UeRDI3iy0/lPBZJqnT4Y7QPxwgn1fxPEOw0h4ZmMVYXI8tywbdaB4NBotEowWAQi8VCY1MT/liKLn+MzpBK+1CM4UgKdYorYCU+hKSa++rqP4pz8BQpZznBxncCo2Lesei2IkK4TdfRZARaJhjSWcC43W5cLheGYeByuVi3bh2GYXDkyJG8218L2ZihoSG+9rWv8a1vfStjqDafXG329ZpiaCTzPkF7dTwep7+/H0VRqKioyNxfqBeqMw5iNm7cyPnz53nsscf4kz/5E37yk59w4cIFNm7cCJilpY9//ONzvtCCJBU3o1rZCrWb6erqQlVVysrKcrQAOa3VI11J9yxwFgZGhkK6zBOBzeHkruvrUQ0ZRyrE3vMjJY50ar7vdCY1HznwXVxSglfst2PUbae3wzzJ1tbWIcsyuq4XXGt1Pm5fU0mVx077UJTLA1evN+nf+El02YrvynPIqfG19nhKo9sfp33YPDn3BeMMhJMMRZMEYiqhvstE/P3EUxopTae2tha32xweGgwGkSSpoFqtDcMYcZ3WiCZVeocDDIdjnDh9FoCqZQ10BVUGQgnCsSR25zQ/E7qGNdKT/iOUnjHb14fXfBhDseWIecetSbETVKXREQSXXrqqfVwsSkpKMtqDTZs2oSgKFy5cwO/3j9v2WsjGvPLKK8TjcQYGBnj00Ud544035i3rmFT1cVmYpfzaXTVTtFdfuXIFXdepqKjIdM9BYc1LymbmvaaYKdCxupi3JZ1vgZ4yFd7OEi69ZZreZethwgk1J3V+sMWs+S90KSlNsctKfyiBLMsUu53gKkaODfHT145z1/V1uIoqTBdHf5tZ9jB0GnueJ2FYOb3m96kEettbAGhsMKdXF5JL72TIssT9G2v41z0t/PJYF5+4tSmjE5oNqruG4TUPUnb2R/hanmVw/W8ip8ablSVV8wMQSmhEU2MO1IFzJD3L0e2mb095XQOR86c5duYCTWvWIxMjaS1CkSQUWcIiSyiKlPuzLE37CknTDTTdQDcM1PT3uvm9bhiZxzXdQBvz89jjvn+wDzWV4uJ5U9BbvXz0yk5RrFht00s9W6O9GTGvq/cQzuFzpFxVBBruAXLFvPlIykXEq7fh4AfQ+ibomjl9fAmR1sf09vbicrm47rrrOH78OAcOHODee+/Nu30gEKC4uHjhF3uVdHd3c/LkSWw2Gzt27ODNN9/kpZde4uLFizzwwANzvk8D4cS4z24spVGAGtX5xzDMafAA5c15N0lngpeCHgZmEcQEAgH+7M/+jFdffZWBgYGciDbdV/62IT25umo9WOw5Ez/Tb/pQVilpMGxOTi52Wdm6vHgRFmwOhXTZFaIJDYvVSuOyGjouDBEc6OaVs33cv7HWzMb426D3FEbLHhR0/lV7F6ubN6DrOr2d5oc83VpdiIr1ifjozuX8654WXj7bx+5zfTRVFLFpmY+Ny4pZX+PFYZ3Zya9/4+9Seu4xSs/+iL5NvweGgazOJMtjYAu1kZQa0W0e6hpX03r+ND2dHdSsaEbTdKThMFbr5K+xLINFllFkUGQZRZJGghAdTScTuMzVBWgiHkPTNDqunCeVTODxePD4Ssw9MgzsrunNVpHUOEp8pKyWrYVp/gjI1vFi3jxoNg8Bz2ocVhf0nzUP0pM4kRYqVqsVn89HIBDI+Ma0trbS2dk5zlA0nY3xer1L4gIimxdfNAd23nTTTWzYsIGmpiaefvppWltb+c53vsN9992XyexfLaqmMxQZL5APJzR8hZ88nnuy26sr80+vzjf0EQpTDwOzKCf97u/+LidPnuQv//IvCQaDfOc732HVqlV87nOfm4/1FTZdh83b6k055kBVVWaWxW6355SSDrWZpaQ7miuxzLNL72SUpktKNgdlIzXPKjnEt3ZfMucLpXUxp59EurKboOHkp66PUFfiZLi/h1QyQUlpGS6XqyBdeidjdZWHP3/PenY0lmC3KFzsC/PE4U7+1y9P8dF/3scXfnacHx9o41RXgNQ0Zi2lvA34V74fWYtTfupfSXobMZSZXrEY2IKtyKkIFTX1KIoF/2A/qWQCRZFJxKbWxei6mTaPJXXCcZVALEU4rhJL6iRVPW8m5WqIRULomkb7RdObJfuAZxjgdE1vCKg13AmMdOd078Phv0jSXUOw/i4mEvMahkEsqWUuoHRrEWFVwqjdaj7XhcKfaj0RHo8Hu92OzWZj+3azdXzfvn15Sy1LURtz+fJlLl++nOnEisViWK1WPvCBD3D99deTSCT4+c9/zhNPPDEnpdSBcDLv5z4cn0Crdq2jJcA/EsSUrhz3sKqq9Pb2IklS5jwGhauHgVlkYl588UXOnj1LaWkpiqLwvve9j+3bt/Oe97yHT3/60/OxxsIlPbl62Xb6+/uJx+P4fD5cLheKopDUclur32oZnVq9mPicVroCMWx2Ox5fKZIkU6lEeL7Lz3Mne3hv2vSu9yQA/6i+lzXNZumoZ6SU1LDc/LlQXXon4+EblnN7cwXhhMq5nhDHOwIc7/BzpjvEqa4gp7qC/OgA2C0y19V62bismI11PpoqilDk8f/I/Zt+n+KLP6fs9HcZ2PBJEr4V2AOXkLSZiAl1bMEWDF8TFbX19LRfYbCni+rlK1CThdVVkc7CdF45TyqZpLSyJieQtdlsUx7wDMNADQ8RDIQIpwxCCY07T5hamF95f419J1MEdQtB7TLhhGp+xdXM97oBNzaV8af3rQNJRlXcJKq34Wh9A67shhsemc+XYF4pLy+nu7ubtWvXcvr0aYaGhjh//nymey2NJElEIhF8Pt+SyMYYhpGThUmlUsiynNFd3HDDDVRVVbFnzx5OnjxJW1sbDzzwACtWrJjV39N0g8FI/v/BhKqh6Ube/+drmlAvJILgrgD7+AuNtrY2VFWloqIiRwNTqHoYmEUQo+t6Zg5KUVERfr+fqqoqLl68OOeLK2iiI8P9rG6ouo6LR8yBgOno1Waz5XQlxVMah9uGsSoSt60pX4wVZ5BHhkIOR1LYbE68JaUEhgYolyP83UsXeM8f3oRkdUEqypBUzL9q7+ILjaZrY0+HOWsp7Q9TqCnGyXBYFZoqzH/gNVUe7llfRSxltmAeaRvmWLsZ1JzrDXG4zZ/x9XHbFTbU+dhYV8zGZT6Wl7qQJIlEyWoCK+6j+MozlJ3+N/q2fpqErwm7/zKSPoNAxtCwB66wbHkjPe1XGOjpoHr5CjQtha7rBXOiioXNLEzbxRFBb9NauoeG6e2PEUnq6FYXsZZOQgmVSFbgERr5PjLyc7bb8f3yPn7N1spFvZY/unIDOkkgv0+OVZFAN9h7eZD2oSj1pS50axGBsi04ANr2QzIGtsKs4U+FJEmUlZXR39/Prl27eO655zh06BBNTU3jLhjS2ZiSkpJFWu30OX36NN3d3ZSWllJfXz/u8yzLMk1NTVRWVvLaa6/R2dnJ9773PW666SbuuOOOcYMzp2IwnMgZ9XJlIMw/vXaZd5VBo2E2XKSH475tSA/3LckfGI6d+wcUtJEpzCKI2bx5M7t37+buu+/m1ltv5ZOf/CRut5s1a9bMx/oKl/b9gAEVzWD35NXDdIdHg5hjHX4Sqs4tq8rxOBb/H6fUbWM4ksJis+ErqyQwNMAqZ4w3+8L88kQf76/bDi2v8fXkB9AtLjbU+YhFIwQG+7HZ7FRUVBT8h3s6OKwKDqtCMVDjc7KuxssHt2rEUxoD4QQHrgxxsGWYY+1+Lg9E2Hd5iH2XTXF2scuaCWhuXv0IN155hrJT/8rA9Z9AtxWR9K3AFrgMMwpkVBp88JYkMdTXkxkImYhHp12imWtSmk5/KEFPIE7HYIiOwSBS/0WqUkk69WL+7ZUQ5qFk+q2yFhmK7RIeu4TPYvDF+BOgw56Kj/BwsQOntxy3y43HYcFtt+B1WqjyOqjxOSkvsvP/vXCOf3z1Ms+d7OaR21ai2TyEbOVUepchBTug8zCsuHneXpP5xm634/F4qK2tpb6+nvb2do4ePcrOnTtztktPPC/0bIymabz8sjmkc/v27ZOutaioiHe/+92cOnWKAwcO8Oabb3LhwgU+9KEP5ZQ4JkPXDQbCuUHwP712mZNdQZxJmVt5mwYxA+np1eNLSbD09DAwiyDm0UcfzdSiv/nNb/Knf/qnhEIhvve978354gqa9KTnqutAVsYNy7JabUQSoczmCznwcTq4bCNDIe02fKVmZmidO8abEfj6C+d5z8e+yslXf8oP39rIjhXFWBWZzhGDu5q6uszU6qUexOQjE9i4bKyq9PDwDQ1mu3QgxpsXB9l/ZYjDrcN0+GO8dqGf1y708/fA953buDXxFv7Xvo164x9S6raR9DZiGb40o79vt0qUFnsZHA4w3N9DeXUdqURi3oIYwzDwR1P0BOP0BOP0BuP0BNLfJxgMJ8iWFdhQ+TVHO0jwVrKGYoeCV1HxeVx4XXZ8bidFdgseu4UiuwW3w7wtslvwOCx4FBVv+FKm4cjT/grVb3WS8DZy0y13cIO9hJS3AYdVxuOwUuSw4LYpOSWq/7KzgX969TIvne3jN29sxGF1YkgWEjXbcAQ74OILSzqIAfB6vYRCIXbt2kVHRwcnTpxg7dq1eSeCF3o25vDhwwwNDVFVVcWyZcumLDdKksT1119PbW0tr7zyCv39/fzzP/8zN998M7fccsuU5Y3BSHLUwBM40RngZJdpHnjGL6GPzOt62zFkjhOgfHx7tWEYGc+e7CBmOuXhxWTGQUxjY2Pm+8rKSh599NG5XM/SofuYeVu7Bb/fTzAYxOl04vV6sVgsRFKjrqS6YbD/yuK2VuejxGUjntTwFpcDEkSHaSxZS8tglJ+3u3lDuw+dLrY3mKWkjEtvVimpkD/cc4nDqrCivIgV5UX8l12mS3HrYITXzvez9/Igh1qG+Vr4fdxqf4s1V77LrWe3UlFSzMZlxWyo8pEI9SMPazgsElZZwiqDVQGrLGGRwSqTU5+vrylncDjAQHc75dV1qKkkhmHM+vWOJTV6g3F6Q9kBSpyeYILeYJzkJOMSnFaFKq+dCreVUgdUxVpRBjTc5bX83xs2YpUN/O3n8dZVU1ZVO+UabYHO0Y5pXaP07I8BGF73MEVOB/balXjcTqyTiN+Xl7m4dU0Fr53v59Xz/bzzumo0WxGhss04eBJaR0YQLOHPpyRJuN1uZFlm/fr1mczE3XffPW67QtbGJJPJzFiarVu35niPTEVpaSkPPPAAhw4d4vjx47z22mtcunSJu+66i6qqKlwu17jfMQyDgXBu9vOxg+YoHAkIqxKX+iM015gjYWyWwnvN5oXs6dUV49ure3p6SCQSlJSUZAa4QmHrYWCWPjH79u3j8OHDhMO5nhh/8id/MieLKngMA3pM0SvLdmT0QNXV1UiSNE4Pc7EvzFAkSXOVh/rS8f90i0WJy0pvMI7D5cTjKyEUGOLXNrr56psxvv7CeaIjbrPbGkrQdZ2+kdbq+vp6DMPI+aC/HWkoc/MbN7r5jRsbMQyDloFddP34OWqH9vFx52t8e/ge2odjPHMCwAInJ+8ykiUzmLHK4FPc3CtBS1s73+2qwqLI2G092CwKFkXGqkhYFRmLnPX9yK1VkdF1g75QYiRQiROYZHqvLEGV106V10H1yFeV10G1z7z1OixIkoR/oI9kMsHeF15DA9ZfvxGbRcbQzREL1mlcsckJf46XTmnXK9giXeilq6jacA+Srw480+t2+9iNDbx2vp9nT3Rz7/oqdKuHQOn1lMsWpO5jEOmHosLIfM4Wj8dDOBxm69atXLx4kStXrtDd3Z3j4QGFrY3Zt28fkUiE+vr6ca3i00FRFG644Qbq6+vZvXs3nZ2dPPbYY9xyyy2sXLkSl8uFx+PJBHDD0VyH6DPdQY51BCh2Wbm5qZRnTvZyuM1Pc00x4YRKqWVpNSbMGi2ZNb16fHt19nksja7reQPFQmLGQcznPvc5/umf/oldu3bl7Nzb5YocMC3jwz3gLIXSlbTt/U8gVw/T5x89aRxqTWdhCuuAalFkPA4LIYuV4vIKQoEhKowhVlcWcaHPPNE0lrmo8NgZ6Ok0O1HKynA6nWiaRlHR4mg0ChFJklhRUQTv/Z/w3ffwJ57neeB3/gd7roTYd2mQ1q4eXB4fqaSZ9UhqkNIMUrphfq8bqDqkdEhoEE7ZGLI7KZVjJCN+OvUiJhK6Tgevw5IJTDJBitdBlc9Buds2Zct/PBZF03XaL51FU1NU1NRT5Ms9YTqncug1dOzRbtx2Cy6bBZfVwPrSfwAg7/htU4g7g6Dj9uZKanwOLg9EONcbYm1lEYbFSariOmy9x+DSK7DpI9N+vkLEYrHgcDhQFIWtW7eyd+9e9u3bxwMPPJBzzC3UbEw0GmXPHnMUxNatW6/qPFFbW8uHPvQh9uzZw6VLl3jppZdob29n165dhEIhnE4nHo9n3IiB/zhonrg/uKWOFWVOM4hp9/PRGyCSUCl1v02CmFQMAp0gKVDcMO7hbF1nNoXefTrjIOZf/uVfOHr06Kzb3q4Juo6at5VrwebK1BHTojNDtpBUR/+RDoyUku5ewKnV06XEbWPA7sBbWg6XzjHQ28Vv7LqLP//lKYBMKalnpJS0fPnSculdcBpvgfpdSO37aO55huZbfovfuKGeZ599lvvuuxGrZMDgRdOvYQRdh5RuermomkFc04inDPYereDsxTY+0einfHkZcUclTm8ZqqaT0nQz8NF0MxjSdFTNIKWbP4NBhScdsNhx2WbvTAymL4ymJum4bLrzNq69ftw2lgkOdk6bTJHdijfVh1O2j1Z4zjxlTvSuXAfLbwTfshmVfxRZ4td3NfC3/3mOZ090s/aeZgzFQbhiK6W9x+Dyy0s+iAFTG9PX18f69es5ffo0AwMDXLhwYVwzRSFmY1577TWSyWSm6+hqsdvt3HnnnSxfvpw33niD8+fP093dzR133EFVVRV9gwEG4joOpxu7w8mFvjCH24bxOiy8+/oaJEPDLhuc7w0TiqfeXi3W/jZQ4+BbDmOyT4Zh5B36WOh6GJhFEFNZWZkzFOptSdfIYLbqjYQjEQYHB7FarZmZSeHE6ATjvmCcS/0RyotsbF5WvDjrnQSP3YLT6cBXah5g/AN93LLMx+b6Yo61+7lppTmqvbfdbL1ryApiBHmQJLjtc/DDD8HrX4PNv577uMUG5avNQEY15y3JMtizAkLvyL+lcn0DZy+2EQsFuKFGIUGYitVr0FFGghUzcEmmAxhNn1NDuzTxWBRdN2i/eBZNVamsXU6RtzjzuD7GFFCWwWO34nFYKHJYTG1LKg79Q6YoAczU9uHvm99v/21wloB9cmfefHxkRz3fePE8r18Y4L/e0kSZzUOoYgul/Bu07gVNBeXqArjFJm1+p2kaN9xwA88//zwHDx5kxYoVOXqFQsvG+P1+Dh06hCzLbNu2bU6fe9WqVVRVVbF79256enp46qmn2LJlC+UNzRg6REIBYpEQP9xrzuR6YEsdDquCRZJY4zM4MSxxtN3PrasriKe0GTt1L0nS06tLxycghoeHiUajeDyenAz7UjjOz/i/++/+7u945JFH+PSnPz2u3S19lX7Nkw5iardy+fJlDMOgsrISWZax2Wz4s1TvB0dKSXc0VyIXYNQvSRKlbjuDLjduj49IKEBgsJ8vvnstPYE4TRVFxCJhAkMD2Oxma/VSc+ldcFbdBTWbofsonPwprP9Q7uOKFcrSgczErqS15cW4nXb84RiBcBSf20AfuICv/voJMxaqpqPqIwENYOhgYLr1GphXXObtmO+zttFHujrS9ydDMWySSscV8yC45vqN2C1y5nctI9Oly4rslBQ5cI3pJALMEmx2j9OZpyHSB1XXm2MufMum99qOobzIzj3rq3j2RA8vnunlweuLSXpXoDlKUPyt5hiC6vFZo6WG2+3G7/ezfPly6urq6Ozs5NixYxlX3zSFlI3ZvXs3mqaxbt26eZnx5PF4uP/++zl+/DhvvfUWhw8fxnullXVbb8RV5KFlKM7hjhBum8xtyx3oWpJl5R7WlRicGIZDrcPcurqCUFx9ewQx6cGPedqrL10yH8vOwmiaVvB6GJjF2IFUKsWLL77IDTfcQGNjY+brbVNeMgzz5ARQv3OcP4zd7shp3cu0VhdgKSlNiduK1WqluMzMxvR2tuKyWTKGcOmupNq6OiRJWpIuvQuKJMFtf2x+//r/Z0YSY1EsZkbGOnEwKEkSq5aZWc/W7kHTWC84PDrALQ8WRcZhVfA4rHgdVnwuK8UuGyVuG6VuG2VFdsqL7FR47FRmCXhrfE5qi53UFTupL3VRX+pieZmLMrtBtddOX+sFNFVl5cqVrGuoMR8vdbGs2M7mlaZYs8prx223jA9gYn7TJTSNmoCj5owkdvxX8Naagd0s+e2bzWPPcye7US0ukBSiFVvMBy88P+vnLSTcbjeSZA763LVrF5Ikcfz48XHNFelszHxNhJ4uvb29HDt2DIvFwtatW+ft78iyzObNm3n/+9+P2+MlODzIod2/oqv1Ek+eNGdyvau5BJukI8eDBIYHWFdsBtOH24bRDSNnNMw1zXC6vXq8p1tLSwuQG8RIklTwnUkwy9lJf/7nf87g4CCxWCzzFY1OPd/lmsDfCtFB88DrrcvUEauqqpAkiSRKJqUfTaocbfdjs8jcunpxXXonw25RKPa4TV0MMNDTmfN4T6a12sy0FbLxUcHQfL853XzgPNLZp/NvIytQthJsEwukr2syA4TWbvOAnEilMKJDZn17AQgGg8TjcU6dOoUkSeNOSE6nc/IDna6PZGGyOP1L83+odouphXFfXXl6W0MJqyqL6A0mONzuR7cWEakcWeeVV6/quQsFSZJwuVwYhkFpaSlr165F0zQOHDiQd9vFnqn00ksvAbBhw4YFuZp3eYvZdts7qVuxGk1TOXf0AL6B45RYNe5Z7cOiSJQWOczXzw71xQ780RSX+yNEEmrOIONrEsOA4ZGLnzzt1enzWHbX21LQw8AsgphIJMLv/d7vUVJSgt1uz/l6W2DzwL1/BRsfIq7q9PX1IUkSlZWVI3qY0aj+SJuflGZw08qyqxZWzjdVJUUUl5knE/9A7+hwPV2jr9M8YdbX118TLr0LgixnsjHKG19jQrGKrJjp3QkmNTfVVWCzKPQOBYklkkhALJEyg4BAZ97fmSvC4TCGYXD8+HHUkSxMdllA1/WpywThXlP/kiYVg2M/Mr/f/vEZi3nzIUkSv77LDLCfPdGNZisiWrkFA8k0pcwynVzK+Hy+TIZl27ZtWK1WLl26RG9vb852i52NaW1t5cKFCzgcjjmbRj0Vw5EUisXCmo3b2bjrHaQkKw2Kn/fZTxEb7qXElXtCvq7SzCQfbhvGMCCS1CZ66muDnPbq3ExMKBQiEAhkfM7SLJVs+4yDmEceeeTta3AH4C6DnZ+Arb9BS0sLmqZRXl6O1WrFZrPlBDFvjUytLiSDu4nwOW0Ueb043R6SiTjBYfPKf7C3GzWVpLSsPOMLI4KYaXLdB6B0JVLfSeqH3pgkkJGhtAns451YFUWmsdbMkLV2DyLLMrHESFAQ6YPg9K3+Z0owGCQWi+XNwhiGgcvlmnyejZowg5hsTv0cYsOwbAeseMesxLz5eHBbPS6bwqGWYbpjVjS7j0TxSkhGoGXPnPyNxUaW5cz/ntPpzLwf+/btG5dJWKxsjGEYvPDCCwBs2bJlQU6EsZRGLDUahMQcZfw0up4OvRhZS3J836ucPHwAVVVBNf93NtaY2aF0uf+an2qdCEOwCyx28NTmPHT5sllmSvucwdLRw8Asgpjdu3fzqU99iqamJm666aacr7cVVjdtbWaGIq2HkSw2Uqp5MNF0g4Pp1uoC84fJhyxLlHrcmWxMWgfTm6e1eimkGAsCWYFbPwvA1rZ/xvIvd8LRH5kn93HbjgQyDt+4h9Y1mnXq1h4zsEwmsw644R4I98350sPhMLquc+zYMTRNY9WqVfh8uWubMgsT6CBHzJuMwDHTF4YdvwPemRufTYTbbuH+jTUYwHNnhzFkG9F0Senii3P2dxYbr9eLppkn7Ouuuy7Tfp0WZqZZrGzM2bNn6ezspKioiHXr1i3I3/RHcv2TfnlqiDhWnE3bad68E0WxcPr0aX7+058w0GoO6t1Y48ZukTjbEyQUT137upihy6Y2r7jBPNZkkW/o43T1MLECyGDNuMbxyU9+cj7WsfSwuTLzktJ6mIQx+uE41xvCH0txXa2XGt/SyFzUlPnwlpbT3XaZwV5z33raWwBobFiOYRgiCzNTNj+MFg8Tf+VvcfeegF/8Lrz4JfMkvv23wZ2llZIkc7rscAvE/Zm7166o5anXj9PZ5yelasiyREpVsaazIMFOkOTc57oKDMPIaGHOnDmTNwtTVFQ0eRtvPJAr5gU4+YR53/IbYeWdVyXmzcfHb2rk8UMdvHCml99cX0ukciul538CrXtMbU4BtB1fLTabDbvdjqqqKIrCzp07efHFFzlw4ACNjY05mTFJkvD7/ZSWli7I2nRd58UXzYBx+/btMxovMFsSKT2nFNQVTLK/LYzdIvHOtaWUWT1srLKx+8Ax+odDPP3aEZY3NNC01sb6ShdHuiLsO9fJPZsa0XTj2vWNybRXN417KD30cTZ6mKFokrpFnhY/4yDmYx/72HysY4khkdAVenpMD4Lq6mqsViuRLH+YQy2FNytpKnxFTiqrqzkHDPX3EI2ECA6brdXl5eXoui5aq2eKJKFv/21e7K3k/pUSloP/CG1vwit/ZXYubfww7Po90/BtZHtKV5givJj5GXLYrNRVFtPeO0xn3zANNWVEYyl8nqx/30C7Gci4rv6Elb6CP3r0KJqmsWbNmpxauSRJeYcQZjCM8XqdRAiO/8T8/obfvWoxbz7W1/rYXO/jaHuA1zs17q1Yi2ZxIvefRQq0Q8l4l9KliNvtZnh4GFmWaWxspKamhu7ubo4fP54TbKYnXBcXFy+Ib0x6yGNpaSmrVo0fMDgfDEdzszBPnR7CAO5eWURpqpcyScXtVHjfrZs4fLaNI+fbaGtrY+O6HjbWmEHMkfYANzUFCMdd+FyF340zKwZHpleX5bZXx2IxBgcHsdlsOW350ykDRpPq0srEPP/81K2K995771UtZsmgWOlq7ySRSODz+XA6nSgWK9H46Bt6YCSIuWcJBTEAy6oqsDtdJGJRLp82h1zWjUydFS69V4EkY6y9DzY8YPoM7f0WnPoZHP6e+bXyTtj1+6bHjCSZJ1xJhugAAKuXV9HeO0xrzyCNteUkkilgzBWQv838HWfxrJeZzsLEYjHOnj2LJEls2bIl87iu6/h8vsmv0sK9OY7EAJx4HJJhaLwVVt05b4MZP7qzgaPtx3n6TIB7q6zEyjdR1LMPLrwAO39nXv7mQuN2uzN6F0mSuPHGG/nZz37GsWPHaG5uzrnQWKhsTCqVYvfu3QDs3LlzQUrOSVXPKQP1hpK82RLCJsP7a0PYDBm33dTxybLM9vWNxBJJzrb28PqeN9m87S6+C5zoiRKLROgb8uNzXaNGrkNmyYjy3M6k1lazY6mqqipzbJ+uHmYwPPsxKHPJtIOY//bf/tukj0uSlBEITUU8Huehhx7i9OnTOJ1OKisr+fa3v82qVavo6+vjN3/zN7l06RJ2u51vfetb3HbbbdNd5sIgSZk3P62H0SQrxogfSHcgRutglCqvnevrJrliLUDKfUUUl1XS29HCpVNHgdHW6qWiVi94arfAh/4Z7vkyHPhneOvf4NLL5ld5M+z6Xdj0EBTXm0FJpI/rm5bx8sGztPUMoRsGSVXNM9XaMEtRUhM4Zve5S2th0lmYtWvX5mRdFEXB45lEjKsmx4t54wE48VNAgpv/cM7EvPl4YHMt/+e5M5zpCXExVIavaqsZxFzefc0EMWAGMuFwGEmSKCsro7m5mXPnznHw4EFuv/32zHYLlY154403iEQi1NTUsGzZ7IwLZ0p2FkbSUjxzvAsDuGc5+Ozgc4zPquy8fiWtfQGGgxF8Fw9R422gO5iiM6Ritw5TW+K6NrPNmenVuYMfs0W9adIXrJOh6QaBWKogTAKn/am+cuXKpF/TDWDSPPLII5w7d45jx47x/ve/n9/5HfMA84UvfIFdu3Zx4cIF/u3f/o2HH36YVGriCbyLRXpeUlrRncx6KQ+OZGHuXFu15ESwniJ3ZsaJpqmAxLJly4RL73zgrYW7/xd8+jTc/zXTxXfgHDz9R/C19fDSX5ri4KIqSrwuyouLiCdT9A4GAYlYPN+VkAHDVyA6ZLYz69NP9xqGQSgUIhqNcvbs2YyRWJp0FmZSgh3jzf2OPwapKKy8A5pun/Z6ZoPdqvD+zaZg+KmLqYxfjN62N9OZci3g8XhyOpK2b9+O1WrlwoUL9Pf352ybzsbMF6FQiH379gELl4VJaSMmdbqGJdqPv7eF1ztSWGV4byNYFRmXbfwJ1qKYJTirRaGlvYtdbvNYfaw7gmpAb/8AiUQe4f1SRs1qry5fnfNQ2qx1rB5mKvzR5LyMOJkNi1IbcDgc3HfffZkP+65duzKOgT/5yU8y4uEdO3ZQW1vLq68WnmFVdhAzdl7SobRL7xLoShqL1WqlYdlo10hZeXlmiq7IxMwTNpfpXPv7B+Dhx80TfWwIXv8qfP16ePHLEB0ede/tGUSWJaKJCU7Khm6aMvafhZ7j0H0c+s+ZHQqBDgj3my66Y4KcdBbmyJEj6LpOc3NzTtbFarVOnmZOhMysSzaxYTj5MzOjdNufzLmYNx+/eWMDsgQvXwoTsFWRdNcgRweg6/C8/+2FIrvdGsDlcmUCzr179+YEOOlszHx1Kr388sskk0kaGxtnNORR13Vz9MUszob+SBw5Oog92IKS9POLy6AZcHc9lDjA65z4c2a327l1s+mVUhK4gE+KcbzbNGtNajAwMFCQF86zJjYEkQGz8zFLM5dMJunr60NRFMrLR5sCpnOcH4oUzgVBQTiwffOb3+T9738/g4ODpFKpnNRWY2NjppV5MhKJRE4EHQyanRGpVGrOP5ChUCgz+t3lcpHSDJIjfyOcUDneEcBhldnZ4Juzv51+noX456qpKsdmd5BMxDNZGJvNNi9/eyH3a7GY0T6uuMP86juNcuAfkU7+FOn4f8Dx/2BDxTvYxzZaugbYvq4RbWSa9dTooE7ytyUFQ7LgH44Qiqc4f/4csiyzYd0aNDUFkoymaVRUVOTdh8z+DXeYfysL+ciPUNQ4+qp3otVsgwV4n+uL7dywopS9l4d4/kqKhoqt2CLPoJ79FUbN7AcRFtpn1el0EgqFMl1A69at48