diff --git "a/code.ipynb" "b/code.ipynb" new file mode 100644--- /dev/null +++ "b/code.ipynb" @@ -0,0 +1,2547 @@ +{ + "nbformat": 4, + "nbformat_minor": 0, + "metadata": { + "colab": { + "provenance": [] + }, + "kernelspec": { + "name": "python3", + "display_name": "Python 3" + }, + "language_info": { + "name": "python" + } + }, + "cells": [ + { + "cell_type": "code", + "source": [ + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "\n", + "# Data from the table\n", + "llms = ['BloomGPT', 'Code Tutor', 'Copilot', 'LLaMa']\n", + "remember = [10, 8, 4, 5]\n", + "understand = [8, 9, 5, 5]\n", + "apply = [9, 9, 9, 3]\n", + "analyze = [9, 8, 7, 2]\n", + "evaluate = [10, 11, 8, 3]\n", + "create = [12, 6, 7, 3]\n", + "\n", + "\n", + "fig, axes = plt.subplots(2, 2, figsize=(13, 10), sharey=True)\n", + "\n", + "# Flatten the axes array for easy iteration\n", + "axes = axes.flatten()\n", + "\n", + "\n", + "# Setting the bar width\n", + "bar_width = 0.15\n", + "index = np.arange(len(llms))\n", + "\n", + "# Plotting the bars for each Bloom's Taxonomy level\n", + "plt.bar(index, remember, bar_width, label='Remember')\n", + "plt.bar(index + bar_width, understand, bar_width, label='Understand')\n", + "plt.bar(index + 2*bar_width, apply, bar_width, label='Apply')\n", + "plt.bar(index + 3*bar_width, analyze, bar_width, label='Analyze')\n", + "plt.bar(index + 4*bar_width, evaluate, bar_width, label='Evaluate')\n", + "plt.bar(index + 5*bar_width, create, bar_width, label='Create')\n", + "\n", + "# Labeling the axes and title\n", + "plt.xlabel('LLMs')\n", + "plt.ylabel('Scores')\n", + "plt.title('LLMs vs. Bloom’s Taxonomy Cognitive Levels')\n", + "\n", + "# Adding the LLMs to the x-axis\n", + "plt.xticks(index + 2.5*bar_width, llms)\n", + "\n", + "# Adding the legend\n", + "plt.legend()\n", + "\n", + "# Displaying the plot\n", + "plt.show()\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 472 + }, + "id": "SzxRMf052Tcs", + "outputId": "cc74dc34-80ce-4300-f602-bd5218c2e3c7" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "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\n" + }, + "metadata": {} + } + ] + }, + { + "cell_type": "code", + "source": [ + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "\n", + "# Data from the table\n", + "bloom_levels = ['Remember', 'Understand', 'Apply', 'Analyze', 'Evaluate', 'Create']\n", + "llms = ['BloomGPT', 'Code Tutor', 'Copilot', 'LLaMa']\n", + "bloomgpt = [10, 8, 9, 9, 10, 12]\n", + "code_tutor = [8, 9, 9, 8, 11, 6]\n", + "copilot = [4, 5, 9, 7, 8, 7]\n", + "llama = [5, 5, 3, 2, 3, 3]\n", + "\n", + "# Setting the bar width\n", + "bar_width = 0.2\n", + "index = np.arange(len(bloom_levels))\n", + "\n", + "# Create a figure and axis\n", + "fig, ax = plt.subplots(figsize=(10, 6))\n", + "\n", + "\n", + "# Plotting the bars for each LLM\n", + "ax.bar(index, bloomgpt, bar_width, label='BloomGPT')\n", + "ax.bar(index + bar_width, code_tutor, bar_width, label='Code Tutor')\n", + "ax.bar(index + 2*bar_width, copilot, bar_width, label='Copilot')\n", + "ax.bar(index + 3*bar_width, llama, bar_width, label='LLaMa')\n", + "\n", + "\n", + "# Labeling the axes and title\n", + "ax.set_ylabel('number of questions', fontsize=12)\n", + "\n", + "# Adding the Bloom's levels to the x-axis\n", + "ax.set_xticks(index + 1.5*bar_width)\n", + "ax.set_xticklabels(bloom_levels, rotation=45, ha='right')\n", + "\n", + "# Adding the legend\n", + "ax.legend()\n", + "\n", + "# Displaying the plot\n", + "plt.tight_layout()\n", + "plt.show()\n", + "\n", + "\n", + "\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 541 + }, + "id": "rSmTNBL825M6", + "outputId": "644df934-4368-44bb-be6c-c6b85b98c6c0" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "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\n" + }, + "metadata": {} + } + ] + }, + { + "cell_type": "code", + "source": [ + "import pandas as pd\n", + "\n", + "# Load the dataset\n", + "# Replace 'your_dataset.csv' with the path to your actual file\n", + "df = pd.read_csv('/content/sample_data/Dataset for analysis-csv.csv')\n", + "\n", + "# Reshape the data from wide to long format\n", + "# This step assumes that each expert's columns are structured in a similar naming pattern\n", + "long_df = pd.melt(df,\n", + " id_vars=['question_num', 'bloom_level', 'llm'],\n", + " value_vars=[\n", + " 'expert1_bloom', 'expert1_subject', 'expert1_persian',\n", + " 'expert2_bloom', 'expert2_subject', 'expert2_persian',\n", + " 'expert3_bloom', 'expert3_subject', 'expert3_persian',\n", + " 'expert4_bloom', 'expert4_subject', 'expert4_persian',\n", + " 'expert5_bloom', 'expert5_subject', 'expert5_persian'\n", + " ],\n", + " var_name='expert_aspect',\n", + " value_name='score')\n", + "\n", + "# Extract expert and aspect from 'expert_aspect' column\n", + "long_df[['expert', 'aspect']] = long_df['expert_aspect'].str.split('_', expand=True)\n", + "\n", + "# Drop the original 'expert_aspect' column\n", + "long_df.drop(columns=['expert_aspect'], inplace=True)\n", + "\n", + "# Calculate the average score for each expert for each aspect for each llm\n", + "average_scores = long_df.groupby(['llm', 'expert', 'aspect'])['score'].mean().reset_index()\n", + "\n", + "# Output the result\n", + "print(average_scores)\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "jUquvmG1trgQ", + "outputId": "38630f74-23ca-428c-dafd-a9c003229f30" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + " llm expert aspect score\n", + "0 1 expert1 bloom 4.137931\n", + "1 1 expert1 persian 4.827586\n", + "2 1 expert1 subject 4.672414\n", + "3 1 expert2 bloom 3.517241\n", + "4 1 expert2 persian 4.517241\n", + "5 1 expert2 subject 3.103448\n", + "6 1 expert3 bloom 4.431034\n", + "7 1 expert3 persian 4.982759\n", + "8 1 expert3 subject 4.931034\n", + "9 1 expert4 bloom 3.500000\n", + "10 1 expert4 persian 3.603448\n", + "11 1 expert4 subject 3.672414\n", + "12 1 expert5 bloom 3.310345\n", + "13 1 expert5 persian 3.586207\n", + "14 1 expert5 subject 3.844828\n", + "15 2 expert1 bloom 4.392157\n", + "16 2 expert1 persian 4.823529\n", + "17 2 expert1 subject 4.784314\n", + "18 2 expert2 bloom 3.588235\n", + "19 2 expert2 persian 4.333333\n", + "20 2 expert2 subject 2.882353\n", + "21 2 expert3 bloom 4.686275\n", + "22 2 expert3 persian 4.941176\n", + "23 2 expert3 subject 4.921569\n", + "24 2 expert4 bloom 3.431373\n", + "25 2 expert4 persian 3.549020\n", + "26 2 expert4 subject 3.647059\n", + "27 2 expert5 bloom 3.313725\n", + "28 2 expert5 persian 3.549020\n", + "29 2 expert5 subject 3.705882\n", + "30 3 expert1 bloom 4.175000\n", + "31 3 expert1 persian 4.825000\n", + "32 3 expert1 subject 4.575000\n", + "33 3 expert2 bloom 3.325000\n", + "34 3 expert2 persian 4.625000\n", + "35 3 expert2 subject 2.900000\n", + "36 3 expert3 bloom 4.725000\n", + "37 3 expert3 persian 4.975000\n", + "38 3 expert3 subject 4.950000\n", + "39 3 expert4 bloom 3.400000\n", + "40 3 expert4 persian 3.800000\n", + "41 3 expert4 subject 3.725000\n", + "42 3 expert5 bloom 2.950000\n", + "43 3 expert5 persian 3.725000\n", + "44 3 expert5 subject 3.800000\n", + "45 4 expert1 bloom 3.523810\n", + "46 4 expert1 persian 1.809524\n", + "47 4 expert1 subject 3.952381\n", + "48 4 expert2 bloom 2.380952\n", + "49 4 expert2 persian 3.333333\n", + "50 4 expert2 subject 2.238095\n", + "51 4 expert3 bloom 4.047619\n", + "52 4 expert3 persian 4.285714\n", + "53 4 expert3 subject 4.952381\n", + "54 4 expert4 bloom 3.333333\n", + "55 4 expert4 persian 3.190476\n", + "56 4 expert4 subject 3.666667\n", + "57 4 expert5 bloom 2.857143\n", + "58 4 expert5 persian 2.952381\n", + "59 4 expert5 subject 3.761905\n" + ] + } + ] + }, + { + "cell_type": "code", + "source": [ + "\n", + "# Reshape the data from wide to long format\n", + "long_df = pd.melt(\n", + " df,\n", + " id_vars=['question_num', 'bloom_level', 'llm'],\n", + " value_vars=[\n", + " 'expert1_bloom', 'expert1_subject', 'expert1_persian',\n", + " 'expert2_bloom', 'expert2_subject', 'expert2_persian',\n", + " 'expert3_bloom', 'expert3_subject', 'expert3_persian',\n", + " 'expert4_bloom', 'expert4_subject', 'expert4_persian',\n", + " 'expert5_bloom', 'expert5_subject', 'expert5_persian'\n", + " ],\n", + " var_name='expert_aspect',\n", + " value_name='score'\n", + ")\n", + "\n", + "# Extract expert and aspect from 'expert_aspect' column\n", + "long_df[['expert', 'aspect']] = long_df['expert_aspect'].str.split('_', expand=True)\n", + "\n", + "# Drop the original 'expert_aspect' column\n", + "long_df.drop(columns=['expert_aspect'], inplace=True)\n", + "\n", + "# Calculate the average score for each expert for each aspect for each llm\n", + "average_scores = long_df.groupby(['llm', 'expert', 'aspect'])['score'].mean().reset_index()\n", + "\n", + "# Create the bar plot\n", + "plt.figure(figsize=(10, 6))\n", + "\n", + "# Use seaborn to create a bar plot\n", + "sns.barplot(\n", + " data=average_scores,\n", + " x='aspect',\n", + " y='score',\n", + " hue='llm',\n", + " ci=None\n", + ")\n", + "\n", + "# Customize the plot\n", + "plt.title('Average Scores by Aspect for Each LLM')\n", + "plt.xlabel('Aspect')\n", + "plt.ylabel('Average Score')\n", + "plt.legend(title='LLM')\n", + "\n", + "# Show the plot\n", + "plt.show()\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 654 + }, + "id": "TzjYs6zJvPxW", + "outputId": "536ca1d8-245f-4fdb-ee6c-2304b0d93e49" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":29: FutureWarning: \n", + "\n", + "The `ci` parameter is deprecated. Use `errorbar=None` for the same effect.\n", + "\n", + " sns.barplot(\n" + ] + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0EAAAIjCAYAAADFthA8AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABLV0lEQVR4nO3dd3RUVf/+/WvSJkBI6JBApEvvStUkFKkSQTooTYEbQxNRb74WRFGKgOG20ERQCASQIqIIEekCQqgiKChNWugpQCDJef7gyfwcEyADMxngvF9rzVrOPnvO/pxxZpgr+5w9FsMwDAEAAACASXi4uwAAAAAAyE6EIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIACA073zzjuyWCw6d+6cu0t5KCUmJurFF19UkSJFZLFYNGTIEHeX5HSzZs2SxWLR9u3b3V0KgIcQIQiA03z22WeyWCyqU6eOu0u571y/fl2TJk1SjRo15O/vrzx58qhSpUrq27evDhw44O7yHgq1a9eWxWLR5MmT3V2Kwz777DPNmjUry/0/+OADzZo1S/3799fs2bP1/PPPu644SSVKlJDFYsn01rx5c5eOfbeyEsTXrl0ri8Wir7/++rb7Sj/WF198MdPtb7zxhq0PwR94MHi5uwAAD4+oqCiVKFFCv/zyiw4dOqQyZcq4u6T7Rrt27bRixQp16dJFffr00Y0bN3TgwAEtX75c9evXV/ny5d1d4gPt4MGD2rZtm0qUKKGoqCj179/f3SU55LPPPlOBAgXUs2fPLPX/6aefVLduXY0YMcK1hf1D9erV9corr2RoDwoKyrYa3MnX11eLFi3SZ599Jh8fH7tt8+bNk6+vr65du+am6gA4ihAEwCkOHz6sn3/+WYsXL1a/fv0UFRWVrV/QJCktLU3Xr1+Xr69vto57J9u2bdPy5cv1/vvv6//+7//stn3yySe6dOlSttVy7do1+fj4yMPj4ToRYM6cOSpUqJAmTJig9u3b68iRIypRooS7y3KZuLg4VaxY0Wn7S0lJUVpaWoYv9/9UtGhRPffcc04b80HTvHlzLVu2TCtWrNAzzzxja//55591+PBhtWvXTosWLXJjhQAc8XD9KwjAbaKiopQ3b161atVK7du3V1RUlG3bjRs3lC9fPvXq1SvD4+Lj4+Xr66thw4bZ2pKTkzVixAiVKVNGVqtVwcHBeu2115ScnGz3WIvFogEDBigqKkqVKlWS1WrVDz/8IEkaP3686tevr/z58ytHjhyqVatWpqe8XL16VYMGDVKBAgWUO3duhYeH68SJE7JYLHrnnXfs+p44cUK9e/dW4cKFZbVaValSJX3xxRd3fG7+/PNPSVKDBg0ybPP09FT+/PkzjPPCCy8oKChIVqtVJUuWVP/+/XX9+nVbn7/++ksdOnRQvnz5lDNnTtWtW1ffffed3X7ST/WJjo7Wm2++qaJFiypnzpyKj4+XJG3dulXNmzdXQECAcubMqdDQUG3atMluHwkJCRoyZIhKlCghq9WqQoUK6amnntKOHTvueNySdO7cOXXs2FH+/v7Knz+/Bg8ebPfX8tDQUFWrVi3Tx5YrV07NmjXL0jhz585V+/bt9fTTTysgIEBz587N0CcrxxIWFqbKlSsrNjZW9evXV44cOVSyZElNmTIlw/6y+jqVboa02rVrK2fOnMqbN69CQkK0atUqSTdPNdu3b5/WrVtnO6UqLCws0+NM/396+PBhfffdd7b+R44ckXQzHL3wwgsqXLiwfH19Va1aNX355Zd2+zhy5IgsFovGjx+vyMhIlS5dWlarVb/99luWnuvb2bNnj3r27KlSpUrJ19dXRYoUUe/evXX+/PkMfbPyOpduPs9Dhw5VwYIFlStXLrVt21Znz56951odVbRoUYWEhGR4bUVFRalKlSqqXLlyttcE4O4xEwTAKaKiovTss8/Kx8dHXbp00eTJk7Vt2zY9/vjj8vb2Vtu2bbV48WJNnTrV7q/NS5cuVXJysjp37izp5mxOeHi4Nm7cqL59+6pChQrau3evPvroI/3xxx9aunSp3bg//fSTFixYoAEDBqhAgQK2v/5PmjRJ4eHh6tatm65fv67o6Gh16NBBy5cvV6tWrWyP79mzpxYsWKDnn39edevW1bp16+y2pztz5ozq1q1rC14FCxbUihUr9MILLyg+Pv62F6YXL17c9hw1aNBAXl63/ug9efKkateurUuXLqlv374qX768Tpw4oa+//lpXrlyRj4+Pzpw5o/r16+vKlSsaNGiQ8ufPry+//FLh4eH6+uuv1bZtW7t9vvfee/Lx8dGwYcOUnJwsHx8f/fTTT2rRooVq1aqlESNGyMPDQzNnzlSjRo20YcMG1a5dW5L0n//8R19//bUGDBigihUr6vz589q4caP279+vmjVr3vI40nXs2FElSpTQ6NGjtWXLFv3vf//TxYsX9dVXX0mSnn/+efXp00e//vqr3ZfIbdu26Y8//tCbb755xzG2bt2qQ4cOaebMmfLx8dGzzz6rqKioDLNuWT2WixcvqmXLlurYsaO6dOmiBQsWqH///vLx8VHv3r0lOfY6HTlypN555x3Vr19f7777rnx8fLR161b99NNPatq0qSIjIzVw4ED5+fnpjTfekCQVLlw402OtUKGCZs+erZdfflnFihWznZ5WsGBBXb16VWFhYTp06JAGDBigkiVLauHCherZs6cuXbqkwYMH2+1r5syZunbtmvr27Sur1ap8+fLd9nm+ceNGpte75MqVSzly5JAkxcTE6K+//lKvXr1UpEgR7du3T9OmTdO+ffu0ZcsWWSwWSVl7nacbOHCg8ubNqxEjRujIkSOKjIzUgAEDNH/+/NvW6wpdu3bV4MGDlZiYKD8/P6WkpGjhwoUaOnQop8IBDxoDAO7R9u3bDUlGTEyMYRiGkZaWZhQrVswYPHiwrc/KlSsNSca3335r99iWLVsapUqVst2fPXu24eHhYWzYsMGu35QpUwxJxqZNm2xtkgwPDw9j3759GWq6cuWK3f3r168blStXNho1amRri42NNSQZQ4YMsevbs2dPQ5IxYsQIW9sLL7xgBAYGGufOnbPr27lzZyMgICDDeP+UlpZmhIaGGpKMwoULG126dDE+/fRT4+jRoxn6du/e3fDw8DC2bduW6X4MwzCGDBliSLJ7jhISEoySJUsaJUqUMFJTUw3DMIw1a9YYkoxSpUrZ1ZeWlmaULVvWaNasmW2f6c9ZyZIljaeeesrWFhAQYERERNzy2G5lxIgRhiQjPDzcrv2ll14yJBm7d+82DMMwLl26ZPj6+hqvv/66Xb9BgwYZuXLlMhITE+841oABA4zg4GDbsaxatcqQZOzcudOuX1aOJf3/04QJE2xtycnJRvXq1Y1ChQoZ169fNwwj66/TgwcPGh4eHkbbtm1t/1/S/fO5r1SpkhEaGnrHY01XvHhxo1WrVnZtkZGRhiRjzpw5trbr168b9erVM/z8/Iz4+HjDMAzj8OHDhiTD39/fiIuLy/J4kjK9jR492tYvs/fBvHnzDEnG+vXrbW1ZeZ3PnDnTkGQ0adLE7rl6+eWXDU9PT+PSpUu3rTn9NXj27Nlb9kl/jyxcuPC2+5JkREREGBcuXDB8fHyM2bNnG4ZhGN99951hsViMI0eOZGk8APcPTocDcM+ioqJUuHBhNWzYUNLN09Q6deqk6OhopaamSpIaNWqkAgUK2P319uLFi4qJiVGnTp1sbQsXLlSFChVUvnx5nTt3znZr1KiRJGnNmjV2Y4eGhmZ6bUT6X6bTx7l8+bKefPJJu1Of0k+de+mll+weO3DgQLv7hmFo0aJFat26tQzDsKurWbNmunz58m1PD7NYLFq5cqVGjRqlvHnzat68eYqIiFDx4sXVqVMn2zVBaWlpWrp0qVq3bq3HHnss0/1I0vfff6/atWvriSeesG3z8/NT3759deTIkQynNfXo0cPu+di1a5cOHjyorl276vz587ZjSUpKUuPGjbV+/XqlpaVJkvLkyaOtW7fq5MmTtzy+24mIiLC7n/7cfv/995KkgIAAPfPMM5o3b54Mw5Akpaamav78+WrTpo1y5cp12/2npKRo/vz56tSpk+35adSokQoVKmR3SqYjx+Ll5aV+/frZ7vv4+Khfv36Ki4tTbGyspKy/TpcuXaq0tDS9/fbbGa7DSq/XWb7//nsVKVJEXbp0sbV5e3tr0KBBSkxM1Lp16+z6t2vXTgULFszy/uvUqaOYmJgMt3+O98/X2bVr13Tu3DnVrVtXkmzvkay+ztP17dvXru3JJ59Uamqqjh49muXanSVv3rxq3ry55s2bJ+nmaZj169e3zfYCeHBwOhyAe5Kamqro6Gg1bNhQhw8ftrXXqVNHEyZM0OrVq9W0aVN5eXmpXbt2mjt3rpKTk2W1WrV48WLduHHDLgQdPHhQ+/fvv+WXs7i4OLv7JUuWzLTf8uXLNWrUKO3atcvuGo1/fpk6evSoPDw8Muzj36vanT17VpcuXdK0adM0bdq0LNX1b1arVW+88YbeeOMNnTp1SuvWrdOkSZO0YMECeXt7a86cOTp79qzi4+PveG3B0aNHM12GvEKFCrbt/9zHv4/v4MGDkm6Go1u5fPmy8ubNq3HjxqlHjx4KDg5WrVq11LJlS3Xv3l2lSpW6bY3pypYta3e/dOnS8vDwsF3DIkndu3fX/PnztWHDBoWEhOjHH3/UmTNnsrTs86pVq3T27FnVrl1bhw4dsrU3bNhQ8+bN09ixY23hI6vHEhQUlCF8Pfroo5JuXk9Tt27dLL9O//zzT3l4eDh1EYNbOXr0qMqWLZshbP3zdfFPt3rv3EqBAgXUpEmT2/a5cOGCRo4cqejo6AzvicuXL0tSll/n6R555BG7+3nz5pV0848b7tC1a1c9//zzOnbsmJYuXapx48a5pQ4A94YQBOCe/PTTTzp16pSio6MVHR2dYXtUVJSaNm0qSercubOmTp2qFStWqE2bNlqwYIHKly9vd2F8WlqaqlSpookTJ2Y6XnBwsN39f/7lOd2GDRsUHh6ukJAQffbZZwoMDJS3t7dmzpyZ6QXzd5I+K/Lcc8/dMjhUrVo1y/sLDAxU586d1a5dO1WqVEkLFixw6DdiHPXv5yj9eD788ENVr14908f4+flJunlNz5NPPqklS5Zo1apV+vDDDzV27FgtXrxYLVq0cLiWzGY/mjVrpsKFC2vOnDkKCQnRnDlzVKRIkTt+4ZZkm+3p2LFjptvXrVtnm6F05rE4+jq9H2X23rlXHTt21M8//6xXX31V1atXl5+fn9LS0tS8eXPb685Rnp6embanzxxmt/DwcFmtVvXo0UPJycm3fO0BuL8RggDck6ioKBUqVEiffvpphm2LFy/WkiVLNGXKFOXIkUMhISEKDAzU/Pnz9cQTT+inn36yXQiernTp0tq9e7caN25816cLLVq0SL6+vlq5cqWsVqutfebMmXb9ihcvrrS0NB0+fNhuxuKfMwrSzYvOc+fOrdTU1Cx9Mc8qb29vVa1aVQcPHtS5c+dUqFAh+fv769dff73t44oXL67ff/89Q3v6j67e6dSc0qVLS5L8/f2zdDyBgYF66aWX9NJLLykuLk41a9bU+++/n6XgcPDgQbsZh0OHDiktLc1u+WpPT0917dpVs2bN0tixY7V06VL16dPnll9+0yUlJembb75Rp06d1L59+wzbBw0apKioKFsIyuqxnDx5UklJSXazQX/88Yck2erO6uu0dOnSSktL02+//XbLwCk559S44sWLa8+ePUpLS7ObDcrq6+JeXbx4UatXr9bIkSP19ttv29rTZx7TFSxYMEuv8/tVjhw51KZNG82ZM0ctWrRQgQIF3F0SgLvANUEA7trVq1e1ePFiPf3002rfvn2G24ABA5SQkKBly5ZJkjw8PNS+fXt9++23mj17tlJSUuxOhZNu/iX5xIkTmj59eqbjJSUl3bEuT09PWSwW2/VI0s3TmP69slz68sufffaZXfvHH3+cYX/pvwGS2Re3Oy3Xe/DgQR07dixD+6VLl7R582blzZtXBQsWlIeHh9q0aaNvv/1W27dvz9A//S/fLVu21C+//KLNmzfbtiUlJWnatGkqUaLEHU+9qlWrlkqXLq3x48crMTHxlseTmppqO4UpXaFChRQUFJTpMtCZ+Xc4Tn9u/x2gnn/+eV28eFH9+vVTYmJiln6PZsmSJUpKSlJERESmr7+nn35aixYtUnJyskPHkpKSoqlTp9ruX79+XVOnTlXBggVVq1YtSVl/nbZp00YeHh569913M8yE/HMmI1euXPf8e1EtW7bU6dOn7a67S0lJ0ccffyw/Pz+Fhobe0/7vJD20/nuGJjIy0u5+Vl/n97Nhw4ZpxIgReuutt9xdCoC7xEwQgLu2bNkyJSQkKDw8PNPtdevWVcGCBRUVFWULO506ddLHH3+sESNGqEqVKrbrFdI9//zzWrBggf7zn/9ozZo1atCggVJTU3XgwAEtWLBAK1euzPRi6n9q1aqVJk6cqObNm6tr166Ki4vTp59+qjJlymjPnj22frVq1VK7du0UGRmp8+fP25bITv+r/z//Oj9mzBitWbNGderUUZ8+fVSxYkVduHBBO3bs0I8//qgLFy7csp7du3era9euatGihZ588knly5dPJ06c0JdffqmTJ08qMjLS9gXygw8+0KpVqxQaGmpbevnUqVNauHChNm7cqDx58ui///2v5s2bpxYtWmjQoEHKly+fvvzySx0+fFiLFi264w+henh46PPPP1eLFi1UqVIl9erVS0WLFtWJEye0Zs0a+fv769tvv1VCQoKKFSum9u3bq1q1avLz89OPP/6obdu2acKECbcdI93hw4cVHh6u5s2ba/PmzZozZ466du2a4beBatSoocqVK9sWHMjK8ttRUVHKnz+/6tevn+n28PBwTZ8+Xd99950aNWqU5WMJCgrS2LFjdeTIET366KOaP3++du3apWnTpsnb21tS1l+nZcqU0RtvvKH33ntPTz75pJ599llZrVZt27ZNQUFBGj16tKSbr8XJkydr1KhRKlOmjAoVKmRbZCGr+vbtq6lTp6pnz56KjY1ViRIl9PXXX2vTpk2KjIxU7ty5Hdrfv504cUJz5szJ0O7n56c2bdrI399fISEhGjdunG7cuKGiRYtq1apVdtcKpsvK69xZJk6cqJw5c9q1eXh42C2hvmjRItuM2T+lX0P2b9WqVbvl71sBeEC4b2E6AA+61q1bG76+vkZSUtIt+/Ts2dPw9va2LS2dlpZmBAcHG5KMUaNGZfqY69evG2PHjjUqVapkWK1WI2/evEatWrWMkSNHGpcvX7b10/+/bG1mZsyYYZQtW9awWq1G+fLljZkzZ9qWsP2npKQkIyIiwsiXL5/h5+dntGnTxvj9998NScaYMWPs+p45c8aIiIgwgoODDW9vb6NIkSJG48aNjWnTpt32eTpz5owxZswYIzQ01AgMDDS8vLyMvHnzGo0aNTK+/vrrDP2PHj1qdO/e3ShYsKBhtVqNUqVKGREREUZycrKtz59//mm0b9/eyJMnj+Hr62vUrl3bWL58ud1+7rT8786dO41nn33WyJ8/v2G1Wo3ixYsbHTt2NFavXm0Yxs2loV999VWjWrVqRu7cuY1cuXIZ1apVMz777LPbHq9h/L/liX/77Tejffv2Ru7cuY28efMaAwYMMK5evZrpY8aNG2dIMj744IM77v/MmTOGl5eX8fzzz9+yz5UrV4ycOXMabdu2zfKxhIaGGpUqVTK2b99u1KtXz/D19TWKFy9ufPLJJxn2n9XXqWEYxhdffGHUqFHD1i80NNS2pLxhGMbp06eNVq1aGblz5zYk3XG57MyWyE5/Xnr16mUUKFDA8PHxMapUqWLMnDnTrk/6Etkffvjhbcf493i6xRLZxYsXt/X7+++/jbZt2xp58uQxAgICjA4dOhgnT57MsOS8Ydz5dZ6+RPa/l9FOf12vWbPmtjWnvwYzu3l6etrt61a39CXQb/dZ8+/xWCIbeDBYDOMBmHcGgGy0a9cu1ahRQ3PmzFG3bt3cXY5pTJo0SS+//LKOHDmSYUWw7BIWFqZz5849sNerAACyhmuCAJja1atXM7RFRkbKw8NDISEhbqjInAzD0IwZMxQaGuq2AAQAMA+uCQJgauPGjVNsbKwaNmwoLy8vrVixQitWrFDfvn0fiGWOH3RJSUlatmyZ1qxZo7179+qbb75xd0kAABMgBAEwtfr16ysmJkbvvfeeEhMT9cgjj+idd97JsHQ3XOPs2bPq2rWr8uTJo//7v/+75SIbAAA4E9cEAQAAADAVrgkCAAAAYCqEIAAAAACm8kBfE5SWlqaTJ08qd+7cdj9qCAAAAMBcDMNQQkKCgoKC7vjD4Q90CDp58iSrNwEAAACwOX78uIoVK3bbPg90CMqdO7ekmwfq7+/v5moAAAAAuEt8fLyCg4NtGeF2HugQlH4KnL+/PyEIAAAAQJYuk2FhBAAAAACmQggCAAAAYCqEIAAAAACm8kBfEwQAAADgJsMwlJKSotTUVHeX4hKenp7y8vJyyk/jEIIAAACAB9z169d16tQpXblyxd2luFTOnDkVGBgoHx+fe9oPIQgAAAB4gKWlpenw4cPy9PRUUFCQfHx8nDJbcj8xDEPXr1/X2bNndfjwYZUtW/aOP4h6O4QgAAAA4AF2/fp1paWlKTg4WDlz5nR3OS6TI0cOeXt76+jRo7p+/bp8fX3vel8sjAAAAAA8BO5lZuRB4axjfPifKQAAAAD4B0IQAAAAAFMhBAEAAAAwFUIQAAAAAPXs2VNt2rTJdFuJEiUUGRmZ6bYjR47IYrHI09NTJ06csNt26tQp22/7HDlyxLkF3wNCEAAAAIB7VrRoUX311Vd2bV9++aWKFi3qpopujRAEAAAA4J716NFDM2fOtGubOXOmevTo4aaKbo0QBAAAAOCehYeH6+LFi9q4caMkaePGjbp48aJat27t5soyIgQBAAAAuGfe3t567rnn9MUXX0iSvvjiCz333HPy9vZ2c2UZEYIAAAAAOEXv3r21cOFCnT59WgsXLlTv3r3dXVKmCEEAAAAAnKJKlSoqX768unTpogoVKqhy5cruLilThCAAAAAATtO7d2+tXbv2vp0FkiQvdxcAAIAzvNV+uLtLcJr3vh7t7hIAmNTly5e1a9cuu7b8+fNLkk6cOJFhW/HixTPso0+fPurQoYPy5MnjoirvHSEIAAAAgCRp7dq1qlGjhl3bCy+8IEkaP368xo8fb7dt9uzZeuKJJ+zavLy8VKBAAdcWeo8IQQAAAAA0a9YszZo1664eaxjGLbdVr179ttvdgWuCAAAAAJgKIQgAAACAqRCCAAAAAJgKIQgAAACAqRCCAAAAAJgKq8MBQBbF7dvt7hKcqlClau4uAQAAt2AmCAAAAICpEIIAAAAAmAohCAAAAICpcE0QAAAA8JDK7utZH5TrTZkJAgAAAOAW69evV+vWrRUUFCSLxaKlS5dmy7iEIAAAAABukZSUpGrVqunTTz/N1nE5HQ4ATGpd5JfuLgEAYHItWrRQixYtsn1cZoIAAAAAmAohCAAAAICpEIIAAAAAmAohCAAAAICpEIIAAAAAmAqrwwEAAABwi8TERB06dMh2//Dhw9q1a5fy5cunRx55xGXjEoIAAACAh1ShStXcXcJtbd++XQ0bNrTdHzp0qCSpR48emjVrlsvGJQQBAAAAcIuwsDAZhpHt43JNEAAAAABTIQQBAAAAMBVCEAAAAABTIQQBAAAAMBVCEAAAAABTIQQBAAAAMBVCEAAAAABTIQQBAAAAMBVCEAAAAABT8XJ3AQAAAABcY13kl9k6XuiQHtk63t0iBAEAADggbt9ud5fgNIUqVXN3CTC50aNHa/HixTpw4IBy5Mih+vXra+zYsSpXrpxLx71vTocbM2aMLBaLhgwZ4u5SAAAAAGSDdevWKSIiQlu2bFFMTIxu3Lihpk2bKikpyaXj3hczQdu2bdPUqVNVtWpVd5cCAAAAIJv88MMPdvdnzZqlQoUKKTY2ViEhIS4b1+0zQYmJierWrZumT5+uvHnzurscAAAAAG5y+fJlSVK+fPlcOo7bQ1BERIRatWqlJk2a3LFvcnKy4uPj7W4AAAAAHnxpaWkaMmSIGjRooMqVK7t0LLeeDhcdHa0dO3Zo27ZtWeo/evRojRw50sVVAQAAAMhuERER+vXXX7Vx40aXj+W2EHT8+HENHjxYMTEx8vX1zdJjhg8frqFDh9rux8fHKzg42FUlAgAAPNSye/lkV/tx4wF3l+BU73092t0lZJsBAwZo+fLlWr9+vYoVK+by8dwWgmJjYxUXF6eaNWva2lJTU7V+/Xp98sknSk5Olqenp91jrFarrFZrdpcKAAAAwAUMw9DAgQO1ZMkSrV27ViVLlsyWcd0Wgho3bqy9e/fatfXq1Uvly5fX66+/niEAAQAAAHi4REREaO7cufrmm2+UO3dunT59WpIUEBCgHDlyuGxct4Wg3LlzZ7jgKVeuXMqfP7/LL4QCAAAAzCB0SA93l3BbkydPliSFhYXZtc+cOVM9e/Z02bj3xe8EAQAAADAfwzDcMu59FYLWrl3r7hIAAAAAPOTc/jtBAAAAAJCdCEEAAAAATIUQBAAAAMBUCEEAAAAATIUQBAAAAMBU7qvV4ZB91kV+6e4SnOZ+X/8eAAAA9xdmggAAAACYCiEIAAAAgKlwOhwAAADwkHqr/fBsHe+9r0dn63h3i5kgAAAAAG4xefJkVa1aVf7+/vL391e9evW0YsUKl