yZM/T29nLp0qWcUTCGYTAwMDDpTKXZ7F93dzcnTpxAkiS2bduWaf+eLpWVlQwODpJMJqdX7jIM1Mgw0d5uZENFBwZisLsTLBK8pxEssoxdkdH08cFR+r766jLWr6jl9JUu7rRd4pkBG5F4irBFwm2T6e7uprKyckE6rOaKCd+/njNYAb24AS3rsdbWVnRdz2RhNE1D0zR8vsnPW7GkSnQkC6ypxryeF6bDogcxX/nKV7h48SIvvfQSsdjEw/Cm4q//+q/58pe/PO7+559/fs5Ne4aHzUyL3W7nyJEjOY8dHpDQdIV1Po2XX/jPOf27QMZIar7xeooYTCZIxGMcPjz/V7ALtV+LyYz3UX4ntnU3smLgJRoHXqKm/1WKWYU/Cm37n8RfdTOH28NTP88MaGtrQ9cNysvLOH+5FZh4TtNYXjh8Jedne8rPPad+gY7My7a7iTz33JyudTI22yX2ovCzMzFua1zDLp5h6MQL7I1tmfqXp6CQP6tlZWWEw2HeeOMNBgcHZ6WDmcn+tbW1oWkaZWVl47xqForHL8tohszNVTpqUqcjqdExnHsStKphGgdeoTJ0nLLqD3GsYy0WTzkulx+iUW6wtHHglMHmcphfH+z5Z+z7Vzu8nx1AR9LDkWefzdyfriboun5Vx/izs/7NiZnJGKNFDWK++tWv8rOf/YwXX3wRl8uVcQDt6enJZGNaWlqmNR37i1/8Ip/5zGcyPweDQerr67n33nsnbwedBc8//zytra00Nzezbt064lYv8ZR5Bfrzly4CA3z0tuu5b/vcCdxSqRQvvPAC99xzz7wP5UqlUrR3dBGIJqgp82ai9fmodS/kfi0WV7+PD4EaRzv1M1a/+AoHEz6M3jPcN/g48ZXvxLr5w1A0+y44wzDoHQwSjMQ4dmwYRZa4a0sTbqc5SkTTdSpLPaO+NGP3T5d44Wg792yqxaqMnjTlvT9HMVLo697POz7w8VmvbzbcldL4xTf30B2I01W+A6PFQnnkPPe9Yxe4Z9epU4if1Vgslmm3BvO9fOaZZ+jr68NqtbJp06bMtoZh4HA4JszGzHT/zp8/z7Fjx1AUhbvvvntGmjlZlseVnoLBIOFweHzgFQ+aJREthW5AdzCa0WMMxWFfHyiSwUPNErVFNio9oyNwLOFOPJeextX+CvJIx9zOK9+g77a/wfDWsbJsHT9/9SirLIP0B8sp3rKJumIXdquMYRgoikJFRcWS0DZO9P7JL7wJLVC3fhc177gvc/93v/tdwHRWrq01XXzT+zsRmm5wvjeUef0dVoWmirnXSqYrKdNhRkGMpmmUlJTQ399/1bOSvva1r/HjH/+YF198Mcf588EHH+Q73/kOX/rSlzh48CCdnZ284x3vmPL5JprfZLVa5/yAkzYHqq2txWqzE9JlZEVG1XTeavUDcM/1NfNyoJuP/cn3NzxFLlwuJ4osYbfb510PsxD7tdhc1T5arbD9YzS5d3LwJz/hjG0ztyYP4Dr3czj/pGnhv+HXoOq6GT91KBJHkuD4hXYMw2Dtilq8bkfmcZfDjss+2ftvBvBWRR4NYsJ9cPZpkBXku/8X8gK/t1arlQe21PHt3Zd4ssXCu8rW4Ro4gdH6KsqmD1/1cxfKZ9VqtRKJRHJ0JTfddBO/+MUvOH78OGvXrs3JRCcSCRRFmTRDM539C4VCHDx4EMMwuP7662d0oajrembWXDZlZWXYbDYCgYC5vmTY1HKkHa5liUg8hYSU6dB/ugVUA+5aBpUuKHZbUSSwDZyg6OIvcfQcAsCQZGJ1N2MYOq6uvVQd+CsGbv9byn1FrGxawaWLFykNnCccXEHC68Q1UkZKWw5ka0YKnXHv38jUe6VyHcrI/Zqm0dvbiyRJVFdXZ8pmbrd70vc+EE4gyRbSIZ1iUebtPDddZpRrVBSFurq6cePfZ0pHRwef/exn8fv93HHHHWzevJkbbrgBgP/7f/8vb775JqtXr+a3fuu3+MEPflAwBwww28N7e3uxWq2UlpaSMuRMVHqmO0goobKpvphKj2PyJypwbDYbiiwJl94CY/WaNdhsNrpTRVx55w8IN74TQ7bC5Vfgyd+HX/weXHoF9OnZqBuGQSgaJxSJc6G9F0WW2bymPvO4bhiUeGZRjj3yfdBTsOHBcQZbC8VDO+qxKTJvdKgMlJhdSqnzLy3KWmaKphv4o9MTT6anW6epqKhg1apVpFIpDh06lLOtoihX3amkqirnz5+ntbUVu92ek+2ZivTxZKJjusfjoazIjj7UYhomZo3oMAwIZc048ifgxXaQJXigCeyyTmnnbipe+TTlb/w5jp5D6FYX4VUP0HfvPzK883MMbv00g+41WCPdlB74G9BVNq0o5zJVKJLBsf2vEwzHM39DkiQSiURGQrAkGQliqBxtr+7q6kJVVSoqKjLvhaZpUx7rhwtI0JtmxuWkP/iDP+DBBx/ki1/8IsuXL89Js61Zs2aS3xxl2bJlEyrSq6qqpmWst1iEw2EqKyux2WwYhkEq6yVMG9zdvXbpdSWNxW63k0gkhEtvgaEoCo2NjZw/f57LURfW2/4C4u0UtfwnnPw59J2Gl74M7kpYeTs03gZV683uoDyEo3F0w+DIuVYMA9Y11eBymBlNwzBwO2dhcBjshrPPgmyFO/7nVe7x7KkvcXHbmnJePNPHs4kNfBJQ2t9cEiMIeoJxYkmNYtfUGVCv10soFMo5Fu/YsYOWlhbOnTvHddddR1lZWeaxSCRyVb4xAwMDmeBo8+bNM87KTyguTkYh1I0zEaSiyEK/P46ctU/hhIqedd546gqkdHhvTZDNfS9Sc+xFLAk/AKq7isjK9xJdfheGNevErFg5sOIPuOfyX2DvP47v+KNE1/0Oroo6envCVMUjHD20l/p33ZN5fdJzqBRFmXNpwryTSpgmmJAzciDb5yyNLMuTZtwjCTUjmygkZhzE/P7v/z5Axp0xjSRJM1amL0XKy8v55Cc/SWdnJ4ZhkDBG044HW5bO1OqpcLlc+P1+4dJbgKxevTpzJbxu3TpijtUUbamEjR8xTd1OPG62VB//ifnlLIXGW2DFrVCzOdPmbGZhEgQjMS6296EoMptWj2ZhDKC4aJZZGEODrR8zDfsWCVmW+MiOel4808e/ttXwO/ZirME24t1ncNTNvOy2UIQTKkMjbqhJVcdmmfz/T5IkXC4X8fhoBqGoqIiNGzdy+PBh9u7dy/33358JctLZmNm4+IZCIa5cuUJPTw9ut5v169dP+3d1Xcfj8Yw/nqTiEOrOmRdmt1mpKvXQNxxGwvwshuKjYt1AAi63d/DXluf4sP8NlBEhb6JsPZFV7yNeswOk/J1FSauXgRv+lMrXv4j7yq8o89SzueIOvtm2kg86T9PX2cbxk6fZvPH6zO/IskwgEEBRlKV1URfuMctyRVVgG113vqGPU1U8CqmtOpsZn510Xc/79XYIYLKRZRkVmXQXX4c/Rqc/Rl2xk3U18+dIulAoiiK8YQqU9evXI8synZ2dJJNJkoYVSleCYodVd8MD34EP/Qts+y3z/tgQnPklPPs5+P4H4OW/giuvEw76zSzM2TYM4LoVtbgc5vutGwYel2PmgsZAB5z/lbmW2z8/5/s+U25aWc6qSjd9cYmWos0AJM/+anEXNQm6btA5PNqlGYxPr9U0e7p1mk2bNuF2u+nu7s5ceaeJRCIzPmarqsrQ0BAHDx4EYNu2bTkDJ6dCluXcTIaumVmC/rM5AUwaq8VCdakHSYJQLIVmGGAY+IaOUX/4//Cs9U/4qOUVJHTCdbfRf/tXGbztK8Rrd00YwKRJeRsY3vFZDGSKT/wLt1lOkJJtvJo029LfOrifgYGBnN9RFIXh4eGcYLHg6T9n3paMttvrup7RdWYHMZNl1NIOvYWIuMS+ClQ9y6X3SnrgY+WSULJPB4vFQlFR0WIvQzAGl8tFXV1d5mCk6zpJyWZqTyTFnEtUttIMYn7tX+ChH8INnzRFv8kwXHwBXvgz3D/9COz5Oy519GJRZDauHu2mkyUpR9w7bd76rmm2t+O/gqd66u3nGbfdwv/f3nuHx1Vf+/rvnq6p6l2yLMtVsiQXuWE7NGMDAUICoR2Cc+CGhCSEtJPDgRxC7oUkpNyQcn8kOcQUB+LQISaAAVMN7nK3cVOXrK6RNH32/v2xNaNeLVmj8fd9Hj+a2X3NjGevWeuz1rq6SG3e+E+3+stac/q9Sbyioal1evAFukP2zhHeOHQ6Xb+bkE6nY8mSJQBs3769l9Oi1WrD85dGSlNTE6dPn6alpYXY2Fhmzpw5/E5dyLJMXFxc7+9Gdwu4mlDjLAOj1WpJiXfg9nhIrHmX+Tt/xJz9v2Cudz+tioVjqVdzZOUfcS75Hv640Q2d9KYuxllwG5IiM+vQo1wWW01l0EFMQlp4Inff7r0ajWZqNcNrOq7+7ZFKqq+vx+fzER8fH/7MDDcvqSWCOvT2ZdROjN/v5ze/+Q3XXHMNF1xwAStWrAj/O5+QZRmv1P0rZFf51JtaPRxxcXERJaoWdBOaElxeXo5Wq6Wzs1MNFyfkgabPr2N7hjqL6Zo/wi3Pw8rvEkhdgCQH2d4Qg4LEkuAOMnY9jPn0G0juJhyWMURhWsvhxNugN8PK7w2//TnimuJ0rAYNTzerKaSYM7vp6Bh5Cee5omcaKYTLFyQwooaGajSmb1feGTNmkJSUhNPp5NChQ73WdXZ2EgiMTADudDrxeDzs3r0bUMcLjCbNbDAY+otG+3Z3HghXE/7tjzN/2zeZ8dn/YHZVUa9N437/V/lu7O9pnXMj5tixaxA7866mc9qlaAMu/rfvV8TSzjElnXiHjfb2dj744IN++k2NRkN9ff2IX7tJJTS9use4gcH0MEN910eioDfEqJ2Ye+65hw0bNnDJJZewb98+brzxRlpbW8+fCdZdBIIywS5JkdPt51CNE4tBy7LciZ0Wey4RDkzkEtIiqM3p5O5fjAazOoNJO0ga0JKIMvdq6pf9N4dXPcZBaTZ6KchyqRRTfSmxpY+R+sYdWLb8QJ155KwZ8TVp9zwFKLDkTrBGzjTgVIeJS+Yk0oSDSkMuWn8n7uORNcqkbxophKKA0zOym2VoPEhPQlOuAfbs2dOroehIozF+vx+n08nRo0fp6OggJSVlRL27QoRac/RCltWo4GA0nYT3fg7P3IjxwEZ0PieuxEIOFvyICz2PsDG4hivzTJj0OozDaIaGRJJoK74Tb0I+8YEz/H/6RznYFKRwznRMBj1lZWX9nD91N4mGhoYJG+UwboSmV/cY/Bga8TNSPcxQgt4dp5vw+CdXSjLqd/+ll17i9ddf5+6770an03H33Xfz8ssvs3Xr1om4voglKOnDv1R3lbcgK7B6VhJG3dRpUy2YuiQkJJCUlITP56O2tha/39/9hao3qamjlALVoYmdpgr7TLGgi8Hp8iIDu082ABL5M3Nov/LPNC/5EZ2Zq9VIypmD8On/B3+/GV64XU0TNZ9isJiyzV2J5vRWMFjhgrvP1cswIow6LdcumgbAG96ulNKJd5AHaEs/WdT1SSP1ZKQpJVAFvX1vrCkpKcyYMQO/3x+OpIRwu93DRhSam5sJBALh7uSjGfKoKMrAvUd87f0HhSoyVHwC//ye+pn77A0UwJl1CeUX/Y7qlQ/zD88SXEEtS1MUsmzgMI1Dv1aNnpalP8Ibk8Jy7WG+E9hAo19PyXw12vnpp59y5syZfrvJskx9ff2YZj+dM0Ll1UmzAfX9CI3xCTkxiqIMqYcZTNBb1tjJfS8d5KrffzTgiIdzxaidGLfbHR61HhMTQ2dnJ7NmzerXfj/akYzd+cNoTCUJIhtJksjNVfPc5eXlaDQaNaXUE60ejFZ16Js9HeKnoyTNptMyjVZNPKdrGtHrdcwvmI9iScQz7XN0LvsB0ldehssfgblXQUyc+qt49wZ4/t9h07/B9j/BmUO9bkJzal9UHyz/Zq8hc5FCYWYsC9NMvO1Te5qYaz6hPUKEip3eAE0dg4frO7yBEd8krFbrgA5GSUkJWq2Wo0eP0tzcHF4eqroZjLa2Nvx+P/v27cPr9ZKdnd3rF/xwKIrSq5lpmJ6pJL8bDr8C/7gN3rhXHdRpcsDCr1BzxQbOLPouPkcunT6FF4+pEcdb8o0YdZphK7d6oTOoDvoAyEY7dSX34pZiuFm3FdOpN7DbrSyaMw1FUXj33Xf7CXpDFbl9BcARQ8CrCu01WohVI2fNzc24XC7sdnu4ykqW5UH1MEMJep/+tBwFuLwgFa1m8nSgo3Zj582bx/bt21m2bBmLFy/mxz/+MXa7nYyMjOF3jhIURcHd9ePFH5TZXd6CJMFFsyMnhC6IfubMmcP27dspLy9n+fLleDyeXlOnB8LpdCLLMrv37gOgoGA+pgT1/24wGCQpNRUkRQ0/F3wJfJ1QtUsVA596D5zVsO9Z9Z85EXJWIsXPIL1tN4rRgbTsrok2e0w4YvRcWZDMz2tn4sJETMtnNDZU4LCMTgw63siyQnXrwDPjgrLS1XBSLS8eSc+YULm1y+Xq5czYbDbmz59PaWkpn376KZdffnl4/WDRGL/fT3t7Ox6Ph4MHDyJJEiUlJaOwTcbhcAysnfE41W68h16CI6+Bt0ujFJejdp7OW0OnrMXV2u04vHzcR4cfVmTomJtsIsmio6PT3auXTD90BjDYwGgDnRGCQaB/VAVAEz+Nd7PuZl35I6xr2cjxhnTyZy6hrslJdUML77//Ppdddlmv17VnM7yhhmtOCs1lEPSpkdiutgp9ozAwtB5mMEHv4VonO8qaccTouWN1bv8NziGjjsT87ne/w2RSqxZ+85vfsH//fv71r3/x5z//edwvLlLxBeXwL6OD1W24fEEWZceRYD27UQwCwWjIysrCbrfT0dFBc3MzXq93yNC2oih0dHTQ0tJCWVkZBoOB+fPnh9eHO6nqDGCygyUR4qbB/C/BtY/Bdw/DHe/ABd+BxFngaoTDL6P76NcAyMvugpjYiTZ7TGg1EhcVZBMbo+fjYD4SMpoTb41YNDtRnGn34B1Ab/D4R6e44c+fUN6kRtdGU95qt9sH/BwUFxcTExNDdXU1lZWV4eUDRWMURQkPkNyzZw+BQICZM2eOqreMVqsduLrR54LOejVlVPo31YHJLFGjf9dtgDmfB52R5s5um11+hReOqdGqW/KNWI06Yi0mkmNt/W3VGbs+u9PVf5ZEdVmvi+v/+10rSSTkLuBnwX9Di0Le4d9BcwUXLZ6N2WSkoqKC/fv399svFAUdbbXXhNPQNZqxR2VSSA8TmlwNDNlGY6BUkqIoPPWJepy7LpyB3TS52skROzF33XUXzz33HFlZWRQXFwNq0623336b7du3s2rVqom6xohmZ5lIJQkmB61Wy/Tpav+HUMVB35LQnoSGqoV0EQUFBeFcuCzLA4f9e6LRQOZiWPNT+NZO+PYeuPRB5IzFNFlmIpd87SwtmlgSHWYun2nmfbkQAGvVh7ROYkrJ5QvQ2N7/JvHGwTpeLq3BG5B5/7MGQG23P1INj1arHXSOXCiS8umnn/bSzvSNxjidTgKBAG1tbRw9ehStVsuiRYtGbFswGCQ2NnZg7YzXCeWfqCmllHzVcbnil5C1hNBQJLcv2Esw+uoJH+0+haXpOmbFa4m3qDdeg0FHSoIddCYUc8hxyQFzguqMD4Y1Feh/bbExWnbY1/FM4GJ0QTfpu3+OBQ8XL56DJEns3LmTurq6fvtpNBra29v7p3QnkwHKq/tWJimKMqgT0+kNDOhg765o4VCNk2SbkdtW5IzvNY+BETsx7e3tfP/73yctLY158+Zx11138Y9//IP6+vqJvL6IpmeX3kvnTv1RA4KpR6jUuqysDI1GM+gI+1AUprGxkfLy8l5RGEVRwhPkR0XCDFh5D8H1b/DRrB+rIfsIxmbUcUVBEh93OTExdTtobR/49ZpoFEWhaoBqpCO1Tv70wcnw89CPJEWBdu/IS3oHan4H6miYhIQE2traOHz4cHi5RqMJO7mhNJJGo2HXrl0oisK8efNG1TPKaDQOPofH44TK7erjeV+A+On9Nml2dTuX7oDC80dDURgDNlNXRZLeArZ0dGnzSZ27FK0tCUU7wqiAPmbACjqjTktxssR/B9ZzVD8Po/sMcdt/QVq8mcXzpqMoCu+8806vKq8QGo2GlpaWAddNCs1dn6NEdRyQ0+mkra0Ns9kcTjsP1R9moCiMrCg89YnqCN19yUxM+skvZBmxE/P0009TUVHB8ePH+eEPf4jb7eY///M/w05NaBzB+UR5k4s6p4dpCWbykkVTOMG5Jy8vD5PJRFNTEx0dHfh8AwtEQzeoPXv2AFBYWBj+BTao+DLKkCSJnLRksjMzKJNTMHgaoWbvpJSInnF6+/3Kberw8rN/HSEgK9xUkoUjRk9Zk4uGdjW6NpoqJaPROKDOoW/JdU+xauhxc3MzGo2GhoYGTp06hV6vD0ffR8KAJdXhlQHwtkH1LkBS00h98PhlXL5uh23zCR9tXoXFqVrmpNqJTcmB5HxImgW2FNAZ0Wg0pKSkYDAYRl72HBPXz/E26jQsTIIAOr7l/w7umBSMTYdwlP6JohnpZKUl4XK52Lp164ApO41GQ3Nz86D/D88pzWXq36S5QO8oTM8RFAN9TgJBecAU5ofHGznd2El2vJkbSyZvpEhPRq2Jyc3N5atf/SobNmzg008/5Ze//CUtLS089thjE3F9EU1o4OMlc1KipkuvYGphMpnCPTvKy8vx+Xz9foGHojANDQ1UVFRgNBrJz88Pr7NarefNfKxYRxxX5Rn5IJRSqniXVte5TSm5fUEaO3qn/XwBmYf/dYQWl59lufHcuCSbRdmqI7C7XI32Oj3+UZXzDlRuDaoeYvr06Xi93rBTC4Q/A8FgEEVR2LFjB6BqaUI6yOEYtKQ6hNepDin1daplvwNoqFp6TO/2BhT+cVR9f25YPouYtLmY4tIGTBVJkkRSUhIxMTEjd2Rsqf2OlRurJTkGTnhsfJj7A4I6M5byt7Ge+icXLsjDYlZ1RYNV5IZ6yEx6M7xwebXa6C4k6h2JHqbV7e8n6A0EZf62XT3m99bMRKeNjO+MUV1FXV0dmzZt4q677iI/P5/Fixeze/duHnjgAQ4cODBR1xixhEqrRSpJMJnMmDED6O7e29HRu4nYUFEYAIfDcY6udPIxGXQsnh7HEZNaaq2UfUyr+9z9albTSK5eNwhFUfh/753gszMdZMWb+e6ls9BIEotzVCcm9D0jy6NLKVkslkGd01DH3cOHD9PS0tJrXWguV01NDWazmYKCggGPMZh9Q0b1PK1QqTpHZC3tt9obkOnwBpH1VvyWDF6pi6fFI1OY6WBuZgLJ9uGLJxISErDZbCNzZCQN2NJ7TXk3G3QUJaqPP+rMoGz+PSiSBvuBJ3A07+OSElUfs2fPHqqrqwc+rCRRX18/ec3wfJ3qUE2dCWyq09K3yd1QepiBUklbjpyhts3D7BQb1xRHTjXyiJ2Y2bNns2LFCt566y2WLl3Ka6+9RkVFBX/729/4+te/PqppptFAi8vH0dp2bCYdJdMjry+G4Pxh5syZaLVaampq8Pl8vcS9oShMfX09lZWVmEymcBRGlmXsdvt5F0W0ORJImbEAv6Iloe0AwY4WOkbhHJwN9e3eft1P/7m/lneO1mMxarn/irmYDao2aUF2HBoJSitbw43wRpNSCpVbDxS9sdvtFBQUoCgK27dv77WuZxRm4cKFI9ZKDVlSrR4YvB09nJglPa8WjHYaNUl44ufic+Ti1sfxwt5aAG4sySbWrB9xM1GHw4HD4RiZE6Ezqs0guzDqNSzoksuUNkCjo4C2+bcjIRO389dkaJtZWjgbRVHYunXroDo0YPKa4TUeV/s4xeWAJOFyuWhsbMRoNIZTfYP1hxlI0OvxB3l2hxrJ+f5lsyLqO2PETkxaWhptbW2Ul5dTXl5ORUXFkJUQ0c6usmYU4MLZyegjJKwmOD+JjY0lIyMDRVGorKzsVWrdtyKpsLAwHOrXarXD9pWJRhyx8Vw4w85eZmHAT/DkVlpdEx+NcfuCYX1LiANVrfzPR6fQSPDDy+aQHtsthrUadcxNs+MNyBysVst3ne7ROVsDzVMKsWDBAkwmE5WVlb1Krk+fPk1TUxMOh4PZs2eP+FzDfp687dDZAI2fqVqUpLlgtIMjG1IK8DpyaJEc4dlfW46cobnTR366ncJMB8m20Q0ktdlsJCQkjMyRMdnVBnuopdYLkjToJDjcDJ4ANGWvpXP6OjQBN/GfPERhmoFpWRm43W7efffdQc8RDAZpaGgY1XWPC6Hp1V2VSaFUUkpKyrB6mIGiMK/tr6HF5WdBViyX5U/+YNeejPju+95771FbW8sDDzyARqPhoYceIi0tjZUrV/Jf//VfvPFG5I63nwhCVQMilSSYbPp275UkCbfbHY7CnDlzhurqamJiYsIR09Cv5vMRndFMkt1MrWMBAG1HttLm9k/oGIKB0kj1Tg8/f+MosgJfWZ7Domn9xbAlOWqUd2dXSikoK6OKGmk0mkGrhAwGA4sXLwa6S65lWQ47vIsXLx6xVio0pXpIvE6o2qk+zlgMlgS1ws2SAFodjR3djdX8QZnnd6uO1U0l2cRZDKPrzttFTEwMSUlJI3NkrClq+gWINWuZGw8BJeTIyLQV3oE3qRCdq574Hb/gcwWZ2KxWamtre2mLeiJJEj6fr1eX5HNCuLxaTTUPpIcZqaC3wxPghT1VgBqFiTRG9akwGAysWrWK+++/ny1btlBVVcW1117Lhg0buPLKKyfqGiMOrz/I3spWtBqJC2cJJ0Yw+cycqYr3Kisr1Y7Sbne4+dZAURi9Xj9oaeX5gM0RT2KeWhmT1roLry9I+wgHLY6Fhj5pJI8/yEP/OoLTE2DVzES+uGBgjcHiLsdmV1lLd3RtlL1tHA7HgOXWoMoE4uPjaW1t5dixYzQ1NdHe3k5SUlK4B9FIMBgMw4t/Pc7uVFL2UjUK04U/KPealPzOkXoaO3zMTbVRlOUg2Tb2RqJGo5GUlJTh0zqSpI7nkDSYdFqKu3QxpY2o6RWNjuYl/0HAmo6x6QgpB//MJcvmo9Fo2Lt3b69oVk9CrQ/OaTO8pq7y6q6ZSQP1hxmol1CLq7+g94U9VXR6g1yQl8AFeYkTd81jZFROjN/v5+OPP+bhhx9m7dq1pKSk8MADD1BQUMBPf/rTibrGiOPTU814/DIlOXE4zGLSs2DySUpKIjk5Gb/fT01NDV6vl46ODurq6qipqekVhQk1IjufscXGkZo9gxbJQR5VHDpY2qsqZjzx+IPUt/fWKf3+3ROcauhkeqKFuy+e2U9joNdJxFn0ZMebSbYZqXN6wuMJnJ7ROTF6vX5QAadGo2HZsmWAKvwODToczZBHWZaH7+Tr94Df1R2JySzp5cQ0dnjDN89AUOa5rijMjSXZJFiNZ52y1+v1pKWlDW+TVg+2VIy63roYX0AmqCgoBitNy+5D1lswV7xDTu1mli8uBtRsRV9RfYhz3gyvpUz9mzQnPCRWp9ORmKg6IbIsDxih6/t/oKnDy6v7a9BIal+YSNLChBjxJ+Piiy8mNjaWdevW8f7773PhhRfy5ptv0traypYtW7jvvvsm8jojiq3H1AZ/l4ouvYIIQa/Xk5OTA6i/umRZRpKkcBSmuLg4LNA0Go1DTq09H5CMdmxmEw2xxQA0H3qbDm9g3McQDJRGemlvNR8cb8Bm0nHfFXP7NQyTJJgWbyHBYkSSJBZ3pZR2dTXW9AeUXn1URoLNNkB7/i4yMjKYNm0aXq+XQCBARkYG6enpI7bPYrEML/71tEHjMTWllDBT1cF0tf4PBOVeAzDfO9ZAfbuXWSlWFk6LJeksojA90Wg0JCerkXOtVhtOn/XDaIOYePLitCSa4Iwbal2Exa5BWwbNS/5DrVg6+CTF0lFyp0/H4/EMqY85p83wQuXVCTPC0dmUlJRwenAgPcxAgt5NuyrxBWQumZvC4mmRWcAyYidm7dq1vPPOO7S2tvLmm29y7733smLFitF3+ZziKIrC1qOqUEuMGhBEEqGUUkVFBZIkUVtbS21tLWazmTlz5gDDNCI7n9BosdnsmHPUVvozOnZT1tg57mMIGtq9uH3dN4Y95S08+UkZGgl+tG4OKfb+KZjMuBhiDFpiDFr0OomSrpRSSBcDoxf4ms3mIfUtPSMvIZ3MSBlRVM/rhIquKqisJaqQtoumzm4tTFBW+EdXFOaGxdkk2s4+CtOTkI1JSUlkZGSQkJCA0WgMT6QOO3qWRExmG8U9ojHuHk0RfclFtBX+LyQU4nf9iotyddjtds6cOcPOnTsHPb9Go6GpqQm/fwJ7E7lawdUEplgwx/dLJcHAepi+gt6aVjdvHT6DXitx14UzJnVS9VCM+NPxox/9iGXLlqHVTn6b4cnkUI2TOqeHGUkWpidaJvtyBIIw6enpxMbG0tnZSWNj44BRGLPZPHgjsvOMGGscSoZ6w16pOcCbByrHtUqpbxqpptXNI2+pQt7bV06nKDO23z6JNkOvadV2k575mQ4MWg2HapzhCMxoBkKGGKzcGlRHZM2aNeTk5JCQkDCi4ymKMrISfTnYNQ29R3+YrlRSUFZ6Nf57/7MGats85CZZWDI9jqQJHKorSRIxMTEkJCSQlpZGamoqVqsVnU5HUJYxxKZTnKTeIvc1gjfQW1fkyr2cjtwr0AQ8pH7831y2NB+tVsv+/fvDjsNAhLohD6ZTOmuaugY/xuUA/SdXD6SHGUjQ+7ftFQRlhSvnp5GfHrlFAKI2eJTMTLHy1L8v4T/WzZnsSxEIehETExPu3vvpp59SV1eHxWIJl8mOaMjj+YTRhsWRSLstl1ipk4Zj22nu8I/LGILQbKSQz+DyBXjo9SN0eoNcPCeZqwr7p2usJh1pjt46BXuM2hulMNNBUFbYW9EKqBqN0V7nUCklgMzMzFF9PkZcou9pA08L1B9V5x2lFoFBFZU3dXoJZV+CssI/dnVrYZJspnPaFVav1+NwOEhOTiYzM5PEpCQWzsxGK8GhZgW3X8HfJ93onH87nuRidO4GZu7+b1YsVZ3i999/n/b29iHPN1E9ZKTGrsqkhBkEAgGqqqp6pdJkWcZi6f0DvK+g91RDBx8cbyBGr+X2ldPHVBl2rojcK4tQjDotq2clsTbCauUFAkmSwt17Q5N2Q1GYkHbhfI+k9sJgwWY24U9dCMBSuZT3PqsflzEEDR1e3D7VyZAVhd++fZyKZhczk61888K8ftELg05Ddnz/ajGLQYtW00MX0yulNLrrHKrcerSMqKQ6hNcJVbsBBTIXgSW+6xhKLy3MRycaqW51k5NgZvmM+HHTwowFSZKwWq3k5s6gIDUGvyxxrE2LN6AQlOVu50OjpaXkB/itGRiajrC06nHy8vLwer288847Q0ZbZFmeEEdGajqhPkjIo6amhmAwSFJSUjgaq9Vq+30P9BX0PvWpGkm6dkEGM1Miu5eUcGIEgigiOzs7fKPqGYU5X4Y8jhZ9jJ1gulpqvVq7n9cP1NLiOrsmnh5/kHpn9zGe21XJJ6eaiDXr+a8r5vb7VavRwLQE84CaA0mSsMfowrqY3eUtyF03vbGklAabbj1aRiwOV5TeU6uzloRTSa1uP4GgaousKGzqEYVJsZkiQoNhM+lYOF3VPh5s0WKzWchIisVuiUGv0yIrCgGdmebl9yHrrVjK32Kd5TCxsbE0NDSEOx8PhCRJ+P1+mpqaxvWapZZT6oPE2QPqYfpWqnX0EfQerG5jd3kLjhg9tyzLjohJ1UMhnBiBIIqw2WzhKqWFCxeGqzBsNltElkdOOkY7xoxCgtoYijUnaG6qZ3+V86zGEPRMI+043cTG7RXoNBL3Xj6XxAE0Hplx5iFvFPYYPcl2E9nxZlpcfk41qGW6Hr/cT6cxHHq9/qwr00YVhfF1guzvUVrdrYfpqYX55GQTlc0usuLNXDAzgYQJ1MKMBrNBy+Lpqq07agO4fAEkSYPNYiIpzkZGUixJsVb08dm0LP0RiqQlvvSPXDnHjE6n4+DBg5w+fXrQ42s0GrxeL62treN30aHy6uQ5A+ph+joxPfvzKIrCU5+o+395cdaA0cFIQzgxAkEUodVqWb58OVdffXU4CiNJEna7fZg9z1OMNqzmGDxJ89GicIHmEP86UDtmga9ajaQ6FpUtLn711mcAfG11LvPS+r8HKXYjjpihhdY2ow6NBkq6BkKGuoXD2KIxIx6OOAAjLqkO4WlT5/i4WyBuutqhV6Oh3eMP//qXFYW/71RvtjcsziLFHhlRGFD/7xSkOUi0GqjpkKl0ynh6OI6SJGEyGoh3WEmYtwp5+beRUMjd+WNWF+YAqj5msEZ4oWN0dHQMq6EZEYqC1KaeS47NCZ83JUWNJvXVw/QV9O4sa+ZIXTvJNiPXLkjHZor8IgDhxAgEUYbFYgnPSFEUBYfDIaIwg6EzoDGYCHZVKV2o3c9HJxopb+oc9RgCbyDIGacHUHtuPLT5CG5/kLX5qVxekNZve3uMjuQBSqz7IkkSNqM+3Kdj11mUWoMqAB/pOIGBGFVa0uvsPbU6HIXpdhK3n26mrMlFRmwMn5uVRKIlMqIwIWwxehZlqw7k9mZL2EkdCG3BF6DgS2gCblYe/z8sKJiD3+/njTfeYOfOnUP2kGlrazvrHjKGQDuSrwNsaZxpduL1esNl5NBfD9NT0BuUFZ76RE0/3bwku9ccr0hGODECQZRhNpvDX5ZarRar1TrJVxThGG3op6lday/V7ycgy7x56MyoxxCE0khBWeFXbx2jutXN3DQ7d67O7betSa8hK27koXpHjJ45qTYsBi3Hz3SEI0VuXzA84Xo0WK3WUQtKR+0QB7wQ8PTWw5jsePxBOrpeW6VHFObLi7NIdZjQREgUJoTNpAvPtdpZ46NdM4zQddk3ILMEyVnDVU1/4Yq1l6HRaCgtLWXz5s2Ddu3VaDQ0Nzef1WBlq1cV9BM3bUR6mJ6C3g+ON1DeldJbk58ybIQwUhBOjEAQZYR+dQWDwfN2yOOoMNqJSczGb0klQW5ihlTDGwfremk2hqOxw4vLq/5Cf2ZHBbvKW4i3GLh33Zx+zdq0GonsBPOobtZWkw69TsOC7DgUVIFviNGOIYDhy60HYtQOscepTq6uP6wOVswsAZ2Rph4ajJ1lLZxq6CTNYeKSuckkWAYejzCZ6LUalkyPR6uR2F/dRpM2BVkzRLRIo4NLH4DYaUg1Oymp/BO333QtNpuNuro6XnzxRaqqqgbcVZIkGhsbCQTGpskKOzHxecP2h+kp6PUHZTZ2VSR9Zdk0UuymKRO9FU6MQBBlSJIUrh4Zr5LaqMZgBUlDIF1NKX3Reoj6di8ffNYwojEE3kCQujY1jfTxiUb+sasSvVbivivmEtfnpixJkBUfg1E3uooPrUbCatR162J6OjFj0MVIkjSqAaBj6vTsdUL1blBkyFgE5gQCPQY99orCLIrMKEyIVEcMc1Nt+AIyB6rb6bBmgTTE7dNghXUPq+mzQy+R/rdVfN3xHjMSdHg8Hv71r3+xa9euAdNLkiRRX18/Jt2SpcuJUeJn9IvEBIPBXu95T0Hvm4fqqG/3MjvFxvIZ8cSbI8+ZHAzhxAgEUYjZbBZRmJGi0YDBii57KQBXmPYDsPlA7YjGEFR3pZFON3byf99Whbx3XZjHrAH6a6TYTWMWS9pj9CyaFo8ElFa0hB0sly84pplPoym3NplMo6tqkoNqFKbPqIFmV/eIgb0VrRyv7yDZZuTSeZEZhQlhNelY1KVJ2l3RQkdAB46soXeyZ8BVv4W8NaAzYq58n1saH2Ed76FBZu/evfzr1RdwNVSqwmdvO/jdIAdQFGVMPWSsHtWJaTLl4HK5cDgcYcdFp9OF9TA9Bb1uX5BNO1UB8G3Lp5FojVxnciDOr8FHAsF5Qt+OnIJhMNrQZy1EkbRM6zxIrCHI7vIWDte0sXpW8qC7NXV46fQGaff4eej1w3gDMlcVpg04HDbWrD+rBm52k45Ys55ZKTaOnWnnSF078zMcKAo4PQHiR+kE6HQ6TCbTsHN8RlVSHcLrVCMwoVED2ctR9BaaW9Qpz4qi8GxXFOb6RVmkx8ZEdPrCYtBSMj2OJz8pY3d5C53eAMTGqyXkrsbBd4zPhYvvU7VBFZ8inXiHJRWfkiVX8xyfp6YBXnr1NS6bZSZp1mLQqA6uhERQo6XRWUVSUqK6XGtQp2xrDV3/+t++Q+mkco/quAymh+kp6H11fw2tbj8Ls2MpzIolwRq5zuRAiEiMQCAQGG1gsBBMmodG9nFHeiUK8EppzaDt/X0Bmdo2D0FZ4ZE3j3HG6aUww8G/XzC937YxBi0ZZ1ntodNq1L4lXSmlXWVj794bwmazDRmNURQlPE9oVHic0HxSHUQYmw2Js2jzBPAH1DvnsTPtHK1rJ9FqYG1ByqgdsHONJEkUpNuJtxiobnVT1uRSRxA4MtVRCsOhM0HuhXDZ/0b6yks4lt7MbYn7mc0JXLKel4/4OPbP32Pb/UcMDftBCSDJAXwuJy1nqqCzHpxV0HJanQZ+5gDUlMKZw9B4AlrKwVmDxXsGRaOloquX0GB6mNCwR6fbz4t7VH3OrctyiDXrx3Xg5rlgal2tQCAQTAT6GNDo0XSllK40lQLw1uEz4bLpvlS1uFAUeGJbGaWVrSTbjPzHujn95v3otBLTRinkHQx7THepdU9dTIc3QHCUJeGgpomGGwg6prRkz9LqzCVgcvQqq377SD0A6wrSyIwzR3QUJoQ9xhAutd5d3qJWWEmSOmhRMwonz2DFUnQN/ov+m9Vrv8DqNC8aSeEDeTEvV9gxffRzUt64A/v+/8HQcpxOl5f2zoFKrxUIesHXDu5mOHMQreIHexblXf1h0tLU0v6eepgObyBc0fb8nipcviAr8xLJS7YO2Iwx0hFOjEAgEAAYbWiylwCQ3ryD+RkO2j0B/rm/tt+moTTSe8fqebm0GoNOw31XzO1XlipJkB1vHrdft3aTntwkC3FmPZXNrrCDpSjQPoYqJVBTjwOJSEMDQ0ftYPg6QQ50l1ZnL8UlxYT7q3gDQT483oAErM1PIc48NUp5rcbuUuvd5c3dXZ11BjXaNEoSY61I5nhmL13DVasXYjXpOCVN40+a26j0WLCe/CdJ7/8Hqe98A2XH/+CuPTrk8SSnGlFptc2kra0Ni8USribT6/VhPUxzlzPZ2OFl8/5aNBL829Jp2Ey6iB8xMBDCiREIBAJQU0oJecimWIzOMr48Q73pvravptcYAl9Aps7p4UR9B79/Vx22952LZ5Kb1L/8OD02Botx/KSHBp0Gi1HX3fjuLLv3gtozZiBHRa/Xj01b5XGqjkzdQdAaIWsZja7uKNEnJ5tw+YIUZcVSkD51GjEadBqWTo9HI8H+qrZwSgYAkwOs/XVQQyFJEklxaql7cryday8uITs1ng4lhqekL7Ml6Q58lgx0nWewH3+BmNe+jvyP9bDnaXDW9D9em+rElOvUIbCpqanh1zYUbQsE5XBJ/t93VOALyqyZl0pGXMykDtw8G4QTIxAIBKCWw0oapEy11PoidhJn1nO0rp0dp7uH9NW2eWju8PHQ60fwBWW+tDCD1bOS+h0u3mqYEK2HPUbXrYvpkVJq9wRG3WUYusute1bCBIPBsQ8M9Tqheg8oQUhfgM+U0KuXzTtHu1JJ+anETpEoTIjUWBNz0+x4AzL7q9p666VsaWAY3cRnnVZLYqwVWZYxGfRctjSfpfnTQYJtjXaetn6d8gt+ScfMLxCISUTTWga7Hoe/3wwvfQMOPAedXcLiLiemwq86uD31MCaT2hk6JOitanGx5cgZDFoNN5VkEWPQjquzfS4RToxAIBCAWu2hNyNlqikle+1HXJav3gg27awMOwhtLh8/f+MojR1eFmbHcuuynH6Hshi1pDuGHykwFuwmPcVZseg0Uq8bqaJA+xgHVzocjl5OTExMzNgGRQb94Hf1GDWwhJagKVwJU9/uYV9lKxaDlisL06ZMFCaEzaTvoYtp7j0oVJIgblq4wmikGA164mxmZEVGkiQKZ2Zx1coiLDEGqupbeG5PE5+lXUv92j/TuOphOnMvRzHFQsMR+OSP8Lfr4Z/fRVOtDtmsbFdf05AeRpblcL+oUPRo4/YKZAU+X5hGgtVI0hTUwoQQToxAIBCEMNqgKxJjqdvFujkJaCR471gDFS2quPLxbeUcqnGS5jDxw8vm9BtWqNdJZMdPnFjVpNcSZzGQn27HF5Q5UN0WXjfWKiWNRhN2WmRZJj4+fmwX52lTvakuPYycvYJGX/cv/K1H61GAVbOSSJsgJ28isRi0LMrpI+7tiVavCn0Z3XtvMZuwxhjDjmRKgoMvXrSIrJR4Oj0+/vnRPkqPV+NNmEtb0Z3UrfsryuWPwKx1oDdDzV6ktko6iaHR6cZkMoUjaaH+MCFB74n6Dj4+0YjFoOW6RZkYdBocUywi1hPhxAgEAkEIox3M8ZCQh9bfzrTOfSydnoA3IPOPXVV8ckbi9YNnMOlVIa/V1DsEL0kwLd7Sr0JpvHHE6Fmc01Wl1LPU2uMfdYO0EDabmgqxWCxjHxDpaYPWcrUk2J5Bmy2PkGZYUZRwKumaovQJf40mAkmSKMp0qMLqFjenGjr6v95GK9jTR33sWJsFg14XPp7JoGftsnxK5qkl+zsPn+bNTw/h8fmRkWiwzIEL/xNufRHW/G/k3IvZ4bgKIDwAFrr1MCFB71OflAHwxYWZ2Ex6EqdYX5i+TL1PkUAgEEwUBgtIWnXOD2CreJcr5qth+Rf31vDcafUr83uXzmJaQn/Ra2ZcDDGGcajw8Hao/wbBbtKzeFq3LiZ045Nleqc4RkGoGZrdbh/T/sgy+Dp6DXxsCnZHWw7XOqlt85AZF8MFeQljO0cEYDPpWZjd/dq7BppqbU1Wxb6jJDG2t8hakiSKZ2Vx5QVFmE0GKs808+LW3dQ3O/H5gzS1dYDOCNNXEbzofo4YFwDdqaSQHiYk6N1X1creylZizXquLkpHq5GIm0IjBgZCODECgUAQQpLUX9JZqi7GWvsxhZkO0h0mmjp9BBWJGxZlsHxGYr9dk2xGYs/mhhAMQEc91B+BpuNqRGOQqEqMQUtOopk0h4mGdi8Vza7wurFWKYUYcxrM16526e3Sw7jTSnDT7ei9feQMoJZV22Om7o3TZtL3KLVuGdxpjJ2mVmeNAkmSSI7vP5wzLdHBFy9aSGZyHJ1uH699tJ+DJ6tweXy0dXT3kOnoUB3fkKhXlmXMZjPNLh+yrISjMDcuzsKk15JgNUypEQMDIZwYgUAg6InRDikFoDNhaj6K3t3E5wvV9EBBnMyNJZn9drGZdKSOVePhbYfm03DmIDir1Rb1AEFfd+XJADhiDL2iMSGc7rFFYs4aj1MV9dbuB62epoQS0KhRKbcvyEcnGtFI8MUF/V+/qYRBp2FpbnepdYvLN/CGGq2qjxlqUOQA9KxY6kmM0cC65QUsnpsDisL2Q6d5e8dhGluddLo8eH1+3G43Op2OhAQ10qXT6dBoNLR0+vn0dDOfnekgxW7ksvxUJImInlc1UqZmTZVAIBBMFEabKtBMX4BU8Qm2qne5svB6suKMmM8cQNMnUmHUa8iKH/lEaECt4nE1q235g97e63wuOPmuWjJb8u9gTlCHVPYhpIt5bX8tO8ua+dJC1TkIygod3gDWc10y63WqrfBlP3LGYtq13eLgbScb8fhlFk+LY1bq6MqQI5H02Bhmp9o5Uutkd1kLM5Nt/QTeABjMkFakptqU4f4p4cdGRSbO2EFLq1N967vWSYrCgvzZpCQl8O6OA1TUNfPy+6VctGgO+q5GdcnJyWFNk8FgoN3jx+0L8nRXFObflk5Dr9UQZzFMSV1SX4QTIxAIBD3RGdU0QGYJVHyCrep9Wmd9mcIMB2UNvTfVaNSOvAPewPqiKOqN3tWkRi1Qeq+rPwxHN6sOTCgao49R0xK21H6Hsxh1FGfGYtJrOFLrpMMTCAuNnW7/uXVifC41ctSVSupMXkiwR8+UUCrp84VpU242z0BYjToWZcdypNbJrvIWPl+U3q9bcy80Gkab+LDYwR/TSmdnZ78UX3pcNl/MmsPWrVupqalh88cHwkM6+85Lau70s/VYPZUtbnISzOGeRlNd0Bti6n+aBAKBYLwx2sK6GFvddhigLT9AVrx5+FbtAR84a+HMIWg+pVbwhBwYTyvsfw6e/yq88k049rrqDGSoZd4cehFaKlS9zAAk2AwUZcYiK7C3skdKaYwjCMaM19mrtLopcSmKTk2v1bV5OFjjxGbUcWXh6Kt2IhGrUReuDtsTmmo9AcTGxmIwGAasODObzVx++eUsWrQIWZZpalIbMvbUwxiMJpo6vTyzQ50YfuuyHDSShCNGj1E39UYMDISIxAgEAkFfjDawZ4AtDW17Lcbmw7jj5vTaJMVuxG4a5Ne3oqgOiqtZvcH3WierHW2Pboayj0Ducjhs6TDnCrX3hyURNv8AqnfB0dfAka5OTO5DaCDk9tPN7CxrZtVM9Ve2P6Dg8gUwG87RV7zHqaa/2msJWlLwxM8Lr3rnqBqFuXBOUtT8+pckicJMB7ExesqbXZxq7CD9LKeUD0ZiYiK1tf3nd4Ha32fhwoWkpKSwdetW/H4/iYmq6Fyn09HqCfD6gVoa2r3MTbNT0tXjJtEWHe8DiEiMQCAQ9MdoUwWZXaXWcVXv9FrtiNGTbB9AyOv3QFu1KtJtKevtwHTUw56n1Jbxr/8ATm1Vl8+4BK78Ddy4ERb8m+rAABTfpP7dvwna69SITh9sRh1LpndXyvScZH3OBL7BAPg7oSqUSlpE0KiWacuKwrtdvWG+tDBzynXoHQpHTHep9acnm8OToccbSZJITk4esv9PRkYG119/PXPmzEGnUx1Xg8FAdYubTTvVida3LZ+mjpgwas+dc3sOiB5LBAKBYLzQaNWeMVklcORVHLUfUzP/mwAYdRoy43r86pblrqhLk9onpSdyACo+UaMulTvUKAxA3HSY+3nIu3TwfiLpCyFpNjQcU3UyliS1rX0PJEkiN8nK9EQLpxs7OV7fzpxU1YFoc/vHXjE1GkKOWlcqqSN5EbJeLa0+UNVGfbuXnAQzy3LH2AU4QrGa1KnW7x6rZ3dXSsmgm5gIh06nIzExkYaGhkEbEep0urADI8syfrQ8t6sSpyfA4mlx5Kern7OpOuhxMIQTIxAIBANhtKmOhKRFX7cXya/2YsmKN6u9NXwu1XFxt6jDDnvSVqU6Lp+9oa4H0Jkg7xKYcyUkzVV70gyFJEHRTfD2T2Dfs6rDY01Wxb49CDW+O93Yya7ylrAT4wvIePzB4TU7Z4unDQJeqClFkXR0pF0QLit+uyuVdPn8NEz66LrdGHValnRNtS6tbKXF5SNuAkuWjUYjcXFxtLS0DNtRWVEU6jpkXtpbDcBXlqvOr1GvGTwFOkWJrk+VQCAQjBdGuxqNSclHqttPYsM2TmPC4GuBjja1J0pPAl44/YHqvNSWdi9Pnqc6LrkXqSW3oyFnlaqFaT6lRjpMDkiY0WsTm0nHkunxPLe7il1lzfzb0u5ojdPtn1gnRlHUPje1pRD04U4sJGBJAaDTG2DbySa0GonrF03t3jCDkREXw8xkG8fOtLOzrJncJOuEns9iseD3+wesWOqFRsuGj8tx+4N8blYS0xPV60qcwoMeB0NoYgQCgWAgDGbQ6MK6mKSad9XlzureDkzTCfj4Udj4Jdj6kHpDN9qh4Dq47q/whf+nOjGjdWBATWsV3qg+3vesmrrpM45Ao5FYmB2HzajjZENneFIxnH333mHxtqtRqFBpdcoi5K7S6o9ONOILyJTkxDM9sf+IhmgglFIC2HG6JTxRfCIZqmIpRHmrj837a9FqJG5Zmg2ATisRN4UHPQ6GcGIEAoFgMIw2VRcD6LqEqwD4OuHwq/DSnfDCHXDoJVUPk7EQLvlvuOU5WPEtiM89+2uYuQZi4qF2n1qm3d6/UiXeYmBhuHtv90BIj1/GG5jAG2uXHkYJOTFpy1G0akrlna7eMNcuSI8qQW9PrAYdi8NTrZtp7zvVeoJITEwcNKUkyzIbdtQRkBUum5dCmkNNPyZYDVH5PkyaE3P33XeTk5ODJEmUlpaGl+fk5DB79myKi4spLi5m06ZNk3WJAoHgfMdoh8RZYHIgtZwitXU32g8eUaMuH/1GFd2aE9WqohufUauMZlysNswbL3RGmH+d+rj0GdVZ8rT12sRm0oXLZ3eVtfRaN6HRGI8TnDVIbZX4TQm4E4sAqGpxcaSuHUeMPjxAMxrRaCSKsmKxm3SUNbk43dh5Ts4rSRJJSUkDRmMOn+lg62dNGHQabizJ7toeEizRl0qCSXRirrvuOj766COmTZvWb92mTZsoLS2ltLSUG264YRKuTiAQCOhXar309KNojr+hNqSbdgGsfRhu/juU3AH2CWzkNu9q0Fug/GNoKVeb5/VAp9WwMi8xLDL1B7vLfSes1NrvgaAXuUKNwrhSFiF3lVa/c0Qtq750XjK2KBOS9qVnqfVHxxuGTPOMJ6GKpb4zlp7Y1YiswDVF6cR3CY3jLYaRdZWegkyaE7N69WoyM6NT7CUQCKIErR50MZC3BoAOQzLBxXeo6aK1D8G0FapuZqIxWFVHBmDf3yHgVhvp9SAz3szsVDtuf5DDNd39ady+4MT0MOlKJQXKPwXAlbIEWWcmKCu8e0x1Ym5YlDX+540wrMZuXcyu8hY6fROviwkRqlgKOTLl7bC9ogOrUccXu2ZpSVJ0CnpDRGR10le+8hUURWHJkiX8/Oc/Jykpadh9vF4vXm/3IDWnU/0P5vf78fvPcQvuCSBkQzTY0pNotasn0W5jtNuH1gwZJfhvepF3jjhZU5Cpzv8JTkxzs0GZ90V0B56HE1sILFwPig501nCptlkHi7MdHKl1suN0I/PTuytlmjvcw04sHvX72NEMPg+6ulIUSUNr6nJkOcjeilaaO33kJVkoyrBF1OdiIj6rWtTXXUKNgjU5XRgnqHvvQBgMBkwmE+3t7bxWocYlvrggDbMO5GAAR4weSQniPwei4/FiNO+PpJyr2Ncg5OTk8PLLL1NcXAxARUUF2dnZ+P1+7r//fg4cOMDrr78+7HF+8pOf8OCDD/Zb/swzz2A2j6EqQCAQCCKMooq/ktP0HieS1nEo8+Z+66s74ZH9OpJNCvctmPibVmL7IS448QsaLbP5eNZ9ADzxmYa9TRq+lBNkddqk3l7OKb85oKW8Q+Lu/AAz7Of+/MdaJf7fES12vcKPFwQxTOHRSC6Xi5tvvpm2tjbs9qFfzIiLxGRnq0IkvV7PPffcw6xZs0a037333sv3vve98HOn00lWVhaXXXbZsC/CVMDv97NlyxbWrFmDXh89OeZotasn0W5jtNuHLEPDYfyBAFv21bCmKH3yJjHP+HeU599nRuv7TFv7DTDFqV19Neodq77dw+MnP6W+04dx+mLSujr2ShLMTLaiG+K6R/U+ulugrRLXBy8BoGQtIadwGR1+OLB9NzoNfPe6i0g5Fx2DR8FEfVbbvX4+8X9G+a5qaozZ3LUu/5xpULz+IFsO1/PCwWOAj5uX5jCrS0xtMeqYljD1fsSHMikjIaKcmM7OTvx+P7GxsQA8++yzLFiwYET7Go1GjMb+eT+9Xh9VX6zRZk+IaLWrJ9FuY1TbF2MHVysAeq1m4p0YSQNI/TsBx2fD9NVw+n30R1+FhbeCtwXs6k0r0a5hcU48bx6qY3dlG1fHd6eU3EGJ+BGIbEf0Pna48CpgqtsNQGfGKjR6Ex8erSUgK6yelUhmom1UJp9LxvuzGqvVsXh6An/fVc2eija8soTDOLH/F9y+IIdq2vjdO8f54HgjALk2hTXzUtBo1Vt7Sqx5Sv6fHM01T5qw98477yQzM5OqqirWrl1LXl4eZ86c4aKLLqKwsJD58+fz/vvv89RTT03WJQoEAoGKcYJuyFqDWsZtSQJHFiTkQXI+pBVBXM7A+4QGQx58Qe0S3FmvDmFEbYW/YkYCADv7lFo7x6vUuqtLr7OhGmN7BQFjLO7kYgDe7uoN86WF51fRhkYjUZwVh82k41RjJ2VNE1dq3eENcPxMO4+8cZRbH9/BB8cbsRi13Lkqh2/nB8PRNpNeE/WVYTCJkZg//elPAy7fu3fvOb4SgUAgGAbjWaSkJY06N0lnBK1R/aszqf+GmoFjsqvn9fYJrSfNUZvqVe+BY/+C/C9AR506ngD43Kwk9FqJg9VtuH1BYrrEER3eAEFZOfs0h6+TQMCP0jXw0ZW8iKAxlvKmTk7UdxBvMbAuP/XszjEFccToWZAVxwfHG/jgWANFmbHjevw2t5+Gdi/bTzXx2AenqGxWu0avmZvCbStysBkkykpPhLePtkGPgxFR6SSBQCCISPQm0Awz3E+j73ZWev09i6GA9gxoaAf6CGSLbladmP2b1GnYnY1qNEdnJMVuYn6Ggz0VreyramVZrhqZURRo9/iJNZ/lkEKvE6fbj/lMVyopdQmKzsTbR04DcHlBKsaJHjoZgdi6RhB8cLyB7aebuTMgY9CdXbJDURRaXX4aOrzUtLj568enu1NHSRa+sXoGc9JUB1sOdvcD0uskHDHRH4UB4cQIBALByDB26Uu0JjCZ+zssmgm4cetNqnPSWd97ecYiSJgJTcfh1PvqdOz2WojLIcagZWluPHsqWtlV1hx2YkD9NX+2TozibqWt0012QykKGtozP0cgKPNeV2+Ym5ZEf2+YgTDptSzNVfvF7K1sodXtI9k2NmGzLCs0u3w0dnhxe4O8uq+Gv++sxO0PYjFquXVZDuvyUweNqiVYjFE5YmAgxOwkgUAgGAm2ro68SbMgfroqpjXHdw2KnMDIgy21f0M9SYLirhLrfc+oYRZ3C/jdAFwyR50kvau8pVcH2XZPAFk+i7LngJf2jk70jYfRBtx44mbit2Wyp6KFVrefuWk2CjJix378KU5WvIW8ZCud3iA7TzUPv0MfgrJCfbuHo3Xt1LZ62F3Wwt2bStmwrQy3P8iauSk8dssirpyfNqgDo9EQ7tR7PiAiMQKBQDAShtKvTOh5tWBLg7bK3sunr1ZHHTSdhKqdkLUEnDWQMIO5aXYyYmOobnVzurGT3CQ1iqQo0O4NjD3V4HHS5vZh6UoluVJKkPVW3j5yDIBrF2SM2cxowGrUsSg7jhP1HXxwopEri0Y2isIflGnq8NHU6UWWoanD2zt1lGjhG5/rTh0NRZzZGLUjBgZCRGIEAoEg0jEnqOMPeqLRQmHXbLnSZ9S/Xid4O7AYdSydHg/AzvLxq1JytTfjCchhJ6YjYyVtniA7ypoxaDVcv+j8qkrqi83YY6p1WQvuYUYQeANBqlvdHKtrp6Hdi88v8+KeKr7xtz3hqqOvf24Gv/ly8YgcGIB4y/mhhQlx3kZiAoEAweDUasOs1Wrxer39Bn5FIlqtFp3uvP14CQTjiySBIwOaTvRePmsd7H4Cakuh/ggkz1WjMUmz+NzsJF7cW82usmZuWNytU3F6/CiKMnrNhCzT1tqM1t2E0XmagMFOZ8pi3v+snqCscPHcZOKidFLySFFLrWOxGnWcaOigrKmDuWmOftt5/EEa2r20uf2Esn37q1p7VR1dOjeZ25bnjFrDNGmNGCeJ88vaLtra2nrNWZoKaLVali5dilY7NVT/Xq+Xtra2yb4MgSB6MNrA1OeGqDNCwXXq49Jn1b/+TnC3siovkRi9lmN17bT1iL7IslpuPVq8rlY6vAEs9XsAcCUvQDY6eLtrYvX1Jed3FCZErNnAguxYAN7/rLHXOpcvQFljJ8fPdNDqUh2Ypg4vv3zzKPe9fJDKZhe5iRZ++aVCvnPJrFE5MJOV7ZxszrufyoFAAJ1Oh8VimexLGRWyLOP1ejEajWimwKfVaDTS2dkZfr0FAsE4YM8Aj5NeJdfzrobSv0HZh9BaAbHZ0F5HXNJsFmbH8vHJJvZWtHDh7OTwLm1u/6gbobU1NwBg7nJiOtOWcaLZz+nGTpJsxrCY+HzHZlJ1MR8eb2TbiUa+tiqXTl+AhnYvnd7u6H8gKI+66mggdFqJRKsRu0Hi1EQYFOFE/t1wnAkGg+Kmeo7Q6XRTKmUnEEQ8OiNYk3svM9pg7lWAAvv+ri4LuJHcLayalQQM1L13dJGYoKzgbGsBORh2YtozL+Sdrg69VxUOXi1zvmHSa1mSq+qR9la0cqTOSVmjq5cDs7+qtVfV0aVzk4etOuqLQachPdbEnFQbSTYjmvP09T/vnBiBQCCY0lhT1MZ6PZl/vbrs+Ba18R1Aey1r5qgOz56KFoI9SquDsjKqlFJLawtK0I+p5TO0/k48sTNxmbN47zM1OnPz0uyzsynKyI43k5tkod0b4HBNd8fl8UgdmfQasuJjmJViJcF6/vSDGQzhxAgEAsFUIlRy3RNLIsxcA7IfDjyvLgv6mB7jIi/ZSoc3wNG63uMLRlOl1NrSBID5zC4AOlMW82m1l3ZPgMJMB3nJkTvscTKwmfQsyu6qUipvIRAcoOpode6oqo7MRi3TEs3MTLERazac985LCOHERAjt7e1YrVZuv/32Ee9TVlbGY489NoFX1Z+XXnqJwsJCiouLmTdvHvfdd1+vZloCgeAcYEkAvbn3sqIbAQmOvAredgA0rnou6Ept7OqbUvKMzIlpc/uRXapIPyTq7chYxTtHVUHvded5WfVAWHuUWr//WcPAqaPC9BGljqwmHdOTLMxIsmI/DwY6jhbhxEQImzZtYtGiRbz44ot0dHSMaJ/JcGIuvfRSSktLw/+2bNnCyy+/fE6vQSAQoIp8exKbDTmrwO+Cw6+qy+QAl05Xb3y7ynt3kPUHFFy+4VNKTW0dSEE3Wm8rptbjBPVWKu0L2V3eglGn4Yvn2cTqkaDVSCzIjsNi0FLb5hlT6sgRoycv2cr0RAtWo9BxDoZwYiKExx9/nB/96EesXr2aTZs2hZc/8cQTXHrppdx8882sWLGCJUuWcOqUqkH/+te/zrFjxyguLubqq68G4Pjx41x55ZWUlJRQWFjIH/7wh/CxJEni4YcfZsmSJUyfPp0NGzaE1/3gBz+gpKSE4uJiVq9ezbFjxwa8TpvNFq6O8ng8eL1eEdYUCCYDoxVMsb2XFd+k/j34PATUNhLLEtzExugoa3LR0N67tcRwAl+PP4inoxUAc/1eQC2tfqfcj6zApXNTxA12EGLNeq4sTCfRahhx6kiS1P1mpljJTjCHJ5ALBue8//Tl//cb+IMTlw7RayUO/XTdkNscPnyYyspK1q5dSyAQ4Oc//3mvtNLOnTvZs2cPCQkJ/OxnP+MXv/gFf/rTn3jssce45557KC0tBdTKq5tuuomNGzcyZ84cXC4Xy5YtY+nSpZSUlABq6fOOHTs4evQoJSUl3Hrrreh0On70ox/xq1/9CoC///3vfOc73+GNN94Y8Hq3bdvGnXfeyfHjx/nGN77BNddcMw6vlEAgGDX2DLVLr9LVADN5LqQVq83vPnsD5l2DXgPLsmJ447N2dpU3c3lBt56mze0n1TH4kMKGdi9ar6qlCelhOlKX8fYRVSNzvg57HAk2o55bl03j1mXTht1WktR5R4lW41lPvj7fEK9WBPD444/zla98Ba1WyxVXXMHp06c5cuRIeP3y5cuZPn06AMuWLePkyZMDHufYsWMcOnSIG2+8keLiYlasWEF7ezuHDx8Ob3PLLbcAMGfOHHQ6HXV1dQBs2bKF5cuXU1BQwE9/+tOwYzQQK1as4MCBA1RWVrJ7924+/PDDs30JBALBWNAZwNKn5Do0GHL/JpDVst6LMtVoaV9djC8g4/EP3AbBH5Rpc3nR+DtACWLpisSUWlZQ2ewizWFixYzEcTQmuogxaNFph45SazSQbDcyJ9VGemyMcGDGwHkfiRkuSjLR+P1+nn76afR6Pc88o84/cblcPP744+HIiMnU/UtJq9USCAwcAlYUhfj4+CEdkIGOVVFRwbe+9S127tzJjBkz2L9/P6tXrx722pOSkrjiiit47rnnRrS9QCCYAKwp4GpSK5MAMksgIU8dUXD6A5hxEZfOiEGz1cm+qlZ8AbnXzdLp9mPS909bNHf6kHwdgIyx9QRanxOPI5dXqyyAmy8syDhve5OMFKtRR6urv4Bap5VIsBpIsJxfwxonAuH2TTKvvvoqubm5VFdXU1ZWRllZGZ9++ilPP/00fv/Q1QN2u71Xa//Zs2djt9t7aV1OnDhBc/PQI+Hb2trQ6/WkpaWhKEovHU1fjh49Gp7d1N7ezubNmyksLByJqQKBYCLQaNRp1iEkCYq6tDGlz4CikBCjpTBJhzcgc6C69ziQtgFKrRVFobnTh9anVjlZznQ1uEtazAcn1GjOzUtEb5jh6FtNFGpQNzvFRrLNJByYcUA4MZPM448/Hk7xhJg7dy4ZGRm89tprQ+5bWFhIfn4+BQUFXH311eh0Ov75z3/y4osvhtfdfvvtuN3uIY8zf/58brzxRvLz8ykpKSE7e/Avp02bNlFQUEBRURHLly/nkksu4Y477hi5wQKBYPwxx4O+xyiV3M+pvWSajkO1OnH6whx1OGPfKiWPX8Yb6J1SanX5CQQVNF1OTEgPs023hE5fkMXT4siK71PiLeiH1aRDkvo3qBMRrPHjvE8nTTavv/76gMv37NkTfrx+/fpw9OPzn/98uBIp5LT0ZMaMGYM6P337uTQ2dg8ne/TRR3n00UfDz++///4Bj/HAAw/wwAMPDGaOQCCYLBwZ0PiZ+lijg8Ib4OPfwr5nIHMxl82I4bc7Otl1uomvrcrtVVXY5vYTZ+pOKTV1epECHiTZh8bnxNRynKDOzMb6XMDDl0uEoHckaDUS0xMtWEQF14QhIjECgUAQDRgsEBPX/Xz25WoJdvUeaDjK3EQ9qRYNde0+qlp7R2d7llp3+gK4fTJan5p2MtfvRUKmNb6IXTUezAYtVxWmIxgZwoGZWIQTIxAIBNGCLR2krq91nREKvqg+Ln0WSZJYna02Wdt9oqbXbm5fEH9QjfY2d/gA+ulhPtYsQgHW5qeK/iWCiEE4MQKBQBAt6AxqtVKI/GtBH6NWKbVVcWluDAC7ypqgT3rZ6VGjMe2eAMgBpIALFBlzvaqp+WtTPgC3iGGPgghCODECgUAQTViSQdvV1t5ogzlXAQrs+zurp5kwauFgvR9Pe2Ov3dp7VCmFojDGtlPovK04zdmUtjvIjjezaFocAkGkIJwYgUAgiCY0mt5Trudfpwp9P3sTk6+ZknQ9QQX2narp7vQLuHs0vdP4Ql161VTSp5oFAHxpYYYYMyKIKIQTIxAIBNGGOR4MVvWxNRny1qjN8A48z8U5asPLndUedO6m8C7h7JKioPWrQ2gt9Wpp9bNtBWgkuKFEpJIEkYVwYiKAnJycIbvsDoff7+fuu+8mPz+foqIi5s2bx29+85uzuqaf/OQneDyeszqGQCCYRHpOuS66Uf17+FXWZasRlx21ATSu+vBoghAavwuUIBpfB6bmo/g1Jj72z2JZbsKQc5YEgslAODFRwKOPPkpNTQ379u1j37597Nmzh7Vr157VMR988EHhxAgEUxmDGWLi1cdx02DaBeDvJL3ydWbEamnxKBxv9qFzN/TaLZxKaihFUmT2avLxoedG0RtGEIEIJyaC+c1vfkNJSQnFxcUsXbqUHTt2DLhdVVUVycnJ6HRqPwKTyUR+fn54/dNPP83SpUtZuHAhq1evZt++fciyzLp168LzmU6ePElmZibHjh3j61//OgCrVq2iuLiY+vp6nnnmGZYuXcqCBQsoKioatpuwQCCIAOw9Sq5DgyEPPM/F2aquZUdNAJ27sXvuEqgDHwFzvaqHedVdiNWo47L81HN33QLBCBFdeB7OgKBv4o6vNcB/VY9p11tvvZXvfe97AGzbto3169dz9OjRftv9r//1v1i3bh1bt25l5cqVXHLJJVx//fVotVo+/vhjnn32WT744AOMRiMffvghN998M4cOHWLjxo2UlJSwePFivv/97/PLX/6S2bNn89hjj/GnP/2JDz/8kNjYWADWrl3LTTfdhCRJlJWVsWzZMsrLyzEajWN+aQQCwQSj1YM1FdprICUf0oqgdh83GT7mLyxle42fWwuM6F31eGPU0mxJ9oEE5jNqafV7chFXFaUPOCRSIJhshBMTwezdu5eHHnqIpqYmdDodx48fx+12Y7FYem2Xn5/PyZMn+eijj9i2bRsPPPAATz/9NJs3b+aVV15h3759LF26NLx9c3MzbrebxMRENm7cyEUXXcT69eu56aabBr2W06dPc8stt1BVVYVOp6O5uZnTp08zZ86cCbNfIBCMA9Zkdcp10KsOhqzdx/SK53EYlnCsWabFIxNHM5IhNryLwVmO3tNEGelUKckilSSIWIQTM8YoyUTj8/n44he/yNatWykpKaG1tZW4uDi8Xm8/JwbAYDBw8cUXc/HFF3PHHXeQlpZGc3MziqJw22238fDDDw94nr1795KQkEB1dTWKogxaPnnjjTfy85//nOuuuw6A+Ph4oZkRCKYCkgT2NGgpg6ylEJ+L1HyKbybu5eGaheysDXDZdA1aV7c2JlSV9G6gkBlJFgozHZN08QLB0AhNTITi8Xjw+XzhidJ/+MMfBt32gw8+oLa2Nvx89+7dxMfHExsby9VXX83GjRupqKgAQJZldu1Sv6D27NnDr371K/bu3QvAI488Ej6GzWajra0t/LylpYXp06cDsHHjRlpaWsbJUoFAMOHExIHBpjo0RWrE9TrfK4DC9hq1U29oVhJ094d5Ty7i+sVZojeMIGIRkZgIYe3atej1+vDzTz/9lP/zf/4PS5YsITExkRtuuGHQfSsqKrjnnnvweDwYDAasViuvvPIKGo2GVatW8cgjj3DttdcSCATw+XxceeWVzJo1ixtvvJG//vWvpKam8tRTT7FkyRJWrlzJBRdcwPe//33WrFmD2Wzmrbfe4tFHH+W6664jNjaWiy++OOxcCQSCKYI9HRqPwYyLYOf/EN9xnAs0h9hVV0BAVgi5KZqAC1PTYTyKnl3KXH61MGPIwwoEk4mkKH0GaEQJTqcTh8NBW1sbdrs9vNzr9QJMOUGqLMs4nU7sdjsazdQIoI3ktfb7/bz++utcccUVvZy4aCLabYx2+0JEhZ2tFao+5uCLsO137NUWcm3nf/Kri8wUJGk5VtnCIt0xMnc+xNZgEU/m/oonvrpksq963IiK93AQosm2we7fAzE17oYCgUAgOHts6SBpYc4VYHKwILiffOk022sD4U0sXQMf35eLhKBXEPEIJ0YgEAjOF7Q6sKWCzgT5XwTgG7rXwroYFAVTlx5mj34RF89JGexIAkFEIJwYgUAgOJ+wJIHWCPlfQNGZuFy7HdprqeuQsXpriPHUUy4ns7B4IQaduEUIIhvxCRUIBILzCUlSRb4mB9Kcz6NF4Wvaf7KjNkBS2wGgqyppybRJvlCBYHiEEyMQCATnGzGxYLRD4ZeRJS3Xaz/gs+pGbM37AThpKyE/XfSGEUQ+wokRCASC8xF7OlhTCOZeglHys7LlJTI9R/EqOmYuWTfZVycQjAjhxAgEAsH5iD4GzAnoF6jN79Zr30BPgJ3KHK5cMneSL04gGBnCiZlkiouLKS4uZt68eWi12vDzwZrbvffeexQXF4/5fDU1NaxatWrM+wsEgijClgYJeZTHdveCqYlbSrzFMIkXJRCMHNGxd5IpLS0FoKysjOLi4vDziSI9PZ0PP/xwQs8hEAimCFod2NLQLbgZtu4AIH3R5ZN8UQLByBGRmAjlzTffZOXKlSxatIglS5awdevWUW+3YcMGiouLKSoqYvHixZSVlVFWVkZsbGx4G0mSePjhh1myZAnTp09nw4YN4XVHjhxh7dq1FBYWUlhYyGOPPQbAiRMnuPTSSyksLKS4uJiXX355Ql4DgUBwDrAkkjF3GW+Z1rJZ+hwLF10w2VckEIyY8z4S87Of/YxgMDhhx9dqtdx7772j2ufUqVP85Cc/4c0338Rut3PixAlWrVrVL0oz2HZlZWV88skn/PSnP2Xbtm2kpaXhcrkAqK+v73c+o9HIjh07OHr0KCUlJdx6660AXHPNNTz44IPcdJOaM29sbATglltu4d///d+58847OX78OMuWLWPBggVMmyZKMgWCKUdXyfVFt/yI1/dUoTdM7Zb1gvOL896JiUTeeOMNTpw4werVq8PLNBoNVVVVI9quoqKCzZs3c+utt5KWlgaA2Wwe9Hy33HILAHPmzEGn01FXV0dbWxsejyfswAAkJibS3t7Onj17+PjjjwGYOXMmK1eu5MMPPxROjEAwVTE51CnXAsEU47x3YkYbJTkXKIrCmjVreOaZZ8LLQgMg9+zZM+R2Y8FkMoUfa7VaAoHAEFv3R5Kk4TcSCASRjS0dODLZVyEQjAqhiYlA1q5dy9tvv83+/fvDy3bs2DGq7a666io2btxIbW0tAC6XK5xSGgmzZ8/GbDbz7LPPhpc1NjZis9lYuHBhWDtz4sQJPvroo17RIIFAMAXRDz5tXiCIVIQTE4Hk5eXxzDPPcOedd1JUVMTcuXN59NFHR7Tdb3/7WwBWr17NAw88wNq1aykqKuJzn/scDQ0NI74GnU7HK6+8woYNG5g/fz5FRUW88MILAPztb39j06ZNFBUVcd111/E///M/ZGdnj4vtAoFAIBCMlPM+nRQp5OTk0NraGn5+6aWXcumll4afh9JJF154YS+Bb9/tenLbbbdx22239Vve8zyKovRaFxLvghqNeeutt/rtn5eXx9tvvz2cSQKBQCAQTCgiEiMQCAQCgWBKIpwYgUAgEAgEUxLhxAgEAoFAIJiSTJoTc/fdd5OTk4MkSb00HsePH2fFihXMmjWLkpISDh06NK7nHUsJsWBsBAIBtFrtZF+GQCAQCKKUSRP2XnfddfzHf/wHK1eu7LX8zjvv5Gtf+xrr16/n+eefZ/369ezcuXPczqvT6ejs7KSzsxOdburommVZxu/34/V60WgiP4AWCAQIBAJYLJbJvhSBQCAQRCmTdhcfqK9IfX09u3btClfEfOlLX+Jb3/oWJ06cIC8vb8jjeb1evF5v+LnT6QTA7/fj9/t7bWs2mwkEAv2WRzKBQIDt27ezcuXKKeF8abVazGbzsK9xaP1Uei9GS7TbGO32hYh2O6PdPohuG6PJttHYEFF3w8rKStLS0sI3aUmSyM7OpqKiYlgn5mc/+xkPPvhgv+VvvfXWkC33pxrvv