49LCAIAAADgFsWKFdOYMWMUGxur7du3q1GjRnrmmWe0b98+l47L6XAAAAAA3KJ169Z2999//31NnjxZW7ZsUaVKlVw2LiEIAAAAgNulpqZq4cKFSkpKUr169Vw6FiEIAAAAgNvs3btX9erV07Vr1+Tn56clS5aoYsWKLh2Ta4IAAAAAuE25cuW0a9cubd26Vf3791ePHj3022+/uXRMZoKyKG7fbneXAAAAADx0fHx8VKZMGUlSrVq1tG3bNk2aNElTp0512ZjMBAEAAAC4b6SlpSk5OdmlYzATBAAAAMAthg8frhYtWuiRRx5RQkKC5s6dq7Vr12rlypUuHZcQBAAAADyk3vt6tLtLuK24uDh1795dp06dUkBAgKpWraqVK1fqqaeecum4hCAAAAAAbjFjxgy3jMs1QQAAAABMhZkgPPDeaj/c3SU41f0+bQ0AALJH1dINstSvSFAhvT5ikFJuGPKw3J9zHJWqlHd3CXbuz2cJAAAAAFyEEAQAAADAVAhBAAAAAEyFEAQAAADAVAhBAAAAAEyFEAQAAADAVAhBAAAAAEyF3wkCAAAAHlJd2ryQrePNWzojW8e7W8wEAQAAAHC7MWPGyGKxaMiQIS4fixAEAAAAwK22bdumqVOnqmrVqtkyHiEIAAAAgNskJiaqW7dumj59uvLmzZstYxKCAAAAALhNRESEWrVqpSZNmmTbmCyMAAAAAMAtoqOjtWPHDm3bti1bxyUEAQAAAMh2x48f1+DBgxUTEyNfX99sHZsQBAAAACDbxcbGKi4uTjVr1rS1paamav369frkk0+UnJwsT09Pl4xNCAIAAACQ7Ro3bqy9e/fatfXq1Uvly5fX66+/7rIAJBGCAAAAALhB7ty5VblyZbu2XLlyKX/+/BnanY0QBAAAADyk5i2d4e4S7kuEIAAAAAD3hbVr12bLOPxOEAAAAABTIQQBAAAAMBVCEAAAAABTIQQBAAAAMBVCEAAAAPAAMwxDMgzJcHclrmcYzjlIQhAAAADwALt8KUE3bqTIMFLdXYrLXblyRZLk7e19T/thiWwAAADgAXbt6jVtWLNZTVs1VN48eWWxeEoWd1dl79q1a/f0eMMwdOXKFcXFxSlPnjzy9PS8p/0RggAAAIAH3PIlMZKkJxvWk7e3l2S5v1KQl7dz6smTJ4+KFClyz/shBAEAAAAPOMMw9O3iVYr5fp0C8vrLcp+FoGUx8+55H97e3vc8A5SOEAQAAAA8JK5dS9a1U2fdXUYGvr6+7i7BDgsjAAAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVL3cXAMBe1dIN3F2CU+35c5O7SwAAALDDTBAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAU/FydwEAAMBe1dIN3F2CU+35c5O7SwAAO8wEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAUyEEAQAAADAVQhAAAAAAU3FrCJo8ebKqVq0qf39/+fv7q169elqxYoU7SwIAAADwkHNrCCpWrJjGjBmj2NhYbd++XY0aNdIzzzyjffv2ubMsAAAAAA8xt/5YauvWre3uv//++5o8ebK2bNmiSpUquakqAAAAAA8zt4agf0pNTdXChQuVlJSkevXqZdonOTlZycnJtvvx8fHZVR4AAACAh4TbF0bYu3ev/Pz8ZLVa9Z///EdLlixRxYoVM+07evRoBQQE2G7BwcHZXC0AAACAB53bQ1C5cuW0a9cubd26Vf3791ePHj3022+/Zdp3+PDhunz5su12/PjxbK4WAAAAwIPO7afD+fj4qEyZMpKkWrVqadu2bZo0aZKmTp2aoa/VapXVas3uEgEAAAA8RNw+E/RvaWlpdtf9AAAAAIAzuXUmaPjw4WrRooUeeeQRJSQkaO7cuVq7dq1WrlzpzrIAAAAAPMTcGoLi4uLUvXt3nTp1SgEBAapatapWrlypp556yp1lAQAAAHiIuTUEzZgxw53DAwAAADCh++6aIAAAAABwJUIQAAAAAFMhBAEAAAAwFUIQAAAAAFMhBAEAAAAwFUIQAAAAAFMhBAEAAAAwFUIQAAAAAFMhBAEAAAAwFUIQAAAAAFMhBAEAAAAwFUIQAAAAAFMhBAEAAAAwFUIQAAAAAFMhBAEAAAAwlbsKQRs2bNBzzz2nevXq6cSJE5Kk2bNna+PGjU4tDgAAAACczeEQtGjRIjVr1kw5cuTQzp07lZycLEm6fPmyPvjgA6cXCAAAAADO5HAIGjVqlKZMmaLp06fL29vb1t6gQQPt2LHDqcUBAAAAgLM5HIJ+//13hYSEZGgPCAjQpUuXnFETAAAAALiMwyGoSJEiOnToUIb2jRs3qlSpUk4pCgAAAABcxeEQ1KdPHw0ePFhbt26VxWLRyZMnFRUVpWHDhql///6uqBEAAAAAnMbL0Qf897//VVpamho3bqwrV64oJCREVqtVw4YN08CBA11RIwAAAAA4jUMhKDU1VZs2bVJERIReffVVHTp0SImJiapYsaL8/PxcVSMAAAAAOI1DIcjT01NNmzbV/v37lSdPHlWsWNFVdQEAAACASzh8TVDlypX1119/uaIWAAAAAHC5u/qdoGHDhmn58uU6deqU4uPj7W4AAAAAcD9zeGGEli1bSpLCw8NlsVhs7YZhyGKxKDU11XnVAQAAAICTORyC1qxZ44o6AAAAACBbOByCQkNDXVEHAAAAAGQLh0OQJF26dEkzZszQ/v37JUmVKlVS7969FRAQ4NTiAAAAAMDZHF4YYfv27SpdurQ++ugjXbhwQRcuXNDEiRNVunRp7dixwxU1AgAAAIDTODwT9PLLLys8PFzTp0+Xl9fNh6ekpOjFF1/UkCFDtH79eqcXCQAAAADO4nAI2r59u10AkiQvLy+99tpreuyxx5xaHAAAAAA4m8Onw/n7++vYsWMZ2o8fP67cuXM7pSgAAAAAcBWHQ1CnTp30wgsvaP78+Tp+/LiOHz+u6Ohovfjii+rSpYsragQAAAAAp3H4dLjx48fLYrGoe/fuSklJkSR5e3urf//+GjNmjNMLBAAAAABncjgE+fj4aNKkSRo9erT+/PNPSVLp0qWVM2dOpxcHAAAAAM7mcAi6fPmyUlNTlS9fPlWpUsXWfuHCBXl5ecnf39+pBQIAAACAMzl8TVDnzp0VHR2doX3BggXq3LmzU4oCAAAAAFdxOARt3bpVDRs2zNAeFhamrVu3OqUoAAAAAHAVh0NQcnKybUGEf7px44auXr3qlKIAAAAAwFUcDkG1a9fWtGnTMrRPmTJFtWrVckpRAAAAAOAqDi+MMGrUKDVp0kS7d+9W48aNJUmrV6/Wtm3btGrVKqcXCAAAAADO5PBMUIMGDbR582YFBwdrwYIF+vbbb1WmTBnt2bNHTz75pCtqBAAAAACncXgmSJKqV6+uqKgoZ9cCAAAAAC6X5RCUkpKi1NRUWa1WW9uZM2c0ZcoUJSUlKTw8XE888YRLigQAAAAAZ8lyCOrTp498fHw0depUSVJCQoIef/xxXbt2TYGBgfroo4/0zTffqGXLli4rFgAAAADuVZavCdq0aZPatWtnu//VV18pNTVVBw8e1O7duzV06FB9+OGHLikSAAAAAJwlyyHoxIkTKlu2rO3+6tWr1a5dOwUEBEiSevTooX379jm/QgAAAABwoiyHIF9fX7sfQ92yZYvq1Kljtz0xMdG51QEAAACAk2U5BFWvXl2zZ8+WJG3YsEFnzpxRo0aNbNv//PNPBQUFOb9CAAAAAHCiLC+M8Pbbb6tFixZasGCBTp06pZ49eyowMNC2fcmSJWrQoIFLigQAAAAAZ8lyCAoNDVVsbKxWrVqlIkWKqEOHDnbbq1evrtq1azu9QAAAAABwJod+LLVChQqqUKFCptv69u3rlIIAAAAAwJWyfE0QAAAAADwMCEEAAAAATIUQBAAAAMBUCEEAAAAATOWuQtClS5f0+eefa/jw4bpw4YIkaceOHTpx4oRTiwMAAAAAZ3NodThJ2rNnj5o0aaKAgAAdOXJEffr0Ub58+bR48WIdO3ZMX331lSvqBAAAAACncHgmaOjQoerZs6cOHjwoX19fW3vLli21fv16pxYHAAAAAM7mcAjatm2b+vXrl6G9aNGiOn36tFOKAgAAAABXcTgEWa1WxcfHZ2j/448/VLBgQacUBQAAAACu4nAICg8P17vvvqsbN25IkiwWi44dO6bXX39d7dq1c3qBAAAAAOBMDoegCRMmKDExUYUKFdLVq1cVGhqqMmXKKHfu3Hr//fddUSMAAAAAOI3Dq8MFBAQoJiZGGzdu1J49e5SYmKiaNWuqSZMmrqgPAAAAAJzK4RCU7oknntATTzzhzFoAAAAAwOUcDkH/+9//Mm23WCzy9fVVmTJlFBISIk9Pz3suDgAAAACczeEQ9NFHH+ns2bO6cuWK8ubNK0m6ePGicubMKT8/P8XFxalUqVJas2aNgoODnV4wAAAAANwLhxdG+OCDD/T444/r4MGDOn/+vM6fP68//vhDderU0aRJk3Ts2DEVKVJEL7/8sivqBQAAAIB74vBM0JtvvqlFixapdOnStrYyZcpo/Pjxateunf766y+NGzeO5bIBAAAA3Jccngk6deqUUlJSMrSnpKTo9OnTkqSgoCAlJCTce3UAAAAA4GQOh6CGDRuqX79+2rlzp61t586d6t+/vxo1aiRJ2rt3r0qWLOm8KgEAAADASRwOQTNmzFC+fPlUq1YtWa1WWa1WPfbYY8qXL59mzJghSfLz89OECROcXiwAAAAA3CuHrwkqUqSIYmJidODAAf3xxx+SpHLlyqlcuXK2Pg0bNnRehQAAAADgRHf9Y6nly5dX+fLlnVkLAAAAALjcXYWgv//+W8uWLdOxY8d0/fp1u20TJ050SmEAAAAA4AoOh6DVq1crPDxcpUqV0oEDB1S5cmUdOXJEhmGoZs2arqgRAAAAAJzG4YURhg8frmHDhmnv3r3y9fXVokWLdPz4cYWGhqpDhw6uqBEAAAAAnMbhELR//351795dkuTl5aWrV6/Kz89P7777rsaOHev0AgEAAADAmRwOQbly5bJdBxQYGKg///zTtu3cuXPOqwwAAAAAXMDha4Lq1q2rjRs3qkKFCmrZsqVeeeUV7d27V4sXL1bdunVdUSMAAAAAOI3DIWjixIlKTEyUJI0cOVKJiYmaP3++ypYty8pwAAAAAO57DoWg1NRU/f3336pataqkm6fGTZkyxSWFAQAAAIArOHRNkKenp5o2baqLFy+6qh4AAAAAcCmHF0aoXLmy/vrrL1fUAgAAAAAu53AIGjVqlIYNG6bly5fr1KlTio+Pt7sBAAAAwP3M4YURWrZsKUkKDw+XxWKxtRuGIYvFotTUVOdVBwAAAABO5nAIWrNmjSvqAAAAAIBs4XAICg0Nddrgo0eP1uLFi3XgwAHlyJFD9evX19ixY1WuXDmnjQEAAAAA/+TwNUGStGHDBj333HOqX7++Tpw4IUmaPXu2Nm7c6NB+1q1bp4iICG3ZskUxMTG6ceOGmjZtqqSkpLspCwAAAADuyOEQtGjRIjVr1kw5cuTQjh07lJycLEm6fPmyPvjgA4f29cMPP6hnz56qVKmSqlWrplmzZunYsWOKjY11tCwAAAAAyJK7Wh1uypQpmj59ury9vW3tDRo00I4dO+6pmMuXL0uS8uXLl+n25ORkVqMDAAAAcE8cDkG///67QkJCMrQHBATo0qVLd11IWlqahgwZogYNGqhy5cqZ9hk9erQCAgJst+Dg4LseDwAAAIA5ORyCihQpokOHDmVo37hxo0qVKnXXhUREROjXX39VdHT0LfsMHz5cly9ftt2OHz9+1+MBAAAAMCeHV4fr06ePBg8erC+++EIWi0UnT57U5s2bNWzYML311lt3VcSAAQO0fPlyrV+/XsWKFbtlP6vVKqvVeldjAAAAAIB0FyHov//9r9LS0tS4cWNduXJFISEhslqtGjZsmAYOHOjQvgzD0MCBA7VkyRKtXbtWJUuWdLQcAAAAAHCIwyHIYrHojTfe0KuvvqpDhw4pMTFRFStWlJ+fn8ODR0REaO7cufrmm2+UO3dunT59WtLN64ty5Mjh8P4AAAAA4E4cviZozpw5unLlinx8fFSxYkXVrl37rgKQJE2ePFmXL19WWFiYAgMDbbf58+ff1f4AAAAA4E4cDkEvv/yyChUqpK5du+r7779XamrqXQ9uGEamt549e971PgEAAADgdhwOQadOnVJ0dLQsFos6duyowMBARURE6Oeff3ZFfQAAAADgVA6HIC8vLz399NOKiopSXFycPvroIx05ckQNGzZU6dKlXVEjAAAAADiNwwsj/FPOnDnVrFkzXbx4UUePHtX+/fudVRcAAAAAuITDM0GSdOXKFUVFRally5YqWrSoIiMj1bZtW+3bt8/Z9QEAAACAUzk8E9S5c2ctX75cOXPmVMeOHfXWW2+pXr16rqgNAAAAAJzO4RDk6empBQsWqFmzZvL09LTb9uuvv6py5cpOKw4AAAAAnM3hEBQVFWV3PyEhQfPmzdPnn3+u2NjYe1oyGwAAAABc7a6uCZKk9evXq0ePHgoMDNT48ePVqFEjbdmyxZm1AQAAAIDTOTQTdPr0ac2aNUszZsxQfHy8OnbsqOTkZC1dulQVK1Z0VY0AAAAA4DRZnglq3bq1ypUrpz179igyMlInT57Uxx9/7MraAAAAAMDpsjwTtGLFCg0aNEj9+/dX2bJlXVkTAAAAALhMlmeCNm7cqISEBNWqVUt16tTRJ598onPnzrmyNgAAAABwuiyHoLp162r69Ok6deqU+vXrp+joaAUFBSktLU0xMTFKSEhwZZ0AAAAA4BQOrw6XK1cu9e7dWxs3btTevXv1yiuvaMyYMSpUqJDCw8NdUSMAAAAAOM1dL5EtSeXKldO4ceP0999/a968ec6qCQAAAABc5p5CUDpPT0+1adNGy5Ytc8buAAAAAMBlnBKCAAAAAOBBQQgCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCpuDUHr169X69atFRQUJIvFoqVLl7qzHAAAAAAm4NYQlJSUpGrVqunTTz91ZxkAAAAATMTLnYO3aNFCLVq0yHL/5ORkJScn2+7Hx8e7oiwAAAAAD7EH6pqg0aNHKyAgwHYLDg52d0kAAAAAHjAPVAgaPny4Ll++bLsdP37c3SUBAAAAeMC49XQ4R1mtVlmtVneXAQAAAOAB9kDNBAEAAADAvSIEAQAAADAVt54Ol5iYqEOHDtnuHz58WLt27VK+fPn0yCOPuLEyAAAAAA8rt4ag7du3q2HDhrb7Q4cOlST16NFDs2bNclNVAAAAAB5mbg1BYWFhMgzDnSUAAAAAMBmuCQIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKoQgAAAAAKZCCAIAAABgKvdFCPr0009VokQJ+fr6qk6dOvrll1/cXRIAAACAh5TbQ9D8+fM1dOhQjRgxQjt27FC1atXUrFkzxcXFubs0AAAAAA8ht4egiRMnqk+fPurVq5cqVqyoKVOmKGfOnPriiy/cXRoAAACAh5CXOwe/fv26YmNjNXz4cFubh4eHmjRpos2bN2fon5ycrOTkZNv9y5cvS5Li4+NdXmtCYqLLx8hOSdeuursEp0m+kXznTg+Q1LQUd5fgVNnx/swufA7c3x6mzwI+B+5vD9NnAZ8D97eH6bMgOz4H0scwDOOOfd0ags6dO6fU1FQVLlzYrr1w4cI6cOBAhv6jR4/WyJEjM7QHBwe7rEYA9yYgIMDdJQBwMz4HAGTn50BCQsIdx3NrCHLU8OHDNXToUNv9tLQ0XbhwQfnz55fFYnFjZXCX+Ph4BQcH6/jx4/L393d3OQDchM8CAHwOwDAMJSQkKCgo6I593RqCChQoIE9PT505c8au/cyZMypSpEiG/larVVar1a4tT548riwRDwh/f38+8ADwWQCAzwGTy+qMk1sXRvDx8VGtWrW0evVqW1taWppWr16tevXqubEyAAAAAA8rt58ON3ToUPXo0UOPPfaYateurcjISCUlJalXr17uLg0AAADAQ8jtIahTp046e/as3n77bZ0+fVrVq1fXDz/8kGGxBCAzVqtVI0aMyHCaJABz4bMAAJ8DcITFyMoacgAAAADwkHD7j6UCAAAAQHYiBAEAAAAwFUIQAAAAAFMhBCHbhYWFaciQIbfcXqJECUVGRmZbPQAeTu+8846qV6/u7jIAOFFWviPwPQJZQQgCADyUhg0bZvc7dADMYdu2berbt6/T9keoeji5fYlsAAAclZqaKovFIg+PW/8tz8/PT35+ftlYFYD7QcGCBd1dAh4AzATBLVJSUjRgwAAFBASoQIECeuutt3Sr1dqPHTumZ555Rn5+fvL391fHjh115swZuz6TJ09W6dKl5ePjo3Llymn27Nl22y0Wi6ZOnaqnn35aOXPmVIUKFbR582YdOnRIYWFhypUrl+rXr68///zTZccMmFlYWJgGDBhwy/d9cnKyhg0bpqJFiypXrlyqU6eO1q5da3v8rFmzlCdPHi1btkwVK1aU1WrVsWPHtHbtWtWuXVu5cuVSnjx51KBBAx09elRSxtPhtm3bpqeeekoFChRQQECAQkNDtWPHDrs6LRaLPv/8c7Vt21Y5c+ZU2bJltWzZMpc/P4CZfP3116pSpYpy5Mih/Pnzq0mTJkpKSsr0dPk2bdqoZ8+edm0JCQnq0qWLcuXKpaJFi+rTTz+12/7vmZtLly7pxRdfVMGCBeXv769GjRpp9+7ddo/59ttv9fjjj8vX11cFChRQ27ZtJd387Dp69KhefvllWSwWWSwWpz0PcC9CENziyy+/lJeXl3755RdNmjRJEydO1Oeff56hX1pamp555hlduHBB69atU0xMjP766y916tTJ1mfJkiUaPHiwXnnlFf3666/q16+fevXqpTVr1tjt67333lP37t21a9culS9fXl27dlW/fv00fPhwbd++XYZhaMCAAS4/dsCsbve+HzBggDZv3qzo6Gjt2bNHHTp0UPPmzXXw4EHb469cuaKxY8fq888/1759+5QvXz61adNGoaGh2rNnjzZv3qy+ffve8ktKQkKCevTooY0bN2rLli0qW7asWrZsqYSEBLt+I0eOVMeOHbVnzx61bNlS3bp104ULF1z3xAAmcurUKXXp0kW9e/fW/v37tXbtWj377LO3/ENoZj788ENVq1ZNO3fu1H//+18NHjxYMTExt+zfoUMHxcXFacWKFYqNjVXNmjXVuHFj2/v6u+++U9u2bdWyZUvt3LlTq1evVu3atSVJixcvVrFixfTuu+/q1KlTOnXq1L09Abh/GEA2Cw0NNSpUqGCkpaXZ2l5//XWjQoUKhmEYRvHixY2PPvrIMAzDWLVqleHp6WkcO3bM1nffvn2GJOOXX34xDMMw6tevb/Tp08dujA4dOhgtW7a03ZdkvPnmm7b7mzdvNiQZM2bMsLXNmzfP8PX1dd6BArC53fv+6NGjhqenp3HixAm7xzRu3NgYPny4YRiGMXPmTEOSsWvXLtv28+fPG5KMtWvXZjrmiBEjjGrVqt2yptTUVCN37tzGt99+a2v792dFYmKiIclYsWKFQ8cLIHOxsbGGJOPIkSMZtoWGhhqDBw+2a3vmmWeMHj162O4XL17caN68uV2fTp06GS1atLDrk/49YsOGDYa/v79x7do1u8eULl3amDp1qmEYhlGvXj2jW7dut6z5n/vDw4OZILhF3bp17f5aW69ePR08eFCpqal2/fbv36/g4GAFBwfb2ipWrKg8efJo//79tj4NGjSwe1yDBg1s29NVrVrV9t+FCxeWJFWpUsWu7dq1a4qPj7/HowOQmVu97/fu3avU1FQ9+uijtut4/Pz8tG7dOrtTVH18fOzex/ny5VPPnj3VrFkztW7dWpMmTbrtX2nPnDmjPn36qGzZsgoICJC/v78SExN17Ngxu37/HCNXrlzy9/dXXFycM54CwPSqVaumxo0bq0qVKurQoYOmT5+uixcvOrSPevXqZbj/73/z0+3evVuJiYnKnz+/3efL4cOHbZ8vu3btUuPGje/ugPDAYmEEmIa3t7ftv9O/iGXWlpaWlr2FASaXmJgoT09PxcbGytPT027bPxc2yJEjR4ZT3WbOnKlBgwbphx9+0Pz58/Xmm28qJiZGdevWzTBOjx49dP78eU2aNEnFixeX1WpVvXr1dP36dbt+//xckG5+NvC5ADiHp6enYmJi9PPPP2vVqlX6+OOP9cYbb2jr1q3y8PDIcFrcjRs37mm8xMREBQYG2l1jmC5PnjySbn62wHyYCYJbbN261e5++vn5//4CVKFCBR0/flzHjx+3tf3222+6dOmSKlasaOuzadMmu8dt2rTJth3A/eFW7/saNWooNTVVcXFxKlOmjN2tSJEid9xvjRo1NHz4cP3888+qXLmy5s6dm2m/TZs2adCgQWrZsqUqVaokq9Wqc+fOOeXYAGSdxWJRgwYNNHLkSO3cuVM+Pj5asmSJChYsaDebm5qaql9//TXD47ds2ZLhfoUKFTIdq2bNmjp9+rS8vLwyfL4UKFBA0s3Z39stp+/j45PhTBU8+JgJglscO3ZMQ4cOVb9+/bRjxw59/PHHmjBhQoZ+TZo0UZUqVdStWzdFRkYqJSVFL730kkJDQ/XYY49Jkl599VV17NhRNWrUUJMmTfTtt99q8eLF+vHHH7P7sADcxq3e948++qi6deum7t27a8KECapRo4bOnj2r1atXq2rVqmrVqlWm+zt8+LCmTZum8PBwBQUF6ffff9fBgwfVvXv3TPuXLVtWs2fP1mOPPab4+Hi9+uqr/AUYyGZbt27V6tWr1bRpUxUqVEhbt27V2bNnVaFCBeXKlUtDhw7Vd999p9KlS2vixIm6dOlShn1s2rRJ48aNU5s2bRQTE6OFCxfqu+++y3S8Jk2aqF69emrTpo3GjRunRx99VCdPnrQthvDYY49pxIgRaty4sUqXLq3OnTsrJSVF33//vV5//XVJN1ebW79+vTp37iyr1WoLT3iwEYLgFt27d9fVq1dVu3ZteXp6avDgwZn+sJnFYtE333yjgQMHKiQkRB4eHmrevLk+/vhjW582bdpo0qRJGj9+vAYPHqySJUtq5syZCgsLy8YjAnAnt3vfz5w5U6NGjdIrr7yiEydOqECBAqpbt66efvrpW+4vZ86cOnDggL788kudP39egYGBioiIUL9+/TLtP2PGDPXt21c1a9ZUcHCwPvjgAw0bNswlxwogc/7+/lq/fr0iIyMVHx+v4sWLa8KECWrRooVu3Lih3bt3q3v37vLy8tLLL7+shg0bZtjHK6+8ou3bt2vkyJHy9/fXxIkT1axZs0zHs1gs+v777/XGG2+oV69eOnv2rIoUKaKQkBDb9cFhYWFauHCh3nvvPY0ZM0b+/v4KCQmx7ePdd99Vv379VLp0aSUnJzu0kh3uXxaD/5MAABcLCwtT9erV+dV1AC4XGBio9957Ty+++KK7S8F9jJkgAAAAPPCuXLmiTZs26cyZM6pUqZK7y8F9joURAAAA8MCbNm2aOnfurCFDhmRYRhv4N06HAwAAAGAqzAQBAAAAMBVCEAAAAABTIQQBAAAAMBVCEAAAAABTIQQBAAAAMBVCEAAAAABTIQQBALLd5s2b5enpqVatWrm7FDtHjhyRxWLRrl273F0KAMCFCEEAgGw3Y8YMDRw4UOvXr9fJkyfdXQ4AwGQIQQCAbJWYmKj58+erf//+atWqlWbNmmXbdvHiRXXr1k0FCxZUjhw5VLZsWc2cOVPS/5uliY6OVv369eXr66vKlStr3bp1dvv/9ddf1aJFC/n5+alw4cJ6/vnnde7cOdv2tLQ0jRs3TmXKlJHVatUjjzyi999/X5JUsmRJSVKNGjVksVgUFhbm2icDAOAWhCAAQLZasGCBypcvr3Llyum5557TF198IcMwJElvvfWWfvvtN61YsUL79+/X5MmTVaBAAbvHv/rqq3rllVe0c+dO1atXT61bt9b58+clSZcuXVKjRo1Uo0YNbd++XT/88IPOnDmjjh072h4/fPhwjRkzxjbW3LlzVbhwYUnSL7/8Ikn68ccfderUKS1evDg7nhIAQDazGOn/8gAAkA0aNGigjh07avDgwUpJSVFgYKAWLlyosLAwhYeHq0CBAvriiy8yPO7IkSMqWbKkxowZo9dff12SlJKSopIlS2rgwIF67bXXNGrUKG3YsEErV660Pe7vv/9WcHCwfv/9dwUGBqpgwYL65JNP9OKLL95yjJ07d6p69eouew4AAO7l5e4CAADm8fvvv+uXX37RkiVLJEleXl7q1KmTZsyYobCwMPXv31/t2rXTjh071LRpU7Vp00b169e320e9evVs/+3l5aXHHntM+/fvlyTt3r1ba9askZ+fX4ax//zzT126dEnJyclq3LixC48SAHC/IwQBALLNjBkzlJKSoqCgIFubYRiyWq365JNP1KJFCx09elTff/+9YmJi1LhxY0VERGj8+PFZ2n9iYqJat26tsWPHZtgWGBiov/76y2nHAgB4cHFNEAAgW6SkpOirr77ShAkTtGvXLttt9+7dCgoK0rx58yRJBQsWVI8ePTRnzhxFRkZq2rRpdvvZsmWL3T5jY2NVoUIFSVLNmjW1b98+lShRQmXKlLG75cqVS2XLllWOHDm0evXqTGv08fGRJKWmprriKQAA3CeYCQIAZIvly5fr4sWLeuGFFxQQEGC3rV27dpoxY4ZOnjypWrVqqVKlSkpOTtby5cttASfdp59+qrJly6pChQr66KOPdPHiRfXu3VuSFBERoenTp6tLly567bXXlC9fPh06dEjR0dH6/PPP5evrq9dff12vvfaafHx81KBBA509e1b79u3TCy+8oEKFCilHjhz64YcfVKxYMfn6+maoFQDw4GMmCACQLWbMmKEmTZpkGiratWun7du3y8vLS8OHD1fVqlUVEhIiT09PRUdH2/UdM2aMxowZo2rVqmnjxo1atmyZbQW5oKAgbdq0SampqWratKmqVKmiIUOGKE+ePPLwuPlP3ltvvaVXXnlFb7/9tipUqKBOnTopLi5O0s1rjP73v/9p6tSpCgoK0jPPPOPiZwUA4A6sDgcAeCCwchsAwFmYCQIAAABgKoQgAAAAAKbC6XAAAAAATIWZIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCqEIAAAAACmQggCAAAAYCr/Hz0v13ftr/Z9AAAAAElFTkSuQmCC\n" + }, + "metadata": {} + } + ] + }, + { + "cell_type": "code", + "source": [ + "\n", + "# Reshape the data from wide to long format\n", + "long_df = pd.melt(\n", + " df,\n", + " id_vars=['question_num', 'bloom_level', 'llm'],\n", + " value_vars=[\n", + " 'expert1_bloom', 'expert1_subject', 'expert1_persian',\n", + " 'expert2_bloom', 'expert2_subject', 'expert2_persian',\n", + " 'expert3_bloom', 'expert3_subject', 'expert3_persian',\n", + " 'expert4_bloom', 'expert4_subject', 'expert4_persian',\n", + " 'expert5_bloom', 'expert5_subject', 'expert5_persian'\n", + " ],\n", + " var_name='expert_aspect',\n", + " value_name='score'\n", + ")\n", + "\n", + "# Extract expert and aspect from 'expert_aspect' column\n", + "long_df[['expert', 'aspect']] = long_df['expert_aspect'].str.split('_', expand=True)\n", + "\n", + "# Drop the original 'expert_aspect' column\n", + "long_df.drop(columns=['expert_aspect'], inplace=True)\n", + "\n", + "# Calculate the average score for each expert for each aspect for each llm\n", + "average_scores = long_df.groupby(['llm', 'expert', 'aspect'])['score'].mean().reset_index()\n", + "\n", + "# Create the bar plot\n", + "plt.figure(figsize=(12, 7))\n", + "\n", + "# Use seaborn to create a bar plot with 'expert' as hue\n", + "sns.barplot(\n", + " data=average_scores,\n", + " x='aspect',\n", + " y='score',\n", + " hue='expert',\n", + " ci=None\n", + ")\n", + "\n", + "# Customize the plot\n", + "plt.title('Average Scores by Aspect for Each Expert')\n", + "plt.xlabel('Aspect')\n", + "plt.ylabel('Average Score')\n", + "plt.legend(title='Expert')\n", + "\n", + "# Show the plot\n", + "plt.show()\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 662 + }, + "id": "ijvIMG88v4J3", + "outputId": "34ece588-5146-404a-8f33-0093aeefe0ef" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":29: FutureWarning: \n", + "\n", + "The `ci` parameter is deprecated. Use `errorbar=None` for the same effect.\n", + "\n", + " sns.barplot(\n" + ] + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "iVBORw0KGgoAAAANSUhEUgAAA9wAAAJwCAYAAAB7+Wk8AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABqTElEQVR4nO3deVhV5f7+8XsDskEQUMERVERyzAFNcQSHckqzY05ZqOVQSeJ4PB6Ps4mV5pClVqZmmJpp5ZhDomHOqThlOJuzOSBaoLB+f/hjf9uBBsVyi71f17Wvy/WsZz/rsxa7HTfPGiyGYRgCAAAAAAA5ysnRBQAAAAAA8CgicAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwBgopEjR8pisejy5cuOLuWRlJSUpO7du6tIkSKyWCzq27evo0vKcXPmzJHFYtHOnTsdXQoAIJsI3ABggvfff18Wi0W1atVydCkPnZSUFE2ZMkXVqlWTl5eXfHx8VLFiRfXs2VM//vijo8t7JNSsWVMWi0XTp093dCnZ9v7772vOnDlZ7j9u3DjNmTNHr776qubNm6cXX3zRvOIklSpVShaLJdNXs2bNTN32X5X+R597vc6fP+/oEu9r/vz5mjx5sqPLAIC/xMXRBQDAoygmJkalSpXS9u3bdeTIEZUpU8bRJT002rZtq1WrVqlTp07q0aOHbt++rR9//FHLly9XnTp1VK5cOUeXmKslJCRox44dKlWqlGJiYvTqq686uqRsef/99+Xr66uuXbtmqf+3336r0NBQjRgxwtzCfqdq1aoaMGBAhvZixYo9sBr+iunTp8vT0zNDu4+Pz4MvJhvmz5+v/fv3P5JnLwB49BG4ASCHHT9+XN9//72WLFmiXr16KSYm5oGGAUlKS0tTSkqK3NzcHuh2/8yOHTu0fPlyvfHGG/rvf/9rt27atGm6du3aA6vlt99+k6urq5ycHq2TvT799FMVKlRIEydO1HPPPacTJ06oVKlSji7LNBcvXlSFChVybLw7d+4oLS1Nrq6u9+xTvHhxvfDCCzm2zQflueeek6+vr6PLyLKbN2/Kw8PD0WUAwN/yaP2WAQAPgZiYGOXPn18tW7bUc889p5iYGNu627dvq0CBAurWrVuG9yUmJsrNzU0DBw60tSUnJ2vEiBEqU6aMrFarAgIC9O9//1vJycl277VYLIqMjFRMTIwqVqwoq9Wq1atXS5ImTJigOnXqqGDBgnJ3d1f16tW1ePHiDNv/9ddf1adPH/n6+ipfvnxq3bq1zpw5I4vFopEjR9r1PXPmjF566SUVLlxYVqtVFStW1Mcff/ynx+bo0aOSpLp162ZY5+zsrIIFC2bYzssvv6xixYrJarUqMDBQr776qlJSUmx9jh07pnbt2qlAgQLKmzevQkNDtWLFCrtxYmNjZbFYtGDBAv3vf/9T8eLFlTdvXiUmJkqStm3bpmbNmsnb21t58+ZVWFiYNm/ebDfGjRs31LdvX5UqVUpWq1WFChXSk08+qR9++OFP91uSLl++rPbt28vLy0sFCxZUVFSUfvvtN9v6sLAwValSJdP3li1bVk2bNs3SdubPn6/nnntOTz/9tLy9vTV//vwMfbKyL+Hh4apUqZJ27dqlOnXqyN3dXYGBgZoxY0aG8bL6OZXu/kGgZs2ayps3r/Lnz68GDRpozZo1ku6ern3gwAFt3LjRdrpzeHh4pvuZ/jM9fvy4VqxYYet/4sQJSXeD+Msvv6zChQvLzc1NVapU0dy5c+3GOHHihCwWiyZMmKDJkycrKChIVqtVBw8ezNKxvp/4+Hh17dpVpUuXlpubm4oUKaKXXnpJv/zyS4a+WfmcS3ePc//+/eXn5ycPDw89++yzunTp0t+uNV2XLl3k5uamQ4cO2bU3bdpU+fPn19mzZyX93zXlmzZtUq9evVSwYEF5eXkpIiJCV69ezTDuqlWrVL9+fXl4eChfvnxq2bKlDhw4YNena9eu8vT01NGjR9WiRQvly5dPnTt3Vnh4uFasWKGTJ0/afsaP8h+QADx6mOEGgBwWExOjf/3rX3J1dVWnTp00ffp07dixQ0888YTy5MmjZ599VkuWLNHMmTPtZtG+/PJLJScnq2PHjpLuzlK3bt1acXFx6tmzp8qXL699+/Zp0qRJ+umnn/Tll1/abffbb7/VokWLFBkZKV9fX9svpVOmTFHr1q3VuXNnpaSkaMGCBWrXrp2WL1+uli1b2t7ftWtXLVq0SC+++KJCQ0O1ceNGu/XpLly4oNDQUFvI9/Pz06pVq/Tyyy8rMTHxvqd9lixZ0naM6tatKxeXe/9v6OzZs6pZs6auXbumnj17qly5cjpz5owWL16sW7duydXVVRcuXFCdOnV069Yt9enTRwULFtTcuXPVunVrLV68WM8++6zdmGPGjJGrq6sGDhyo5ORkubq66ttvv1Xz5s1VvXp1jRgxQk5OTpo9e7YaNWqk7777TjVr1pQkvfLKK1q8eLEiIyNVoUIF/fLLL4qLi9OhQ4cUEhJyz/1I1759e5UqVUrR0dHaunWrpk6dqqtXr+qTTz6RJL344ovq0aOH9u/fr0qVKtnet2PHDv3000/63//+96fb2LZtm44cOaLZs2fL1dVV//rXvxQTE5PhbIKs7svVq1fVokULtW/fXp06ddKiRYv06quvytXVVS+99JKk7H1OR40apZEjR6pOnToaPXq0XF1dtW3bNn377bd66qmnNHnyZL3++uvy9PTU0KFDJUmFCxfOdF/Lly+vefPmqV+/fvL397ed4u3n56dff/1V4eHhOnLkiCIjIxUYGKjPP/9cXbt21bVr1xQVFWU31uzZs/Xbb7+pZ8+eslqtKlCgwH2P8+3btzO9CZ6Hh4fc3d0lSWvXrtWxY8fUrVs3FSlSRAcOHNAHH3ygAwcOaOvWrbJYLJKy9jlP9/rrryt//vwaMWKETpw4ocmTJysyMlILFy68b73prly5kqHNxcXFdkr5lClT9O2336pLly7asmWLnJ2dNXPmTK1Zs0bz5s3LcMp8ZGSkfHx8NHLkSB0+fFjTp0/XyZMnbX8MkaR58+apS5cuatq0qd58803dunVL06dPV7169bR792678Hznzh01bdpU9erV04QJE5Q3b14VKVJE169f188//6xJkyZJUqanxQPAQ8sAAOSYnTt3GpKMtWvXGoZhGGlpaYa/v78RFRVl6/PNN98Ykoxly5bZvbdFixZG6dKlbcvz5s0znJycjO+++86u34wZMwxJxubNm21tkgwnJyfjwIEDGWq6deuW3XJKSopRqVIlo1GjRra2Xbt2GZKMvn372vXt2rWrIckYMWKEre3ll182ihYtaly+fNmub8eOHQ1vb+8M2/u9tLQ0IywszJBkFC5c2OjUqZPx3nvvGSdPnszQNyIiwnBycjJ27NiR6TiGYRh9+/Y1JNkdoxs3bhiBgYFGqVKljNTUVMMwDGPDhg2GJKN06dJ29aWlpRnBwcFG06ZNbWOmH7PAwEDjySeftLV5e3sbvXv3vue+3cuIESMMSUbr1q3t2l977TVDkrF3717DMAzj2rVrhpubmzF48GC7fn369DE8PDyMpKSkP91WZGSkERAQYNuXNWvWGJKM3bt32/XLyr6k/5wmTpxoa0tOTjaqVq1qFCpUyEhJSTEMI+uf04SEBMPJycl49tlnbT+XdL8/9hUrVjTCwsL+dF/TlSxZ0mjZsqVd2+TJkw1JxqeffmprS0lJMWrXrm14enoaiYmJhmEYxvHjxw1JhpeXl3Hx4sUsb09Spq/o6Ghbv8z+O/jss88MScamTZtsbVn5nM+ePduQZDRp0sTuWPXr189wdnY2rl27dt+a0z+Dmb3Kli1r1zf9+2ns2LHGsWPHDE9PT6NNmzZ2fdLrqV69uu1zYBiG8dZbbxmSjK+++sowjLv/Lfr4+Bg9evSwe//58+cNb29vu/YuXboYkoz//Oc/Gepv2bKlUbJkyfvuIwA8rDilHAByUExMjAoXLqyGDRtKunuqd4cOHbRgwQKlpqZKkho1aiRfX1+7WamrV69q7dq16tChg63t888/V/ny5VWuXDldvnzZ9mrUqJEkacOGDXbbDgsLy/Ra1vQZt/TtXL9+XfXr17c7fTj99PPXXnvN7r2vv/663bJhGPriiy/UqlUrGYZhV1fTpk11/fr1+55ibbFY9M0332js2LHKnz+/PvvsM/Xu3VslS5ZUhw4dbNdwp6Wl6csvv1SrVq1Uo0aNTMeRpJUrV6pmzZqqV6+ebZ2np6d69uypEydOZDg1uEuXLnbHY8+ePUpISNDzzz+vX375xbYvN2/eVOPGjbVp0yalpaVJuntjqW3bttlOq82u3r172y2nH9uVK1dKkry9vfXMM8/os88+k2EYkqTU1FQtXLhQbdq0+dNrWe/cuaOFCxeqQ4cOtuPTqFEjFSpUyO6yhuzsi4uLi3r16mVbdnV1Va9evXTx4kXt2rVLUtY/p19++aXS0tI0fPjwDNfNp9ebU1auXKkiRYqoU6dOtrY8efKoT58+SkpK0saNG+36t23bVn5+flkev1atWlq7dm2G1++39/vP2W+//abLly8rNDRUkmz/jWT1c56uZ8+edm3169dXamqqTp48maW6v/jiiww1z549267PU089pV69emn06NH617/+JTc3N82cOTPT8Xr27Kk8efLYll999VW5uLjYPtNr167VtWvX1KlTJ7vPhrOzs2rVqpXhOyx9DAB4lHBKOQDkkNTUVC1YsEANGzbU8ePHbe21atXSxIkTtX79ej311FNycXFR27ZtNX/+fCUnJ8tqtWrJkiW6ffu2XeBOSEjQoUOH7hkELl68aLccGBiYab/ly5dr7Nix2rNnj901tb//xf3kyZNycnLKMMYf765+6dIlXbt2TR988IE++OCDLNX1R1arVUOHDtXQoUN17tw5bdy4UVOmTNGiRYuUJ08effrpp7p06ZISExPtTq3OzMmTJzN99Fr58uVt638/xh/3LyEhQdLdIH4v169fV/78+fXWW2+pS5cuCggIUPXq1dWiRQtFRESodOnS960xXXBwsN1yUFCQnJycbNccS1JERIQWLlyo7777Tg0aNNC6det04cKFLD3qas2aNbp06ZJq1qypI0eO2NobNmyozz77TG+++aYt6GZ1X4oVK5Yh6D/22GOS7l7/HBoamuXP6dGjR+Xk5JSjNzi7l5MnTyo4ODhDsP/95+L37vXfzr34+vqqSZMm9+1z5coVjRo1SgsWLMjw38T169clKcuf83QlSpSwW86fP78kZXrddGYaNGiQpZumTZgwQV999ZX27Nmj+fPnq1ChQpn2++Nn2tPTU0WLFrV9ptP/+0r/48sfeXl52S27uLjI39//T+sDgNyEwA0AOeTbb7/VuXPntGDBAi1YsCDD+piYGD311FOSpI4dO2rmzJlatWqV2rRpo0WLFqlcuXJ2N81KS0vT448/rnfeeSfT7QUEBNgt/35GLd13332n1q1bq0GDBnr//fdVtGhR5cmTR7Nnz870Zlp/Jn2294UXXrhnSK1cuXKWxytatKg6duyotm3bqmLFilq0aFG2nsGcXX88Run78/bbb6tq1aqZvif9etH27durfv36Wrp0qdasWaO3335bb775ppYsWaLmzZtnu5bMZnWbNm2qwoUL69NPP1WDBg306aefqkiRIn8a7iTZZrHbt2+f6fqNGzfazrzIyX3J7uf0YZTZfzt/V/v27fX9999r0KBBqlq1qjw9PZWWlqZmzZrZPnfZ5ezsnGl7+hkROWX37t22PxLs27fPbuY+O9L3c968eSpSpEiG9X+8h4PVan3knhoAAARuAMghMTExKlSokN57770M65YsWaKlS5dqxowZcnd3V4MGDVS0aFEtXLhQ9erV07fffmu7SVS6oKAg7d27V40bN/7Lp9x+8cUXcnNz0zfffCOr1Wpr/+NppCVLllRaWpqOHz9uN2v1+5lS6e4NqfLly6fU1NQshcCsypMnjypXrqyEhARdvnxZhQoVkpeXl/bv33/f95UsWVKHDx/O0P7jjz/a1t9PUFCQpLszbVnZn6JFi+q1117Ta6+9posXLyokJERvvPFGlkJqQkKC3UzqkSNHlJaWZnfTKGdnZz3//POaM2eO3nzzTX355Zfq0aPHPYNWups3b+qrr75Shw4d9Nxzz2VY36dPH8XExNgCd1b35ezZsxkezfTTTz9Jkq3urH5Og4KClJaWpoMHD97zjxtSzpxeXrJkScXHxystLc0uwGX1c/F3Xb16VevXr9eoUaM0fPhwW3v6jG86Pz+/LH3OH6SbN2+qW7duqlChgurUqaO33npLzz77rJ544okMfRMSEuw+U0lJSTp37pxatGgh6f/++ypUqNDf+r7I6UsOAOBB4s+IAJADfv31Vy1ZskRPP/20nnvuuQyvyMhI3bhxQ19//bUkycnJSc8995yWLVumefPm6c6dO3ank0t3Z8jOnDmjDz/8MNPt3bx580/rcnZ2lsVisV0/Lt09FfiPdzhPf+TU+++/b9f+7rvvZhivbdu2+uKLLzINCX/2iKKEhASdOnUqQ/u1a9e0ZcsW5c+fX35+fnJyclKbNm20bNky7dy5M0P/9Bm9Fi1aaPv27dqyZYtt3c2bN/XBBx+oVKlSf3r6cvXq1RUUFKQJEyYoKSnpnvuTmppqOw04XaFChVSsWLFMH32VmT/+ISb92P4xrL/44ou6evWqevXqpaSkpCw973np0qW6efOmevfunenn7+mnn9YXX3yh5OTkbO3LnTt37K7fTUlJ0cyZM+Xn56fq1atLyvrntE2bNnJyctLo0aMzzPD+fobWw8Pjbz+PvUWLFjp//rzdfRLu3Lmjd999V56engoLC/tb4/+Z9D+Q/HHmefLkyXbLWf2cP0iDBw/WqVOnNHfuXL3zzjsqVaqUunTpkunn/IMPPtDt27dty9OnT9edO3dsn+mmTZvKy8tL48aNs+uXLquPNPPw8MjwmQWA3IIZbgDIAV9//bVu3Lih1q1bZ7o+NDRUfn5+iomJsQXrDh066N1339WIESP0+OOP264vTffiiy9q0aJFeuWVV7RhwwbVrVtXqamp+vHHH7Vo0SJ98803md5o6fdatmypd955R82aNdPzzz+vixcv6r333lOZMmUUHx9v61e9enW1bdtWkydP1i+//GJ7LFj6bObvZ5jGjx+vDRs2qFatWurRo4cqVKigK1eu6IcfftC6desyffRQur179+r5559X8+bNVb9+fRUoUEBnzpzR3LlzdfbsWU2ePNkWVsaNG6c1a9YoLCzM9ripc+fO6fPPP1dcXJx8fHz0n//8R5999pmaN2+uPn36qECBApo7d66OHz+uL7744k9PT3VyctJHH32k5s2bq2LFiurWrZuKFy+uM2fOaMOGDfLy8tKyZct048YN+fv767nnnlOVKlXk6empdevWaceOHZo4ceJ9t5Hu+PHjat26tZo1a6YtW7bo008/1fPPP5/h2dvVqlVTpUqVbDcjy8ojx2JiYlSwYEHVqVMn0/WtW7fWhx9+qBUrVqhRo0ZZ3pdixYrpzTff1IkTJ/TYY49p4cKF2rNnjz744APbzbKy+jktU6aMhg4dqjFjxqh+/fr617/+JavVqh07dqhYsWKKjo6WdPezOH36dI0dO1ZlypRRoUKF7nkN8L307NlTM2fOVNeuXbVr1y6VKlVKixcv1ubNmzV58mTly5cvW+P90ZkzZ/Tpp59maPf09FSbNm3k5eWlBg0a6K233tLt27dVvHhxrVmzxu7eDumy8jnPKYsXL870kVpPPvmkChcurG+//Vbvv/++RowYYfvczZ49W+Hh4Ro2bJjeeustu/elpKSocePGat++vQ4fPqz3339f9erVs30Penl5afr06XrxxRcVEhKijh07ys/PT6dOndKKFStUt25dTZs27U/rrl69uhYuXKj+/fvriSeekKenp1q1apUDRwQAHgDH3SAdAB4drVq1Mtzc3IybN2/es0/Xrl2NPHny2B6nlZaWZgQEBNgewZOZlJQU48033zQqVqxoWK1WI3/+/Eb16tWNUaNGGdevX7f1k3TPxzzNmjXLCA4ONqxWq1GuXDlj9uzZtscE/d7NmzeN3r17GwUKFLA9Cujw4cOGJGP8+PF2fS9cuGD07t3bCAgIMPLkyWMUKVLEaNy4sfHBBx/c9zhduHDBGD9+vBEWFmYULVrUcHFxMfLnz280atTIWLx4cYb+J0+eNCIiIgw/Pz/DarUapUuXNnr37m0kJyfb+hw9etR47rnnDB8fH8PNzc2oWbOmsXz5crtx0h8L9vnnn2da1+7du41//etfRsGCBQ2r1WqULFnSaN++vbF+/XrDMO4+DmvQoEFGlSpVjHz58hkeHh5GlSpVjPfff/+++2sY//dIpoMHDxrPPfeckS9fPiN//vxGZGSk8euvv2b6nvTHK40bN+5Px79w4YLh4uJivPjii/fsc+vWLSNv3rzGs88+m+V9CQsLMypWrGjs3LnTqF27tuHm5maULFnSmDZtWobxs/o5NQzD+Pjjj41q1arZ+oWFhdkeo2cYdx8Z1bJlSyNfvnyGpD99RFhmjwVLPy7dunUzfH19DVdXV+Pxxx83Zs+ebdcn/bFgb7/99n238cft6R6P2Pr9o6t+/vln49lnnzV8fHwMb29vo127dsbZs2czPGbPMP78c57+GK4/Pjos/XO9YcOG+9Z8v8eCpb8/MTHRKFmypBESEmLcvn3b7v39+vUznJycjC1bttjVs3HjRqNnz55G/vz5DU9PT6Nz587GL7/8kmH7GzZsMJo2bWp4e3sbbm5uRlBQkNG1a1dj586dtj5dunQxPDw8Mq0/KSnJeP755w0fH58MxxkAHnYWw3DA+UoAgFxhz549qlatmj799FN17tzZ0eX8Y0yZMkX9+vXTiRMnMtyZ+kEJDw/X5cuXH6rri/FwmDNnjrp166YdO3b86Vk2APBPxzXcAABJd6+3/aPJkyfLyclJDRo0cEBF/0yGYWjWrFkKCwtzWNgGAAA5g2u4AQCS7j6bedeuXWrYsKFcXFy0atUqrVq1Sj179swVj3bK7W7evKmvv/5aGzZs0L59+/TVV185uiQAAPA3EbgBAJKkOnXqaO3atRozZoySkpJUokQJjRw5MsPjymCOS5cu6fnnn5ePj4/++9//3vMGfAAAIPfgGm4AAAAAAEzANdwAAAAAAJiAwA0AAAAAgAly9TXcaWlpOnv2rPLlyyeLxeLocgAAAAAAjzjDMHTjxg0VK1ZMTk73n8PO1YH77Nmz3DkXAAAAAPDAnT59Wv7+/vftk6sDd758+STd3VEvLy8HVwMAAAAAeNQlJiYqICDAlkfvJ1cH7vTTyL28vAjcAAAAAIAHJiuXNXPTNAAAAAAATEDgBgAAAADABARuAAAAAABMkKuv4QYAAACA3MwwDN25c0epqamOLgX/n7Ozs1xcXHLk0dMEbgAAAABwgJSUFJ07d063bt1ydCn4g7x586po0aJydXX9W+MQuAEAAADgAUtLS9Px48fl7OysYsWKydXVNUdmVPH3GIahlJQUXbp0ScePH1dwcLCcnP76ldgEbgAAAAB4wFJSUpSWlqaAgADlzZvX0eXgd9zd3ZUnTx6dPHlSKSkpcnNz+8tjcdM0AAAAAHCQvzN7CvPk1M+Fny4AAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAMA/TNeuXWWxWDK8mjVr5ujS1LVrV7Vp08bRZeQIhwbukSNHZvgBlytXzpElAQAAAMA/QrNmzXTu3Dm712effeawelJTU5WWluaw7ZvB4TPcFStWtPsBx8XFObokAAAAAHjkWa1WFSlSxO6VP39+xcbGytXVVd99952t71tvvaVChQrpwoULkqTw8HBFRkYqMjJS3t7e8vX11bBhw2QYhu09ycnJGjhwoIoXLy4PDw/VqlVLsbGxtvVz5syRj4+Pvv76a1WoUEFWq1UvvfSS5s6dq6+++so2Kfv79+Q2Dn8Ot4uLi4oUKeLoMgAAAAAAuhum+/btqxdffFF79+7VsWPHNGzYMH3++ecqXLiwrd/cuXP18ssva/v27dq5c6d69uypEiVKqEePHpKkyMhIHTx4UAsWLFCxYsW0dOlSNWvWTPv27VNwcLAk6datW3rzzTf10UcfqWDBgipatKh+/fVXJSYmavbs2ZKkAgUKPPiDkEMcHrgTEhJUrFgxubm5qXbt2oqOjlaJEiUy7ZucnKzk5GTbcmJi4oMqEwAAAAAeKcuXL5enp6dd23//+1/997//1dixY7V27Vr17NlT+/fvV5cuXdS6dWu7vgEBAZo0aZIsFovKli2rffv2adKkSerRo4dOnTql2bNn69SpUypWrJgkaeDAgVq9erVmz56tcePGSZJu376t999/X1WqVLGN6+7uruTk5EdiYtahgbtWrVqaM2eOypYtq3PnzmnUqFGqX7++9u/fr3z58mXoHx0drVGjRjmgUgAAAAB4tDRs2FDTp0+3a0ufTXZ1dVVMTIwqV66skiVLatKkSRneHxoaKovFYluuXbu2Jk6cqNTUVO3bt0+pqal67LHH7N6TnJysggUL2pZdXV1VuXLlnNyth4pDA3fz5s1t/65cubJq1aqlkiVLatGiRXr55Zcz9B8yZIj69+9vW05MTFRAQMADqRUAAAAAHiUeHh4qU6bMPdd///33kqQrV67oypUr8vDwyPLYSUlJcnZ21q5du+Ts7Gy37vez6u7u7nah/VHj8FPKf8/Hx0ePPfaYjhw5kul6q9Uqq9X6gKsCAAAAgH+Wo0ePql+/fvrwww+1cOFCdenSRevWrZOT0//dd3vbtm1279m6dauCg4Pl7OysatWqKTU1VRcvXlT9+vWztW1XV1elpqbmyH44msPvUv57SUlJOnr0qIoWLeroUgAAAADgkZacnKzz58/bvS5fvqzU1FS98MILatq0qbp166bZs2crPj5eEydOtHv/qVOn1L9/fx0+fFifffaZ3n33XUVFRUmSHnvsMXXu3FkRERFasmSJjh8/ru3btys6OlorVqy4b12lSpVSfHy8Dh8+rMuXL+v27dumHQOzOXSGe+DAgWrVqpVKliyps2fPasSIEXJ2dlanTp0cWRYAAAAAPPJWr16dYbKzbNmyev7553Xy5EktX75cklS0aFF98MEH6tSpk5566inbDc4iIiL066+/qmbNmnJ2dlZUVJR69uxpG2v27NkaO3asBgwYoDNnzsjX11ehoaF6+umn71tXjx49FBsbqxo1aigpKUkbNmxQeHh4zu78A2Ixfv+gtAesY8eO2rRpk3755Rf5+fmpXr16euONNxQUFJSl9ycmJsrb21vXr1+Xl5eXydUCAAAAQM747bffdPz4cQUGBsrNzc3R5WRbeHi4qlatqsmTJzu6FFPc7+eTnRzq0BnuBQsWOHLzAAAAAB6Quu/WdXQJOWLz65sdXQJykYfqGm4AAAAAAB4VD9VdygEAAAAAD7/Y2FhHl5ArMMMNAAAAAIAJCNwAAAAAAJiAwA0AAAAAgAkI3AAAAAAAmIDADQAAAACACQjcAAAAAACYgMeCAQAAAMBDpPqgTx7o9na9HfFAt/dPwgw3AAAAACDXO3HihCwWi/bs2WPXfuDAAbVt21alSpWSxWLR5MmTH1hNBG4AAAAAQK6WkpJyz3W3bt1S6dKlNX78eBUpUuQBVkXgBgAAAABkQ1pamqKjoxUYGCh3d3dVqVJFixcvlmEYatKkiZo2bSrDMCRJV65ckb+/v4YPHy5Jio2NlcVi0YoVK1S5cmW5ubkpNDRU+/fvt9tGXFyc6tevL3d3dwUEBKhPnz66efOmbX2pUqU0ZswYRUREyMvLSz179lRgYKAkqVq1arJYLAoPD5ckPfHEE3r77bfVsWNHWa3WB3CE/g+BGwAAAACQZdHR0frkk080Y8YMHThwQP369dMLL7ygTZs2ae7cudqxY4emTp0qSXrllVdUvHhxW+BON2jQIE2cOFE7duyQn5+fWrVqpdu3b0uSjh49qmbNmqlt27aKj4/XwoULFRcXp8jISLsxJkyYoCpVqmj37t0aNmyYtm/fLklat26dzp07pyVLljyAo3F/3DQNAAAAAJAlycnJGjdunNatW6fatWtLkkqXLq24uDjNnDlT8+fP18yZMxUREaHz589r5cqV2r17t1xc7KPniBEj9OSTT0qS5s6dK39/fy1dulTt27dXdHS0OnfurL59+0qSgoODNXXqVIWFhWn69Olyc3OTJDVq1EgDBgywjens7CxJKliw4AM/dfxeCNwAADyE6r5b19El5IjNr292dAkAgBx05MgR3bp1yxaW06WkpKhatWqSpHbt2mnp0qUaP368pk+fruDg4AzjpId1SSpQoIDKli2rQ4cOSZL27t2r+Ph4xcTE2PoYhqG0tDQdP35c5cuXlyTVqFEjx/cvpxG4AQAAAABZkpSUJElasWKFihcvbrcu/froW7duadeuXXJ2dlZCQsJf2kavXr3Up0+fDOtKlChh+7eHh0e2x37QCNwAAAAAgCypUKGCrFarTp06pbCwsEz7DBgwQE5OTlq1apVatGihli1bqlGjRnZ9tm7dagvPV69e1U8//WSbuQ4JCdHBgwdVpkyZbNXm6uoqSUpNTc3ubpmGwA0AAAAAyJJ8+fJp4MCB6tevn9LS0lSvXj1dv35dmzdvlpeXl3x9ffXxxx9ry5YtCgkJ0aBBg9SlSxfFx8crf/78tnFGjx6tggULqnDhwho6dKh8fX3Vpk0bSdLgwYMVGhqqyMhIde/eXR4eHjp48KDWrl2radOm3bO2QoUKyd3dXatXr5a/v7/c3Nzk7e2tlJQUHTx4UNLdU9/PnDmjPXv2yNPTM9uhPrsI3AAAAADwENn1doSjS7ivMWPGyM/PT9HR0Tp27Jh8fHwUEhKiIUOGqEOHDho5cqRCQkIkSaNGjdKaNWv0yiuvaOHChbYxxo8fr6ioKCUkJKhq1apatmyZbYa6cuXK2rhxo4YOHar69evLMAwFBQWpQ4cO963LxcVFU6dO1ejRozV8+HDVr19fsbGxOnv2rO36cunu3c0nTJigsLAwxcbG5vwB+h2Lkf6AtFwoMTFR3t7eun79ury8vBxdDgAAOYabpgF41PC9Zu+3337T8ePHFRgYaLvr9j9BbGysGjZsqKtXr8rHx8fR5dzT/X4+2cmhPIcbAAAAAAATELgBAAAAADAB13ADAAAAAB6I8PBw5eKrmrONGW4AAAAAAExA4AYAAAAAwAQEbgAAAAAATEDgBgAAAADABARuAAAAAABMQOAGAAAAAMAEPBYMAAAAAB4ip0Y//kC3V2L4vge6vX8SZrgBAAAAALneiRMnZLFYtGfPHrv2Dz/8UPXr11f+/PmVP39+NWnSRNu3b38gNRG4AQAAAAC5WkpKyj3XxcbGqlOnTtqwYYO2bNmigIAAPfXUUzpz5ozpdRG4AQAAAABZlpaWpujoaAUGBsrd3V1VqlTR4sWLZRiGmjRpoqZNm8owDEnSlStX5O/vr+HDh0u6G34tFotWrFihypUry83NTaGhodq/f7/dNuLi4lS/fn25u7srICBAffr00c2bN23rS5UqpTFjxigiIkJeXl7q2bOnAgMDJUnVqlWTxWJReHi4JCkmJkavvfaaqlatqnLlyumjjz5SWlqa1q9fb/qxInADAAAAALIsOjpan3zyiWbMmKEDBw6oX79+euGFF7Rp0ybNnTtXO3bs0NSpUyVJr7zyiooXL24L3OkGDRqkiRMnaseOHfLz81OrVq10+/ZtSdLRo0fVrFkztW3bVvHx8Vq4cKHi4uIUGRlpN8aECRNUpUoV7d69W8OGDbOdJr5u3TqdO3dOS5YsybT+W7du6fbt2ypQoEBOH5oMuGkaAAAAACBLkpOTNW7cOK1bt061a9eWJJUuXVpxcXGaOXOm5s+fr5kzZyoiIkLnz5/XypUrtXv3brm42EfPESNG6Mknn5QkzZ07V/7+/lq6dKnat2+v6Ohode7cWX379pUkBQcHa+rUqQoLC9P06dPl5uYmSWrUqJEGDBhgG9PZ2VmSVLBgQRUpUuSe+zB48GAVK1ZMTZo0ybHjci8EbgAAAABAlhw5ckS3bt2yheV0KSkpqlatmiSpXbt2Wrp0qcaPH6/p06crODg4wzjpYV2SChQooLJly+rQoUOSpL179yo+Pl4xMTG2PoZhKC0tTcePH1f58uUlSTVq1Mh2/ePHj9eCBQsUGxtrC+5mInADAAAAALIkKSlJkrRixQoVL17cbp3VapV095TtXbt2ydnZWQkJCX9pG7169VKfPn0yrCtRooTt3x4eHtkad8KECRo/frzWrVunypUrZ7uuv4LADQAAAADIkgoVKshqterUqVMKCwvLtM+AAQPk5OSkVatWqUWLFmrZsqUaNWpk12fr1q228Hz16lX99NNPtpnrkJAQHTx4UGXKlMlWba6urpKk1NTUDOveeustvfHGG/rmm2/+0sz4X0XgBgAAAABkSb58+TRw4ED169dPaWlpqlevnq5fv67NmzfLy8tLvr6++vjjj7VlyxaFhIRo0KBB6tKli+Lj45U/f37bOKNHj1bBggVVuHBhDR06VL6+vmrTpo2ku9dYh4aGKjIyUt27d5eHh4cOHjyotWvXatq0afesrVChQnJ3d9fq1avl7+8vNzc3eXt7680339Tw4cM1f/58lSpVSufPn5ckeXp6ytPT09TjReAGAAAAgIdIieH7HF3CfY0ZM0Z+fn6Kjo7WsWPH5OPjo5CQEA0ZMkQdOnTQyJEjFRISIkkaNWqU1qxZo1deeUULFy60jTF+/HhFRUUpISFBVatW1bJly2wz1JUrV9bGjRs1dOhQ1a9fX4ZhKCgoSB06dLhvXS4uLpo6dapGjx6t4cOHq379+oqNjdX06dOVkpKi5557zq7/iBEjNHLkyJw9OH9gMdIfkJYLJSYmytvbW9evX5eXl5ejywEAIMfUfbeuo0vIEZtf3+zoEgA8JPhes/fbb7/p+PHjCgwMfCA373pYxMbGqmHDhrp69ap8fHwcXc493e/nk50cynO4AQAAAAAwAYEbAAAAAAATcA03AAAAAOCBCA8PVy6+qjnbmOEGAAAAAMAEzHADAADgkVV90CeOLiFH7Ho7wtElAPgLmOEGAAAAAMAEBG4AAAAAAEzAKeV4JPBcRwAAAAAPG2a4AQAAAAAwATPcAAAAAPAQedBnb3KWpXmY4QYAAAAA5HonTpyQxWLRnj177NqXLFmiGjVqyMfHRx4eHqpatarmzZv3QGpihhsAAAAAkKulpKTcc12BAgU0dOhQlStXTq6urlq+fLm6deumQoUKqWnTpqbWxQw3AAAAACDL0tLSFB0drcDAQLm7u6tKlSpavHixDMNQkyZN1LRpUxmGIUm6cuWK/P39NXz4cElSbGysLBaLVqxYocqVK8vNzU2hoaHav3+/3Tbi4uJUv359ubu7KyAgQH369NHNmzdt60uVKqUxY8YoIiJCXl5e6tmzpwIDAyVJ1apVk8ViUXh4uCQpPDxczz77rMqXL6+goCBFRUWpcuXKiouLM/1YEbgBAAAAAFkWHR2tTz75RDNmzNCBAwfUr18/vfDCC9q0aZPmzp2rHTt2aOrUqZKkV155RcWLF7cF7nSDBg3SxIkTtWPHDvn5+alVq1a6ffu2JOno0aNq1qyZ2rZtq/j4eC1cuFBxcXGKjIy0G2PChAmqUqWKdu/erWHDhmn79u2SpHXr1uncuXNasmRJhtoNw9D69et1+PBhNWjQwIzDY4dTygEAAAAAWZKcnKxx48Zp3bp1ql27tiSpdOnSiouL08yZMzV//nzNnDlTEREROn/+vFauXKndu3fLxcU+eo4YMUJPPvmkJGnu3Lny9/fX0qVL1b59e0VHR6tz587q27evJCk4OFhTp05VWFiYpk+fLjc3N0lSo0aNNGDAANuYzs7OkqSCBQuqSJEidtu7fv26ihcvruTkZDk7O+v999+3bd9MBG4AAAAAQJYcOXJEt27dyhBWU1JSVK1aNUlSu3bttHTpUo0fP17Tp09XcHBwhnHSw7p09xrrsmXL6tChQ5KkvXv3Kj4+XjExMbY+hmEoLS1Nx48fV/ny5SVJNWrUyHLd+fLl0549e5SUlKT169erf//+Kl26tO20c7MQuAEAAAAAWZKUlCRJWrFihYoXL263zmq1SpJu3bqlXbt2ydnZWQkJCX9pG7169VKfPn0yrCtRooTt3x4eHlke08nJSWXKlJEkVa1aVYcOHVJ0dDSBGwAAAADwcKhQoYKsVqtOnTqlsLCwTPsMGDBATk5OWrVqlVq0aKGWLVuqUaNGdn22bt1qC89Xr17VTz/9ZJu5DgkJ0cGDB20BOatcXV0lSampqX/aNy0tTcnJydka/68gcAMAAAAAsiRfvnwaOHCg+vXrp7S0NNWrV0/Xr1/X5s2b5eXlJV9fX3388cfasmWLQkJCNGjQIHXp0kXx8fHKnz+/bZzRo0erYMGCKly4sIYOHSpfX1+1adNGkjR48GCFhoYqMjJS3bt3l4eHhw4ePKi1a9dq2rRp96ytUKFCcnd31+rVq+Xv7y83Nzd5e3srOjpaNWrUUFBQkJKTk7Vy5UrNmzdP06dPN/twEbgBAAAA4GGy+fXNji7hvsaMGSM/Pz9FR0fr2LFj8vHxUUhIiIYMGaIOHTpo5MiRCgkJkSSNGjVKa9as0SuvvKKFCxfaxhg/fryioqKUkJCgqlWratmyZbYZ6sqVK2vjxo0aOnSo6tevL8MwFBQUpA4dOty3LhcXF02dOlWjR4/W8OHDVb9+fcXGxurmzZt67bXX9PPPP8vd3V3lypXTp59++qfj5QQCNwAAAAAgyywWi6KiohQVFZVh3fnz5+2W8+TJo507d2boV69evQzP3v69J554QmvWrLnn+hMnTmTa3r17d3Xv3t2ubezYsRo7duw9xzITz+EGAAAAAMAEBG4AAAAAAEzAKeUAAAAAgAciPDxchmE4uowHhhluAAAAAABMQOAGAAAAAMAEBG4AAAAAAExA4AYAAAAAwAQEbgAAAAAATEDgBgAAAADABDwWDAAAAAAeIhsbhD3Q7YVt2vhAt/dPwgw3AAAAACDXO3HihCwWi/bs2XPPPgsWLJDFYlGbNm0eSE0EbgAAAABArpaSkvKnfU6cOKGBAweqfv36D6CiuwjcAAAAAIAsS0tLU3R0tAIDA+Xu7q4qVapo8eLFMgxDTZo0UdOmTWUYhiTpypUr8vf31/DhwyVJsbGxslgsWrFihSpXriw3NzeFhoZq//79dtuIi4tT/fr15e7uroCAAPXp00c3b960rS9VqpTGjBmjiIgIeXl5qWfPngoMDJQkVatWTRaLReHh4bb+qamp6ty5s0aNGqXSpUubfIT+D4EbAAAAAJBl0dHR+uSTTzRjxgwdOHBA/fr10wsvvKBNmzZp7ty52rFjh6ZOnSpJeuWVV1S8eHFb4E43aNAgTZw4UTt27JCfn59atWql27dvS5KOHj2qZs2aqW3btoqPj9fChQsVFxenyMhIuzEmTJigKlWqaPfu3Ro2bJi2b98uSVq3bp3OnTunJUuW2PqOHj1ahQoV0ssvv2zmocmAm6YBAB45p0Y/7ugS/r78Xo6uAACADJKTkzVu3DitW7dOtWvXliSVLl1acXFxmjlzpubPn6+ZM2cqIiJC58+f18qVK7V79265uNhHzxEjRujJJ5+UJM2dO1f+/v5aunSp2rdvr+joaHXu3Fl9+/aVJAUHB2vq1KkKCwvT9OnT5ebmJklq1KiRBgwYYBvT2dlZklSwYEEVKVLE1h4XF6dZs2bd99pusxC4AQAAAABZcuTIEd26dcsWltOlpKSoWrVqkqR27dpp6dKlGj9+vKZPn67g4OAM46SHdUkqUKCAypYtq0OHDkmS9u7dq/j4eMXExNj6GIahtLQ0HT9+XOXLl5ck1ahR40/rvXHjhl588UV9+OGH8vX1zf4O/00EbgAAAABAliQlJUmSVqxYoeLFi9uts1qtkqRbt25p165dcnZ2VkJCwl/aRq9evdSnT58M60qUKGH7t4eHx5+OdfToUZ04cUKtWrWytaWlpUmSXFxcdPjwYQUFBWW7xqwicAMAAAAAsqRChQqyWq06deqUwsIyf174gAED5OTkpFWrVqlFixZq2bKlGjVqZNdn69attvB89epV/fTTT7aZ65CQEB08eFBlypTJVm2urq6S7t4gLV25cuW0b98+u37/+9//dOPGDU2ZMkUBAQHZ2kZ2EbgBAAAAAFmSL18+DRw4UP369VNaWprq1aun69eva/PmzfLy8pKvr68+/vhjbdmyRSEhIRo0aJC6dOmi+Ph45c+f3zbO6NGjVbBgQRUuXFhDhw6Vr6+v7dnYgwcPVmhoqCIjI9W9e3d5eHjo4MGDWrt2raZNm3bP2goVKiR3d3etXr1a/v7+cnNzk7e3typVqmTXz8fHR5IytJuBwA0AAAAAD5GwTRsdXcJ9jRkzRn5+foqOjtaxY8fk4+OjkJAQDRkyRB06dNDIkSMVEhIiSRo1apTWrFmjV155RQsXLrSNMX78eEVFRSkhIUFVq1bVsmXLbDPUlStX1saNGzV06FDVr19fhmEoKChIHTp0uG9dLi4umjp1qkaPHq3hw4erfv36io2NNe04ZAWBGwAAAACQZRaLRVFRUYqKisqw7vz583bLefLk0c6dOzP0q1evXoZnb//eE088oTVr1txz/YkTJzJt7969u7p3737P90nSnDlz7rs+J/EcbgAAAAAATEDgBgAAAADABJxSDgAAAAB4IMLDw2UYhqPLeGCY4QYAAAAAwAQEbgAAAAAATEDgBgAAAADABARuAAAAAABMwE3TAAA21Qd94ugScsTSfI6uAAAAgMD9tzwqv5juejvC0SUAAAAAwCOHwA0AAAAAD5FpA5Y90O1FTmz1QLf3T/LQXMM9fvx4WSwW9e3b19GlAAAAAABymRMnTshisWjPnj127XPmzJHFYrF7ubm5PZCaHooZ7h07dmjmzJmqXLmyo0sBAAAAAOQyKSkp913v5eWlw4cP25YtFovZJUl6CGa4k5KS1LlzZ3344YfKnz+/o8sBAAAAANxHWlqaoqOjFRgYKHd3d1WpUkWLFy+WYRhq0qSJmjZtKsMwJElXrlyRv7+/hg8fLkmKjY2VxWLRihUrVLlyZbm5uSk0NFT79++320ZcXJzq168vd3d3BQQEqE+fPrp586ZtfalSpTRmzBhFRETIy8tLPXv2VGBgoCSpWrVqslgsCg8Pt/W3WCwqUqSI7VW4cGGTj9JdDg/cvXv3VsuWLdWkSZM/7ZucnKzExES7FwAAAADgwYmOjtYnn3yiGTNm6MCBA+rXr59eeOEFbdq0SXPnztWOHTs0depUSdIrr7yi4sWL2wJ3ukGDBmnixInasWOH/Pz81KpVK92+fVuSdPToUTVr1kxt27ZVfHy8Fi5cqLi4OEVGRtqNMWHCBFWpUkW7d+/WsGHDtH37dknSunXrdO7cOS1ZssTWNykpSSVLllRAQICeeeYZHThwwMxDZOPQU8oXLFigH374QTt27MhS/+joaI0aNcrkqgAAAAAAmUlOTta4ceO0bt061a5dW5JUunRpxcXFaebMmZo/f75mzpypiIgInT9/XitXrtTu3bvl4mIfPUeMGKEnn3xSkjR37lz5+/tr6dKlat++vaKjo9W5c2fb/b2Cg4M1depUhYWFafr06bbrrxs1aqQBAwbYxnR2dpYkFSxYUEWKFLG1ly1bVh9//LEqV66s69eva8KECapTp44OHDggf39/046V5MDAffr0aUVFRWnt2rVZvmB9yJAh6t+/v205MTFRAQEBZpUIAAAAAPidI0eO6NatW7awnC4lJUXVqlWTJLVr105Lly7V+PHjNX36dAUHB2cYJz2sS1KBAgVUtmxZHTp0SJK0d+9excfHKyYmxtbHMAylpaXp+PHjKl++vCSpRo0aWaq5du3adturU6eOypcvr5kzZ2rMmDFZ3PO/xmGBe9euXbp48aJCQkJsbampqdq0aZOmTZum5ORk218o0lmtVlmt1gddKgAAAABAd0/NlqQVK1aoePHiduvSs9qtW7e0a9cuOTs7KyEh4S9to1evXurTp0+GdSVKlLD928PDI9tjS1KePHlUrVo1HTly5C+9PzscFrgbN26sffv22bV169ZN5cqV0+DBgzOEbQAAAACAY1WoUEFWq1WnTp1SWFhYpn0GDBggJycnrVq1Si1atFDLli3VqFEjuz5bt261heerV6/qp59+ss1ch4SE6ODBgypTpky2anN1dZV0dyL3flJTU7Vv3z61aNEiW+P/FQ4L3Pny5VOlSpXs2jw8PFSwYMEM7QAAAAAAx8uXL58GDhyofv36KS0tTfXq1dP169e1efNmeXl5ydfXVx9//LG2bNmikJAQDRo0SF26dFF8fLzdU6lGjx6tggULqnDhwho6dKh8fX3Vpk0bSdLgwYMVGhqqyMhIde/eXR4eHjp48KDWrl2radOm3bO2QoUKyd3dXatXr5a/v7/c3Nzk7e2t0aNHKzQ0VGXKlNG1a9f09ttv6+TJk+revbvZh+vheA43AAAAAOCuyImtHF3CfY0ZM0Z+fn6Kjo7WsWPH5OPjo5CQEA0ZMkQdOnTQyJEjbZcOjxo1SmvWrNErr7yihQsX2sYYP368oqKilJCQoKpVq2rZsmW2GerKlStr48aNGjp0qOrXry/DMBQUFKQOHTrcty4XFxdNnTpVo0eP1vDhw1W/fn3Fxsbq6tWr6tGjh86fP6/8+fOrevXq+v7771WhQgXzDlJ6TaZvIRtiY2MdXQIAAAAA4D4sFouioqIUFRWVYd358+ftlvPkyaOdO3dm6FevXr0Mz97+vSeeeEJr1qy55/oTJ05k2t69e/cMM9eTJk3SpEmT7jmWmRz+HG4AAAAAAB5FBG4AAAAAAEzwUJ1SDgAAAAB4dIWHh8swDEeX8cAwww0AAAAAgAkI3AAAAADgIGlpaY4uAZnIqZ8Lp5QDAAAAwAPm6uoqJycnnT17Vn5+fnJ1dZXFYnF0Wf94hmEoJSVFly5dkpOTk+1RZX8VgRsAAAAAHjAnJycFBgbq3LlzOnv2rKPLwR/kzZtXJUqUkJPT3zspnMANAAAAAA7g6uqqEiVK6M6dO0pNTXV0Ofj/nJ2d5eLikiNnHBC4AQAAAMBBLBaL8uTJozx58ji6FJiAm6YBAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACnsMNAAAAAP8w0wYsc3QJf1vkxFaOLuFPMcMNAAAAAIAJCNwAAAAAAJiAwA0AAAAAgAkI3AAAAAAAmIDADQAAAACACQjcAAAAAACYgMANAAAAAIAJCNwAAAAAAJiAwA0AAAAAgAlcHF0AAAAAAOQWGxuEObqEnPHEQEdX8I/ADDcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAIXRxcAAAAeXRsbhDm6hBwRtmmjo0sAAORCzHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACZwcXQBAAAAD7tpA5Y5uoS/LXJiK0eXAAD/OMxwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACruEGHiIbG4Q5uoQcEbZpo6NLAAAAAByOGW4AAAAAAExA4AYAAAAAwAQEbgAAAAAATEDgBgAAAADABARuAAAAAABMQOAGAAAAAMAEBG4AAAAAAExA4AYAAAAAwAQEbgAAAAAATEDgBgAAAADABARuAAAAAABMQOAGAAAAAMAEBG4AAAAAAExA4AYAAAAAwAQEbgAAAAAATEDgBgAAAADABARuAAAAAABMQOAGAAAAAMAEBG4AAAAAAExA4AYAAAAAwAQEbgAAAAAATODi6ALgeKdGP+7oEv6+/F6OrgAAAAAA7DDDDQAAAACACQjcAAAAAACYgMANAAAAAIAJCNwAAAAAAJiAwA0AAAAAgAkI3AAAAAAAmIDADQAAAACACQjcAAAAAACYgMANAAAAAIAJXBxdAIBHz7QByxxdwt8WObGVo0sAAABALscMNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGCCvxS4v/vuO73wwguqXbu2zpw5I0maN2+e4uLicrQ4AAAAAAByq2wH7i+++EJNmzaVu7u7du/ereTkZEnS9evXNW7cuBwvEAAAAACA3CjbgXvs2LGaMWOGPvzwQ+XJk8fWXrduXf3www/ZGmv69OmqXLmyvLy85OXlpdq1a2vVqlXZLQkAAAAAgIdOtgP34cOH1aBBgwzt3t7eunbtWrbG8vf31/jx47Vr1y7t3LlTjRo10jPPPKMDBw5ktywAAAAAAB4q2Q7cRYoU0ZEjRzK0x8XFqXTp0tkaq1WrVmrRooWCg4P12GOP6Y033pCnp6e2bt2a3bIAAAAAAHiouGT3DT169FBUVJQ+/vhjWSwWnT17Vlu2bNHAgQM1bNiwv1xIamqqPv/8c928eVO1a9fOtE9ycrLtmnFJSkxM/MvbAwAAAADATNkO3P/5z3+Ulpamxo0b69atW2rQoIGsVqsGDhyo119/PdsF7Nu3T7Vr19Zvv/0mT09PLV26VBUqVMi0b3R0tEaNGpXtbQAAAAAA8KBl65Ty1NRUfffdd+rdu7euXLmi/fv3a+vWrbp06ZLGjBnzlwooW7as9uzZo23btunVV19Vly5ddPDgwUz7DhkyRNevX7e9Tp8+/Ze2CQAAAACA2bI1w+3s7KynnnpKhw4dko+Pzz1norPD1dVVZcqUkSRVr15dO3bs0JQpUzRz5swMfa1Wq6xW69/eJgAAAAAAZsv2TdMqVaqkY8eOmVGLJCktLc3uOm0AAAAAAHKjbF/DPXbsWA0cOFBjxoxR9erV5eHhYbfey8sry2MNGTJEzZs3V4kSJXTjxg3Nnz9fsbGx+uabb7JbFgAAAAAAD5VsB+4WLVpIklq3bi2LxWJrNwxDFotFqampWR7r4sWLioiI0Llz5+Tt7a3KlSvrm2++0ZNPPpndsgAAAAAAeKhkO3Bv2LAhxzY+a9asHBsLAAAAAICHSbYDd1hYmBl1AAAAAADwSMl24Jaka9euadasWTp06JAkqWLFinrppZfk7e2do8UBAAAAAJBbZfsu5Tt37lRQUJAmTZqkK1eu6MqVK3rnnXcUFBSkH374wYwaAQAAAADIdbI9w92vXz+1bt1aH374oVxc7r79zp076t69u/r27atNmzbleJEAAAAAAOQ22Q7cO3futAvbkuTi4qJ///vfqlGjRo4WBwAAAABAbpXtU8q9vLx06tSpDO2nT59Wvnz5cqQoAAAAAAByu2wH7g4dOujll1/WwoULdfr0aZ0+fVoLFixQ9+7d1alTJzNqBAAAAAAg18n2KeUTJkyQxWJRRESE7ty5I0nKkyePXn31VY0fPz7HCwQAAAAAIDfKduB2dXXVlClTFB0draNHj0qSgoKClDdv3hwvDgAAAACA3Crbgfv69etKTU1VgQIF9Pjjj9var1y5IhcXF3l5eeVogQAAAAAA5EbZvoa7Y8eOWrBgQYb2RYsWqWPHjjlSFAAAAAAAuV22A/e2bdvUsGHDDO3h4eHatm1bjhQFAAAAAEBul+3AnZycbLtZ2u/dvn1bv/76a44UBQAAAABAbpftwF2zZk198MEHGdpnzJih6tWr50hRAAAAAADkdtm+adrYsWPVpEkT7d27V40bN5YkrV+/Xjt27NCaNWtyvEAAAAAAAHKjbM9w161bV1u2bFFAQIAWLVqkZcuWqUyZMoqPj1f9+vXNqBEAAAAAgFwn2zPcklS1alXFxMTkdC0AAAAAMnFq9ON/3ulhl5/HB+OfJ8uB+86dO0pNTZXVarW1XbhwQTNmzNDNmzfVunVr1atXz5QiAQAAAADIbbIcuHv06CFXV1fNnDlTknTjxg098cQT+u2331S0aFFNmjRJX331lVq0aGFasQAAAAAA5BZZvoZ78+bNatu2rW35k08+UWpqqhISErR37171799fb7/9tilFAgAAAACQ22Q5cJ85c0bBwcG25fXr16tt27by9vaWJHXp0kUHDhzI+QoBAAAAAMiFshy43dzc9Ouvv9qWt27dqlq1atmtT0pKytnqAAAAAADIpbIcuKtWrap58+ZJkr777jtduHBBjRo1sq0/evSoihUrlvMVAgAAAACQC2X5pmnDhw9X8+bNtWjRIp07d05du3ZV0aJFbeuXLl2qunXrmlIkAAAAAAC5TZYDd1hYmHbt2qU1a9aoSJEiateund36qlWrqmbNmjleIAAAAAAAuVGWA7cklS9fXuXLl890Xc+ePXOkIAAAAAAAHgVZvoYbAAAAAABkHYEbAAAAAAATELgBAAAAADABgRsAAAAAABP8pcB97do1ffTRRxoyZIiuXLkiSfrhhx905syZHC0OAAAAAIDcKlt3KZek+Ph4NWnSRN7e3jpx4oR69OihAgUKaMmSJTp16pQ++eQTM+oEAAAAACBXyfYMd//+/dW1a1clJCTIzc3N1t6iRQtt2rQpR4sDAAAAACC3ynbg3rFjh3r16pWhvXjx4jp//nyOFAUAAAAAQG6X7cBttVqVmJiYof2nn36Sn59fjhQFAAAAAEBul+3A3bp1a40ePVq3b9+WJFksFp06dUqDBw9W27Ztc7xAAAAAAAByo2wH7okTJyopKUmFChXSr7/+qrCwMJUpU0b58uXTG2+8YUaNAAAAAADkOtm+S7m3t7fWrl2ruLg4xcfHKykpSSEhIWrSpIkZ9QEAAAAAkCtlO3Cnq1evnurVq5eTtQAAAAAA8MjIduCeOnVqpu0Wi0Vubm4qU6aMGjRoIGdn579dHAAAAAAAuVW2A/ekSZN06dIl3bp1S/nz55ckXb16VXnz5pWnp6cuXryo0qVLa8OGDQoICMjxggEAAAAAyA2yfdO0cePG6YknnlBCQoJ++eUX/fLLL/rpp59Uq1YtTZkyRadOnVKRIkXUr18/M+oFAAAAACBXyPYM9//+9z998cUXCgoKsrWVKVNGEyZMUNu2bXXs2DG99dZbPCIMAAAAAPCPlu0Z7nPnzunOnTsZ2u/cuaPz589LkooVK6YbN278/eoAAAAAAMilsh24GzZsqF69emn37t22tt27d+vVV19Vo0aNJEn79u1TYGBgzlUJAAAAAEAuk+3APWvWLBUoUEDVq1eX1WqV1WpVjRo1VKBAAc2aNUuS5OnpqYkTJ+Z4sQAAAAAA5BbZvoa7SJEiWrt2rX788Uf99NNPkqSyZcuqbNmytj4NGzbMuQoBAAAAAMiFsh2405UrV07lypXLyVoAAAAAAHhk/KXA/fPPP+vrr7/WqVOnlJKSYrfunXfeyZHCAAAAAADIzbIduNevX6/WrVurdOnS+vHHH1WpUiWdOHFChmEoJCTEjBoBAAAAAMh1sn3TtCFDhmjgwIHat2+f3Nzc9MUXX+j06dMKCwtTu3btzKgRAAAAAIBcJ9uB+9ChQ4qIiJAkubi46Ndff5Wnp6dGjx6tN998M8cLBAAAAAAgN8p24Pbw8LBdt120aFEdPXrUtu7y5cs5VxkAAAAAALlYtq/hDg0NVVxcnMqXL68WLVpowIAB2rdvn5YsWaLQ0FAzagQAAAAAINfJduB+5513lJSUJEkaNWqUkpKStHDhQgUHB3OHcgAAAAAA/r9sBe7U1FT9/PPPqly5sqS7p5fPmDHDlMIAAAAAAMjNsnUNt7Ozs5566ildvXrVrHoAAAAAAHgkZPumaZUqVdKxY8fMqAUAAAAAgEdGtgP32LFjNXDgQC1fvlznzp1TYmKi3QsAAAAAAPyFm6a1aNFCktS6dWtZLBZbu2EYslgsSk1NzbnqAAAAAADIpbIduDds2GBGHQAAAAAAPFKyHbjDwsLMqAMAAAAAgEdKtq/hlqTvvvtOL7zwgurUqaMzZ85IkubNm6e4uLgcLQ4AAAAAgNwq24H7iy++UNOmTeXu7q4ffvhBycnJkqTr169r3LhxOV4gAAAAAAC50V+6S/mMGTP04YcfKk+ePLb2unXr6ocffsjR4gAAAAAAyK2yHbgPHz6sBg0aZGj39vbWtWvXcqImAAAAAAByvWwH7iJFiujIkSMZ2uPi4lS6dOkcKQoAAAAAgNwu24G7R48eioqK0rZt22SxWHT27FnFxMRo4MCBevXVV82oEQAAAACAXCfbjwX7z3/+o7S0NDVu3Fi3bt1SgwYNZLVaNXDgQL3++utm1AgAAAAAQK6T7cBtsVg0dOhQDRo0SEeOHFFSUpIqVKggT09PM+oDAAAAACBXyvYp5Z9++qlu3bolV1dXVahQQTVr1iRsAwAAAADwB9kO3P369VOhQoX0/PPPa+XKlUpNTTWjLgAAAAAAcrVsB+5z585pwYIFslgsat++vYoWLarevXvr+++/N6M+AAAAAABypWwHbhcXFz399NOKiYnRxYsXNWnSJJ04cUINGzZUUFCQGTUCAAAAAJDrZPumab+XN29eNW3aVFevXtXJkyd16NChnKoLAAAAAIBcLdsz3JJ069YtxcTEqEWLFipevLgmT56sZ599VgcOHMjp+gAAAAAAyJWyPcPdsWNHLV++XHnz5lX79u01bNgw1a5d24zaAAAAAADItbIduJ2dnbVo0SI1bdpUzs7Oduv279+vSpUq5VhxAAAAAADkVtkO3DExMXbLN27c0GeffaaPPvpIu3bt4jFhAAAAAADoL17DLUmbNm1Sly5dVLRoUU2YMEGNGjXS1q1bc7I2AAAAAAByrWzNcJ8/f15z5szRrFmzlJiYqPbt2ys5OVlffvmlKlSoYFaNAAAAAADkOlme4W7VqpXKli2r+Ph4TZ48WWfPntW7775rZm0AAAAAAORaWZ7hXrVqlfr06aNXX31VwcHBZtYEAAAAAECul+UZ7ri4ON24cUPVq1dXrVq1NG3aNF2+fNnM2gAAAAAAyLWyHLhDQ0P14Ycf6ty5c+rVq5cWLFigYsWKKS0tTWvXrtWNGzfMrBMAAAAAgFwl23cp9/Dw0EsvvaS4uDjt27dPAwYM0Pjx41WoUCG1bt3ajBoBAAAAAMh1/vJjwSSpbNmyeuutt/Tzzz/rs88+y6maAAAAAADI9f5W4E7n7OysNm3a6Ouvv86J4QAAAAAAyPVyJHADAAAAAAB7BG4AAAAAAExA4AYAAAAAwAQEbgAAAAAATEDgBgAAAADABA4N3NHR0XriiSeUL18+FSpUSG3atNHhw4cdWRIAAAAAADnCoYF748aN6t27t7Zu3aq1a9fq9u3beuqpp3Tz5k1HlgUAAAAAwN/m4siNr1692m55zpw5KlSokHbt2qUGDRo4qCoAAAAAAP4+hwbuP7p+/bokqUCBApmuT05OVnJysm05MTHxgdQFAAAAAEB2PTQ3TUtLS1Pfvn1Vt25dVapUKdM+0dHR8vb2tr0CAgIecJUAAAAAAGTNQxO4e/furf3792vBggX37DNkyBBdv37d9jp9+vQDrBAAAAAAgKx7KE4pj4yM1PLly7Vp0yb5+/vfs5/VapXVan2AlQEAAAAA8Nc4NHAbhqHXX39dS5cuVWxsrAIDAx1ZDgAAAAAAOcahgbt3796aP3++vvrqK+XLl0/nz5+XJHl7e8vd3d2RpQEAAAAA8Lc49Bru6dOn6/r16woPD1fRokVtr4ULFzqyLAAAAAAA/jaHn1IOAAAAAMCj6KG5SzkAAAAAAI8SAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAKHBu5NmzapVatWKlasmCwWi7788ktHlgMAAAAAQI5xaOC+efOmqlSpovfee8+RZQAAAAAAkONcHLnx5s2bq3nz5o4sAQAAAAAAUzg0cGdXcnKykpOTbcuJiYkOrAYAAAAAgHvLVTdNi46Olre3t+0VEBDg6JIAAAAAAMhUrgrcQ4YM0fXr122v06dPO7okAAAAAAAylatOKbdarbJarY4uAwAAAACAP5WrZrgBAAAAAMgtHDrDnZSUpCNHjtiWjx8/rj179qhAgQIqUaKEAysDAAAAAODvcWjg3rlzpxo2bGhb7t+/vySpS5cumjNnjoOqAgAAAADg73No4A4PD5dhGI4sAQAAAAAAU3ANNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmeCgC93vvvadSpUrJzc1NtWrV0vbt2x1dEgAAAAAAf4vDA/fChQvVv39/jRgxQj/88IOqVKmipk2b6uLFi44uDQAAAACAv8zhgfudd95Rjx491K1bN1WoUEEzZsxQ3rx59fHHHzu6NAAAAAAA/jIXR248JSVFu3bt0pAhQ2xtTk5OatKkibZs2ZKhf3JyspKTk23L169flyQlJiaaX2wmUpN/dch2c9qNPKmOLuFvu/PrHUeXkCNuPhq7oV+Tbzm6hL/NUd8rjsb32sOD77WHC99ruRffaw8PvtceLnyv/f3tGobxp30tRlZ6meTs2bMqXry4vv/+e9WuXdvW/u9//1sbN27Utm3b7PqPHDlSo0aNetBlAgAAAABg5/Tp0/L3979vH4fOcGfXkCFD1L9/f9tyWlqarly5ooIFC8pisTiwMjzqEhMTFRAQoNOnT8vLy8vR5QDA38b3GoBHDd9reFAMw9CNGzdUrFixP+3r0MDt6+srZ2dnXbhwwa79woULKlKkSIb+VqtVVqvVrs3Hx8fMEgE7Xl5efIEDeKTwvQbgUcP3Gh4Eb2/vLPVz6E3TXF1dVb16da1fv97WlpaWpvXr19udYg4AAAAAQG7j8FPK+/fvry5duqhGjRqqWbOmJk+erJs3b6pbt26OLg0AAAAAgL/M4YG7Q4cOunTpkoYPH67z58+ratWqWr16tQoXLuzo0gAbq9WqESNGZLikAQByK77XADxq+F7Dw8ihdykHAAAAAOBR5dBruAEAAAAAeFQRuAEAAAAAMAGBGwAAAAAAExC48cgKDw9X375977m+VKlSmjx58gOrBwByg5EjR6pq1aqOLgMAMpWV39/4HQ8PEwI3AACwGThwoNavX+/oMgDgL9uxY4d69uyZY+MR4PF3OPyxYAAA4MFITU2VxWKRk9O9/97u6ekpT0/PB1gVAOQsPz8/R5cA2DDDjUfanTt3FBkZKW9vb/n6+mrYsGG615PwTp06pWeeeUaenp7y8vJS+/btdeHCBbs+06dPV1BQkFxdXVW2bFnNmzfPbr3FYtHMmTP19NNPK2/evCpfvry2bNmiI0eOKDw8XB4eHqpTp46OHj1q2j4DeHSEh4crMjLynt9jycnJGjhwoIoXLy4PDw/VqlVLsbGxtvfPmTNHPj4++vrrr1WhQgVZrVadOnVKsbGxqlmzpjw8POTj46O6devq5MmTkjKeUr5jxw49+eST8vX1lbe3t8LCwvTDDz/Y1WmxWPTRRx/p2WefVd68eRUcHKyvv/7a9OMDIHdavHixHn/8cbm7u6tgwYJq0qSJbt68menlgG3atFHXrl3t2m7cuKFOnTrJw8NDxYsX13vvvWe3/o8z0teuXVP37t3l5+cnLy8vNWrUSHv37rV7z7Jly/TEE0/Izc1Nvr6+evbZZyXd/R4+efKk+vXrJ4vFIovFkmPHAf8MBG480ubOnSsXFxdt375dU6ZM0TvvvKOPPvooQ7+0tDQ988wzunLlijZu3Ki1a9fq2LFj6tChg63P0qVLFRUVpQEDBmj//v3q1auXunXrpg0bNtiNNWbMGEVERGjPnj0qV66cnn/+efXq1UtDhgzRzp07ZRiGIiMjTd93AI+G+32PRUZGasuWLVqwYIHi4+PVrl07NWvWTAkJCbb337p1S2+++aY++ugjHThwQAUKFFCbNm0UFham+Ph4bdmyRT179rznL5E3btxQly5dFBcXp61btyo4OFgtWrTQjRs37PqNGjVK7du3V3x8vFq0aKHOnTvrypUr5h0YALnSuXPn1KlTJ7300ks6dOiQYmNj9a9//eueEyKZefvtt1WlShXt3r1b//nPfxQVFaW1a9fes3+7du108eJFrVq1Srt27VJISIgaN25s+45asWKFnn32WbVo0UK7d+/W+vXrVbNmTUnSkiVL5O/vr9GjR+vcuXM6d+7c3zsA+OcxgEdUWFiYUb58eSMtLc3WNnjwYKN8+fKGYRhGyZIljUmTJhmGYRhr1qwxnJ2djVOnTtn6HjhwwJBkbN++3TAMw6hTp47Ro0cPu220a9fOaNGihW1ZkvG///3PtrxlyxZDkjFr1ixb22effWa4ubnl3I4CeGTd73vs5MmThrOzs3HmzBm79zRu3NgYMmSIYRiGMXv2bEOSsWfPHtv6X375xZBkxMbGZrrNESNGGFWqVLlnTampqUa+fPmMZcuW2dr++N2XlJRkSDJWrVqVrf0F8OjbtWuXIck4ceJEhnVhYWFGVFSUXdszzzxjdOnSxbZcsmRJo1mzZnZ9OnToYDRv3tyuT/rveN99953h5eVl/Pbbb3bvCQoKMmbOnGkYhmHUrl3b6Ny58z1r/v14QHYxw41HWmhoqN2sTe3atZWQkKDU1FS7focOHVJAQIACAgJsbRUqVJCPj48OHTpk61O3bl2799WtW9e2Pl3lypVt/y5cuLAk6fHHH7dr++2335SYmPg39w7AP8G9vsf27dun1NRUPfbYY7brrj09PbVx40a7y1ZcXV3tvpcKFCigrl27qmnTpmrVqpWmTJly3xmbCxcuqEePHgoODpa3t7e8vLyUlJSkU6dO2fX7/TY8PDzk5eWlixcv5sQhAPAIqVKliho3bqzHH39c7dq104cffqirV69ma4zatWtnWP7j72Pp9u7dq6SkJBUsWNDuu/L48eO278o9e/aocePGf22HgD/BTdOAHJYnTx7bv9N/Sc6sLS0t7cEWBuCRkpSUJGdnZ+3atUvOzs52635/0zN3d/cMp4vPnj1bffr00erVq7Vw4UL973//09q1axUaGpphO126dNEvv/yiKVOmqGTJkrJarapdu7ZSUlLs+v3+e066+13H9xyAP3J2dtbatWv1/fffa82aNXr33Xc1dOhQbdu2TU5OThlOLb99+/bf2l5SUpKKFi1qd3+LdD4+PpLufk8CZmGGG4+0bdu22S2nX3/4x19Oy5cvr9OnT+v06dO2toMHD+ratWuqUKGCrc/mzZvt3rd582bbegAww72+x6pVq6bU1FRdvHhRZcqUsXsVKVLkT8etVq2ahgwZou+//16VKlXS/PnzM+23efNm9enTRy1atFDFihVltVp1+fLlHNk3AP9MFotFdevW1ahRo7R79265urpq6dKl8vPzszvjJjU1Vfv378/w/q1bt2ZYLl++fKbbCgkJ0fnz5+Xi4pLhu9LX11fS3TN07vc4RFdX1wxnRwJZxQw3HmmnTp1S//791atXL/3www969913NXHixAz9mjRposcff1ydO3fW5MmTdefOHb322msKCwtTjRo1JEmDBg1S+/btVa1aNTVp0kTLli3TkiVLtG7duge9WwD+Qe71PfbYY4+pc+fOioiI0MSJE1WtWjVdunRJ69evV+XKldWyZctMxzt+/Lg++OADtW7dWsWKFdPhw4eVkJCgiIiITPsHBwdr3rx5qlGjhhITEzVo0CBmgwD8Zdu2bdP69ev11FNPqVChQtq2bZsuXbqk8uXLy8PDQ/3799eKFSsUFBSkd955R9euXcswxubNm/XWW2+pTZs2Wrt2rT7//HOtWLEi0+01adJEtWvXVps2bfTWW2/pscce09mzZ203SqtRo4ZGjBihxo0bKygoSB07dtSdO3e0cuVKDR48WNLdu55v2rRJHTt2lNVqtQV1ICsI3HikRURE6Ndff1XNmjXl7OysqKgo9ezZM0M/i8Wir776Sq+//roaNGggJycnNWvWTO+++66tT5s2bTRlyhRNmDBBUVFRCgwM1OzZsxUeHv4A9wjAP839vsdmz56tsWPHasCAATpz5ox8fX0VGhqqp59++p7j5c2bVz/++KPmzp2rX375RUWLFlXv3r3Vq1evTPvPmjVLPXv2VEhIiAICAjRu3DgNHDjQlH0F8Ojz8vLSpk2bNHnyZCUmJqpkyZKaOHGimjdvrtu3b2vv3r2KiIiQi4uL+vXrp4YNG2YYY8CAAdq5c6dGjRolLy8vvfPOO2ratGmm27NYLFq5cqWGDh2qbt266dKlSypSpIgaNGhgu9dOeHi4Pv/8c40ZM0bjx4+Xl5eXGjRoYBtj9OjR6tWrl4KCgpScnJytO6oDFoNPDAAAD6Xw8HBVrVrV7nmyAID7K1q0qMaMGaPu3bs7uhSAGW4AAAAAud+tW7e0efNmXbhwQRUrVnR0OYAkbpoGAAAA4BHwwQcfqGPHjurbt2+GR4cBjsIp5QAAAAAAmIAZbgAAAAAATEDgBgAAAADABARuAAAAAABMQOAGAAAAAMAEBG4AAAAAAExA4AYAAAAAwAQEbgAAcoktW7bI2dlZLVu2dHQpdk6cOCGLxaI9e/Y4uhQAAB4qBG4AAHKJWbNm6fXXX9emTZt09uxZR5cDAAD+BIEbAIBcICkpSQsXLtSrr76qli1bas6cObZ1V69eVefOneXn5yd3d3cFBwdr9uzZkv5v9nnBggWqU6eO3NzcVKlSJW3cuNFu/P3796t58+by9PRU4cKF9eKLL+ry5cu29WlpaXrrrbdUpkwZWa1WlShRQm+88YYkKTAwUJJUrVo1WSwWhYeHm3swAADIJQjcAADkAosWLVK5cuVUtmxZvfDCC/r4449lGIYkadiwYTp48KBWrVqlQ4cOafr06fL19bV7/6BBgzRgwADt3r1btWvXVqtWrfTLL79Ikq5du6ZGjRqpWrVq2rlzp1avXq0LFy6offv2tvcPGTJE48ePt21r/vz5Kly4sCRp+/btkqR169bp3LlzWrJkyYM4JAAAPPQsRvr/rQEAwEOrbt26at++vaKionTnzh0VLVpUn3/+ucLDw9W6dWv5+vrq448/zvC+EydOKDAwUOPHj9fgwYMlSXfu3FFgYKBef/11/fvf/9bYsWP13Xff6ZtvvrG97+eff1ZAQIAOHz6sokWLys/PT9OmTVP37t3vuY3du3eratWqph0DAAByGxdHFwAAAO7v8OHD2r59u5YuXSpJcnFxUYcOHTRr1iyFh4fr1VdfVdu2bfXDDz/oqaeeUps2bVSnTh27MWrXrm37t4uLi2rUqKFDhw5Jkvbu3asNGzbI09Mzw7aPHj2qa9euKTk5WY0bNzZxLwEAePQQuAEAeMjNmjVLd+7cUbFixWxthmHIarVq2rRpat68uU6ePKmVK1dq7dq1aty4sXr37q0JEyZkafykpCS1atVKb775ZoZ1RYsW1bFjx3JsXwAA+CfhGm4AAB5id+7c0SeffKKJEydqz549ttfevXtVrFgxffbZZ5IkPz8/denSRZ9++qkmT56sDz74wG6crVu32o25a9culS9fXpIUEhKiAwcOqFSpUipTpozdy8PDQ8HBwXJ3d9f69eszrdHV1VWSlJqaasYhAAAg12KGGwCAh9jy5ct19epVvfzyy/L29rZb17ZtW82aNUtnz55V9erVVbFiRSUnJ2v58uW2MJ3uvffeU3BwsMqXL69Jkybp6tWreumllyRJvXv31ocffqhOnTrp3//+twoUKKAjR45owYIF+uijj+Tm5qbBgwfr3//+t1xdXVW3bl1dunRJBw4c0Msvv6xChQrJ3d1dq1evlr+/v9zc3DLUCgDAPxEz3AAAPMRmzZqlJk2aZBpg27Ztq507d8rFxUVDhgxR5cqV1aBBAzk7O2vBggV2fcePH6/x48erSpUqiouL09dff227k3mxYsW0efNmpaam6qmnntLjjz+uvn37ysfHR05Od39VGDZsmAYMGKDhw4erfPny6tChgy5evCjp7jXhU6dO1cyZM1WsWDE988wzJh8VAAByB+5SDgDAI4w7iAMA4DjMcAMAAAAAYAICNwAAAAAAJuCUcgAAAAAATMAMNwAAAAAAJiBwAwAAAABgAgI3AAAAAAAmIHADAAAAAGACAjcAAAAAACYgcAMAAAAAYAICNwAAAAAAJiBwAwAAAABggv8HHDLD8o6MJBoAAAAASUVORK5CYII=\n" + }, + "metadata": {} + } + ] + }, + { + "cell_type": "code", + "source": [ + "\n", + "# Reshape the data from wide to long format\n", + "long_df = pd.melt(\n", + " df,\n", + " id_vars=['question_num', 'bloom_level', 'llm'],\n", + " value_vars=[\n", + " 'expert1_bloom', 'expert1_subject', 'expert1_persian',\n", + " 'expert2_bloom', 'expert2_subject', 'expert2_persian',\n", + " 'expert3_bloom', 'expert3_subject', 'expert3_persian',\n", + " 'expert4_bloom', 'expert4_subject', 'expert4_persian',\n", + " 'expert5_bloom', 'expert5_subject', 'expert5_persian'\n", + " ],\n", + " var_name='expert_aspect',\n", + " value_name='score'\n", + ")\n", + "\n", + "# Extract expert and aspect from 'expert_aspect' column\n", + "long_df[['expert', 'aspect']] = long_df['expert_aspect'].str.split('_', expand=True)\n", + "\n", + "# Drop the original 'expert_aspect' column\n", + "long_df.drop(columns=['expert_aspect'], inplace=True)\n", + "\n", + "# Calculate the average score for each aspect for each llm across all experts\n", + "average_scores = long_df.groupby(['llm', 'aspect'])['score'].mean().reset_index()\n", + "\n", + "# Calculate the overall average score across the three aspects for each LLM\n", + "overall_average_scores = average_scores.groupby('llm')['score'].mean().reset_index()\n", + "\n", + "# Rename columns for clarity\n", + "overall_average_scores.columns = ['llm', 'overall_average_score']\n", + "\n", + "# Display the overall average scores\n", + "print(overall_average_scores)\n", + "\n", + "# Plot the overall average scores for each LLM\n", + "plt.figure(figsize=(10, 6))\n", + "sns.barplot(\n", + " data=overall_average_scores,\n", + " x='llm',\n", + " y='overall_average_score',\n", + " ci=None\n", + ")\n", + "\n", + "# Customize the plot\n", + "plt.title('Overall Average Scores for Each LLM Across All Aspects')\n", + "plt.xlabel('LLM')\n", + "plt.ylabel('Overall Average Score')\n", + "\n", + "# Show the plot\n", + "plt.show()\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 744 + }, + "id": "QYodMxtaxHne", + "outputId": "1202c33f-7c94-4dd5-cfa1-9ca067a3d999" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + " llm overall_average_score\n", + "0 1 4.042529\n", + "1 2 4.036601\n", + "2 3 4.031667\n", + "3 4 3.352381\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":36: FutureWarning: \n", + "\n", + "The `ci` parameter is deprecated. Use `errorbar=None` for the same effect.\n", + "\n", + " sns.barplot(\n" + ] + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "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\n" + }, + "metadata": {} + } + ] + }, + { + "cell_type": "code", + "source": [ + "pip install statsmodels\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "v85_J2BH2D2S", + "outputId": "9717c4e1-54f1-4680-a857-ae7471fdd90b" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Requirement already satisfied: statsmodels in /usr/local/lib/python3.10/dist-packages (0.14.2)\n", + "Requirement already satisfied: numpy>=1.22.3 in /usr/local/lib/python3.10/dist-packages (from statsmodels) (1.26.4)\n", + "Requirement already satisfied: scipy!=1.9.2,>=1.8 in /usr/local/lib/python3.10/dist-packages (from statsmodels) (1.13.1)\n", + "Requirement already satisfied: pandas!=2.1.0,>=1.4 in /usr/local/lib/python3.10/dist-packages (from statsmodels) (2.1.4)\n", + "Requirement already satisfied: patsy>=0.5.6 in /usr/local/lib/python3.10/dist-packages (from statsmodels) (0.5.6)\n", + "Requirement already satisfied: packaging>=21.3 in /usr/local/lib/python3.10/dist-packages (from statsmodels) (24.1)\n", + "Requirement already satisfied: python-dateutil>=2.8.2 in /usr/local/lib/python3.10/dist-packages (from pandas!=2.1.0,>=1.4->statsmodels) (2.8.2)\n", + "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas!=2.1.0,>=1.4->statsmodels) (2024.1)\n", + "Requirement already satisfied: tzdata>=2022.1 in /usr/local/lib/python3.10/dist-packages (from pandas!=2.1.0,>=1.4->statsmodels) (2024.1)\n", + "Requirement already satisfied: six in /usr/local/lib/python3.10/dist-packages (from patsy>=0.5.6->statsmodels) (1.16.0)\n" + ] + } + ] + }, + { + "cell_type": "code", + "source": [ + "pip install pingouin" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "ukWHZ030N8iG", + "outputId": "7d791929-a60a-4dcc-fd09-840292c50c68" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Collecting pingouin\n", + " Downloading pingouin-0.5.4-py2.py3-none-any.whl.metadata (1.1 kB)\n", + "Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from pingouin) (1.26.4)\n", + "Requirement already satisfied: scipy in /usr/local/lib/python3.10/dist-packages (from pingouin) (1.13.1)\n", + "Requirement already satisfied: pandas>=1.5 in /usr/local/lib/python3.10/dist-packages (from pingouin) (2.1.4)\n", + "Requirement already satisfied: matplotlib in /usr/local/lib/python3.10/dist-packages (from pingouin) (3.7.1)\n", + "Requirement already satisfied: seaborn in /usr/local/lib/python3.10/dist-packages (from pingouin) (0.13.1)\n", + "Requirement already satisfied: statsmodels in /usr/local/lib/python3.10/dist-packages (from pingouin) (0.14.2)\n", + "Requirement already satisfied: scikit-learn in /usr/local/lib/python3.10/dist-packages (from pingouin) (1.3.2)\n", + "Collecting pandas-flavor (from pingouin)\n", + " Downloading pandas_flavor-0.6.0-py3-none-any.whl.metadata (6.3 kB)\n", + "Requirement already satisfied: tabulate in /usr/local/lib/python3.10/dist-packages (from pingouin) (0.9.0)\n", + "Requirement already satisfied: python-dateutil>=2.8.2 in /usr/local/lib/python3.10/dist-packages (from pandas>=1.5->pingouin) (2.8.2)\n", + "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas>=1.5->pingouin) (2024.1)\n", + "Requirement already satisfied: tzdata>=2022.1 in /usr/local/lib/python3.10/dist-packages (from pandas>=1.5->pingouin) (2024.1)\n", + "Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pingouin) (1.2.1)\n", + "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pingouin) (0.12.1)\n", + "Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pingouin) (4.53.1)\n", + "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pingouin) (1.4.5)\n", + "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pingouin) (24.1)\n", + "Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pingouin) (9.4.0)\n", + "Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->pingouin) (3.1.2)\n", + "Requirement already satisfied: xarray in /usr/local/lib/python3.10/dist-packages (from pandas-flavor->pingouin) (2024.6.0)\n", + "Requirement already satisfied: joblib>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from scikit-learn->pingouin) (1.4.2)\n", + "Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn->pingouin) (3.5.0)\n", + "Requirement already satisfied: patsy>=0.5.6 in /usr/local/lib/python3.10/dist-packages (from statsmodels->pingouin) (0.5.6)\n", + "Requirement already satisfied: six in /usr/local/lib/python3.10/dist-packages (from patsy>=0.5.6->statsmodels->pingouin) (1.16.0)\n", + "Downloading pingouin-0.5.4-py2.py3-none-any.whl (198 kB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m198.9/198.9 kB\u001b[0m \u001b[31m1.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", + "\u001b[?25hDownloading pandas_flavor-0.6.0-py3-none-any.whl (7.2 kB)\n", + "Installing collected packages: pandas-flavor, pingouin\n", + "Successfully installed pandas-flavor-0.6.0 pingouin-0.5.4\n" + ] + } + ] + }, + { + "cell_type": "code", + "source": [ + "import pandas as pd\n", + "import numpy as np\n", + "import matplotlib.pyplot as plt\n", + "import seaborn as sns\n", + "from scipy.stats import f_oneway, ttest_ind\n", + "from statsmodels.stats.multicomp import pairwise_tukeyhsd\n", + "from sklearn.metrics import cohen_kappa_score\n", + "from statsmodels.stats.inter_rater import fleiss_kappa\n" + ], + "metadata": { + "id": "Kdmk6o4DqbHT" + }, + "execution_count": null, + "outputs": [] + }, + { + "cell_type": "code", + "source": [ + "df = pd.read_csv('/content/sample_data/Dataset for analysis-csv.csv')\n", + "\n", + "for col in ['expert1_bloom', 'expert1_subject', 'expert1_persian',\n", + " 'expert2_bloom', 'expert2_subject', 'expert2_persian',\n", + " 'expert3_bloom', 'expert3_subject', 'expert3_persian',\n", + " 'expert4_bloom', 'expert4_subject', 'expert4_persian',\n", + " 'expert5_bloom', 'expert5_subject', 'expert5_persian']:\n", + " df[col] -= 3\n", + "\n", + "# Calculate the average scores for each aspect per LLM\n", + "df['bloom_mean'] = df[[f'expert{i}_bloom' for i in range(1, 6)]].mean(axis=1)\n", + "df['subject_mean'] = df[[f'expert{i}_subject' for i in range(1, 6)]].mean(axis=1)\n", + "df['persian_mean'] = df[[f'expert{i}_persian' for i in range(1, 6)]].mean(axis=1)\n" + ], + "metadata": { + "id": "oRiPCblqyESg" + }, + "execution_count": null, + "outputs": [] + }, + { + "cell_type": "code", + "source": [ + "df" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 443 + }, + "id": "Sg6Q4n9X4hs6", + "outputId": "2b4da6ff-9a17-46dd-aed1-aa0a1bec1c95" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + " question_num expert1_bloom expert1_subject expert1_persian \\\n", + "0 1 2 0 1 \n", + "1 1 2 0 1 \n", + "2 1 2 0 1 \n", + "3 1 2 0 1 \n", + "4 2 1 1 2 \n", + ".. ... ... ... ... \n", + "165 118 2 0 2 \n", + "166 119 -1 1 2 \n", + "167 120 -1 1 -2 \n", + "168 121 0 -1 -2 \n", + "169 122 -1 1 -2 \n", + "\n", + " expert2_bloom expert2_subject expert2_persian expert3_bloom \\\n", + "0 -1 -1 0 2 \n", + "1 -1 -1 0 2 \n", + "2 -1 -1 0 2 \n", + "3 -1 -1 0 2 \n", + "4 0 0 2 2 \n", + ".. ... ... ... ... \n", + "165 0 -1 2 2 \n", + "166 -1 1 1 2 \n", + "167 -2 -2 -2 2 \n", + "168 -1 -2 -2 1 \n", + "169 -1 -1 -1 0 \n", + "\n", + " expert3_subject expert3_persian ... expert4_subject expert4_persian \\\n", + "0 2 2 ... 1 1 \n", + "1 2 2 ... 1 1 \n", + "2 2 2 ... 1 1 \n", + "3 2 2 ... 1 1 \n", + "4 2 2 ... 1 1 \n", + ".. ... ... ... ... ... \n", + "165 2 2 ... 1 1 \n", + "166 2 2 ... 1 1 \n", + "167 2 1 ... 0 -1 \n", + "168 2 1 ... 1 1 \n", + "169 2 1 ... 1 1 \n", + "\n", + " expert5_bloom expert5_subject expert5_persian bloom_level llm \\\n", + "0 -1 1 0 1 1 \n", + "1 -1 1 0 1 2 \n", + "2 -1 1 0 1 3 \n", + "3 -1 1 0 1 4 \n", + "4 0 1 0 1 1 \n", + ".. ... ... ... ... ... \n", + "165 1 1 1 6 3 \n", + "166 0 1 1 6 3 \n", + "167 -1 0 -1 6 4 \n", + "168 1 1 -1 6 4 \n", + "169 -1 0 -1 6 4 \n", + "\n", + " bloom_mean subject_mean persian_mean \n", + "0 0.6 0.6 0.8 \n", + "1 0.6 0.6 0.8 \n", + "2 0.6 0.6 0.8 \n", + "3 0.6 0.6 0.8 \n", + "4 0.6 1.0 1.4 \n", + ".. ... ... ... \n", + "165 0.8 0.6 1.6 \n", + "166 -0.2 1.2 1.4 \n", + "167 -0.6 0.2 -1.0 \n", + "168 0.6 0.2 -0.6 \n", + "169 -0.2 0.6 -0.4 \n", + "\n", + "[170 rows x 21 columns]" + ], + "text/html": [ + "\n", + "
\n", + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
question_numexpert1_bloomexpert1_subjectexpert1_persianexpert2_bloomexpert2_subjectexpert2_persianexpert3_bloomexpert3_subjectexpert3_persian...expert4_subjectexpert4_persianexpert5_bloomexpert5_subjectexpert5_persianbloom_levelllmbloom_meansubject_meanpersian_mean
01201-1-10222...11-110110.60.60.8
11201-1-10222...11-110120.60.60.8
21201-1-10222...11-110130.60.60.8
31201-1-10222...11-110140.60.60.8
42112002222...11010110.61.01.4
..................................................................