//+ZF/ChLBly5bJvoQJJ9ptjHb7QkS7ndFuH0S3jdFg22gas0aUE3M23HvvvXzve98LP3c6nWRlZXHZZZdht9sn8crGB7/fz5YtW1izZg16ffQMaItWu3oS7TZGu30hot3OaLcPotvGaLItlEkZCRHlxGRlZVFbW0sgEECn06EoChUVFSPqBms0GjEa+7fN1uv1U/4N7Um02RMiWu3qSbTbGO32hYh2O6PdPohuG6PBttFcf0QpRJOTk1m4cCEbN24E4IUXXiAzM3PYVJJAIBAIBILzj0mLxNx5551s3ryZuro61q5di81m48SJE/zpT39i/fr1PPzww9jt9vCgwdESaqc/mrBUJOP3+3G5XDidzinvZfckWu3qSbTbGO32hYh2O6PdPohuG6PJttB9u+9YnIGQlJFsNQWpqqoiKytrsi9DIBAIBALBGKisrCQzM3PIbaLWiZFlmZqaGmw2G5IkTfblnDUhoXJlZWVUCJVDRKtdPYl2G6PdvhDRbme02wfRbWM02aYoCu3t7aSnpw/bFy2ihL3jiUajGdaDm4rY7fYp/wEdiGi1qyfRbmO02xci2u2Mdvsgum2MFtscDseItosoYa9AIBAIBALBSBFOjEAgEAgEgimJcGKmCEajkQceeGDAXjhTmWi1qyfRbmO02xci2u2Mdvsgum2MZtuGImqFvQKBQCAQCKIbEYkRCAQCgUAwJRFOjEAgEAgEgimJcGIEAoFAIBBMSYQTIxAIBAKBYEoinBiBQCAQCARTEuHECAQCgUAgmJIIJ2ac8Xg8fOELX2DWrFkUFRWxZs0aTpw4AUB9fT3r1q1j5syZFBQU8MEHH4T3++pXvxre54ILLmDnzp3hdffddx/z58+nuLiY4uJi/v73vw96fpfLxU033UReXh6zZs3i+eefD6/761//yvz589HpdPz2t7+NKttGc5ypaN/69evJyMgIH+eHP/zhqO2LdBuvvfba8DGKi4vRaDS8+uqrUWXj8ePHWbNmDUVFReTn57Np06ZR2xcJNg71XXK23zMTZV+II0eOYDabueeeewY9vyzLfPvb32bGjBnk5eXxhz/8Ibxu8+bNLFq0CKPROOQxpqp9f/zjH8Ofg4KCAn73u9+NycZzhiIYV9xut7J582ZFlmVFURTl97//vfK5z31OURRF+epXv6o88MADiqIoyo4dO5SMjAzF5/MpiqIor7zyiuL3+xVFUZTXXntNmTZtWviYLS0t4cdVVVWKzWZTGhoaBjz/gw8+qNx2222KoijKqVOnlKSkJKWxsVFRFEUpLS1VDh8+rNx6663K//2//zeqbBvNcaaifbfddtuY3rOpZGNPdu7cqSQkJCherzeqbFyxYoXyl7/8RVEURamvr1eys7OVqqqqKWfjUN8lZ/s9M1H2KYqi+Hw+ZeXKlcrNN9+sfOc73xn0/E8++aRy8cUXK4FAQGlqalKys7OVgwcPKoqiKMeOHVNKS0uV++67b8hjTFX7Wltbw9u1tbUpWVlZyp49e8Zk57lAODETzM6dO8MfNIvFotTW1obXlZSUKFu2bOm3T0NDg6LT6cIf1p4cPXpUsVqtypkzZwY837x585RPPvkk/Pz6668Pf2mGGK8bYiTaNpLjjJRIsm+83rO+RJKNPfn6178+5htEXyLJRovFolRXV4fXXXXVVcqvf/3rMdnVk3NtY4ihPpfj+ZkdL/vuv/9+5dFHH1UeeOCBIT9fV1xxhfLss8+Gn//whz9U7rvvvl7bDHeM0RCJ9imKotTU1CipqakR7cSIdNIE8+ijj3LNNdfQ1NSE3+8nNTU1vC4nJ4eKiooB97niiivQ6bqHjP/ud79j9uzZLFy4kD//+c8kJycPeL6KigqmTZs27DnGg0izbaTHmar2PfrooxQWFvL5z3+e0tLSs7Kt5zEjyUYAt9vNs88+y+2333625oWvN1JsXLRoERs3bgTg1KlTbNu2jbKysiln47lmPOzbvn07n3zyCd/+9reHPd+5/B4NXWsk2ff888+Tn59PTk4OP/jBD1iwYMHZmDehCCdmAnn44Yc5ceIEP/vZz0a8z8aNG/nHP/7Bn//8517L7777bo4dO8a2bdt4+OGHaWpqGu/LHRWRaNt4vkaRZt9DDz3EyZMn2b9/P7fffjuXX345HR0doz5OTyLNxhDPP/88s2bNYv78+WM+RohIs/HJJ59k+/btFBcX893vfpdLLrmklxMxFiLNxvFmPOxzuVzcdddd/OUvf0GSpIm61DERifZdd911HDp0iGPHjrFx40aOHTt21secKIQTM0H86le/4sUXX+Rf//oXZrOZhIQEdDoddXV14W3KysrIzs4OP9+0aRMPPvggW7ZsISUlZcDjFhUVkZGRwXvvvcfhw4fDIrxvfvObAGRnZ1NeXj7oOc4H23oeJ1rsy8jIQKNR/7tee+212O32s/piiUQbQzz++OPjEoWJRBtzcnJ44YUXKC0t5ZVXXqGtrY38/PwpZ+O5YrzsO3nyJBUVFVx00UXk5OTw29/+lr/+9a/cdttttLa2hu279tprgXPzPToV7MvJyWHp0qX885//HHfbx43JzmdFI7/+9a+VhQsXKs3Nzb2W33bbbb0EW+np6WHB1qZNm5S8vDylrKys3/EOHToUfnzixAklOTlZOXz48IDnfuCBB/qJCvuK884mVx2pto3mOFPRvsrKyvB2n3zyiZKQkNBLgBcNNiqKohw/flyxWCxKW1vbmGwLEak21tXVKcFgUFEURXnjjTeUrKwsxeVyTTkbe55rojQx421fT4bTjGzYsKGf8HX//v2jOsZwRKp9PT8H9fX1ysyZM5W33nprDBaeG4QTM85UVlYqgJKbm6sUFRUpRUVFypIlSxRFUb/A1qxZo+Tl5Snz5s1T3n333fB+Op1OyczMDO9TVFQUrmi48sorlXnz5ilFRUXKokWLlH/84x+Dnr+jo0P58pe/rOTm5iozZ85UNm3aFF63YcMGJSMjQzGbzYrD4VAyMjJGJdiKZNtGc5ypaN8ll1yiFBQUKEVFRcqyZct6nT9abFQURbn33nuVr3zlK2OybSrY+Je//EWZMWOGMmvWLOXCCy9UDhw4MCVtHOq75Gy/ZybKvp4Md5MPBALKXXfdpUyfPl3Jzc1Vfvvb34bXvf3220pGRoZis9kUq9WqZGRkKK+88krU2Pe1r31NmTt3rlJUVKQUFhYqf/zjH0dl27lGUhRFmbw4kEAgEAgEAsHYEJoYgUAgEAgEUxLhxAgEAoFAIJiSCCdGIBAIBALBlEQ4MQKBQCAQCKYkwokRCAQCgUAwJRFOjEAgEAgEgimJcGIEAkHUU1ZWhiRJeDyeyb4UgUAwjggnRiAQnFMuvPBCJEni448/7rX8nnvuQZIkHnvssbM+hyRJHD169KyPIxAIIhvhxAgEgnPOrFmzePLJJ8PP/X4/mzZtIi8vbxKvSiAQTDWEEyMQCM45t9xyCy+++CJutxuAzZs3U1hYSEZGRnibJ598kjlz5hAbG8vFF1/cK7Jy4YUX8uMf/5iLLroIm83GBRdcEB5ot2LFCgAWLVqE1WrtNan5ueeeY/r06cTFxfHd7373XJgqEAgmEOHECASCc05ycjIrVqzg5ZdfBuCJJ55g/fr14fXvvfce99xzD0888QT19fVccsklXHnllXi93vA2Tz31FL///e9pamoiLS2NH//4xwBs27YNgN27d9PR0cHXvva18D5vv/02Bw8eZPfu3WzYsIF333134o0VCAQThnBiBALBpLB+/XqefPJJGhoa2LZtG9dee2143caNG1m/fj3Lli3DYDBw77334na7+eijj8LbfPWrX6WgoACDwcDNN9/Mnj17hj3ngw8+iMViITc3l9WrV49oH4FAELkIJ0YgEEwKn//859m7dy+PPPII1157LSaTKbyuurqanJyc8HONRkN2djbV1dXhZampqeHHZrOZjo6OYc85ln0EAkHkopvsCxAIBOcnBoOBG264gV//+tf9KpUyMjIoKysLP5dlmcrKyl6aGYFAIBCRGIFAMGncf//9vP322yxfvrzX8ltuuYUnnniCHTt24Pf7+cUvfoHBYGDlypUjOm5KSgonT56ciEsWCAQRhHBiBALBpJGcnMzFF1/cb/lFF13Er3/9a2699VaSkpJ466232Lx5M0ajcUTHffDBB7njjjuIjY3lL3/5y3hftkAgiBAkRVGUyb4IgUAgEAgEgtEiIjECgUAgEAimJMKJEQgEAoFAMCURToxAIBAIBIIpiXBiBAKBQCAQTEmEEyMQCAQCgWBKIpwYgUAgEAgEUxLhxAgEAoFAIJiSCCdGIBAIBALBlEQ4MQKBQCAQCKYkwokRCAQCgUAwJRFOjEAgEAgEgimJcGIEAoFAIBBMSf5/aVEs4dcUGdIAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "a3_results['year_month'] = pd.to_datetime(a3_results['year_month'])\n",
    "la6_results['year_month'] = pd.to_datetime(la6_results['year_month'])\n",
    "t5_results['year_month'] = pd.to_datetime(t5_results['year_month'])\n",
    "\n",
    "# Filter out May 2024 data\n",
    "a3_results_filtered = a3_results[a3_results['year_month'] != '2024-05']\n",
    "la6_results_filtered = la6_results[la6_results['year_month'] != '2024-05']\n",
    "t5_results_filtered = t5_results[t5_results['year_month'] != '2024-05']\n",
    "\n",
    "# Plotting\n",
    "plt.figure(figsize=(6.4, 4))\n",
    "\n",
    "# Plot share_class_one with error area from min to max for each channel\n",
    "plt.plot(a3_results_filtered['year_month'], a3_results_filtered['share_class_one'] * 100, label='Antena 3')\n",
    "plt.plot(la6_results_filtered['year_month'], la6_results_filtered['share_class_one'] * 100,label='La Sexta')\n",
    "plt.plot(t5_results_filtered['year_month'], t5_results_filtered['share_class_one'] * 100, label='Telecinco', color='grey')\n",
    "plt.fill_between(a3_results_filtered['year_month'], a3_results_filtered['share_class_one_10th'] * 100, a3_results_filtered['share_class_one_90th'] * 100, alpha=0.2) # , label='Antena 3 Range (10th to 90th Percentile)'\n",
    "plt.fill_between(la6_results_filtered['year_month'], la6_results_filtered['share_class_one_10th'] * 100, la6_results_filtered['share_class_one_90th'] * 100, alpha=0.2) # , label='La Sexta Range (10th to 90th Percentile)'\n",
    "plt.fill_between(t5_results_filtered['year_month'], t5_results_filtered['share_class_one_10th'] * 100, t5_results_filtered['share_class_one_90th'] * 100, alpha=0.2, color='grey') # , label='Telecinco Range (10th to 90th Percentile)'\n",
    "\n",
    "# Formatting the plot\n",
    "plt.xlabel('Month')\n",
    "plt.ylabel('War Images (%)')\n",
    "# plt.title('Share of Class One with Error Area')\n",
    "plt.legend()\n",
    "plt.grid(True)\n",
    "\n",
    "\n",
    "plt.savefig(\n",
    "    'report_charts/bias_corrected_trend.pdf',\n",
    "    bbox_inches='tight',\n",
    ")\n",
    "\n",
    "\n",
    "# Show plot\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}