1651182020-12222...11111630.80.61.6
166119-112-111222...1101163-0.21.21.4
167120-11-2-2-2-2221...0-1-10-164-0.60.2-1.0
1681210-1-2-1-2-2121...1111-1640.60.2-0.6
169122-11-2-1-1-1021...11-10-164-0.20.6-0.4
\n", + "

170 rows × 21 columns

\n", + "
\n", + "
\n", + "\n", + "
\n", + " \n", + "\n", + " \n", + "\n", + " \n", + "
\n", + "\n", + "\n", + "
\n", + " \n", + "\n", + "\n", + "\n", + " \n", + "
\n", + "
\n", + "
\n" + ], + "application/vnd.google.colaboratory.intrinsic+json": { + "type": "dataframe", + "variable_name": "df" + } + }, + "metadata": {}, + "execution_count": 4 + } + ] + }, + { + "cell_type": "code", + "source": [ + "# ANOVA test for each aspect\n", + "anova_bloom = f_oneway(*[df[df['llm'] == llm]['bloom_mean'] for llm in df['llm'].unique()])\n", + "anova_subject = f_oneway(*[df[df['llm'] == llm]['subject_mean'] for llm in df['llm'].unique()])\n", + "anova_persian = f_oneway(*[df[df['llm'] == llm]['persian_mean'] for llm in df['llm'].unique()])\n", + "\n", + "print(\"ANOVA Results for Bloom Alignment: \", anova_bloom)\n", + "print(\"ANOVA Results for Subject Suitability: \", anova_subject)\n", + "print(\"ANOVA Results for Persian Clarity: \", anova_persian)\n", + "\n", + "# Post-hoc test\n", + "for aspect in ['bloom_mean', 'subject_mean', 'persian_mean']:\n", + " tukey = pairwise_tukeyhsd(endog=df[aspect], groups=df['llm'], alpha=0.05)\n", + " print(f\"\\nPost-hoc test for {aspect}:\\n\", tukey)\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "tuyZX0qEyGWV", + "outputId": "dc5442d0-41c5-4220-bcd4-921aadf346dd" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "ANOVA Results for Bloom Alignment: F_onewayResult(statistic=6.589776827731827, pvalue=0.0003103608601602181)\n", + "ANOVA Results for Subject Suitability: F_onewayResult(statistic=7.333640807977937, pvalue=0.0001202604615788771)\n", + "ANOVA Results for Persian Clarity: F_onewayResult(statistic=56.63213624145737, pvalue=2.8998031966902094e-25)\n", + "\n", + "Post-hoc test for bloom_mean:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n", + "====================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "----------------------------------------------------\n", + " 1 2 0.103 0.7898 -0.1851 0.3912 False\n", + " 1 3 -0.0643 0.9488 -0.3728 0.2442 False\n", + " 1 4 -0.5507 0.0014 -0.933 -0.1685 True\n", + " 2 3 -0.1674 0.5199 -0.4844 0.1497 False\n", + " 2 4 -0.6538 0.0001 -1.0429 -0.2646 True\n", + " 3 4 -0.4864 0.0113 -0.8909 -0.082 True\n", + "----------------------------------------------------\n", + "\n", + "Post-hoc test for subject_mean:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n", + "====================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "----------------------------------------------------\n", + " 1 2 -0.0566 0.7191 -0.1962 0.083 False\n", + " 1 3 -0.0548 0.7768 -0.2043 0.0947 False\n", + " 1 4 -0.3305 0.0 -0.5158 -0.1453 True\n", + " 2 3 0.0018 1.0 -0.1519 0.1554 False\n", + " 2 4 -0.2739 0.0013 -0.4625 -0.0854 True\n", + " 3 4 -0.2757 0.002 -0.4717 -0.0797 True\n", + "----------------------------------------------------\n", + "\n", + "Post-hoc test for persian_mean:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n", + "====================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "----------------------------------------------------\n", + " 1 2 -0.0642 0.8325 -0.2613 0.1329 False\n", + " 1 3 0.0866 0.7116 -0.1245 0.2976 False\n", + " 1 4 -1.1892 0.0 -1.4507 -0.9277 True\n", + " 2 3 0.1508 0.2749 -0.0661 0.3676 False\n", + " 2 4 -1.1249 0.0 -1.3911 -0.8587 True\n", + " 3 4 -1.2757 0.0 -1.5524 -0.999 True\n", + "----------------------------------------------------\n" + ] + } + ] + }, + { + "cell_type": "code", + "source": [ + "# swarmplots\n", + "\n", + "# Define the y-axis limits\n", + "y_limits = (-2, 2)\n", + "\n", + "# Create subplots\n", + "fig, axes = plt.subplots(1, 3, figsize=(18, 6))\n", + "\n", + "# Define a color palette\n", + "palette = sns.color_palette(\"tab10\")\n", + "\n", + "# Plot each boxplot with a palette\n", + "sns.boxplot(x='llm', y='bloom_mean', data=df, ax=axes[0], palette=palette)\n", + "sns.boxplot(x='llm', y='subject_mean', data=df, ax=axes[1], palette=palette)\n", + "sns.boxplot(x='llm', y='persian_mean', data=df, ax=axes[2], palette=palette)\n", + "\n", + "# Add grid lines\n", + "for ax in axes:\n", + " ax.grid(True, linestyle='--', alpha=0.7)\n", + "\n", + "\n", + "# Set the same y-axis limits for all plots\n", + "for ax in axes:\n", + " ax.set_ylim(y_limits)\n", + " ax.set_yticks([-2, -1, 0, 1, 2])\n", + " ax.set_yticklabels([-2, -1, 0, 1, 2])\n", + "\n", + "axes[0].set_ylabel('Bloom Alignment Score', fontsize=12)\n", + "axes[1].set_ylabel('Subject Suitability Score', fontsize=12)\n", + "axes[2].set_ylabel('Persian Clarity Score', fontsize=12)\n", + "\n", + "\n", + "# Set titles with larger font size\n", + "axes[0].set_title('Bloom Alignment by LLMs', fontsize=14)\n", + "axes[1].set_title('Subject Suitability by LLMs', fontsize=14)\n", + "axes[2].set_title('Persian Clarity by LLMs', fontsize=14)\n", + "\n", + "# Set x-axis labels with manual values and rotate labels if needed\n", + "llm_labels = ['BloomGPT', 'Code Tutor', 'Copilot', 'LLaMa']\n", + "for ax in axes:\n", + " ax.set_xticks(range(len(llm_labels)))\n", + " ax.set_xticklabels(llm_labels, rotation=45, ha='right')\n", + "\n", + "# Adjust layout to prevent clipping\n", + "plt.tight_layout()\n", + "plt.show()" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 705 + }, + "id": "1-QEPCgcyIKw", + "outputId": "d2bad048-af7b-4a1f-b569-6a05559a26bb" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":13: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='llm', y='bloom_mean', data=df, ax=axes[0], palette=palette)\n", + ":13: UserWarning: The palette list has more values (10) than needed (4), which may not be intended.\n", + " sns.boxplot(x='llm', y='bloom_mean', data=df, ax=axes[0], palette=palette)\n", + ":14: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='llm', y='subject_mean', data=df, ax=axes[1], palette=palette)\n", + ":14: UserWarning: The palette list has more values (10) than needed (4), which may not be intended.\n", + " sns.boxplot(x='llm', y='subject_mean', data=df, ax=axes[1], palette=palette)\n", + ":15: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='llm', y='persian_mean', data=df, ax=axes[2], palette=palette)\n", + ":15: UserWarning: The palette list has more values (10) than needed (4), which may not be intended.\n", + " sns.boxplot(x='llm', y='persian_mean', data=df, ax=axes[2], palette=palette)\n" + ] + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "iVBORw0KGgoAAAANSUhEUgAABv4AAAJOCAYAAAB/dnBOAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAADu+ElEQVR4nOzdd3gU5f7+8Xt2QwolCQQIIBB6VZoUUaRZAEEOAirSQakqdr82qhzsWCkK0gQsiKKcY1dALKgoyEGK0kLvJkFayO78/uC3CzspJMtmZ7N5v66La8ns7M5nsvc882SenRnDNE1TAAAAAAAAAAAAAAo0h90FAAAAAAAAAAAAALh4DPwBAAAAAAAAAAAAYYCBPwAAAAAAAAAAACAMMPAHAAAAAAAAAAAAhAEG/gAAAAAAAAAAAIAwwMAfAAAAAAAAAAAAEAYY+AMAAAAAAAAAAADCAAN/AAAAAAAAAAAAQBhg4A8AAAAAAAAAAAAIAwz8AQHUtm1bGYZhdxkhI6vfx/Lly2UYhsaNG2dPUWGM3y0AIJwNHDhQhmFox44duZp/x44dMgxDAwcOzNe6CoI5c+bIMAzNmTMn35Yxbtw4GYah5cuX5/o1hmGobdu2uXqfrOYNBvpXAAAEnl379UDyp+/jD/pAAPzBwB+QBc+BIuu/YsWKqUGDBho/frz++ecfu8u0zeDBg2UYhhISEnT69Gm7ywk7BfFApWEYqlOnzgXn86xbx44dLzivp5NpGIaaNm2a7Xyffvqpd76C/ocDABRkx48f16RJk9SkSRMVL15cUVFRqlixoq6++mo9+uij2rp1q90l5jt/90UZGRl67bXX1LJlS8XFxSkyMlLly5dXixYtdN9992nNmjWBL1YF/6BbXgeDCxr6VwCAvMrqeFZkZKQqVaqk3r17a926dXaXGNJ+/fVX3X777apZs6aKFSummJgYVa9eXf369dOXX35pd3le9IHOog8EZC/C7gKAUFa9enX17dtXkmSapg4dOqRPP/1U48aN02effabvvvtOTqfT5iqD69ixY3rvvfdkGIaOHj2qJUuW6NZbb83165s3b66NGzeqdOnS+VglwklERIR+/fVXrVu3Tg0aNMj0/JtvvqmIiAhlZGTYUB0AQDrbP2jVqpXWrVunGjVqqG/fvkpISNDhw4f1888/6+mnn1b16tVVvXr1oNV0ySWXaOPGjYqLiwvaMv3hcrnUqVMnffXVV6pQoYJuvvlmJSYmKiUlRb/99pteeeUVFStWTI0bN/Z7GTfddJOuuOIKlS9fPoCVX7yNGzeqaNGiAZ8XF0b/CgDC2/nHs/755x+tWrVKb7/9tj744AN9/fXXuuqqq2ytL9T26263Ww8++KBefPFFRUREqH379uratauKFCmibdu26b///a/mz5+vCRMmaPTo0UGtLdR+VwUdfSAUFgz8ATmoUaNGplPaT58+rZYtW2rVqlVasWKF2rdvb09xNnn33Xd1/Phx3X///XrppZf05ptv5mngr2jRorn61g7g0aFDB3366aeaNWuWXnrpJZ/nDh8+rKVLl+qGG27Qxx9/bE+BAAC99NJLWrdune644w698cYbmS71vX379qBfJaBIkSIFos+xcOFCffXVV+rYsaM+/vhjFSlSxOf5/fv3a+/evRe1jLi4uJAcAM3L51MQPsuChP4VAIS3rI5nPfHEE/r3v/+txx9/PN8vT3khobZff+KJJ/Tiiy+qUaNGev/99zN9We3kyZN67bXXdOTIkaDXFmq/q4KOPhAKCy71CeRRVFSU2rVrJ+nsDiE3MjIyNHnyZDVs2FAxMTGKi4tTu3bttHTp0oue//x7tixdulQtWrRQ0aJFdckll2j06NFyu92SpLlz53rfr3Llynruuef8Wn/PN18efvhhtWvXTl9//bWSk5Nz/fqcrhG+YsUKtW7dWsWKFVNCQoJuvfVW7dq1K8t7BZ5/LfWFCxeqUaNGiomJUfny5XXPPffo5MmT2S73hx9+ULt27VSiRAmVKVNGI0eO9M7/3//+Vy1btlSxYsWUmJiohx9+ONtv+Xz00Ue65pprVLJkSUVHR+vSSy/V888/L5fL5TPf+Z/RF198oSuvvFJFixZVQkKCBgwY4NNxnDNnjqpWrSrp7Gd2/uU58tIx/+6779S2bVuVKFFC8fHx6tGjh7Zs2eJ93u12KykpKcfLtbZu3VoRERHavXt3rpebHypWrKjrrrtOCxYsUHp6us9z8+fPV3p6ugYPHpzla1NTUzVmzBjVq1dPxYsXV2xsrGrUqKEBAwbkKbcAgJz9+OOPkqQ777wzy/sdV61aNdNBi5wun1OlShVVqVIly+fcbreeffZZ1axZU9HR0apataomTJigM2fO+MyX06Wzjx07prFjx6p+/fqKiYlRfHy8OnTooO+++y7LZR47dkzjx49XgwYNVLRoUcXFxalx48YaPXq0zpw54+1nSGf7M+fvvy90Xz3P727YsGGZBv0kqVy5cmrSpEmufz9Z9Zus9/jLTb2pqal65pln1KZNG1WoUEGRkZGqUKGC+vfvf8HLtr755pu67LLLFB0drUsuuUT33Xefjh07lmm+vFxCyTpvlSpVNHfuXEln83X+JZlSU1NVrFgx1a9fP8v3crvdqlKlikqWLJmpz5gT+ldn0b8CgILp7rvvliT98ssvPtP9ObaxdOlSXXXVVSpRooRPn2Tx4sVq06aNypYtq+joaFWoUEHXXnutFi9e7PNeWfUB/vzzTz388MNq0qSJEhISFB0drVq1aumRRx7J8nY7nj7PmTNnNG7cOFWpUkVRUVGqVauWpk6dmuvfy5YtW/Tss88qISFBn332WZZXqIiJidFDDz2k8ePHX/D9Zs2apX/961+qUqWKoqOjVapUKXXo0EHLli3LNK/1WNX111+v+Ph4n74cfaDAog+EwoKBPyCP0tPTvTvmRo0aXXB+0zTVs2dPPfDAAzp16pTuvPNO9e7dW7///ru6du2qF1988aLm9/jwww91yy23qFq1aho+fLiKFy+uiRMnasyYMXruuec0atQoNWzYUEOHDpXb7dbDDz+sefPm5WndN2zYoFWrVun6669XYmKi+vfvL7fbrdmzZ+fpfbLyxRdf6Nprr9XPP/+snj17aujQoUpOTlarVq2UkpKS7etee+01DR06VPXr19eIESNUsmRJvfLKK7rjjjuynP+nn37SNddco7i4OA0bNkyVK1fWtGnTNGTIEL377rvq2bOnkpKSNGzYMMXHx+u5557TpEmTMr3Po48+qm7dumnz5s3q3r27Ro4c6e0I9urVK8tlf/zxx7rxxhtVoUIFjRw5UtWrV9e8efP0r3/9yztPo0aNdM8990iSGjZsqLFjx3r/ZXeAz2rVqlXedbz77rvVpk0bffjhh7ryyiu1bds2SZLD4dAdd9yho0ePZuqAS9LmzZu1cuVKdezYURUrVszVcvPT4MGDvd+8Ot+sWbNUv359tWjRItNrTNNUhw4d9OSTT6pUqVIaOnSohg4dqsaNG+vjjz/WX3/9FazyASDsJSQkSDp7wCa/3XvvvXrmmWd07bXX6u6771ZUVJTGjh2r2267LVevP3r0qFq2bKkJEyaoZMmSGj58uHr06KFff/1V7dq105IlS3zmP3jwoJo3b65x48bJ6XRqxIgRGjx4sMqVK6dnnnlGx48fV5UqVTR27FhJUlJSks/++0L9xWD+7jxyU+/GjRs1ZswYxcTE6KabbtK9996rpk2bauHChWrevHm2BzcmT56sUaNGqVmzZrr33ntVvnx5vfTSS7r++uszDc5ejHvvvVcNGzaUJN1zzz3e+gcOHKi4uDj16tVLGzZs0A8//JDptV9++aWSk5PVp08fxcTE5Gp59K/Oon8FAAXf+YNK/hzbWLRokbp3766yZctq5MiR6tSpkyRp2rRp6tmzp/766y/ddNNNuv/++9WxY0ft379fH3744QXr+uCDD/Tmm2+qWrVqGjBggIYPH65SpUrpmWee0XXXXZdtP+K2227TrFmz1KFDB91+++06evSo7rzzTs2YMSNXv485c+bI5XJp2LBhSkxMzHHeqKioC77fnXfeqQMHDujaa6/Vfffdpy5duujHH3/Utddeq48++ijL1/zwww/egcyhQ4fmeGUt+kAXjz4QCgUTQCbbt283JZnVq1c3x44da44dO9YcM2aMOXLkSLN69epmdHS0+dxzz2V6XZs2bUzrZjV37lxTktmmTRvz9OnT3unJyclm6dKlzYiICHPr1q1+zz979mxTklmkSBHz559/9k5PS0szy5YtaxYtWtQsV66cz2t27txpRkZGmpdddlmefi/333+/Kcl8++23TdM0zWPHjpnFihUzK1eubLpcrlz9PpYtW2ZKMseOHeudlpGRYSYlJZmGYZgrV670mb9///6mpEzvM3bsWFOSGRcXZ27atMk7/cSJE2atWrVMh8Nh7tmzJ9NyJZlLlizxTk9PTzcbNGhgGoZhli5dOsvfYalSpcz09HTv9C+++MKUZHbo0MH8559/vNPdbrc5fPhwU5L5/vvve6d7PqOIiAjzu+++81nvtm3bmpLMH3/80Tvdk78BAwZk+p3m5Px1nD59us9z06dPNyWZXbp08U7bs2ePGRERYbZt2zbTez344IOZflc5kWTWrl37gvN51q1Dhw4XnNezPsOGDTNPnz5tJiQkmDfccIP3+Z9//tmUZL7wwgvmvn37vNuNx7p160xJZrdu3TK996lTp8xjx47lat0AABf20UcfmZLMEiVKmA888ID5+eefm4cPH87xNdZ2+3xJSUlmUlKSz7QBAwaYkswyZcqYu3bt8k4/ffq02bp160z73+z2p7179zYlmTNmzPCZfuDAAbNSpUpmmTJlzJMnT3qn9+jRw5RkPvbYY5nq3L9/v3nmzJlcrVN2fv31VzMiIsKMjIw0hw0bZn788cfm3r17c3xNVr8fj6z6X56+yOzZs32m51RvSkqKeeTIkUzTv/nmG9PhcJh33HGHz3RP3ywyMtL8/fffvdPdbrf3d/78889fcPme91m2bNkF5/VkYvv27Znq/Omnn0xJ5sCBAzM917NnT1OSuXbt2izW3Bf9K/pXAFCQ5LRPGDNmjCnJbNeunWma/h/bcDgc5pdffpnp/Zs0aWJGRkaaBw4cyPSctV+Y1X599+7dPsfBPMaPH29KMufPn+8z3dPnadGihZmamuqdvmnTJjMiIiJX+1DTNL3HZb766qtcze+RXZ9l27Ztmebdu3evWaFCBbNmzZo+08/vZ8yaNSvL5dAHyh59ICB7nPEH5GDr1q0aP368xo8frwkTJmjq1KnaunWrrr32Wl177bW5eg/P6ffPPvusIiMjvdMrV66s++67TxkZGVqwYIHf83v07dtXzZo18/5cokQJdenSRSdOnNCIESNUrVo173OVKlVSq1attGHDhlzfrPbMmTN66623FBsbq27dukmSihcvrptuukk7d+7UV199lav3ycp3332n5ORk3XjjjWrVqpXPcxMnTpTT6cz2tffcc49q167t/TkmJka33Xab3G63fv3110zzt2vXzucMuyJFiqhnz54yTVM33nhjlr/Do0eP+lyK4LXXXpMkvfHGGypWrJh3umEYevrpp2UYht5+++1My+7du7fPDbSdTqcGDBggKfOlNi5GrVq1NGTIEJ9pQ4YMUc2aNfXf//5Xhw4dkiRVqFBBN954o1asWOFziYYzZ85o3rx5Kl++vDp37hywui5GZGSk+vTpo88//9x7n6NZs2apSJEi6tevX46vzepbbFFRUSpevHi+1AoAhVHXrl31wgsvyDRNvfDCC+rQoYNKly6tGjVq6K677groN2Dvuecen28KR0ZG6t///rckXfCymocPH9a7776r9u3bZ7o6QNmyZfXQQw/p0KFD3n7N/v379cEHH6h69epZXqY8MTFREREXd9v0Jk2aaO7cuYqNjdXrr7+url27qkKFCqpUqZIGDRqUZX8mGOLi4lSqVKlM09u1a6f69etn2/fr37+/GjRo4P3ZMAxNmjRJTqfzgp9PIDVv3lyNGzfWokWLlJaW5p1+6NAhffzxx2rWrJn32/K5Qf/KF/0rAAhtW7Zs0bhx4zRu3Dg99NBDat26tSZMmKDo6Ghvv8nfYxv/+te/sj0mVqRIkSwvXe65wkFOLrnkEp/jYB533XWXJGXb93jqqacUGxvr/bl27dq66qqrtHnz5iwvNW61f/9+SQrYmWie27ecr3z58urRo4f++uuvLK+a0KRJEw0aNCggy6cPdGH0gVAYMPAH5KBDhw4yTdP77/Dhw/roo4+0fv16XXXVVfrpp58u+B5r1qxR0aJF1bx580zPee4VuHbtWr/n98jqMlLly5fP8TmXy6UDBw5ccB2ks9d8P3TokG6++WZFR0d7p/fv31/S2Xu5+Ov333+XpEyDftLZQcrKlStn+9rLL7880zRPZy2rS4T683uS5O0ISGcvc1CsWDHNmjXL25H2/HvxxRcVExOjTZs2XXSt/rrqqqvkcPg27w6HQ1dddZVM0/T+vqWz9xMyTVMzZ870Tvv444918OBBDRo06KIPZgbS4MGD5XK5NHfuXJ06dUrvvPOOunTpojJlymQ5f926ddWgQQO9/fbbat26tSZPnqzffvvNe99LAEBg3X///dq7d6/ee+893XvvvWrVqpV27typKVOmqEGDBvr4448Dspyrr74607SWLVsqIiJCa9asyfG1v/zyi1wul06fPp1pHz5u3DitWrVKkrz78dWrV8s0TbVr1y7Lg1iB0rt3b+3atUsff/yxHn74YV1zzTU6cuSI5syZo+bNm2v69On5tuycLF++XN26dVP58uVVpEgR7z1k/ve///n0jc6X1eeTlJSkSpUq6Y8//sh0L5X8NGzYMB0/flwLFy70Tps3b57S09MzHcC6EPpXZ9G/AoCC4fwvsr/88svavn27evfurZ9//lktW7aU5P+xjayOV0lSr169dPz4cV166aV66KGH9Mknn/gMPF2IaZqaNWuWWrdurVKlSsnpdMowDO+gYXZ9j2Ada8mtbdu2aciQIapevbqio6O9/adXX31VUtbrcf6X0AOBPtCF0QdCuAudrQ0oABISEtS1a1cVLVpU1113nZ544gl9+eWXOb4mLS1NlSpVyvI5z6DS+R2hvM7vcf63mzw8O9ScnsvtvVY8A3uegT6Pa665Rpdccok++ugjHT16NMtvhl+IZ33Kli2b5fOJiYnavn17ls/ltG7WG1FfaP7c/p6OHj2qjIyMHG/qfPz48Yuu1V/ZXZPeMz01NdU77frrr1fVqlU1d+5cTZw4UREREZo5c6YMw9Dtt98esJoCoWHDhmrSpIlmz56typUrKyUlJdsbLktnf7fffPONxo0bp8WLF+uBBx6QJJUpU0Z33XWXHn/88RzPJgUA5F2JEiV088036+abb5Z0dp/z2GOPaerUqbr99tu1Z8+eLL/JnRdZ7eecTqcSEhJ89nFZOXr0qCTp+++/1/fff5/tfJ79uOf9LrnkEn/LzbXo6GjdeOONuvHGGyVJp06d0vPPP6/Ro0frnnvuUbdu3VSuXLl8r8Nj0aJFuvXWW1W8eHF16NBBVapUUdGiRWUYhubMmZPtPf5y6ofs2LFDx44dy9W3/gOhd+/eevDBBzVz5kwNHz5c0tk+bfHixXN9T0gP+ldn0b8CgIKhQ4cO+uyzz3Kcx99jG9ntEx988EElJCRo2rRpeuGFF/T8888rIiJCnTt31osvvpjlmXDnGzVqlF577TVVqlRJXbt2Vfny5b331Bs/frxOnz6d5esu9lhLuXLltGnTJu3Zs8fnilL+2LJli5o3b660tDS1a9dON954o2JjY+VwOLR8+XKtWLEiy/W40L0F84o+0IXRB0K444w/wA+em7zm5vKMsbGxOnjwYJbPeS4ncH4nJa/zB8OuXbv0xRdfSJLatGnj/baSYRhyOp3as2ePTp8+rfnz5/v1/p71yW69c3tWYrDExsYqISHB52xQ67/sBiqDIbvfl2d6XFycd5rnxtH79+/X0qVLvZ/1Nddc43N52FBx++2366+//tL//d//qUKFCt6biGcnISFBr776qvbs2aMNGzbotddeU6lSpTR27Fg9++yzQaoaAAqvuLg4vfbaa0pKStLhw4f1v//9z/ucYRjZXnI8pwG8rPZzLpdLR44c8dnHZcXT53jggQdy3I+PHTtWkhQfHy9J2rNnT47vmx+io6P1xBNPqHXr1kpPT/cZqHQ4HH797vJi3Lhxio6O1q+//qpFixbpueee0/jx473Ts5NTP8QwDJUoUSIg9eVGiRIl1KdPH/36669au3atvv/+e23cuFG9evXK8+WY6F+dQ/8KAMKDv8c2DMPI8v0Mw9DgwYP1yy+/6NChQ/rwww/VvXt3ffTRR+rSpUuOg3AHDx70XiVi06ZNmjNnjp566imNGzfOO3CVXzy3ZPn6668v+r1efPFF/f3335ozZ46+/PJLvfTSS5owYYLGjRunOnXqZPu67H6n/qIPlDv0gRDOGPgD/PD3339LUq5O527cuLFOnDihn3/+OdNzy5cvl+R7icm8zh8Mc+bMkdvtVqtWrXT77bdn+ue5T52/l/v0XFs8q2/e7969Wzt37vS/+HzQokULHTlyJKD3Kzqf5xtC/p4F+P3332fKptvt1g8//CDDMDJdy33QoEEqUqSIZs6cqVmzZsntduf50g/B0rt3b0VHR2vPnj3q379/rr9NZRiG6tatqzvvvNN7lm6gLjkHAMiZYRg+943xKFmyZJYDajt27MjxskwrV67MNO3HH39URkaGGjdunGMtzZo1k2EY+vHHHy9cuKSmTZvK4XBo2bJlubpKgsPhCOhZ/JKyPDhTsmRJHTx4MNPg3/Hjx/PUP8mp3q1bt6pu3bqqWbOmz/R9+/Zp27Zt2b5nVp9PcnKydu3apfr161/0GZ/ny02fadiwYZKkGTNmeC875U8/h/5VZvSvAKBgy89jGwkJCerWrZv33sobNmzwue+b1bZt22Sapq699loVLVrU57ms+haBNHDgQDmdTr3xxhve+9VlJ7uzDj22bt0q6ex9EM9nmmaOV5vIK/pAgUEfCOGMgT/AD5MnT5YktW7d+oLzegbFHn30UZ8DRrt27dLkyZMVERGhPn36+D1/fjNNU7Nnz5ZhGJo7d65mzpyZ6d+cOXPUsmVLrVu3TqtXr87zMlq1aqXKlStr6dKlmQ7EjR49OuAH0C7WqFGjJJ29HviRI0cyPb9//35t3LjR7/cvWbKkDMPQrl27/Hr9n3/+qRkzZvhMmzFjhv7880917tw50/XKExMT1a1bN3322WeaNm2aSpcurW7duvlbfr6Kj4/X559/rg8//FD33XdfjvPu2LFDO3bsyDTd8620nM5WAADkzeuvv57tlRCWLFmijRs3Kj4+Xpdeeql3erNmzbRjxw6tWLHCOy09PV33339/jst6+eWXtXv3bp/XPP7445LOHrjJSbly5XTLLbfohx9+0HPPPSfTNDPN89NPP+nEiROSzu4je/To4b1PjpV18K1UqVI+teXGO++8o2+++SbLWlatWqVly5YpIiJCV1xxhXd6s2bNdObMGS1YsMA7zTRNPfroo1lekis7OdWblJSkLVu2+HzL+9SpUxoxYkSOg6Dz5s3TunXrfOp67LHH5HK5Lvj55JXnEvM59ZkaN26sZs2aacGCBVq0aJEaNGiQ7b2JckL/6iz6VwAQPgJ9bGP58uWZ+jNnzpzxXmo9p31EUlKSJOmHH37wGWTavXu3Hn300VzX4I8aNWro4Ycf1uHDh9WpU6csz3I8deqUJk+erHHjxuX4Xp71+O6773ymP/3001q/fn3AaqYPFBj0gRDOuMcfkIMtW7b47NSPHj2q77//Xr/99ptKliypZ5555oLv0a9fP33wwQf66KOP1KBBA3Xp0kXHjx/Xu+++q6NHj+qFF17wOd09r/Pnt2+++Ubbt29XmzZtclzuoEGD9OOPP+rNN99U06ZN87QMp9Op6dOnq2vXrmrfvr1uvfVWlS9fXitWrNCePXvUsGFDnwNIduvYsaNGjx6tJ598UjVq1FDHjh2VlJSkI0eOaMuWLVq5cqUmTpyounXr+vX+xYsXV7NmzfTtt9+qX79+qlmzphwOh/r16+ftROakQ4cOGjVqlD755BPVr19ff/zxh5YuXarSpUvr5ZdfzvI1w4cP16JFi3TgwAE98MADfn0bf9++fdke0CtdurSef/5578//+9//sp23Tp06euSRR7JdTm4G3CVp7dq16t69u5o3b6569eqpXLly2rNnj5YsWSKHw3HBTh0AIPc+/fRTDR8+XDVq1NBVV12lChUq6Pjx41qzZo1Wrlwph8OhqVOneu/TIkn333+/vvjiC91www267bbbVLRoUX355ZeKj4/33tc4K1dccYUaNmyoW2+9VcWKFdPSpUu1efNmde/eXT169LhgrVOnTtXmzZv18MMP66233lLLli0VHx+vXbt2afXq1frrr7+0b98+77fNp06dqvXr1+vf//63PvnkE7Vv316maerPP//UF198oQMHDngvCdq+fXu999576tatmxo3biyn06muXbuqQYMG2dazatUqvfzyy7rkkkvUunVrVa5cWenp6dq4caO++OILud1uPf300z73Gbzrrrs0e/Zs3XHHHfryyy9VpkwZrVy5UikpKWrYsKF+//33C/4eLlTv3XffrbvvvluNGzdWz549lZGRoS+//FKmaea4jA4dOqhly5bq1auXypQpo6+//lqrV6/WFVdcobvvvjtXdeVW+/bt9fzzz2vo0KHq0aOHihUrpqSkJPXr189nvuHDh3vvK+PvN87pX51F/woAwkegj21069ZNsbGxuuKKK5SUlKQzZ87oyy+/1IYNG9SzZ88cj2eUL19ePXr00OLFi9W0aVNdc801OnDggP7zn//ommuu8Z5Jl18mTpyoU6dO6cUXX1Tt2rXVvn17XXrppSpSpIi2b9+ur776SkeOHNHEiRNzfJ/hw4dr9uzZ6tGjh2655RYlJCRo1apV+u2339S5c2f997//DUi99IHoAwEXZALIZPv27aakTP+ioqLM6tWrmyNGjDCTk5Mzva5NmzZmVpvVmTNnzOeff9687LLLzKioKLNEiRJmmzZtzI8++ijL5edl/tmzZ5uSzNmzZ2d6buzYsaYkc9myZZmeGzBggCnJ3L59e46/i9tuuy3b9z9famqqGRMTY8bFxZknTpwwTTPr38eyZctMSebYsWMzvcc333xjtmrVyoyJiTFLlSpl3nzzzebOnTvNSy+91IyLi8v1umX1O8lpuf7+Dr/88kvzxhtvNMuUKWMWKVLELFeunNmyZUvzySefNHfu3Jmr98+urs2bN5s33HCDGR8fbxqGkW0N2b3XypUrzTZt2pjFihUzY2NjzZtuusn866+/sn2t2+02K1eubEoyN27cmONyspLV9nL+v6SkJNM0s9+2zv/Xpk0bn/UZNmzYBZe/b98+n9eapmnu2rXLfOSRR8wrrrjCLFu2rBkZGWlWrlzZ7N69u/njjz/meR0BANnbtGmT+eyzz5rXXXedWbVqVTM6OtqMjo42q1evbg4YMMBcvXp1lq9btGiRedlll5mRkZFmuXLlzLvvvts8duyYmZSU5N13eHj6Llu3bjWffvpps0aNGmZkZKSZlJRkjhs3zjx9+rTP/J59zoABAzIt98SJE+azzz5rXn755WaxYsXMmJgYs2rVqma3bt3MefPmmWfOnPGZPzU11Rw9erRZp04dMyoqyoyLizMbNWpkjhkzxkxPT/fOt2/fPvOWW24xS5cubTocjlz1oXbu3Gm++uqr5o033mjWqFHDLFasmHefdfPNN5tff/11lq/75ptvzBYtWphRUVFmQkKC2a9fP/PAgQNZ9r+y64vkVK/b7TanT59u1q9f34yOjjbLlStn3n777ebBgwezXMb5faYZM2aY9evXN6Oioszy5cub99xzj5mWlpZpHaz7buv7XGhe0zTNZ5991qxZs6ZZpEiRbOc5fvy4GRUVZcbExJh///13Vr/ObNG/on8FAAWJZ5/QoUOHXL8mEMc2TNM0p06danbt2tVMSkoyo6OjzYSEBLN58+bmtGnTfPpLppn1fv3YsWPmAw88YFapUsWMiooya9asaT755JNmenp6lvNndwzONHN/zMvql19+MQcPHmzWqFHDjImJMaOioswqVaqYvXv3Nr/88kufebPrsyxbtsy86qqrzBIlSpjx8fHmDTfcYP76669Zzp/TsSoP+kD0gQB/GKaZxTVlACBEHDt2TImJibrsssv0008/2V1O2Nq3b58qV66sli1b6ttvv7W7HAAALtqmTZtUt25dDR06VK+//rrd5cBGq1evVrNmzdSvXz/NmzcvaMulfwUAAOxEHwgovLjHH4CQcPz4cR07dsxnmsvl0kMPPaSTJ0+G7PXAw8VLL72kjIwMjRgxwu5SAAAIiC1btkiSKlasaHMlsNtzzz0nSUHv59C/AgAAdqIPBBRenPEHICSsXbtWrVq1UocOHVStWjUdO3ZMK1eu1IYNG1S/fn399NNPKlasmN1lhpXU1FRNmzZNycnJmjlzpmrVqqV169bJ6XTaXRoAAH77888/NWvWLC1cuFB79uzRmjVrcrzHHsLTzp07tXDhQv3xxx+aP3++OnTooM8++yzfl0v/CgAA2Ik+EACJgT8AIeLQoUN6+OGHtWLFCh04cEAZGRmqXLmyunXrpscff1zx8fF2lxh2duzYoapVqyo6OlpXXHGFpk+frtq1a9tdFgAAF2XJkiW67bbbVKtWLY0fP56rBhRSy5cvV7t27VS8eHG1a9dOb7zxhsqVK5fvy6V/BQAA7EQfCIBUAAb+nnrqKX3wwQfatGmTYmJidOWVV+qZZ56h4QAAAMgF+lIAAAD+oy8FAAAKmpC/x9+KFSt05513atWqVfryyy915swZXX/99Tp+/LjdpQEAAIQ8+lIAAAD+oy8FAAAKmpA/48/q0KFDKlu2rFasWKHWrVvbXQ4AAECBQl8KAADAf/SlAABAqIuwu4C8Sk1NlSSVKlUqy+dPnz6t06dPe392u906evSoEhISZBhGUGoEAADhxzRNHTt2TBUqVJDDEfIXTcgWfSkAAGAH+lL0pQAAgP/y0pcqUGf8ud1ude3aVSkpKfruu++ynGfcuHEaP358kCsDAACFxa5du1SxYkW7y/ALfSkAAGA3+lIAAAD+y01fqkAN/I0YMUKffvqpvvvuu2xXzPrNqtTUVFWuXFnbt29XbGysJMnhcMjhcMjtdsvtdnvn9Ux3uVw6/9eS3XSn0ynDMJSRkeFTg9PplCS5XK5cTY+IiJBpmj7TDcOQ0+nMVGN203OzThkZGfr999/VsGFDFSlSJCzWKRw/p2CukyT99ttvatiwoXeegr5O4fg5BXOdXC6Xfv/9dzVq1EiRkZFhsU4Xms465bxOZ86c0Zo1a7ztRDisk7+fU1pamqpWraqUlBTFxcWpIKIvRV/Kuk7h+DnRl+Jzoi/F5xRK60Rfir4UfSn6UqwTfSnWib5UTtNZJ/pS+dGXKjADf3fddZc++ugjffvtt6patWquX5eWlqa4uDilpqZ6O1iFlSdQnpACZAJWZAJWZOKcgt6noC918dgeYEUmYEUmYEUmzinofQr6UheP7QFWZAJWZAJWZOKcvPQpQv6i6qZp6q677tKHH36ob775Jk+dK2SWnp5udwkIMWQCVmQCVmSiYKMvFVhsD7AiE7AiE7AiEwUbfanAYnuAFZmAFZmAFZnIu5Af+Lvzzjs1f/58LVy4UCVKlND+/fu1f/9+nTx50u7SChyXy6V169ZleVo9CicyASsyASsyUfDRlwoctgdYkQlYkQlYkYmCj75U4LA9wIpMwIpMwIpM+CfkB/6mTZum1NRUtW3bVuXLl/f+e/fdd+0uDQAAIOTRlwIAAPAffSkAAFDQRNhdwIUUkFsQAgAAhCT6UgAAAP6jLwUAAAqakD/jD4HldDrtLgEhhkzAikzAikwA57A9wIpMwIpMwIpMAOewPcCKTMCKTMCKTOSdYYb5V5fS0tIUFxen1NRUxcbG2l0OAAAooAprn6KwrjcAAAiswtqnKKzrDQAAAisvfQrO+CtETNNUSkoKl6mAF5mAFZmAFZkAzmF7gBWZgBWZgBWZAM5he4AVmYAVmYAVmfAPA3+FiMvl0qZNm+RyuewuBSGCTMCKTMCKTADnsD3AikzAikzAikwA57A9wIpMwIpMwIpM+IeBPwAAAAAAAAAAACAMMPAHAAAAAAAAAAAAhAEG/goRwzAUExMjwzDsLgUhgkzAikzAikwA57A9wIpMwIpMwIpMAOewPcCKTMCKTMCKTPjHMMP8rohpaWmKi4tTamqqYmNj7S4HAAAUUIW1T1FY1xsAAARWYe1TFNb1BgAAgZWXPgVn/BUibrdbBw8elNvttrsUhAgyASsyASsyAZzD9gArMgErMgErMgGcw/YAKzIBKzIBKzLhHwb+ChG3261t27axkcCLTMCKTMCKTADnsD3AikzAikzAikwA57A9wIpMwIpMwIpM+IeBPwAAAAAAAAAAACAMMPAHAAAAAAAAAAAAhAEG/goRwzAUFxcnwzDsLgUhgkzAikzAikwA57A9wIpMwIpMwIpMAOewPcCKTMCKTMCKTPjHME3TtLuI/JSWlqa4uDilpqYqNjbW7nIAAEABVVj7FIV1vQEAQGAV1j5FYV1vAAAQWHnpU3DGXyHidru1e/duboQJLzIBKzIBKzIBnMP2ACsyASsyASsyAZzD9gArMgErMgErMuEfBv4KETYSWJEJWJEJWJEJ4By2B1iRCViRCViRCeActgdYkQlYkQlYkQn/MPAHAAAAAAAAAAAAhAEG/gAAAAAAAAAAAIAwwMBfIeJwOFSmTBk5HHzsOItMwIpMwIpMAOewPcCKTMCKTMCKTADnsD3AikzAikzAikz4xzBN07S7iPyUlpamuLg4paamKjY21u5yAABAAVVY+xSFdb0BAEBgFdY+RWFdbwAAEFh56VMwTFqIuN1ubd26lRthwotMwIpMwIpMAOewPcCKTMCKTMCKTADnsD3AikzAikzAikz4h4G/QsTtduvQoUNsJPAiE7AiE7AiE8A5bA+wIhOwIhOwIhPAOWwPsCITsCITsCIT/mHgDwAAAAAAAAAAAAgDDPwBAAAAAAAAAAAAYYCBv0LE4XCoYsWKcjj42HEWmYAVmYAVmQDOYXuAFZmAFZmAFZkAzmF7gBWZgBWZgBWZ8I9hmqZpdxH5KS0tTXFxcUpNTVVsbKzd5QAAgAKqsPYpCut6AwCAwCqsfYrCut4AACCw8tKnYJi0EHG5XNq4caNcLpfdpSBEkAlYkQlYkQngHLYHWJEJWJEJWJEJ4By2B1iRCViRCViRCf8w8FeImKap1NRUhflJnsgDMgErMgErMgGcw/YAKzIBKzIBKzIBnMP2ACsyASsyASsy4R8G/gAAAAAAAAAAAIAwwMAfAAAAAAAAAAAAEAYY+CtEHA6HqlWrJoeDjx1nkQlYkQlYkQngHLYHWJEJWJEJWJEJ4By2B1iRCViRCViRCf9E2F0AgsfhcKhs2bJ2l4EQQiZgRSZgRSaAc9geYEUmYEUmYEUmgHPYHmBFJmBFJmBFJvzDMGkh4nK59Pvvv8vlctldCkIEmYAVmYAVmQDOYXuAFZmAFZmAFZkAzmF7gBWZgBWZgBWZ8A8Df4WIaZo6efKkTNO0uxSECDIBKzIBKzIBnMP2ACsyASsyASsyAZzD9gArMgErMgErMuEfBv4AAAAAAAAAAACAMMDAHwAAAAAAAAAAABAGGPgrRJxOp+rUqSOn02l3KQgRZAJWZAJWZAI4h+0BVmQCVmQCVmQCOIftAVZkAlZkAlZkwj8RdheA4DEMQ/Hx8XaXgRBCJmBFJmBFJoBz2B5gRSZgRSZgRSaAc9geYEUmYEUmYEUm/MPAn81OnTql5OTkoCzL5XJp8+bNql27dlBGyJOSkhQdHZ3vywk34ZwJiVz4g0zAThkZGVqzZo0aN26siAi6DQg94dxG0j76J5wzIZELf5AJ2Im+FEJdOLeR4dg+HjhwQCkpKfm6jNOnT2v//v35ugwPt9utHTt2qEqVKnI4gnNhunLlyikqKipflxEfH6/ExMR8XYbHpk2btHPnznxdxpkzZ3T48OF8XYaHaZrau3evKlSoIMMwgrLM0qVLq0iRIvm6jMqVK6tOnTr5uoxwRV/KP/ymbJacnKwhQ4bYXUa+mDFjhmrXrm13GQVOOGdCIhf+IBOwm8vlsrsEIFvh3EbSPvonnDMhkQt/kAnYjb4UQlk4t5Hh1j4eOHBAffv00en0dLtLwQVERUZq/oIF+T74d+DAAQ0fMVxulztfl4OL53A69O477wZtQDjc0JfKOwb+bJaUlKQZM2YEZVnbt2/XpEmT9Nhjj6lq1ar5vrykpKR8X0Y4CudMSOTCH2QCALIXzm0k7aN/wjkTErnwB5kAgOyFcxsZbu1jSkqKTqena0T946pQLP8Ogqe7pMOngnP2nR1KR7sVmY8nnO497tS0P85+Xvk9yJOSkiK3yy13fbfMYmb+Lcgt6WT+vb3tYiTlY+SN44YUpEwAHgz82Sw6Ojpo3z7yjIxXrlw5rL7xFG7IBKzIBABkjzYSVmQCVmQCALJHG1nwVCjmUtXY/D37pbY4u6YgMcubUkm7q0B2zL9N6Q+7q0BhE75f30AmnuunB+teEwh9ZAJWZAJWTqdTDRo0IBOAaCORGZmAFZmAFX0p4BzaSABAXtGX8g9n/AEAgBxFRkbaXQIAAECBRV8K/jhw4IBSUlLsLiOgkpOTfR7DRXx8PJfvA4B8RF8q7xj4K0Q8l1TgZpjwIBOwIhOwcrlcWr16tZo2baqICLoNKNxoI2FFJmBFJmBFXwr+OHDggPr06av09NN2l5IvJk6caHcJARUZGaUFC+Yz+AcA+YC+lH/4TQEAAAAAAAAhIiUlRenpp3WqeluZMfF2l4McGCdTpK3LlZKSwsAfACBkMPAHAAAAAAAAhBgzJl7uYqXtLgM5cNhdAAAAWWD/BAAAAAAAAAAAAIQBBv4KEafT6fMIkAlYkQlYOZ1ONW3alEwAoo1EZmQCVmQCVvSlAAAA/Edfyj8M/AEAgBylp6fbXQIAAECBRV8KAADAf/Sl8o6Bv0LE5XL5PAJkAlZkAlYul0vr1q0jE4BoI5EZmYAVmYAVfSkAAAD/0ZfyDwN/AAAAAAAAAAAAQBhg4A8AAAAAAAAAAAAIAwz8AQCAHHEDZQAAAP/RlwIAAPAffam8i7C7AASPZwNhQ4EHmYAVmYBVRESEmjVrZncZQEigjYQVmYAVmYAVfSkAAAD/0ZfyD2f8AQCAbJmmqZSUFJmmaXcpAAAABQ59KQAAAP/Rl/IPA3+FiMvl8nkEyASsyASsXC6XNm3aRCYA0UYiMzIBKzIBK/pSAAAA/qMv5R8G/gAAAAAAAAAAAIAwwMAfAAAAAAAAAAAAEAYY+CtEDMPweQTIBKzIBKwMw1BMTAyZAEQbiczIBKzIBKzoSwEAAPiPvpR/IuwuAMHjcDh8HgEyASsyASun06mGDRvaXQYQEmgjYUUmYEUmYEVfCgAAwH/0pfzDwF8hYpqmzyNAJmBFJgqOU6dOKTk5Od+XY5qmUlJSFB8fH7RvVyUlJSk6OjooywLygjYSVmQCVmQCVm63W4cPH1bp0qUZEEaeGSdTuFRXiDNOpgR9mXuPk4pQZsfnY6QZMkXfI1QZaZypdjHoS/mHgb9CxO12+zwCZAJWZKLgSE5O1pAhQ+wuI1/MmDFDtWvXtrsMIBPaSFiRCViRCVi53W5t27ZNpUqV4mAV8ix663K7S0AImvZHcbtLQIiIj49XZFSk0n9Ot7sUXEBkVKTi4+PtLqNAoi/lHwb+AAAogJKSkjRjxox8X8727ds1adIkPfbYY6patWq+L086u24AAABAYXeqeluZMfF2l4EcGCdTgj5AO6L+P6pQjC+YhKq9xx1BG5xNTEzUgvkLlJKSEpTlBYMdxyCCIT4+XomJiXaXgUKEgT8AAAqg6OjooJwV53K5JEmVK1fmLDwAAAAgiMyYeLmLlba7DOTAjnNPKhRzq2qsy4YlIxQlJiaG1YASxyCAwODcyELEc2+mYN2jCaGPTMCKTMCKTADnsD3AikzAikzAyjAMxcXFkQkAAHKBvhSs6Ev5hzP+snHgwIGwOk1aknbt2uV9DKfr4XKqtP88OQinPODikAlYkQngHLYHWJEJWJEJWDmdTtWtW9fuMgAAKBDoS8GKvpR/GPjLwoEDB9SnT1+lp5+2u5R8MXHiRLtLCKjIyCgtWDCfwT8/mKbp8wiQCViRCeActgdYkQlYkQlYud1u7d27VxUqVOAgJgAAF0BfClb0pfzDwF8WUlJSlJ5+mpsoFwDGyRRp63KlpKQw8OcHt9vt8wiQCViRCeActgdYkQlYkQlYud1u7d69W+XKleNgFQAAF0BfClb0pfzDwF8OuIly6GNTBwAAAAAAAAAAOItxEwAAAAAAAAAAACAMMPAHFGLcMBdWZAJWZAI4h+0BVmQCVmQCVg6HQ2XKlCETAADkAn0pWNGX8g+X+gQKMcMwfB4BMgErMgGcw/YAKzIBKzIBK4fDoerVq9tdBgAABQJ9KVjRl/IPw6RAIWaaps8jQCZgRSaAc9geYEUmYEUmYOV2u7V161a53W67SwEAIOTRl4IVfSn/MPAHFGKeBpOGEx5kAlZkAjiH7QFWZAJWZAJWbrdbhw4dIhMAAOQCfSlY0ZfyDwN/AAAAAAAAAAAAQBhg4A8AAAAAAAAAAAAIAwz8AYWYw+HweQTIBKzIBHAO2wOsyASsyASsHA6HKlasSCYAAMgF+lKwoi/lnwi7CwBgH8MwfB4BMgErMgGcw/YAKzIBKzIBK8/BKgAAcGH0pWBFX8o/DJMChRg3zIUVmYAVmQDOYXuAFZmAFZmAlcvl0saNG+VyuewuBQCAkEdfClb0pfzDwB9QiJmm6fMIkAlYkQngHLYHWJEJWJEJWJmmqdTUVDIBAEAu0JeCFX0p/zDwBwAAAAAAAAAAAIQBBv4AAAAAAAAAAACAMBBhdwFAQXLgwAGlpKTYXUbA7Nq1y/vodDptriZw4uPjlZiYGJRlkYmCIZiZCDcOh8PnEcitcGsfJdrIQAi3XJCJi0cmCoZw7EudOnVKycnJ+b4cz2WptmzZIsMw8n15kpSUlKTo6OigLAv5yziZwjf2Q5xxMsXuEoB8F6x9pmRPX4r9ZmhzOByqVq0ax6XyiIE/IJcOHDigvn366HR6ut2lBNy///1vu0sIqKjISM1fsCDfD04cOHBAffr2UfppMhHqIqMitWB+/mciHHkOUAXrQBXCw4EDB9SnT1+lp5+2u5R8EXZtZGSUFiyYH5T9Jn2pgoG+1MULt0yEY18qOTlZQ4YMsbuMfDFjxgzVrl3b7jJwEeLj4xUZGSVtXW53KciFyMgoxcfH210GkG/s2GcGsy/FfjO0ORwOlS1b1u4yChwG/oBcSklJ0en0dI2of1wVirnsLgfZ2HvcqWl/nP288vvAREpKitJPp8vd3C0zlhvMhiojzVD6z+lByUQ4crvdPo9AbqSkpCg9/bROVW8rMybe7nKQA+NkirR1edD2m/SlQh99KViFa18qKSlJM2bMyPfl7NixQ//+97/1+OOPq0qVKvm+POnsuqFgS0xM1IIF88PqjGjJnu0hGMLxrGjgfMHaZ0pnjz1s3bpV1atXD9oZXuw3Q5vL5dL69et16aWXhtUVNfIbA39AHlUo5lLVWA5W4Rwz1pRK2l0FsmOKA4kXw3N5Ks8jkBdmTLzcxUrbXQZyYMfFUuhLwYq+VGgL175UdHR0UL7d73Kdbe8qVarE2QTIk8TExLAbTGJ7AAqmYO0zJSkjI0OpqamqWbOmIiIYusDZ41EnT57kuFQecWFUAAAAAAAAAAAAIAww8AcAAAAAAAAAAACEAQb+AABAtjzXT+c66gAAAHlHXwo4h+0BwIU4nU7VqVOHdgJeZMI/XCgXAAAAAAAAAADYyjAMxcfH210GQgiZ8A9n/AEAgGy5XC6fRwAAAOQefSngHLYHABeSkZGhX375RRkZGXaXghBBJvzDwB8AAAAAAAAAALAdXw6AFZnIOwb+AAAAAAAAAAAAgDDAwB8AAAAAAAAAAAAQBhj4AwAA2XI6nT6PAAAAyD36UsA5bA8ALsTpdKpBgwa0E/AiE/5h4A8AAAAAAAAAANguMjLS7hIQYshE3jHwBwAAsuW5gTI3UgYAAMg7+lLAOWwPAC7E5XJp9erVtBPwIhP+YeAPAAAAAAAAAAAACAMM/AEAAAAAAAAAAABhgIE/AAAAAAAAAAAAIAww8AcAALLldDp9HgEAAJB79KWAc9geAFyI0+lU06ZNaSfgRSb8E/IDf99++61uvPFGVahQQYZhaMmSJXaXBAAAUGDQlwIAAPAffSkACI709HS99957evHFF/Xee+8pPT3d7pIQIshC3oX8wN/x48fVsGFDTZkyxe5SAAAodFwul88jCh76UgAA2Ie+VMFHXypw2B4AZGfq1Knq0KGDpkyZoqVLl2rKlCnq0KGDpk6dandpsJnL5dK6devYd+RRhN0FXEinTp3UqVMnu8sAAAAokOhLAQAA+I++FADkr6lTp+qdd95RyZIlNXjwYMXExOjkyZOaNWuW3nnnHUnSyJEjba4SKFhC/ow/AAAAAAAAAAAQXtLT07Vo0SKVLFlSixcvVufOnRUbG6vOnTtr8eLFKlmypBYtWsSlHoE8Cvkz/vLq9OnTOn36tPfntLQ0SVJGRoYyMjIkSQ6HQw6HQ263W2632zvv+dNRsLhcLrndbjkcDrlcLpmm6X3O6XTKMAzv53/+dM9rczPdY+9xxstDmefzcblc3s/cMAw5nc5M23x20y/URngyxinmBYunXcivNiIiIiJTLvIrexeaHuh1kiS3262MjIywWSd/PifrssNVIPtSxskUvmUW4oyTKZLoS+EcW/pSaUFYMfjv/38+wepL7d27V6mpqZLOZszhcMg0zUzZy2q6w+GQYRjZTne73Zn6HVlNz2vtOU1PTk6WJO3YsSNs1skzPS4uTomJifSlLALRlypIfWRrjTmtk2e6Z55wWKdgfk4ee487s5yO0OD5fILRl7rQ9IKwPX344YdyuVy6/fbbvcv27MciIyM1aNAgTZ48WR988IF69uxZINYpHD+nrGo/deqUdu/ene/9joyMDO3Zs0fFixf39nPyuy9VpUoVFSlSJOQ+p7z0pcJu4O+pp57S+PHjM01fs2aNihUrJkkqU6aMqlevru3bt+vQoUPeeSpWrKiKFStq586dQasXgbFhwwaVLl1aZcuW1fr163Xy5Envc3Xq1FF8fLzWrFnjs0E1aNBAkZGRWr16tc97NW3aVOnp6Vq3bp13mtPpVGxsrCRp2h/F83ltEAgbNmzw/oEVExOjhg0b6vDhw9q2bZt3nri4ONWtW1d79+7V7t27vdMv1Eb8+eefSk1N9XkNQl9KSook5Vsb0axZM6WmpmrTpk3e6fmVPY9q1arla7vXrFkznTp1SpK0adMm/fPPP2GxTv5+TsePH1dhEMi+VPTW5UGpGRePvhSsgtmXcv7MAcyCIBh9qS1btmj4sGE6E4YDRE8//bTdJQRckYgIPfjQQ6pQoQJ9qfMEoi9VkPrIHrlZJ8/7b9q0SWXLlg2LdQrm51S5cmVFRRbRtD+KCaEtKrKIkpOT870v5VGQ24i//vpLklS2bFmfelJSUlS2bFmVLFlS0tk2tEqVKgVincLxc8pqnXbv3q1XXnlF4WjGjBk6depUyH1OeelLGeb5w5YhzjAMffjhh+rWrVu282T1zapKlSrpyJEj3oMNFxq53rhxo4YNG6aTl3aTu1jpfFsfXDzH8cOKWb9E06dPV506dfL1WxNbt27VkCFDNKL+P6pQjLNCQ9Xe4w5N+6O4pk+frlq1aknKv2+3/Pnnnxo+fLhc17qkksFbR+TR35LzK6feeOMN1alTJ2y+WZXT9ECu06ZNmzR06FBNmzZNtWvXDot18vdzSktLU0JCglJTU719ioIm2H2pU9XbyoyJz6/VQQAYJ1MUvXU5fSl42dKXau6SCmazWjiknR2cDUZfavPmzRoyZIh6SiqTD6uCwDkk6X1J06dPV+3atelLnScQfamC1Ee21pjTOm3atEkjRozQtGnTVLdu3bBYp2B/Tvv379fRo0d9ai/IZxHv3r1bEydO1KOPPqqkpKSwWCen06m4uDiVLn3umHI4ZC+/tqfFixdrypQpevDBB3XDDTfINE2lpaUpPj5eTqdTS5Ys0eTJkzVy5EjO+AuxdQrWGX87duzQv//9b287UZjP+MtLXyrszviLiopSVFRUpukRERGKiPBdXU9grbKahtDmdDq9n5tno7Kyfv7+Tq9QzK2qsVziMdQ5nc48bfN5me7JWHZZQ2gyDENS/rYRhmFkOT3Q2cvt9ECsk6fDYV23grxO/n5O2S0j3ASyL2XGxPMlqhDn+fToS8EqqH2pWPElqgIgGH0pjzKSKsjIW4EIsrMHwpxOpzcT9KXOCkRfqiD1kXM73XPA1LOsYPQ7wvHvs3LlyqlcuXJZzl+QJSUlqV69enaXke8Kcvbya3u66aabNH36dL355pu64YYbZBiG/vrrLzVt2lQZGRmaPXu2nE6nunfv7rP8UF4nj3D6nDzOr7148eKqU6dOlrUFkmeALFTaiYJyXCrke13//POPtmzZ4v15+/btWrt2rUqVKqXKlSvbWBkAAEDooy8FAADgP/pSAJB/IiMjdfPNN+udd95Rjx49NGjQIMXExOg///mPZs+erb///lu9evVSZGSk3aUCBUrID/ytXr1a7dq18/58//33S5IGDBigOXPm2FQVAABAwUBfCgAAwH/0pQAgf40cOVKStGjRIk2ePNk73el0qlevXt7nAeReyA/8tW3b1udaqgAAIHjOvxwPCib6UgAAAP6jLxU4/G0BKzIBj5EjR+qOO+7Qhx9+qHXr1qlBgwa66aabONMPtBN+CvmBPwAAYB/PtcW5/y0AAACAi8HfFrAiEzhfZGSkbr31Vt166612l4IQQjvhH35bAAAgW55vN/MtZwAAAAAXg78tYEUmYOV2u3Xw4EG53W67S0GIoJ3wDwN/AAAgW57ONp1uAAAAABeDvy1gRSZg5Xa7tW3bNjIBL9oJ/zDwBwAAAAAAAAAAAIQBBv4AAAAAAAAAAACAMMDAHwAAyJZhGD6PAAAAAOAP/raAFZmAlWEYiouLIxPwop3wT4TdBQAAgNDlcDh8HgEAAADAH/xtASsyASun06m6devaXQZCCO2Ef/htAQCAbJmm6fMIAAAAAP7gbwtYkQlYud1u7d69W2632+5SECJoJ/zDwB8AAMiWp7NNpzt/paamyuVy2V0GAAAAkG/42wJWZAJWDPzBinbCPwz8AQAA2GD16tXq2LGjihYtqoSEBK1YsUKSdPjwYf3rX//S8uXL7S0QAAAAAAAABQ73+AMAAAiyH374Qe3bt9cll1yivn37aubMmd7nSpcurdTUVL3++utq27atfUUGgHEyhW+ZhTjjZIrdJQAAAAAAgABi4A8AAGSLmyjnj8cee0x169bVqlWrdOzYMZ+BP0lq166d5s6da1N1Fy8+Pl6RkVHS1uV2l4JciIyMUnx8vN1lAACAMMffFrAiE7ByOBwqU6YMmYAX7YR//B74c7lcWrRokZYtW6aDBw9qwoQJuuyyy5Samqqvv/5aV111lRITEwNZKwAACDLDMHweERi//PKLnnrqKUVFRemff/7J9Pwll1yi/fv321BZYCQmJmrBgvlKSUmxu5SASk5O1sSJE/XEE08oKSnJ7nICJj4+nn47AADId/xtASsygfOlp6dryZIl2rt3rypUqKBu3bopMjLS7rJgM9oJ//g18JeSkqKOHTvq559/VvHixXX8+HHdfffdkqTixYtr1KhR6t+/vyZNmhTQYgEAQHCZpunziMAoUqRIjjem3rNnj4oXLx7EigIvMTEx7AaTPNtB5cqVVbt2bZurAQAAKFj42wJWZAIeU6dO1aJFi+RyubzTpk2bpptvvlkjR460sTLYjXbCP36dH/nII4/ojz/+0Oeff65t27b5/NKdTqd69uypTz75JGBFAgAAe3gGp3IapELeXXHFFXr//fezfO748eOaPXu22rRpE+SqcCFsDwAAAP6jLwUrMgHp7KDfO++8o9jYWD3wwAN64okn9MADDyg2NlbvvPOOpk6daneJsBHthH/8GvhbsmSJ7r77bl133XVZnmJZq1Yt7dix42JrAwAACEvjx4/X6tWr1blzZ3366aeSpN9//10zZ87U5ZdfrkOHDmn06NE2VwkAAAAAQP5JT0/XokWLVLJkSS1evFidO3dWbGysOnfurMWLF6tkyZJatGiR0tPT7S4VKFD8utRnamqqqlatmu3zZ86cUUZGht9FhQrjZIp/I6MIGuNkit0lAFKa3QUgR3w+CEEtWrTQJ598ohEjRqh///6SpAceeECSVL16dX3yySdq0KCBnSUC+WbvcafdJSAHdnw+RpohU1y6J1QZacG/n8ohSSITIe2Q3QUgYE6dOqXk5OSgLGvnzp3eR6cz//c3SUlJio6OzvflAPDfkiVL5HK5dMcddygiIsJnTCEiIkK33367nn/+eS1ZskS33HKLjZUWHAcOHFBKSordZQRMsPcdwRQfH59vt0jxa+CvevXq+u2337J9/osvvlC9evX8LipURG9dbncJAAoA58/htdMBzudwOHwecfFM09SxY8d05ZVXavPmzVq7dq3++usvud1uVa9eXZdffjk3rQ5RbA8XJz4+XlGRkZr2h92V4EKiIiMVHx+f78uJj49XZFSk0n/mG9yhLjIqOJnwyPpi2ADyQ3JysoYMGRLUZU6aNCkoy5kxYwb3ZQ5x9K+xd+9eSdKVV14p6WwWKlas6M2EZ7pnPuTswIED6tunj06H4RmSwdp3BFNUZKTmL1iQL4N/fg383XHHHfq///s/tW3bVtdcc40kyTAMnT59WhMmTNBnn32mN954I6CF2uFU9bYyY+LtLgM5ME6mMEAL27mau6RYu6tAttIYnL0YngEoBqICJz09XaVKldKkSZP08MMPq1GjRmrUqJHdZSEX2B4uTmJiouYvWBBW3z5NTk7WxIkT9cQTTygpKcnucgImP795er7ExEQtmE8mCoJgZcKjp6QyQVsa/HFIDNCGi6SkJM2YMcPuMvJFOLXD4Yr+NSpUqCBJ+uGHH3TjjTd6B/48fvjhB5/5kLOUlBSdTk+nL1UAHJL0fnq6UlJSQmfg75577tEff/yh2267zfutv969e+vIkSPKyMjQsGHDdPvttweyTluYMfFyFyttdxnIAd8HQkiIlVTS7iKA/MFNlAMvKipK5cqVU1RUlN2lII/YHi5eYmJiUAcP8psnC5UqVeKMAj+RCWSljKQK4iBwaONSrOEiOjo6aO2Vy+XSn3/+qVq1aoXd5drgH/rX6Natm6ZNm6aZM2eqU6dOMgzD206Ypqk333xTTqdT3bp1s7vUAoW+VEGQv30pv8ZNDMPQjBkz9O2336p///7q1KmTGjVqpKFDh2r58uWaNm1aoOsEAAA2ME3T5xGBMXDgQM2bN48blBcwbA+wIhOwIhMAkD3TNJWamkobCS/2m4iMjNTNN9+sv//+Wz169NDSpUu1c+dOLV26VD169NDff/+tm2++WZGRkXaXChQoeT7j78SJE+rbt6969OihPn36qFWrVvlRFwAAQNi67LLLtGTJEtWvX18DBw5UlSpVFBMTk2m+7t2721AdAAAAAADBMXLkSEnSokWLNHnyZO90p9OpXr16eZ8HkHt5HvgrWrSovvrqK3Xq1Ck/6gEAAAh7t912m/f/o0ePznIewzDkcrmCVRIAAAAAALYYOXKk7rjjDn3wwQdas2aNGjdurO7du3OmH+Anv+7x16pVK/34448aMmRIoOsBAAAhxOFw+DwiMJYtW2Z3CfAD2wOsyASsyAQQPGlpaZo6daqWLVumgwcP6vXXX1fz5s119OhRzZkzR127dlWNGjXsLhPncTgcqlatGm0kvNhv4nyRkZG65ZZb1L59e5UuXZpcABfBr4G/1157TR06dNATTzyh4cOHq2LFioGuCwAAhADDMHweERht2rSxuwT4ge0BVmQCVmQCCI7du3erTZs22rVrl2rWrKlNmzbpn3/+kSSVKlVKr7/+upKTk/Xyyy/bXCnO53A4VLZsWbvLQAhhvwkr2gkgMPwaNm/YsKF2796tp556SklJSYqKilJsbKzPv7i4uEDXCgAAgsztdvs8IvA2bNigTz/9VJ9++qk2bNhgdznIAdsDrMgErMgEEBwPPfSQjh07prVr12rFihUyTdPn+W7duumrr76yqTpkx+Vy6ffff+dy9vBivwkr2gkgMPw6469Hjx58EwMAgELAcxDFejAFF++jjz7S/fffrx07dvhMr1q1qiZPnqyuXbvaUxiyxfYAKzIBKzIBBMcXX3yh++67T/Xq1dORI0cyPV+tWjXt2rXLhsqQE9M0dfLkSdpIeLHfhBXtBBAYfg38zZkzJ8BlAAAAFB6ffPKJevTooaSkJE2aNEl169aVJG3cuFFvvPGGunfvrv/85z/q2LGjzZUCAACEnpMnT6pMmTLZPn/s2LEgVgMAABBa/Br4AwAAgP+efPJJNWjQQCtXrlSxYsW807t27aq77rpLrVq10vjx4xn4AwAAyEK9evX07bffatiwYVk+v2TJEjVu3DjIVQEAAIQGv+7xJ0lpaWkaP368mjdvrsTERCUmJqp58+aaMGGC0tLSAlkjAACwidPp9HlEYKxbt04DBgzwGfTzKFasmAYOHKh169bZUBlywvYAKzIBKzIBBMe9996rd955R88884xSU1Mlnb1H2JYtW9SvXz/9+OOPuu+++2yuElZOp1N16tShjYQX+01Y0U4AgeHXGX979+7V1Vdfre3bt6tOnTq66qqrJEmbN2/WuHHjNG/ePK1cuVLly5cPaLEAAADhIDo6WkePHs32+aNHjyo6OjqIFQEAABQcffv2VXJysp544gk9/vjjkqSOHTvKNE05HA5NmjRJ3bp1s7dIZGIYhuLj4+0uA0AIo50AAsOvgb//+7//0/79+/Wf//xHN9xwg89zn376qW6++WY98sgjmjt3bkCKBAAA9nC5XD6PCIz27dvr5ZdfVseOHdWyZUuf53766Se98soruv76622qDtlhe4AVmYAVmQCC5/HHH1e/fv20ePFibdmyRW63W9WrV1f37t1VrVo1u8tDFjIyMrRmzRo1btxYERHcfQjsN5EZ7QQQGH5tPZ999pnuvffeTIN+ktSpUyeNGjVKM2bMuOjiAAAAwtGzzz6rli1bqlWrVmrevLlq164t6ezVE37++WeVLVtWzzzzjM1VAgAAhJ4TJ07o6quv1pAhQzR8+HAu6VnAMMAD4EJoJ4CL59c9/o4fP67ExMRsny9XrpyOHz/ud1EAAADhrGrVqlq3bp1GjRqlv//+W++++67effdd/f3337rnnnv0+++/q0qVKnaXCQAAEHKKFi2q7du3yzAMu0sBAAAISX4N/NWrV09vv/220tPTMz135swZvf3226pXr95FFwcAABCuypYtqxdffFGbNm3SyZMndfLkSW3atEmTJ09W2bJl7S4PAAAgZHXs2FGff/653WUAAALI5XJp7dq1WrNmjdauXcuZf8BF8Psef7feequaN2+ukSNHqlatWpLOXp5q+vTpWrdund59992AFgoAAILP6XT6PCIwMjIydOLECcXGxmb5fFpamooWLco9DUIM2wOsyASsyAQQHKNHj9bNN9+sfv36adiwYapatapiYmIyzVeqVCkbqkN2nE6nGjRoQBsJL/ab8FixYoWmTJmi/fv3S5LefvttlStXTnfeeafatGljc3VAwePX0aSbb75Zx48f1yOPPKLhw4d7L69gmqbKli2rWbNmqWfPngEtFAAAIFyMGjVK3377rdavX5/l81dddZXat2+vl19+OciVAQAAhL769etLkjZs2KCFCxdmOx9ni4SeyMhIu0sAEGJWrFihMWPGqGXLlhozZoySkpKUnJys+fPna8yYMZowYQKDf0Ae+f018oEDB6pv375avXq1kpOTJUlJSUlq2rQp304HACBMeA6WcNAksD777DP1798/2+d79uyp+fPnM/AXYtgeYEUmYEUmgOAYM2YM9/grgFwul1avXs2xQ3ix34TL5dKUKVPUsmVLTZo0SW6329tOTJo0SY899pimTp2qVq1acWYokAcXtZeNiIjQFVdcoSuuuCJQ9YQU42SKfzdBRNAYJ1OCvsy9x9nJhDI+H9jtwIEDSklJsbuMgNm5c6f3Mdw62fHx8UpMTLRl2Xv37tUll1yS7fMVKlTQnj17glgRAACh45AkybS5CuTkkM3LHzdunM0VAAACYd26ddq/f7/Gjh0rh8Mht9vtfc7hcKhv374aOXKk1q1bp8aNG9tYKVCw+DXw9/bbb+vzzz/XnDlzsnx+0KBB6tSpk2655ZaLqc028fHxioyMkrYut7sU5EJkZJTi4+PzfTnx8fGKiozUtD/yfVG4SFGRkUHJBGB14MAB9e3TR6fT0+0uJeAmTZpkdwkBFxUZqfkLFtgy+JeQkKDNmzdn+/zGjRuzvf8fAADhyvM31/th2JcKR6H0d9fJkyclKcv7/AEAQteRI0ckSVWrVs3y+WrVqvnMByB3/Br4e/HFF3McYY+JidGLL75YYAf+EhMTtWDB/LA6Y0OStm/f7j1FOrvGtCAK1hkbiYmJmr9gQVjlgkwAgZWSkqLT6enqKamM3cUgR4ckvZ+erpSUFFvai44dO+r1119Xnz59MvWpfvvtN73xxhu6+eabg14XAAB24m+ugsXuv7t27typsWPH6pNPPtHhw4clSaVLl1bnzp01duxYJSUl2VYbACB3EhISJJ3dX3ru33q+bdu2+cwHIHf8GvjbvHmzBg8enO3zDRs21Ntvv+13UaEgMTExbAcOqlatqtq1a9tdRoEUrrkgE0BglZFUQdxzJLTZe/mwJ598Up999pmaN2+url27ev/AWb9+vZYuXaqyZcvqySeftLVGZOa53G24XfYW/iMTsCITF4+/uZAbmzZtUqtWrZSSkqLrrrtOdevW9U6fN2+eli5dqu+++47feYhxOp1q2rQpbSS82G+iQYMGKleunN566y1NmjTJp51wu92aP3++ypcvrwYNGthdKlCg+DXwZ5pmjt/A+/vvv3XmzBl/awIAAAhrFSpU0OrVq/XII4/oo48+0ocffihJio2NVZ8+fTRp0iRVqFDB5ioBAABC0yOPPCKHw6E1a9bosssu83lu/fr1uuaaa/TII494+1gIHenp6VySFYCX0+nUnXfeqTFjxuixxx5Tnz59VL58ee3bt08LFizQjz/+qAkTJjA4DOSRXwN/jRs31ttvv637779fkZGRPs+dPn1aCxcu5GabIcjlcvk8AmQCAOxTvnx5zZ07V6Zp6tChQ5KkMmXKyDA4WzRUsd+EFZmAFZmAFZnIHytWrNADDzyQadBPki699FLdddddmjx5sg2VIScul0vr1q1T06ZNFRHh1yFJhBnaSEhSmzZtNGHCBE2ZMkV33nmnd3r58uU1YcIEtWnTxsbqgILJr73sI488oi5duqhdu3Z65JFHfC5P9dRTT+mPP/7Qxx9/HNBCAQAAwpFhGCpbtqzcbrcOHTrE4B8AAMAFnDlzJsezxooWLcqVqACgAGnTpo1atWqlNWvW6JdfflGzZs3UuHFjzvQD/OTw50WdOnXSm2++qfXr16tbt26qWbOmatasqW7dumnDhg2aMWOGOnfuHOhaAQAACqw///xT8+bN099//+0zPTU1Vf3791fRokVVvnx5lSlTRq+99ppNVQIAAIS+xo0ba+bMmUpNTc30XFpamt588001adLEhsoAAP5yOp1q1KiRGjdurEaNGjHoB1wEv8+rHzhwoLp3764vv/xSW7dulSRVr15d119/vUqUKBGwAgEAAMLBCy+8oM8++0z9+vXzmT5s2DC99957qlmzpho0aKAffvhB99xzjypWrKhu3brZUywAAEAIGz9+vDp27Kg6depo0KBBqlWrliRp8+bNmjt3ro4cOaIpU6bYXCWywoF8ABdCOwFcvIu6oHZsbKx69OgRqFqQzzyNJo0nPMgEAATP999/ry5duvhcxnPXrl1677331LJlS61YsUIRERFKSUlRs2bNNGXKFAb+Qgz7TViRCViRCViRifzRvn17ffLJJ3rooYf09NNP+zzXqFEjvfXWW2rXrp1N1SE7ERERatasmd1lIITQRsKKdgIIjIDcSfebb77RggULtG/fPtWpU0f33HOPkpKSAvHWAAAAYWHPnj2qU6eOz7T//Oc/MgxD99xzjyIiznbL4uPj1b9/f7388st2lAkAAFAgXHvttVqzZo3279+v5ORkSVJSUpLKlStnc2XIjmmaSk1NVVxcHPe0BpAl2gkgMHJ9j79x48apaNGiOnz4sM/0mTNn6rrrrtPs2bP12Wef6aWXXlKzZs20Y8eOQNeKi+RyuXweATIBAMHjdrtVpEgRn2nfffedpLM3Mj9fxYoVdezYsaDVhtxhvwkrMgErMgErMpH/ypUrpxYtWqhFixYM+oU4l8ulTZs2sT3AizYSVrQTQGDkeuBv2bJl6tSpk0qXLu2ddvLkSd1///2Kj4/XsmXLdOzYMb3zzjv6559/NHHixHwpGAAAoCCqXr26Vq1a5f3Z5XLpm2++UZ06dZSYmOgz79GjR1WmTJlglwgAAFAgvPLKK+rQoUO2z3fq1EnTpk0LYkUAAAChI9cDf3/++aeaNGniM+3LL7/UP//8o4cfflht2rRRsWLFdMstt6hfv376+uuvA14sAABAQTVgwAAtXLhQzzzzjFauXKk777xTBw8eVN++fTPNu3LlStWqVcuGKgEAAELfm2++qXr16mX7fL169fTGG28EsSIAAIDQkeuBv5SUFJUvX95n2rJly2QYhrp06eIz/fLLL9e+ffsCUyECxnNdZK6PDA8yAQDBM3LkSHXo0EGPPvqo2rZtqzfeeEOtW7fWgw8+6DPfrl279Omnn6pTp042VYrssN+EFZmAFZmAFZnIH1u3blXdunWzfb5OnTraunVrECtCbhiGoZiYGLYHeNFGwop2AgiMiNzOeMkll2S6b9+KFSsUHx+f5besihYtetHFIbAcDofPI0AmACB4ihQpoqVLl2r16tXaunWrkpKSdMUVV2Sa7/Tp01q4cKFat25tQ5XICftNWJEJWJEJWJGJ/BEZGan9+/dn+/y+ffv4nYcgp9Ophg0b2l0GQghtJKxoJ4DAyHWrevXVV2vWrFnavXu3pLNn+61du1ZdunTJNAK/bt06VapUKbCV4qKZpunzCJAJAAi+pk2b6tZbb81y0E+SatSooR49enCPvxDEfhNWZAJWZAJWZCJ/XHHFFZozZ46OHTuW6bnU1FTNnj07274W7ON2u3Xw4EG53W67S0GIoI2EFe0EEBi5PuNv3LhxWrJkiapXr66KFStq165dKlq0qEaPHu0zX0ZGhj744AP17Nkz4MXi4ngaTBpOeJAJAAByj/0mrMgErMgErMhE/hg7dqzatGmjRo0a6d5771X9+vUlSevXr9dLL72kffv2aeHChTZXCSu3261t27apVKlSnOEFSbSRyIx2AgiMXA/8JSUlafXq1Zo8ebK2bdum6667TqNGjVKNGjV85lu1apUuv/xy9e7dO+DFAgAAAAAAoHBr0aKFli5dqmHDhumee+7xXonKNE1VrVpVH3/8sVq2bGlzlQAAAPbI9cCfJFWvXl1TpkzJcZ5WrVqpVatWF1UUAAAAAAAAkJ3rrrtOW7Zs0Zo1a7R161ZJZ49bNWnSJNMtaQAAAAqTPA38oWDzdHzpAMODTAAAkHvsN2FFJmBFJmBFJvKXw+HQ5Zdfrssvv9zuUpALhmEoLi6O7QFetJGwop0AAoML5RYinusic31keJAJAAByj/0mrMgErMgErMhE4Jw4cUI7d+5Uenp6pudmzZqla665RvXq1VP37t31yy+/2FAhLsTpdKpu3bpyOp12l4IQQRsJK9oJIDBoVQsR0zR9HgEyAQD2GDFihH744Qe7y0Aesd+EFZmAFZmAFZkInAkTJqhBgwaZBv4mTpyoIUOGaMWKFTp06JCWLFmitm3b6vfff7epUmTH7XZr9+7dcrvddpeCEEEbCSvaCSAwuNRnIeJpMGk44UEmAMAeCxcu1BtvvKEqVaqob9++6tu3r2rWrGl3WQXSqVOnlJycHJRl7dixw/sYjEvPJCUlKTo6Ot+XE27CORMSufAHmYCd+JsrcJYtW6YuXbqoePHi3mlpaWmaOHGiLrnkEq1YsUJVq1bVzz//rA4dOujpp5/W22+/bWPFsPIc0C9XrhxneEESbSQyo50AAoOBPwAAgCA7ePCgPv74Y82fP19PP/20Jk6cqKZNm6p///669dZbVbp0abtLLDCSk5M1ZMiQoC5z0qRJQVnOjBkzVLt27aAsK5yEcyYkcuEPMgGEhx07dqhHjx4+0z755BOlp6fr//7v/1S1alVJUvPmzTVo0CC99957dpQJAABgO78G/iZMmKDu3bvr0ksvzfL5P/74Q4sXL9aYMWMuqjgAAIBwFBUVpZtvvlk333yz/v77b7333ntasGCBRo0apfvvv1/XXXed+vfvr65du3IWxwUkJSVpxowZQVmWy+XShg0bVK9evaDccyIpKSnflxGOwjkTErnwB5kAwsOxY8eUkJDgM+3bb7+VYRjq0KGDz/R69erp0KFDwSwPAAAgZPg18Ddu3DjVqFEj24G/9evXa/z48Qz8hRhumAsrMgEA9itZsqSGDRumYcOGaefOnXrooYe0aNEiffrppypRooR69uypUaNGqUGDBnaXGpKio6ODdqaL2+1WVFSUqlatyr4zhJEJWJEJ2Im/uQInKSlJmzZt8pm2fPlyJSYmqkaNGj7T09PTFRsbG8zykAsOh0NlypRhe4AXbSSsaCeAwMiXLejo0aOKjIzMj7fGRfDcYyJY95pA6CMTABAadu3apaefflqdO3fWokWLlJCQoBEjRmjw4MH6+OOP1aRJE02bNs3uMgs9h8Oh6tWr80covMgErMgErPibK3Cuv/56zZo1Sz/99JMkad68edq0aZNuuummTPP++uuvqlKlSpArxIXQRsKKNhJWtBNAYOT6jL9vv/1Wy5cv9/78wQcfaMuWLZnmS0lJ0bvvvqvLLrssIAUicEzT9HkEyAQA2CclJcV7ic/vv/9eERER6ty5s5588kl17txZRYoUkSQ99dRTuu222zRhwgSNGDHC5qoLN7fbre3bt3MmD7zIBKzIBKz4mytwRo8erSVLlujKK6+U0+lURkaGypQpk+lqUydOnNCHH36ooUOH2lQpskMbCSvaSFjRTgCBkeuBv2XLlmn8+PGSzn4L44MPPtAHH3yQ5bz16tXTq6++GpgKETBut9vnESATAGCPm266SZ9++qnS09PVokULvfrqq+rVq5dKliyZad6oqCj17NlTS5YsCX6h8OF2u3Xo0CElJSXxRygkkQlkRiZgxd9cgVO6dGmtXbtWM2fO1LZt25SUlKTBgwerbNmyPvOtX79effr0Ub9+/WyqFNmhjYQVbSSsaCeAwMj1wN/DDz+su+66S6ZpqmzZspo+fbp69OjhM49hGCpatKiio6MDXigAAEC4WLNmjR566CH1799fNWvWvOD81113nZYtWxaEygAAAEJXyZIl9dBDD+U4T/PmzdW8efMgVQQAABB6cj3wFxMTo5iYGEnS9u3bVaZMGRUtWjTfCgMAAAhXO3bsyNP8ZcqUUZs2bfKnGAAAAAAAAIQNv86XTUpKYtCvAPKcHs1p0vAgEwBgD6fTqbfffjvb59999105nc4gVoTccDgcqlixIvtNeJEJWJEJWPE3F3AObSSsaCNhRTsBBIZfW5Bpmnr99dfVvHlzlS5dWk6nM9O/iIhcn0yIIDEMw+cRIBMAYA/TNHO8gb3L5aJtDkH8EQorMgErMgEr/uYCzqGNhBVtJKxoJ4DA8Gt07uGHH9bkyZPVqFEj9e3bVyVLlgx0XcgH3DAXVmQCAOyT3R+3aWlp+vzzz1W6dOkgV4QLcblc+vPPP1WrVi3OyIQkMoHMyASs+JsLOIc2Ela0kbCinQACw6+Bv7lz56pHjx567733Al0P8pHnzIKczjBA4UImACB4xo8frwkTJkg6O+jXt29f9e3bN8t5TdPUqFGjglkecsE0TaWmprLfhBeZgBWZgBV/cwHn0EbCijYSVrQTQGD4NfB38uRJXXvttYGuBQAAIGw1b95cI0eOlGmamjp1qq677jrVqlXLZx7DMFSsWDFdfvnl6t69u02VAgAAhLZ3331X//rXvxQdHW13KQAAACHHr4G/a665Rr/88ouGDh0a6HoAAADCUqdOndSpUydJ0vHjxzV8+HC1aNHC5qoAAAAKnttuu02xsbHq0aOH+vbtq3bt2tldEgAAQMjw6y6ZU6dO1apVqzRp0iQdOXIk0DUhn3huisrNUeFBJgDAHrNnz2bQrwByOByqVq0a+014kQlYkQlY8TdX/vjuu+/Up08fLV26VNdee60qV66sRx55ROvXr7e7NOSANhJWtJGwop0AAsOvM/5q164tt9ut0aNHa/To0YqOjs50s03DMJSamhqQIhEYhmH4PAJkAgCCY968eZKkfv36yTAM788X0r9///wsC3nkcDhUtmxZu8tACCETsCITsOJvrvxx5ZVX6sorr9TLL7+szz77TAsWLNBrr72m5557Tpdddpn69eun3r17q3z58naXivPQRsKKNhJWtBNAYPg18NejRw8a5ALI7Xb7PAJkAgCCY+DAgTIMQ7169VJkZKQGDhx4wdcYhsHAX4hxuVxav369Lr300kxfekPhRCZgRSZgxd9c+SsiIkJdunRRly5d9M8//+jDDz/UnDlz9PDDD+uRRx5R27ZtNWDAAN1yyy2KjIy0u9xCjzYSVrSRsKKdAALDr4G/OXPmBLgMBINpmj6PAJkAgODYvn27JHkPOHl+RsFimqZOnjzJfhNeZAJWZAJW/M0VPOvXr9fPP/+s//3vfzJNU3Xq1NGRI0fUv39/Pfroo3r77bfVqlUru8ss1GgjYUUbCSvaCSAw/Br4AwAAQO4lJSXl+DOAgsflcmnt2rVas2aNIiIi1LhxY76VXMiRCSD4/vzzTy1YsEALFy7Utm3bVLp0afXu3Vv9+/dXkyZNJEmrV6/W7bffruHDh3MPQBvRRgK4ENoJIHD8HvjbuXOnJk2apGXLlungwYP66KOP1Lp1ax0+fFgTJkzQoEGD1Lhx40DWCgAAAAC2W7FihaZMmaL9+/dLkt5++22VK1dOd955p9q0aWNzdbADmQCC6+WXX9aCBQv066+/KioqSjfeeKNeeukldezYMdNB4qZNm+r+++/X7bffblO1oI0EcCG0E0Bg+TXwt2HDBl199dVyu91q0aKFtmzZooyMDElS6dKl9d133+n48eN68803A1osLo6n88s3JeBBJgAgONq3b5/n1xiGoa+//jofqoG/nE6n6tSpw36zkFuxYoXGjBmjli1basyYMUpISNCRI0c0f/58jRkzRhMmTODgRCFDJpAT/ubKH/fdd5+uuuoqTZ8+Xbfccovi4uJynL9p06YaPXp0kKrD+WgjkRPaSEi0E0B+8Gvg7+GHH1Z8fLxWrVolwzBUtmxZn+c7d+6sd999NyAFAgAAFHRut1uGYeTpNdzTIPQYhqH4+Hi7y4CNXC6XpkyZopYtW2rSpElyOBySpPLly2vSpEl67LHHNHXqVLVq1YoDWIUEmQDssXXrVlWtWjXX89evX1/169fPx4qQFdpIABdCOwHkD78G/r799luNGTNGZcqU0ZEjRzI9X7lyZe3Zs+eii0NguVwun0eATASGkWbIFAfoQ5WRlrfBlkA4JElkIqQdCvLyli9fHuQlIj9kZGRozZo1aty4sSIiuFV2YbRu3Trt379fY8eOlcPhyJSJvn37auTIkVq3bh23PSgkyETBderUKSUnJ+f7crZv3+7zGAxJSUmKjo4O2vLscPvtt+vxxx/XNddck+Xzy5Yt05NPPqlvvvkmyJXhfLSRBVOw2keJNhK0E/mF41KhL7+PS/l1xMLtdqto0aLZPn/o0CFFRUX5XRQAFATx8fGKjIpU+s/pdpeCC4iMigzqWTrvB21JAIKNL8sUbp4vPZ5/lsn5mahWrZrPfAh/ZKLgSk5O1pAhQ4K2vEmTJgVtWTNmzFDt2rWDtjw7LF++XHfccUe2zx88eFArVqwIYkXICm1kwRTs9lGijSzMaCfyB8el4NfAX5MmTfTf//5XI0eOzPRcRkaG3nnnHV1xxRUXXRwAhLLExEQtmL9AKSkpdpcSMNu3b/deSiEvl84JdfHx8UpMTAza8npKKhO0pcEfhxTcjvDOnTslnb0qwvk/X4hnfgChISEhQdLZ/WVWl4zbtm2bz3wIf2Si4EpKStKMGTPyfTkul0sbNmxQvXr1gnaJsqSkpKAsx245XUZ9y5YtKlGiRBCrQVZoIwumYLWPEm0kaCfyC8elQl9+H5fya+Dv0UcfVZcuXTRixAj16tVLknTgwAF99dVXmjRpkjZu3KjXXnstoIUCQChKTEwM6oBSfvN8q6py5cp8A+4ilJFUQcG/xCjyIriXvKhSpYoMw9DJkycVGRnp/flCOLsMCC0NGjRQuXLl9NZbb2X6Zrrb7db8+fNVvnx5NWjQwKYKEWxkouCKjo4OSn83IyNDaWlpqlWrFpeJvkhz587V3LlzvT9PnDgxy8GJlJQUrVu3TjfccEMwy0MWaCMLpmC1jxJtJGgn8gvHpQqC/D0u5VeL2qlTJ82ZM0f33HOP3njjDUlS3759ZZqmYmNjNW/ePLVu3TqgheLieb45w41Q4UEmYEUmgPwxa9YsGYahIkWK+PyMgsXpdKpBgwa0kYWY0+nUnXfeqTFjxuixxx5Tnz59VKNGDW3cuFELFizQjz/+qAkTJpCRQoRM4ELYdwTOiRMndOjQuTviHDt2TA6Hw2cewzBUrFgxDR8+XGPGjAl2ibCgjcSF0EaCdgLIH35/laJfv37q3r27vvjiC23ZskVut1vVq1dXhw4duJwCAADAeQYOHJjjzyg4IiMj7S4BNmvTpo0mTJigKVOm6M477/ROL1++vCZMmKA2bdrYWB3sQCZwIew7AmPEiBEaMWKEpLP3gnr55ZfVtWtXm6vChdBG4kJoI0E7AQTeRZ1DXaxYMd10002BqgX5zHO5MC4bBg8yASsyAQSfaZreb6+XKVOGMwFDmMvl0urVq9W0aVMuRVTItWnTRq1atdKaNWv0yy+/qFmzZmrcuDHfRC7EyASyw74jf2zfvt3uEpAHtJHIDm0kPGgngMC6qBb1zJkz2rNnj/7++2+ZZuZrkjZp0uRi3h4AACBsbdiwQWPGjNHnn3+uEydOSJKKFi2qDh06aNy4cbr00kttrhBATpxOpxo1aqSMjAw1atSIgxIgE0A+2rlzp6Sz9yI//+cL8cwP+9FGArgQ2gkgcPwa+EtJSdGDDz6oBQsWKD09PdPzpmnKMAzOGAEAAMjCypUr1alTJ7ndbv3rX/9SrVq1JEmbN2/Wxx9/rE8//VSfffaZrr76apsrBQAAsF+VKlVkGIZOnjypyMhI788XwnEpAABQGPk18Ddw4EAtXbpUvXr1UosWLRQXFxfougAAAMLWfffdp7Jly2rFihWqVKmSz3O7du1S69atdf/99+uXX36xqUIAAIDQMWvWLBmGoSJFivj8DAAAgMz8Gvj74osvNGrUKL344ouBrgf5yHN6NKdJw4NMwIpMAMHxxx9/6Mknn8w06CdJlSpV0ogRIzRu3LjgF4YcOZ1ONW3alDYSXmQCVmQCVmQiMAYOHJjjzygY2B5gRSZgRSaAwHD486KEhATVqFEj0LUAAAAUCklJSTp9+nS2z6enp2c5KAj7ZXWZexRuZAJWZAJWZCKwTpw4oYSEBD333HN2lwI/sD3AikzAikwAF8+vgb+hQ4fqnXfekdvtDnQ9yEeea9tzjXt4kAlYkQkgOMaMGaNXXnlFa9euzfTcmjVr9Oqrr3LGXwhyuVxat24dbSS8yASsyASsyETgFS1aVBERESpWrJjdpSCP2B5gRSZgRSaAwPDrUp+jR4/W6dOn1bRpU/Xr108VK1bM8vTb7t27X3SBAAAABd2oUaMyTUtMTNTll1+uK6+80nslhb/++ks//vijLr30Uq1atUq33XZbsEsFAAAIeT169ND777+vESNGcK8/AAAAC78G/vbs2aNvvvlGa9euzfKb6pJkGAYj8wAAAJJee+21bJ/7/vvv9f333/tM+9///qf169fr5Zdfzu/SAAAACpxevXpp5MiRateunYYMGaIqVaooJiYm03xNmjSxoToAAAB7+TXwN3jwYP3222969NFH1aJFC8XFxQW6LgAAgLDB5dHDBzeZhxWZgBWZgBWZCLy2bdt6/79y5cpMz5umyRfSQxTbA6zIBKzIBHDx/Br4++677/R///d/Gj9+fKDrQT7yNJo0nvAgE7AiEwCQvYiICDVr1szuMhBCyASsyASsyET+mD17tt0lwA9sD7AiE7AiE0Bg+DXwV65cOZUqVSrQtQAAAAAhyzRNpaamKi4ujvsJQRKZQGZkAlZkIn8MGDDA7hLgB7YHWJEJWJEJIDAc/rzogQce0MyZM/XPP/8Euh7kI88lLrjUBTzIBKzIBBA8n376qa677jolJCQoIiJCTqcz0z+EFpfLpU2bNtFGwotMwIpMwIpMAOewPcCKTMCKTACB4dcZf6dOnVKRIkVUo0YN3XLLLapUqVKmg1OGYei+++4LSJFTpkzRc889p/3796thw4Z69dVX1bx584C8NwAAQLAtXrxYt9xyi+rXr69evXpp2rRp6t27t0zT1EcffaSaNWuqW7duAVsefSkAABBuTp06pcWLF+u3335TampqpnsqG4ahN998MyDLoi8FAAAKEr8G/h588EHv/1977bUs5wnUwN+7776r+++/X9OnT1eLFi300ksvqUOHDtq8ebPKli170e8PAAAQbE899ZSaN2+u7777Tn///bemTZumwYMHq3379tqxY4euuOIKVa1aNSDLoi8FAADCTXJystq1a6cdO3YoPj5eqampKlWqlFJSUuRyuVS6dGkVL148IMuiLwUAAAoavy71uX379gv+27ZtW0AKnDx5soYMGaJBgwapXr16mj59uooWLapZs2YF5P0LE891kbk+MjzIBKzIBBAcGzZsUK9eveR0OhURcfZ7WGfOnJEkValSRSNHjtQzzzwTkGXRlwocwzAUExNDGwkvMgErMgErMpE/HnroIaWmpmrVqlX6888/ZZqm3n33Xf3zzz965plnFBMTo88//zwgy6IvFThsD7AiE7AiE0Bg+HXGX1JSUqDryFJ6erp+/fVXPfroo95pDodD1157rX788ceg1BBOHA6HzyNAJmBFJoDgKFq0qCIjIyVJ8fHxioqK0r59+7zPJyYmavv27Re9HPpSgeV0OtWwYUO7y0AIIROwIhOwIhP545tvvtHIkSPVvHlzHT16VJJkmqaioqL00EMPaePGjbr33nv13//+96KWQ18qsNgeYEUmYEUmgMDwa+AvWA4fPiyXy6XExESf6YmJidq0aVOWrzl9+rROnz7t/TktLU2SlJGRoYyMDElnO2kOh0Nut9vnGvCe6S6XS6ZpXnC60+mUYRje9z1/uqRMNyHNbnpERIRM0/SZbhiGnE5nphqzm56bdfLUmZGRIbfbHRbrFI6f08mTJ7Vz507vdMMw5Ha7M61TXmrPbvquXbskSdu2bfM+ZxiGHA6HTNPMtE5ZTffUmN10a+1VqlRR0aJFC/znFI7ZM03TW6tnWeGwTheaHsh1QsHjcrnkcrlylT1rRi5G7dq1tWHDBu/PjRo10ltvvaW+ffsqIyNDCxcuVOXKlS96OfSlAtumuFwuHTlyRAkJCYqIiAiLdQrHzymY6+RwOHTo0CGVLFnS+6WZgr5O4fg5BXOd3G63jhw5ojJlyigiIiIs1ulC01mnnNfJ5XLp4MGDSkhI8P6NVNDXyd/PKZB9qRMnTqhKlSqSpNjYWBmGodTUVO/zLVu29LlNjb/oS9GXKoyfE30pPif6UgX3c0LBk1/HpfxKg2maeuONN/Tmm29q27Zt+vvvvzPNk1VYg+Gpp57S+PHjM01fs2aNihUrJkkqU6aMqlevru3bt+vQoUPeeSpWrKiKFSvqzz//9OkwVqtWTWXLltX69et18uRJ7/Q6deooPj5ea9as8dnQGjRooMjISK1evdqnhqZNmyo9PV3r1q3zTnM6nWrWrJlSU1N9Oo0xMTFq2LChDh8+7HPZ1Li4ONWtW1d79+7V7t27vdNzs04bN26UJG3cuFFlypQJi3UKx8/pjz/+0CuvvKJgeuqpp4K2rGeffVZXXHFFgf+cwjF7qamp3td42vVwWCePYHxO//zzj1CwbNiwQWfOnMlV9o4fPx6w5d5000165ZVX9PzzzysqKkqPP/64/vWvfyk+Pl6GYej48eO2XT6KvlT265SSkqKUlBTFx8erevXqYbFO4fg5BXOdGjdurD///FMRERHeyxEV9HUKx88pmOtkmqZSUlLUqFEjlS9fPizWKRw/p2CuU0pKitauXevdx4fDOvn7OQWyL1W5cmXv7ykiIkKXXHKJVq1ape7du0s628eLjo4O2PLygr4UfSnWib5UuH1O9KUKzufEcamCJ7+OSxnm+cPLufTQQw9p8uTJatSoka6++mqVLFkyy/nGjh2b17f2kZ6erqJFi+r9999Xt27dvNMHDBiglJQUffTRR5lek9U3qypVqqQjR44oNjZWUmiOxgfjGwYbN27UyJEjNXXqVNWrVy8s1ikcP6dgnvEnnW1c6tSp4/PNKoeDM/4KY/ZM09SmTZs0cuRITZs2TfXr1w+LdbrQ9ECu0+bNmzVkyBCNkFRBXI8+lO2VqWmSpk+frtq1a+cqe2lpaUpISFBqaqq3TxFIK1eu1AcffCCn06nOnTurXbt2F/2e9KUCf/WE3377TU2aNFGRIkXCYp3C8XMK5jpJ0i+//KImTZp45yno6xSOn1Mw18nlcum3337T5ZdfrsjIyLBYpwtNZ51yXqczZ85o9erV3nYiHNbJ388pkH2pkSNHatWqVfrtt98kSWPGjNFTTz2lgQMHyu1266233lL//v01c+bMi1oOfSn6UoXxc6IvxedEX6rgfk4clyo48vu4lF9n/M2dO1c9evTQe++958/Lcy0yMlKXX365vv76a28Hy+126+uvv9Zdd92V5WuioqIUFRWVaXpERESm0109G5uVZ+PJ7fTsTqPNy3TDMLKcnl2NeZ3udDq908+fp6CvU1YK+joVL15c9erVy7KmQMvIyFBaWprq1KkT9NPBC/rnlJVwWCfPc55v2oXDOuV2eiDWCQWP0+n0yX9OGcvvz/zqq6/W1VdfHdD3pC8V2DbFNE1vx5u+lH/Tw22dMjIyvJnI7fYR6uskhd/nJAV3nTxfmLNO97f27KbzORWcdcqqnSjo6+TP5xTIvtQjjzyiX375RadPn1ZUVJQee+wx7d27V++//76cTqd69+6tyZMnX/Ry6EvRl8pperh+TlmhL1W4PyeJvpQ/0zkuhdzIr+NSfqXj5MmTuvbaa/15aZ7df//9GjBggJo2barmzZvrpZde0vHjxzVo0KCgLD+ceA7kex4BwzAUFxdHJuBFOwEEh9Pp1FtvvaXevXtn+fy7776r3r17Z/mN2LyiLxU47DdhRSZgRSZgRSbyR+XKlX3uhxwdHa2ZM2de9Bl+WaEvFThsD7AiE7AiE0Bg+DXwd8011+iXX37R0KFDA11PJrfeeqsOHTqkMWPGaP/+/WrUqJE+++yzTDdWxoV5RouzGjVG4eR0OlW3bl27y0AIoZ0AguNCV1p3uVwB+0OHvlTgsN+EFZmAFZmAFZko+OhLBQ7bA6zIBKzIBBAYfg38TZ06VR06dNCkSZM0bNgwJSQkBLouH3fddVe2l1BA7nkOMvpxW0eEKbfbrb1796pChQoM9EAS7QQQTNkN7KWlpenzzz9X6dKlA7Ys+lKBwX4TVmQCVmQCVmQiMCZMmJDn1xiGodGjRwdk+fSlAoPtAVZkAlZkAggMvwb+ateuLbfbrdGjR2v06NGKjo7OdG1awzCUmpoakCIRGJ4bQ55/g0gUbm63W7t371a5cuXYmUIS7QSQn8aPH+89aGUYhvr27au+fftmOa9pmho1alQwy0MusN+EFZmAFZmAFZkIjHHjxuX5NYEc+ENgsD3AikzAikwAgeHXwF+PHj24zi4AAEAeNG/eXCNHjpRpmpo6daquu+461apVy2cewzBUrFgxXX755erevbtNlQIAAIQWvpgIAACQe34N/M2ZMyfAZQAAAIS3Tp06qVOnTpKk48ePa/jw4WrRooXNVQEAAAAAACCc+DXwh4LJc3o0p0nDw+FwqEyZMmQCXrQTQHDMnj3b7hLgB/absCITsCITsCITgeN2u/Xss8+qQoUK6t+/f7bzzZs3T/v27dP//d//BbE65AbbA6zIBKzIBBAYfg38zZs3L8fnDcNQdHS0KlasqCZNmigqKsqv4hBYnsuzcplWeDgcDlWvXt3uMhBCaCeA/OHpO/Xr10+GYVywL+WR00EtBB/7TViRCViRCViRicCZN2+ennjiCf388885zle/fn0NHjxYFStWVJ8+fYJUHXKD7QFWZAJWZAIIDL8G/gYOHOg9KGyaps9z5083DEOxsbF69NFH9fDDD19kqbhYns/K+pmh8HK73dq+fbuqVq3KN2kgiXYCyC+evlOvXr0UGRmpgQMHXvA1hmEw8Bdi2G/CikzAikzAikwEzoIFC9S5c2c1adIkx/kuv/xyde3aVXPnzmXgL8SwPcCKTMCKTACB4dfWs3btWjVo0EDt2rXT4sWL9fvvv+v333/X+++/r7Zt26pRo0b6/vvvtXjxYjVp0kSPPvqopk2bFujakUeem2FzU2x4uN1uHTp0iEzAi3YCyB/bt2/Xtm3bFBkZ6f35Qv+2bdtmc9WwYr8JKzIBKzIBKzIROL/99puuueaaXM3btm1b/fbbb/lcEfKK7QFWZAJWZAIIDL/O+HvxxReVmJiozz77zGf6ZZddpptuukmdOnXSm2++qZkzZ6pr1666+uqrNXXqVI0YMSIgRQMAABQkSUlJOf4MAACAnB0/flwlSpTI1bwlSpTQP//8k88VAQAAhCa/zvhbsmSJ/vWvf2X5nGEY6tq1qz744IOzC3A41KNHD23ZssX/KgEAAAAAAFBolS1bVn/99Veu5v3rr79UpkyZfK4IAAAgNPl1xp/b7dbmzZuzfX7Tpk0+p+NGRUUpOjran0UhgDzXReb6yPBwOByqWLEimYAX7QQQHO3bt7/gPIZh6Ouvvw5CNcgt9puwIhOwIhOwIhOB07p1a7311lt64oknVLRo0WznO378uN566y21bds2eMUhV9geYEUmYEUmgMDwawvq2rWrpk6dqtdee02nTp3yTj916pReffVVTZ8+XTfeeKN3+o8//qgaNWpcfLW4KIZh+DwC7ExhRTsBBIfb7ZZpmj7/MjIytHXrVi1fvly7d+/mngYhiP0mrMgErMgErMhE4Dz44IPav3+/brjhBu3ZsyfLefbs2aMbb7xR+/fv1wMPPBDkCnEhbA+wIhOwIhNAYPh1xt/LL7+srVu3atSoUXrwwQdVvnx5SdK+ffuUnp6u5s2b6+WXX5Z0djAwJiZG999/f+Cqhl88BxA5kAgPl8ulP//8U7Vq1ZLT6bS7HIQA2gkgOJYvX57tc//5z380dOhQTZ48OXgFIVfYb8KKTMCKTMCKTAROo0aNNG3aNI0YMULVqlVT69atddlll6lEiRI6duyY/ve//+nbb7+V2+3WlClT1KhRI7tLhgXbA6zIBKzIBBAYfg38lSpVSt9//70+/PBDff7550pOTpYkXX/99erQoYO6devmHZWPjo7WjBkzAlcx/Gaaps8jYJqmUlNTyQS8aCcA+3Xp0kV9+/bVvffeqxUrVthdDs7DfhNWZAJWZAJWZCKw7rjjDl166aUaP368vvnmG5/LokdERKh9+/YaO3asWrZsaWOVyA7bA6zIBKzIBBAYfg38SWcvA9e9e3d17949kPUAAAAUetWrV9drr71mdxkAAAAh54orrtCnn36qkydPasuWLUpLS1NsbKxq1KihmJgYu8sDAACwnd8DfwAAAAi8jIwMvffeeypdurTdpQAAAISsmJgYXXbZZXaXAQAAEHJyNfBXtWpVORwObdq0SUWKFFHVqlVlGEaOrzEMQ1u3bg1IkQgMz+VXuTkqPBwOh6pVq0Ym4EU7ERiHJElcliKUHbJ5+YMHD85yekpKilatWqX9+/dzj78QxH4TVmQCVmQCVmQCOIftAVZkAlZkAgiMXA38tWnTRoZheDc4z88oWDyfGZ8dPBwOh8qWLWt3GQghtBMXJz4+XlGRkXo/Pd3uUpALUZGRio+Pt2XZ33zzTabtzDAMlSxZUq1atdIdd9yh66+/3pbakD32m7AiE7AiE7AiE8A5bA+wIhOwIhNAYORq4G/OnDk5/oyCwe12+zwCLpdL69ev16WXXiqn02l3OQgBtBMXJzExUfMXLFBKSordpQTMjh079O9//1uPP/64qlSpYnc5ARUfH6/ExERblr1jxw5blouLw34TVmQCVmQCVmQCOIftAVZkAlZkAggM7vFXiJim6fMImKapkydPkgl40U5cvMTERNsGk/KDy+WSJFWqVEm1a9e2uRrAXuw3YUUmYEUmYEUmgHPYHmBFJmBFJoDAyNXA37fffuvXm7du3dqv1wEAAISbEydO6PDhwypXrpwiIyN9nps1a5YWLFigffv2qU6dOnr00UfVrFkzmyoFAAAAAABAQZWrgb+2bdvm6X5PpmnKMAzvWQIAAACF3YQJEzR9+nTt3r3bZ+Bv4sSJGjt2rPcef5s2bdLnn3+uH374QQ0bNrSxYgAAgNB27NgxJScn6++//87y7BC+kA4AAAqjXA38LVu2LL/rQBB4rovM9ZHh4XQ6VadOHTIBL9oJWJGJwFm2bJm6dOmi4sWLe6elpaVp4sSJuuSSS7RixQpVrVpVP//8szp06KCnn35ab7/9to0Vw4r9JqzIBKzIBKzIRP44cuSI7rrrLi1evDjLL53zhfTQxPYAKzIBKzIBBEauBv7atGmT33UAsIFhGIqPj7e7DAAoFHbs2KEePXr4TPvkk0+Unp6u//u//1PVqlUlSc2bN9egQYP03nvv2VEmcsB+E1ZkAlZkAlZkIn8MGTJES5cu1ahRo3T11VerZMmSdpeEXGB7gBWZgBWZAAIjVwN/ebF//3698847WrBggX755ZdAvz0uguebbnzjDR4ZGRlas2aNGjdurIiIgDcHKIBoJ2BFJgLn2LFjSkhI8Jn27bffyjAMdejQwWd6vXr1dOjQoWCWh1xgvwkrMgErMgErMpE/vvjiC91333169tln7S4FecD2ACsyASsyAQRGQLaef/75R4sXL9aCBQu0bNkyuVwuVatWLRBvDSCfcTAfAIIjKSlJmzZt8pm2fPlyJSYmqkaNGj7T09PTFRsbG8zykEvsN2FFJmBFJmBFJgKvaNGiqlKlit1lwA9sD7AiE7AiE8DFc/j7woyMDH388ce69dZblZiYqMGDB+vgwYOaMGGC/ve//2nLli2BrBMAAKBAu/766zVr1iz99NNPkqR58+Zp06ZNuummmzLN++uvv3IwCwAAIBt9+/bVhx9+aHcZAAAAISnPZ/ytXLlSCxYs0Pvvv6+jR4/q8ssv18iRI/XCCy9ozJgx6t69e37UCQAAUKCNHj1aS5Ys0ZVXXimn06mMjAyVKVNGY8aM8ZnvxIkT+vDDDzV06FCbKgUAAAhtPXv21IoVK9SxY0cNHTpUlSpVktPpzDRfkyZNbKgOAADAXrke+Hv00Uf19ttva+fOnapbt67uvfde9erVSzVq1NDWrVv1wgsv5GedCABPJzirzjAKJ6fTqQYNGpAJeNFOwIpMBE7p0qW1du1azZw5U9u2bVNSUpIGDx6ssmXL+sy3fv169enTR/369bOpUmSH/SasyASsyASsyET+aNWqlff/X375ZabnTdOUYRhcLi7EsD3AikzAikwAgZHrgb9nnnlGVatW1bJly9SmTZv8rAlAEEVGRtpdAgAUGiVLltRDDz2U4zzNmzdX8+bNg1QR8or9JqzIBKzIBKzIRODNnj3b7hLgJ7YHWJEJWJEJ4OLleuCvZ8+e+s9//qMOHTro+uuv12233aauXbuqWLFi+VkfAsjzTTe+8QYPl8ul1atXq2nTpoqIyPOVfxGGaCdgRSaAc9hvwopMwIpMwIpM5I8BAwbYXQL8wPYAKzIBKzIBBIYjtzO+9957OnDggKZNm6aTJ0+qX79+SkxMVK9evfTf//5XhmHkZ50AAAAAAAAAAAAAcpCnYfMSJUpo0KBBGjRokPbt26eFCxdq4cKFeu+99yRJU6dO1alTp3TDDTcoPj4+P+oFAAAAAABAIXfq1CktXrxYv/32m1JTU+V2u32eNwxDb775pk3VAQAA2Mfv82XLl/9/7d13fJPl/sbx60lKoYwW2TIsQ5ChDEWUoaCCco4bBJkOUFFcBxThgICigqhHBBVU3AsVUfCoKHpAxYGyQdmrbGS2IKNtcv/+4JeE3hRsS9onTT7v16uvypMA34dcd3vZO3lyuh544AE98MADWrFihd555x1NmjRJPXr0UFxcnFq2bKlZs2aFc1YAAAAAAADEuJSUFF1yySXasGGDSpcurdTUVJUpU0b79u2Tz+dTuXLlVLJkSbfHBAAAcEWOL/V5MnXr1tUTTzyhdevW6YcfflDv3r31+++/h+OPRhh5vd4snwGv16umTZuSCQTxdQI2MgGE8H0TNjIBG5mAjUzkjwEDBig1NVVz5szRqlWrZIzRhx9+qAMHDmj06NFKSEjQ119/7faYsLAeYCMTsJEJIDzCsvF3rFatWmnChAnatm1buP9oAPkgPT3d7REAIOb06tVLv/766wlv/+2339SrV68CnAg5xfdN2MgEbGQCNjIRfjNnzlTfvn3VrFkzeTxHf7RljFHRokU1YMAAXXbZZfrXv/7l7pDIFusBNjIBG5kATl3YN/4C4uLyfBVR5BOfz5flM+Dz+bRkyRIygSC+TsBGJvLHm2++qbVr157w9vXr1+utt94qwImQE3zfhI1MwEYmYCMT+ePgwYOqXr26JCkxMVGO4yg1NTV4e/PmzfXjjz+6NB1OhPUAG5mAjUwA4ZFvG38AAADIm61btyohIcHtMQAAACLSGWecoc2bN0s6+sTzKlWqaM6cOcHbly1bpmLFirk1HgAAgKt4WR4AAEABmDZtmqZNmxb89SuvvKJvv/32uPvt27dP3377rc4///yCHA8AAKDQuPTSSzVt2jQNHz5cknTLLbdo1KhR2rt3r/x+v9555x3ddNNNLk8JAADgDjb+XHb48GGlpKQUyN+1cePG4OeCeIPU5ORknmFXCPBmuQBQMJYtW6bJkydLkhzH0a+//qr58+dnuY/jOCpRooQuvvhiPfvss26Mib/B903YyARsZAI2MhF+gwYN0ty5c3XkyBEVLVpUgwcP1tatW/Xxxx/L6/WqW7dudKkIxXqAjUzARiZO3U5JknF5CpzMznz+8x1jTFQnIC0tTUlJSUpNTVViYqLb4xxn5cqVuv32290eI19MnDhRZ511lttjAMiFwNck1i8CyERIODuFx+PRu+++q27duoVpuvwT6V0KAAAUDrHaKWL1vAEABW/Hjh3q0b27jqSnuz0KcqBofLzefe89VaxYMUf3z02nOOVX/Pn9fqWmpiq7/cMyZcqc6h8f9ZKTkzVx4sQC+/sOHDigkiVLFsjflZycXCB/D/LOGKPU1FQlJSXJcRy3xwGAmOH3+90eAXnA903YyARsZAI2MgGEsB5gIxOwkYlTU7FiRb373nvat2+f26OETUpKih5//HE9/PDDUbffULp06Rxv+uVWnjb+MjIyNHr0aL3++uvatGnTCX945fP5Tmm4WFCsWLECewVFZmam5s2bp1q1aikujqu84ugaXbFihZo2bUomICn0dZuv3wggE/ljwYIFmjNnjvr27Zvt7ePHj1eLFi3UuHHjgh0MJ8X3TdjIBGxkAjYyER4jRoyQ4zgaMmSIPB6PRowY8be/x3EcDR06tACmQ06xHmAjE7CRiVNXsWLFfNtMckPg51FVq1aN+StR5UaeVk+fPn301ltv6cILL9R1112npKSkcM8FAAAQtYYMGaKEhIQTbvzNnDlTX375pT7//PMCngwAACDyPPLII3IcRwMHDlR8fLweeeSRv/09bPwBAIBYlaeNv8mTJ6tnz5568803wzwOAABA9Js/f77+/e9/n/D2iy66SKNGjSrAiQAAACKXfaUpLpsOAABwYp68/KbixYvrwgsvDPcsyGeO4yghIYHrIyOITMAWyAKZQACZyB/79+8/6WVLPB6PUlNTC3Ai5ATfN2EjE7CRCdjIBBDCeoCNTMBGJmDj51J5k6eNv65du3LpqULI6/WqUaNG8nq9bo+CCEEmYPN4PFk+A2Qif9SuXVszZsw44e1fffWVatasWYATISf4vgkbmYCNTMBGJgrOwYMH9frrr2vChAlKSUlxexxkg/UAG5mAjUzAxs+l8iZP/1pPPfWUSpcurauuukqffPKJ5s6dqwULFhz3gcji9/v1559/ckkMBJEJ2IwxWT4DZCJ/9O7dW1988YX69++vffv2BY/v27dP/fr101dffaXevXu7NyCyxfdN2MgEbGQCNjKRP3r37q2zzz47+Ov09HRdeOGFuu2223T33XercePGWrhwoYsTIjusB9jIBGxkAjZ+LpU3eXqPvyNHjsjv92v69OmaPn36cbcbY+Q4jnw+3ykPiPDx+/1at26dypQpww45JJEJHC9QrChYCCAT+eO+++7TokWL9Nxzz2ncuHGqXLmyJGnr1q3y+/3q2bOn+vXr5/KUsPF9EzYyARuZgI1M5I9Zs2apR48ewV+///77+v333/Xee++pUaNG6tixox599FFNnTrVvSFxHNYDbGQCNjIBGz+Xyps8bfz16tVLn376qbp06aILLrhASUlJ4Z4LAAAgajmOozfeeEM33XSTpkyZonXr1kmSrr32WnXs2FFt2rRxd0AAAIAItn37dlWvXj3466lTp6pp06bq2rWrJOn222/X008/7dJ0AAAA7srTxt/XX3+te++9V2PGjAn3PAAAADHjkksu0SWXXOL2GAAAAIVKiRIlgpdLz8zM1Hfffad77703eHupUqWUmprq0nQAAADuytPGX2Jios4888xwz4J85jiOkpKS5DiO26MgQpAJ2AJZIBMIIBP5a8uWLfrhhx/0559/qmPHjqpatap8Pp9SU1OVlJTEG5pHGL5vwkYmYCMTsJGJ/HHuuedq4sSJuuSSS/TZZ59p//79uvrqq4O3r127VhUrVnRxQmSH9QAbmYCNTMDGz6XyJk8Xyr399ts1adIk3sOvkPF6vapXrx4/REQQmYAtcP10rqOOADKRP4wx6t+/v2rUqKHu3burf//+WrVqlSTpwIEDql69up5//nmXp4SN75uwkQnYyARsZCJ/PPHEE9qxY4eaNm2qRx99VB07dlSzZs2Ct3/66adq2bKlixMiO6wH2MgEbGQCNn4ulTd5esVf/fr1NW3aNJ177rm6+eabVa1atWwXY4cOHU55QISP3+/X1q1bVblyZRYKJJEJHM8Yk+UzQCbyx9NPP62xY8dq4MCBuuyyy9SuXbvgbUlJSerQoYOmTJmif/3rX+4NiePwfRM2MgEbmYCNTOSPpk2bauXKlfr5559VunRptW7dOnjbvn371Ldv3yzHEBlYD7CRCdjIBGz8XCpv8rTxd+ONNwb/+8EHH8z2Po7j8IrACOP3+7V582ZVqlSJL5yQRCZwPL/fn+UzQCbyx8SJE3XTTTdp5MiR2r1793G3N2zYUNOnT3dhMpwM3zdhIxOwkQnYyET4HTp0SEOGDNEll1yia6+99rjbS5curfvvv9+FyfB3WA+wkQnYyARs/Fwqb/K08Tdr1qxwzwEAABAzNm3apBYtWpzw9hIlSigtLa0AJwIAACgcEhIS9PLLL6t+/fpujwIAABCR8rTxx+USAAAA8q5ChQratGnTCW+fP3++zjjjjAKcCAAAoPA477zz9Pvvv7s9BgAAQEQ65dfLLlu2TNOnT9f06dO1bNmycMyEfOLxeFS+fHleJo0gMgEbb5gLG5nIHx06dNBLL72kdevWBY85jiNJmjFjht5880116tTJrfFwAnzfhI1MwEYmYCMT+eO5557TBx98oFdffVWZmZluj4McYj3ARiZgIxOw8XOpvMnTK/4kadq0aerfv782bNiQ5XiNGjX07LPP6pprrjnV2RBGPp9PS5Ys0e7du5WWlqaGDRvK6/W6PRZclJ6erqlTpwbfMPe6665TfHy822PBZYGNh8BngEzkj0cffVSzZs1S48aNddFFF8lxHI0ePVpDhw7VL7/8oiZNmmjw4MFujwmLx+NRrVq13B4DEYRMwEYmYCMT+eOWW26Rx+NRnz59dN9996lKlSpKSEjIch/HcbR48WKXJkR2WA+wkQnYyARs/Fwqb/K08ffll1+qY8eOSk5O1siRI1WvXj1J0vLly/XKK6+oQ4cO+vzzz9W+ffuwDou8+f777/Xiiy9q+/btwWOVKlXS3XffzWVbY9T48eM1efJk+Xy+4LEJEyaoU6dO6tu3r4uTwW3GmCyfATKRP5KSkjRnzhz95z//0ccff6xixYrp+++/V61atTR8+HANGDDguB9ewX1+v1/r169XjRo1eLYhJJEJHI9MwEYm8keZMmVUtmxZnXXWWW6PglxgPcBGJmAjE7Dxc6m8ydPG32OPPaaGDRtq9uzZKlGiRPD4Nddco3vuuUetWrXSo48+ysZfBPj+++81bNgwNW/eXA8//LD27t2r0047Te+//76GDRumESNGsPkXY8aPH68PPvhAp512mnr16qWEhAQdOnRIr7/+uj744ANJYvMvhvn9/iyfATKRfxISEvTwww/r4YcfdnsU5JDf79fOnTuVnJzM/4RCEpnA8cgEbGQif3z33Xduj4A8YD3ARiZgIxOw8XOpvMnT6lmyZIluvvnmLJt+ASVKlNAtt9yiJUuWnPJwODU+n08vvviimjdvrpEjR6p+/foqWrSo6tevr5EjR6p58+YaP358lld9Ibqlp6dr8uTJOu200zRlyhRdeeWVSkxM1JVXXqkpU6botNNO0+TJk5Wenu72qAAAAAAAAAAAIJfy9Iq/YsWKac+ePSe8fc+ePSpWrFieh0J4LFmyRNu3b9fw4cPl8Xiy7Ip7PB716NFDffv21ZIlS9SkSRMXJ0VBmTp1qnw+n2677TbFxcVleRP0uLg49e7dW88884ymTp2qzp07uzgpjnX48GGlpKQUyN+1cePG4OeCeh/Q5ORkvmfkQUHlgkyER69eveQ4jl555RV5vV716tUrR78vLi5O5cqV02WXXabLLrssn6cEAAAoPNLS0jR+/HjNmjVLf/75p15++WU1a9ZMe/bs0ZtvvqlrrrlGZ555pttjAgAAFLg8bfxdeumlGjt2rNq3b6/mzZtnue3XX3/VuHHjdPnll4dlQOTd7t27JUk1atSQdHSzr2rVqsGXSdesWTPL/RD9tm7dKklq0aKFpOMzETgeuB8iQ0pKim6//fYC/TtHjhxZYH/XxIkTeW+OPCjoXJCJUzNz5szgk3C8Xq9mzpyZozem9vl82r17t0aPHq2RI0dq4MCBBTAtTsT+vgmQCdjIBGxkIn9s3rxZrVu31qZNm1S7dm2tWLFCBw4ckHT0/f9efvllpaSkaOzYsS5PimOxHmAjE7CRCdgCWSATuZOnjb+nnnpKzZs3V6tWrdSsWbPgD+dWrlyp3377TRUqVNDo0aPDOihyr2zZspKk9evXq0GDBsEvnAHr1q3Lcj9Ev8qVK0uSfv75Z1199dXHZeLnn3/Ocj9EhuTkZE2cONHtMfJNcnKy2yMUStGci2jMxIYNG07665PJzMzU7bffrvHjx7Px5zL7+yZAJmAjE7CRifwxYMAA7d+/X4sWLVKFChVUoUKFLLdfd911+vzzz12aDifCeoCNTMBGJmALPGk6J0+eRkieNv5q1KihJUuWaNSoUZo+fbo+/PBDSUd/UHf//fdr0KBBx5UuFLyGDRuqUqVKeueddzRy5EgZY7Rq1SrVqVNHjuPo3Xff1emnn66GDRu6PSoKyHXXXacJEybo1Vdf1T/+8Q85jhPMhDFGr732mrxer6677jq3R8UxihUrVmCvfvL5fMFMFNRlHZE3BZULMuG+uLg4de3atcAu+YsTYz3ARiZgIxOwkYn8MWPGDPXr10/169fP9ipGNWvW1KZNm1yYDCfDeoCNTMBGJmALvH3ZsW9jhr+Xp40/SapQoYLGjBmjMWPGhHMehJHX69Xdd9+tYcOGafDgweratav27t2r9PR0TZo0Sb/88otGjBjBF9EYEh8fr06dOumDDz5Qx44ddeuttyohIUErV67UG2+8ob1796pLly6Kj493e1S4xBij1NRUGWPcHgURgkzkr99//11ffvll8BWA1atX1z/+8Q+dc845We53+eWXcxn1CMB6gI1MwEYmYCMT+ePQoUMqX778CW/fv39/AU6DnGI9wEYmYCMTsAWyQCZyJ88bfwEHDhwIPouqWrVqKlmy5CkPhfBp3bq1RowYoRdffFH33ntv8Pjpp5+uESNGqHXr1i5OBzf07dtXkjR58mQ9++yzweNer1ddunQJ3g4AyD9HjhxRnz599M4778gYE7xWvd/v17///W91795dr776Kk/EAAAAyEb9+vX1ww8/qE+fPtnePnXqVDVp0qSApwIAAIgMed74mzt3rh566CH9+OOPwZdZejweXXTRRXrqqafUtGnTsA2JU9O6dWu1atVKCxcu1Ny5c3X++eerSZMmvNIvhvXt21e33XabPvnkEy1cuFBNmjRRhw4d+AEzABSQgQMH6u2331bfvn117733qlatWnIcR2vWrNG4ceM0YcIElSlTRs8995zbowIAAEScf/3rX7r55pvVsGFDderUSdLRJ1CtWbNGjz76qH755RdNmTLF5SkBAADc4Zg8vEby119/VZs2bRQfH69u3bqpXr16kqTly5dr0qRJSk9P13fffadmzZqFfeDcSktLU1JSklJTU5WYmOj2OK7y+/3atWuXypUrF3xlAWIbmYCNTMBGJkLC2SnKlSunK6+8Um+99Va2t/fs2VPTp0/Xrl27TunvCQe6VAjrATYyARuZgI1MhIS7UzzxxBN65JFHZIyR3++Xx+MJXknh8ccf18CBA8Mw9amjS4WwHmAjE7CRCdhWrFihO+64Q6+88orq1q3r9jiuyk2nyNMr/oYMGaIqVaroxx9/VKVKlbLc9sgjj6hly5YaMmSIvvnmm7z88cgnHo9HFSpUcHsMRBAyARuZgI1M5I+MjAxdeOGFJ7y9RYsW+u9//1uAEyEnWA+wkQnYyARsZCL/DBkyRD179tSUKVO0Zs0a+f1+1apVSx06dFDNmjXdHg/ZYD3ARiZgIxOwOY6T5TNyJk/b5r/++qv69Olz3KafJFWsWFF33HGH5syZc8rDIbx8Pp8WL14sn8/n9iiIEGQCNjIBG5nIH1dccYW+/vrrE97+1Vdf6fLLLy/AiZATrAfYyARsZAI2MhFehw8f1ocffqgnn3xSEydOVJEiRdSvXz+9+OKLmjBhgh588EE2/SIY6wE2MgEbmYAt8DZzgc/ImTy94s/j8SgzM/OEt/t8Pl6KG4GMMTp06JDycHVXRCkyARuZgI1MhMeePXuy/Pqxxx5T586d1aFDB919990688wzJUmrV6/Wiy++qJSUFH344YdujIqTYD3ARiZgIxOwkYnw+fPPP9WiRQutX78++O9ZvHhxTZ06VW3btnV5OuQE6wE2MgEbmYAtkAUykTt52vhr0aKFXnzxRXXr1k3JyclZbtu4caPGjx+vli1bhmVAAACAwq5cuXLHXZbCGKOlS5dq2rRpxx2XpAYNGpz0iVYAAACx5LHHHtOGDRvUr18/XXrppVqzZo0ee+wx9enTR2vXrnV7PAAAgIiRp42/kSNH6uKLL1bdunV1/fXXq06dOpKklStXatq0aYqLi9OoUaPCOigAAEBhNWzYMK5HDwAAcApmzJihm266Sc8880zwWMWKFdWtWzetXLlSZ511lovTAQAARI48bfw1adJEv/76q4YMGaLPPvtMBw8elHT0Egvt27fX448/rvr164d1UJw6r9erunXryuv1uj0KIgSZgI1MwEYmwuORRx5xewSEAesBNjIBG5mAjUyEz8aNGzVw4MAsx1q1aiVjjHbs2MHGXyHAeoCNTMBGJmALZIFM5E6eNv4kqX79+vr000/l9/u1c+dOSVL58uV5b78I5jiOSpcu7fYYiCBkAjYyARuZAEJYD7CRCdjIBGxkInyOHDmiYsWKZTkW+DWXRy8cWA+wkQnYyAQQHnne+AvweDyqWLFiOGZBPsvMzNTChQvVpEkTxcWd8kOPKEAmYCMTsJGJ/DFixIi/vY/jOBo6dGgBTIOcYj3ARiZgIxOwkYnw2rBhgxYsWBD8dWpqqiRp9erV2f6g+Nxzzy2o0ZADrAfYyARsZAI2n8+X5TNyJkerJyc/nLLxw6rIxAKBjUzARiZgIxPhd7JLfzqOI2MMXSpCsR5gIxOwkQnYyET4DB06NNt+1Ldv3yy/DnQp/u0jD48JbGQCNjIBnLocbfzl5X1p+GEVAABA9vx+f7bHUlJS9OKLL+qHH37Q9OnTXZgMAAAgMr3xxhtujwAAAFAo5GjjL7sfTgEAACB8PB6PatSooWeeeUbdu3fXvffeq/fff9/tsQAAACLCzTff7PYIAAAAhYLH7QFQMNLT0zVlyhTNnj1bU6ZMUXp6utsjIQJ4vV41bNhQXq/X7VEQIcgEbGTCHRdffLG+/PJLt8eAhfUAG5mAjUzARiaAENYDbGQCNjIBWyALZCJ3wvIOmZmZmVq9erUOHDigevXqqWTJkuH4YxEm48eP1+TJk7NcH/mll15Sp06djrsOPmJPfHy82yMgwpAJ2MhEwZs3b548Hp6fFYlYD7CRCdjIBGxkAghhPcBGJmAjE8Cpy9XG35dffqlJkyYpLi5OPXv21KWXXqqpU6fqnnvu0bZt2yQdXZgPPPCAHn/88XwZGLkzfvx4ffDBBzrttNPUq1cvJSQk6NChQ3r99df1wQcfSDr+TbARO3w+n+bNm6emTZsqLi4szwNAIUcmYCMT+ePtt9/O9vi+ffv0ww8/6JNPPtFtt91WwFPh77AeYCMTsJEJ2MgEEMJ6gI1MwEYmYAu8mOnYFzXh7+V49Xz11Ve66qqrVKRIESUkJOjdd9/V66+/rt69e6t+/frq1KmTMjMz9fXXX2vUqFFKTk7W7bffnp+z42+kp6dr8uTJOu200zRlyhRJR19BcOmll+rKK69Ux44dNXnyZN122208kwIAgAJ0yy23nPC2cuXKadCgQRo2bFjBDQQAAAAAAICokOONv6eeekpnn322fvjhB5UuXVp33nmn+vTpo3bt2unzzz+X4ziSjl7288ILL9RLL73Exp/Lpk6dKp/Pp9tuu01xcXHKzMwM3hYXF6fevXvrmWee0dSpU9W5c2cXJwUAILasX7/+uGOO4+i0005TqVKlXJgIAAAAAAAA0SDHbx7zxx9/6JZbblHp0qUlSffdd58OHz6sHj16BDf9pKMbSt27d9eKFSvCPixyZ+vWrZKkFi1aZHt74HjgfgAAoGAkJycf93HGGWew6QcAAAAAAIBTkuNX/O3cuVMVK1YM/rpChQqSlOXYsbcdPnw4DOPhVFSuXFmS9PPPP+vqq6+W1+tV06ZN5fV6g8ePvR9ij50JgEzARiYKxooVKzR58mRt27ZNZ511lm699VYlJia6PRYsrAfYyARsZAI2MgGEsB5gIxOwkQnYAlkgE7mT41f8Scryyr5j/xuR6brrrpPX69Wrr74avMxnenq6pKOXZH3ttdfk9Xp13XXXuTgl3BbIBBBAJmAjE+HxwgsvqE6dOtq1a1eW4//973/VuHFjDR8+XC+99JL69eunc88997j7ITKwHmAjE7CRCdjIBBDCeoCNTMBGJoBTl6uNvw0bNmjBggVasGCBlixZIklavXp18FjgI7v3rUHBi4+PV6dOnbR371517NhR06ZN0+zZszVt2jR17NhRe/fuVadOnRQfH+/2qHCJz+fTkiVL5PP53B4FEYJMwEYmwuezzz5TrVq1VK5cueCxzMxM3XbbbfJ6vXrjjTe0dOlSPfnkk0pJSdETTzzh4rTIDusBNjIBG5mAjUwAIawH2MgEbGQCtkAWyETu5PhSn5I0dOhQDR06NMuxvn37Hnc/YwyvCIwQgcdn8uTJevbZZ4PHvV6vunTpku3jBwAAwm/ZsmW6/fbbsxybNWuWdu7cqcGDB+vmm2+WJDVo0ECLFy/Wl19+qTFjxrgxKgAAAAAAAAqpHG/8vfHGG/k5B/JR3759ddttt+mTTz7RwoUL1aRJE3Xo0IFX+gEAUIB2796tatWqZTn2v//9T47j6Prrr89yvGXLlvrkk08KcjwAAAAAAABEgRxv/AWehY7CKT4+XjfccINq1aqlJk2aKC4uVy/2RBTjjVFhIxOwkYnwqFixorZv357l2OzZs1W8eHE1atQoy/H4+HieoBOhWA+wkQnYyARsZAIIYT3ARiZgIxPAqcvVe/yhcIuLi9P555/Pph+CyARsZAI2MhE+TZs21VtvvaX9+/dLkv744w/99ttvuuKKK477912xYoWqVq3qxpg4CdYDbGQCNjIBG5kAQlgPsJEJ2MgEbIGNYDaEc4eNvxhijNG+fftkjHF7FEQIMgEbmYCNTITP8OHDlZKSotq1a+uyyy5Ty5Yt5TiO/v3vfx93308//VQtWrRwYUqcDOsBNjIBG5mAjUwAIawH2MgEbGQCCA82/mKIz+fTihUr5PP53B4FEYJMwEYmYCMT4XPOOedo5syZOu+887R161ZdeOGF+vLLL3Xeeedlud93332n4sWLq1OnTi5NihNhPcBGJmAjE7CRCSCE9QAbmYCNTMAWyAKZyB1eMwsAAFBAWrRooS+++OKk92nTpo2WLl1aQBMBAAAAAAAgmvCKPwAAAAAAAAAAACAKsPEXQxzHUUJCghzHcXsURAgyARuZgI1MACGsB9jIBGxkAjYyAYSwHmAjE7CRCdgCWSATucOlPmOI1+tVo0aN3B4DEYRMwEYmYCMTQAjrATYyARuZgI1MACGsB9jIBGxkAjaPx5PlM3KGf60Y4vf79eeff8rv97s9CiIEmYCNTMBGJoAQ1gNsZAI2MgEbmQBCWA+wkQnYyARsxpgsn5EzbPzFEL/fr3Xr1vGFE0FkAjYyARuZAEJYD7CRCdjIBGxkAghhPcBGJmAjE7AFskAmcoeNPwAAAAAAAAAAACAKsPEXI3w+nxYtWqSFCxdq0aJF8vl8bo8EAAAAAAAAAACAMIrojb9PPvlEl19+ucqWLSvHcbRo0SK3RyqUvv/+e3Xt2lX9+/fXpEmT1L9/f3Xt2lXff/+926PBZY7jKCkpSY7juD0KIgSZgI1MFG50qfBiPcBGJmAjE7CRicKNLhVerAfYyARsZAK2QBbIRO5E9MbfX3/9pVatWmn06NFuj1Joff/99xo2bJhq1qypCRMm6KuvvtKECRNUs2ZNDRs2jM2/GOf1elWvXj15vV63R0GEIBOwkYnCjS4VXqwH2MgEbGQCNjJRuNGlwov1ABuZgI1MwObxeLJ8Rs7EuT3AyfTs2VOStGHDBncHKaR8Pp9efPFFNW/eXCNHjpQkbd26VfXq1dPIkSM1ePBgjR8/Xq1ateKLaYzy+/3aunWrKleuzBdPSCITOB6ZKNzoUuHFeoCNTMBGJmAjE4UbXSq8WA+wkQnYyARsxpgsn5EzEb3xlxdHjhzRkSNHgr9OS0uTJGVmZiozM1PS0d1hj8cjv98vv98fvG/guM/nyxKkEx33er1yHCf45x57XNJx76N3ouNxcXEyxmQ57jiOvF7vcTOe6Hh257Ro0SJt375dw4cPlzFGmZmZ2rRpk8qXL68iRYqoR48e6tu3rxYuXKjGjRsXinM69ni0PE5unpOkYCYC9yns5xSNj1NBnpPP59OmTZtUoUIFxcfHR8U5/d1xzunk5xTIRODrRDScU14fJ/vvjlZ0qROfk92louGcovFxokvxONGleJwi6ZzoUnQpuhRdinOiS3FOdKmTHeec/r5LSaHvo9FwTgXRpaJu42/UqFF69NFHjzu+cOFClShRQpJUvnx51apVS+vXr9fOnTuD96lataqqVq2qVatWKTU1NXi8Zs2aqlChgn7//XcdOnQoeLxu3boqXbq0Fi5cmOWBatiwoeLj4zVv3rwsMzRt2lTp6elasmRJ8JjX69X555+v1NRUrVixIng8ISFBjRo10q5du7Ru3brg8aSkJNWrV09bt27V5s2bg8ezO6eFCxdKkmrUqKFVq1Zp37592rdvnxYsWKBatWqpZs2akqS5c+cGQxPp5xSNj5Ob59SkSRNlZmZqwYIFweskF/ZzisbHqSDPyRijffv2affu3Tr99NOj4pyi8XEqyHNKS0sLfu9wHCcqzimvj9Nff/2lWECXOvE52V0qGs4pGh8nuhSPE12KxymSzokuRZcKoEvRpTgnuhTnRJeKlsepIM8p0B9WrFihAwcORMU5FUSXckyEvEbyvffeU58+fYK/nj59ui666CJJRy+pUKNGjSyvTDuR7J5ZVa1aNe3evVuJiYmSomuX157x2OOLFi1S//79NWHCBNWtWzf4jfTcc89VkSJFtGzZMvXt21fPPvssr/iL0XOSjm78nnvuuTyzinMK/h0LFizQeeedxzOrOCfFxcUpIyND8+bNC36diIZzyuvjlJaWprJlyyo1NTXYKSIJXapgnqV+bJeKhnOKxseJLsXjRJficYqkc6JL0aXoUnQpzokuxTnRpU52nHM6+Tn98ccfuuuuuzR+/HjVrVs3Ks6pILpUxLzi75prrtEFF1wQ/HWVKlXy9OcULVpURYsWPe54XFyc4uKynm7gQbQFHpScHrf/3Lwcdxwn2+MnmjEnx5s0aaJKlSrpnXfe0ciRI1WkSBFVqFBBRYoUkSS9++67Ov3009WkSZPjzi1Sz+lY0fI4Haugz8nv9wczYc9TWM9Jir7HSSq4c/J4PKpQoULw74qGc8rpcc4p++NerzfbrxOF+Zzy+jid6O+IFHSp/F9/juMctx4K+zllh3OiS0Xb4yTRpfJyPJofp5wep0vRpfKCLkWX4pzoUtH2OEl0qbwcj+bHKafHw9WlArcde3thPqeC6FIR07pKlSqlUqVKuT1GVPF6vbr77rs1bNgwDR48WD169FDNmjW1bNkyvfvuu/rll180YsSIE4YY0c/j8ahWrVpuj4EIQiZgIxOFB10q/7EeYCMTsJEJ2MhE4UGXyn+sB9jIBGxkArbAZYADn5Ezx28fRpA9e/Zo0aJFWrZsmSRp5cqVWrRokbZv3+7yZIVH69atNWLECK1bt059+/ZV+/bt1bdvX61fv14jRoxQ69at3R4RLvL7/Vq7dm2WlxAjtpEJ2MhE4UaXCi/WA2xkAjYyARuZKNzoUuHFeoCNTMBGJmALXG4zQt6xrtCI6I2/zz77TE2aNNGVV14pSerSpYuaNGmil156yeXJCpfWrVtr0qRJevbZZ9W1a1c9++yzev/999n0g/x+v3bu3Mk3UwSRCdjIROFGlwov1gNsZAI2MgEbmSjc6FLhxXqAjUzARiZgC2SBTOROxFzqMzu33HKLbrnlFrfHiAper1eNGzdWZmamGjduzOU9AQCIAXQpAACAvKNLAQCAwiiiX/EHAAAAAAAAAAAAIGfY+IshHo9HVatWlcfDw46jyARsZAI2MgGEsB5gIxOwkQnYyAQQwnqAjUzARiZgC2SBTORORF/qE+EV+MIJBJAJ2MgEbGQCCGE9wEYmYCMTsJEJIIT1ABuZgI1MwOY4TpbPyBm2SWOIz+fT8uXL5fP53B4FEYJMwEYmYCMTQAjrATYyARuZgI1MACGsB9jIBGxkAja/35/lM3KGjb8YYoxRamqqjDFuj4IIQSZgIxOwkQkghPUAG5mAjUzARiaAENYDbGQCNjIBWyALZCJ32PgDAAAAAAAAAAAAogAbfwAAAAAAAAAAAEAUYOMvhng8HtWsWVMeDw87jiITsJEJ2MgEEMJ6gI1MwEYmYCMTQAjrATYyARuZgC2QBTKRO3FuD4CC4/F4VKFCBbfHQAQhE7CRCdjIBBDCeoCNTMBGJmAjE0AI6wE2MgEbmYDNcZwsn5EzbJPGEJ/Pp8WLF8vn87k9CiIEmYCNTMBGJoAQ1gNsZAI2MgEbmQBCWA+wkQnYyARsfr8/y2fkDBt/McQYo0OHDskY4/YoiBBkAjYyARuZAEJYD7CRCdjIBGxkAghhPcBGJmAjE7AFskAmcoeNPwAAAAAAAAAAACAKsPEHAAAAAAAAAAAARAE2/mKI1+tV3bp15fV63R4FEYJMwEYmYCMTQAjrATYyARuZgI1MACGsB9jIBGxkArZAFshE7sS5PQAKjuM4Kl26tNtjIIKQCdjIBGxkAghhPcBGJmAjE7CRCSCE9QAbmYCNTADhwSv+YkhmZqbmzp2rzMxMt0dBhCATsJEJ2MgEEMJ6gI1MwEYmYCMTQAjrATYyARuZgM3n82X5jJxh4y/GsEBgIxOwkQnYyAQQwnqAjUzARiZgIxNACOsBNjIBG5kATh0bfwAAAAAAAAAAAEAUYOMPAAAAAAAAAAAAiAJs/MUQr9erhg0byuv1uj0KIgSZgI1MwEYmgBDWA2xkAjYyARuZAEJYD7CRCdjIBGyBLJCJ3GHjL8bEx8e7PQIiDJmAjUzARiaAENYDbGQCNjIBG5kAQlgPsJEJ2MgEcOrY+IshPp9P8+bN4w1SEUQmYCMTsJEJIIT1ABuZgI1MwEYmgBDWA2xkAjYyAVsgC2Qid9j4AwAAAAAAAAAAAKIAG38AAAAAAAAAAABAFGDjDwAAAAAAAAAAAIgCbPzFEK/Xq6ZNm8rr9bo9CiIEmYCNTMBGJoAQ1gNsZAI2MgEbmQBCWA+wkQnYyARsgSyQidxh4y/GpKenuz0CIgyZgI1MwEYmgBDWA2xkAjYyARuZAEJYD7CRCdjIBHDq2PiLIT6fT0uWLJHP53N7FEQIMgEbmYCNTAAhrAfYyARsZAI2MgGEsB5gIxOwkQnYAlkgE7nDxh8AAAAAAAAAAAAQBdj4AwAAAAAAAAAAAKIAG38xhjfBhI1MwEYmYCMTQAjrATYyARuZgI1MACGsB9jIBGxkAjh1jjHGuD1EfkpLS1NSUpJSU1OVmJjo9jgAAKCQitVOEavnDQAAwitWO0WsnjcAAOGwcuVK3X777Zo4caLOOusst8dxVW46Ba/4iyHGGO3bt09RvteLXCATsJEJ2MgEEMJ6gI1MwEYmYCMTQAjrATYyARuZAMKDjb8Y4vP5tGLFCvl8PrdHQYQgE7CRCdjIBBDCeoCNTMBGJmAjE0AI6wE2MgEbmYAtkAUykTts/AEAAAAAAAAAAABRgI0/AAAAAAAAAAAAIAqw8RdDHMdRQkKCHMdxexRECDIBG5mAjUwAIawH2MgEbGQCNjIBhLAeYCMTsJEJ2AJZIBO5E+f2ACg4Xq9XjRo1cnsMRBAyARuZgI1MACGsB9jIBGxkAjYyAYSwHmAjE7CRCdg8Hk+Wz8gZNv5iiN/v165du1SuXDkWCiSRCRyPTMBGJoAQ1gNsZAI2MgEbmQBCWA+wkQnYyEThcfjwYaWkpOT737Nhw4YsnwtCcnKyihUrVmB/X35g4y+G+P1+rVu3TmXKlOELJySRCRyPTMBGJoAQ1gNsZAI2MgEbmQBCWA+wkQnYyEThkZKSottvv73A/r4nnniiwP6uiRMn6qyzziqwvy8/sPEHAAAAAAAAAACAHElOTtbEiRPz/e/x+XxatmyZ6tevL6/Xm+9/n3T03Ao7Nv4AAAAAAAAAAACQI8WKFSuQV8VlZmYqLS1NderUUVwc21k5xetlY4jjOEpKSpLjOG6PgghBJmAjE7CRCSCE9QAbmYCNTMBGJoAQ1gNsZAI2MgEbmcgbxxhj3B4iP6WlpSkpKUmpqalKTEx0exwAAFBIxWqniNXzBgAA4RWrnSJWzxsAAIRXbjoFr/iLIX6/X5s3b5bf73d7FEQIMgEbmYCNTAAhrAfYyARsZAI2MgGEsB5gIxOwkQnYyETesPEXQ1gksJEJ2MgEbGQCCGE9wEYmYCMTsJEJIIT1ABuZgI1MwEYm8oaNPwAAAAAAAAAAACAKsPEHAAAAAAAAAAAARAE2/mKIx+NR+fLl5fHwsOMoMgEbmYCNTAAhrAfYyARsZAI2MgGEsB5gIxOwkQnYyETeOMYY4/YQ+SktLU1JSUlKTU1VYmKi2+MAAIBCKlY7RayeNwAACK9Y7RSxet4AACC8ctMp2CaNIX6/X2vXruWNMBFEJmAjE7CRCSCE9QAbmYCNTMBGJoAQ1gNsZAI2MgEbmcgbNv5iiN/v186dO1kkCCITsJEJ2MgEEMJ6gI1MwEYmYCMTQAjrATYyARuZgI1M5A0bfwAAAAAAAAAAAEAUYOMPAAAAAAAAAAAAiAJs/MUQj8ejqlWryuPhYcdRZAI2MgEbmQBCWA+wkQnYyARsZAIIYT3ARiZgIxOwkYm8cYwxxu0h8lNaWpqSkpKUmpqqxMREt8cBAACFVKx2ilg9bwAAEF6x2ili9bwBAEB45aZTsE0aQ3w+n5YvXy6fz+f2KIgQZAI2MgEbmQBCWA+wkQnYyARsZAIIYT3ARiZgIxOwkYm8YeMvhhhjlJqaqih/kSdygUzARiZgIxNACOsBNjIBG5mAjUwAIawH2MgEbGQCNjKRN2z8AQAAAAAAAAAAAFGAjT8AAAAAAAAAAAAgCrDxF0M8Ho9q1qwpj4eHHUeRCdjIBGxkAghhPcBGJmAjE7CRCSCE9QAbmYCNTMBGJvImzu0BUHA8Ho8qVKjg9hiIIGQCNjIBG5kAQlgPsJEJ2MgEbGQCCGE9wEYmYCMTsJGJvGGbNIb4fD4tXrxYPp/P7VEQIcgEbGQCNjIBhLAeYCMTsJEJ2MgEEMJ6gI1MwEYmYCMTecPGXwwxxujQoUMyxrg9CiIEmYCNTMBGJoAQ1gNsZAI2MgEbmQBCWA+wkQnYyARsZCJv2PgDAAAAAAAAAAAAogAbfwAAAAAAAAAAAEAUYOMvhni9XtWtW1der9ftURAhyARsZAI2MgGEsB5gIxOwkQnYyAQQwnqAjUzARiZgIxN5E+f2ACg4juOodOnSbo+BCEImYCMTsJEJIIT1ABuZgI1MwEYmgBDWA2xkAjYyARuZyBte8RdDMjMzNXfuXGVmZro9CiIEmYCNTMBGJoAQ1gNsZAI2MgEbmQBCWA+wkQnYyARsZCJv2PiLMT6fz+0REGHIBGxkAjYyAYSwHmAjE7CRCdjIBBDCeoCNTMBGJmAjE7nHxh8AAAAAAAAAAAAQBdj4AwAAAAAAAAAAAKKAY4wxbg+Rn9LS0pSUlKTU1FQlJia6PY6rjDE6dOiQEhIS5DiO2+MgApAJ2MgEbGQiJFY7Rayed3ZYD7CRCdjIBGxkIiRWO0Wsnnd2WA+wkQnYyARsZCIkN52CV/zFmPj4eLdHQIQhE7CRCdjIBBDCeoCNTMBGJmAjE0AI6wE2MgEbmYCNTOQeG38xxOfzad68ebwZJoLIBGxkAjYyAYSwHmAjE7CRCdjIBBDCeoCNTMBGJmAjE3nDxh8AAAAAAAAAAAAQBdj4AwAAAAAAAAAAAKIAG38AAAAAAAAAAABAFHCMMcbtIfJTWlqakpKSlJqaqsTERLfHcZUxRj6fT16vV47juD0OIgCZgI1MwEYmQmK1U8TqeWeH9QAbmYCNTMBGJkJitVPE6nlnh/UAG5mAjUzARiZCctMpeMVfjElPT3d7BEQYMgEbmYCNTAAhrAfYyARsZAI2MgGEsB5gIxOwkQnYyETusfEXQ3w+n5YsWSKfz+f2KIgQZAI2MgEbmQBCWA+wkQnYyARsZAIIYT3ARiZgIxOwkYm8YeMPAAAAAAAAAAAAiAJs/AEAAAAAAAAAAABRgI2/GOP1et0eARGGTMBGJmAjE0AI6wE2MgEbmYCNTAAhrAfYyARsZAI2MpF7jjHGuD1EfkpLS1NSUpJSU1OVmJjo9jgAAKCQitVOEavnDQAAwitWO0WsnjcAAAiv3HQKXvEXQ4wx2rdvn6J8rxe5QCZgIxOwkQkghPUAG5mAjUzARiaAENYDbGQCNjIBG5nIGzb+YojP59OKFSvk8/ncHgURgkzARiZgIxNACOsBNjIBG5mAjUwAIawH2MgEbGQCNjKRN2z8AQAAAAAAAAAAAFGAjT8AAAAAAAAAAAAgCrDxF0Mcx1FCQoIcx3F7FEQIMgEbmYCNTAAhrAfYyARsZAI2MgGEsB5gIxOwkQnYyETeOCbK3xUxLS1NSUlJSk1NVWJiotvjAACAQipWO0WsnjcAAAivWO0UsXreAAAgvHLTKXjFXwzx+/36888/5ff73R4FEYJMwEYmYCMTQAjrATYyARuZgI1MACGsB9jIBGxkAjYykTds/MUQv9+vdevWsUgQRCZgIxOwkQkghPUAG5mAjUzARiaAENYDbGQCNjIBG5nIGzb+AAAAAAAAAAAAgCjAxh8AAAAAAAAAAAAQBdj4iyGO4ygpKUmO47g9CiIEmYCNTMBGJoAQ1gNsZAI2MgEbmQBCWA+wkQnYyARsZCJvHGOMcXuI/JSWlqakpCSlpqYqMTHR7XEAAEAhFaudIlbPGwAAhFesdopYPW8AABBeuekUvOIvhvj9fm3evJk3wkQQmYCNTMBGJoAQ1gNsZAI2MgEbmQBCWA+wkQnYyARsZCJv2PiLISwS2MgEbGQCNjIBhLAeYCMTsJEJ2MgEEMJ6gI1MwEYmYCMTecPGHwAAAAAAAAAAABAF2PgDAAAAAAAAAAAAogAbfzHE4/GofPny8nh42HEUmYCNTMBGJoAQ1gNsZAI2MgEbmQBCWA+wkQnYyARsZCJvIvZfKyMjQwMHDtQ555yjEiVKqHLlyrrpppu0detWt0crtDwej2rVqsUiQRCZgI1MwEYmCi+6VPixHmAjE7CRCdjIROFFlwo/1gNsZAI2MgEbmcibiP3XOnjwoBYsWKChQ4dqwYIF+uSTT7Ry5Updc801bo9WaPn9fq1du5Y3wkQQmYCNTMBGJgovulT4sR5gIxOwkQnYyEThRZcKP9YDbGQCNjIBG5nImzi3BziRpKQkffPNN1mOvfDCC2rWrJk2btyoM844w6XJCi+/36+dO3cqOTmZHXJIIhM4HpmAjUwUXnSp8GM9wEYmYCMTsJGJwosuFX6sB9jIBGxkAjYykTeF6l8qNTVVjuOodOnSbo8CAABQ6NClAAAA8o4uBQAACoOIfcWf7fDhwxo4cKC6du2qxMTEE97vyJEjOnLkSPDXqampkqQ9e/YoMzNT0tHrwno8Hvn9/iwvEQ0c9/l8Msb87XGv1yvHcYJ/7rHHJcnn8+XoeFxcnIwxWY47jiOv13vcjCc6npNzyszM1IEDB7R3714VKVIkKs4pGh+ngjwnScFMBO5T2M8pGh+ngjwnn8+nAwcOaN++fYqPj4+Kc/q745zTyc8pIyMjy9eJaDinvD5OaWlpkpRlrsKELnXy43Sp6Hmc6FI8TnQpHqdIOie6FF2KLkWX4pzoUpwTXepkxzknulS+dCkTId59911TokSJ4McPP/wQvC09Pd1cffXVpkmTJiY1NfWkf87w4cONJD744IMPPvjgg498+di0aVN+16I8oUvxwQcffPDBBx+F4YMuxQcffPDBBx988JH3j5x0KceYyHiq1f79+7Vjx47gr6tUqaKEhARlZGSoc+fOWrdunWbOnKmyZcue9M+xn1nl9/u1Z88elS1bVo7j5Nv8hUFaWpqqVaumTZs2nfTZaYgdZAI2MgEbmQgxxmj//v2qXLmyPJ7Iu1o6XSr/sR5gIxOwkQnYyEQIXYouxXqAjUzARiZgIxMhuelSEXOpz1KlSqlUqVJZjgXK1erVqzVr1qy/LVeSVLRoURUtWjTLMa69nlViYmLMLxJkRSZgIxOwkYmjkpKS3B7hhOhSBYf1ABuZgI1MwEYmjqJLQWI94HhkAjYyARuZOCqnXSpiNv5sGRkZuuGGG7RgwQJ9/vnn8vl82r59uySpTJkyio+Pd3lCAACAyEWXAgAAyDu6FAAAKKwiduNvy5Yt+uyzzyRJjRs3znLbrFmz1KZNm4IfCgAAoJCgSwEAAOQdXQoAABRWEbvxV716dUXI2w9GjaJFi2r48OHHXXICsYtMwEYmYCMThRddKvxYD7CRCdjIBGxkovCiS4Uf6wE2MgEbmYCNTOSNY2gxAAAAAAAAAAAAQKHncXsAAAAAAAAAAAAAAKeOjT8AAAAAAAAAAAAgCrDxBwAAAAAAAAAAAEQBNv4AAAAAAAAAAACAKMDGXyHm9/slSQcOHHB5EkQaY4wkadmyZS5PgkgRyAQAIIQuhROhS8FGlwKA49GlcCJ0KdjoUkDBYuOvkPL7/fJ4PFq8eLE6d+6stWvXuj0SIojjOPrvf/+rRo0aac6cOW6PgwjgOI4kaefOnS5PAgCRgS6Fk6FLwUaXAoCs6FI4GboUbHQpoGCx8VcIHVuumjZtqvPOO0+1atVyeyxEkM2bN+vrr7/WCy+8oAsvvNDtceCiadOm6bfffpMkDRw4UE8//bTS09NdngoA3EWXwt+hSyGALgUAx6NL4e/QpRBAlwLcEef2AMidQLlasWKFWrRooeHDh+vhhx92eyxEkIULF2rIkCHasWOHevfuLenoy+kDz6xB7EhNTdXbb7+tGTNm6Nprr9WUKVM0Z84cxcfHuz0aALiGLoW/Q5dCAF0KAI5Hl8LfoUshgC4FuIdX/BUigXK1ZMkStWzZUocOHdKDDz4oSfL5fC5Ph0ixadMmpaamatmyZdqyZYskUa5iVFJSksaNG6eyZcvqww8/1KuvvqpGjRopMzPT7dHgosB19RctWqTPP/9c7733no4cOXLc7UA0okshJ+hSCKBLITt0KcQyuhRygi6FALoUskOXKhiO4V+yUDj2MgrNmzdX9+7dtWLFCu3cuVOzZ89W+fLlg/cBZsyYoccee0wZGRl6+umnddFFF7k9EgpY4Nl0mzdvVq9evWSM0dKlSzV16lRdeOGF8vl88ng8lO8YE8jFp59+qjvvvFPJyclauXKlzj//fA0cOFCXXXaZPB4Pz8ZEVKJLITfoUqBLITt0KcQyuhRygy4FuhSyQ5cqQAaFxh9//GG8Xq8ZMmSIMcaYFStWmPPPP9+cddZZZufOncYYY3w+n5sjooD5/X5jjDGrVq0y8+fPN7Nnzw7e9tVXX5nLL7/ctG/f3vz4449ujYgClt3XgIyMDLN+/Xpzww03mPLly5tffvkly+2bN28uqPEQAWbOnGnKlStnXnvtNWOMMcuWLTOO45jWrVubL774IpihwNcXIJrQpWCjS8FGl8LfoUshltGlYKNLwUaXwt+hSxUMXvEX4Y59ttTcuXM1c+ZMDRw4MHj7ypUr1bNnT6WlpenHH39UuXLleIZVjDD//8yHKVOm6MEHH5TP51NGRoZq1qypN998U7Vr19aXX36pcePGKS4uTg8++KDatGnj9tjIR8eu/S+//FK7d+9WZmamOnTooKSkJK1fv14PPfSQZs+erSlTpqhly5bq3LmzzjnnHA0dOtTl6ZFfMjIyVKRIERljdPjwYY0aNUoZGRkaNWqU1q5dqyuuuELNmzfX77//Hnw25uWXXy6v1+v26EBY0KVwInQp2OhSyA5dCrGOLoUToUvBRpdCduhSLnFvzxF/J7C7vXnzZjNlyhQzdepUs27duuPut3LlSp5hFaN++uknU7JkSfPaa6+ZhQsXmvnz55smTZqYOnXqBLPy2WefmebNm5sbbrjBHDx40OWJURAeeOABU6FCBdOkSRNTvHhx06xZM/PBBx8Yv99v1q5da7p162YcxzHnnnuuqVmzpklPT3d7ZOST5cuXmxtuuMH88ccfxpijz7KbPXu2WbFihUlNTTUXXHCB6d27tzHGmKVLl5pixYqZpk2bmunTp7s5NhA2dCn8HboUskOXQgBdCrGOLoW/Q5dCduhSCKBLuYeNvwgVKEiLFy821atXN2eddZZxHMc0bNjQvPfee8fdf+XKlaZZs2amQYMGZseOHQU9Llwybtw406ZNG5ORkRE8lpmZaRo2bGjatGkTPPb111+blJQUN0ZEATj2pe/vvfeeqVixolm4cKE5cOCA2bdvn7nmmmtMy5YtzRdffGGMMWbXrl3mo48+Ms8991wwO8dmCNHhwIEDpnnz5sZxHNO+fXuzaNEiY4wxR44cMcYc/brQuHFjs2LFCmOMMd99951p27atadu2rdmwYYNrcwPhQpdCTtClYAxdCtmjSyHW0aWQE3QpGEOXQvboUu7idfcRKPCy6KVLl+rCCy/UTTfdpOnTp2vp0qU6dOiQXnjhBf3555+Sjr6sXpLq1Kmjd955R+np6brqqqvk9/vdPAWEUXaP5eHDhyVJmzdv1pYtWxQXFxc87vV6NW7cOK1Zs0aLFy+WJF1++eU644wzCm5oFIi3335bGzdulOM4wa8F69atU7169XT22WeraNGiSkpK0ltvvSWv16vnnntOklS2bFl16tRJ999/v+Li4uTz+YIZQvRwHEfVq1dX9erVVaFCBT3wwANaunSp4uPjJUm7du1SamqqDhw4IEmaOXOmGjVqpM8//1zJyclujg6cMroUjkWXwonQpXAydCnEMroUjkWXwonQpXAydCl3sfEXgTwejzZs2KDmzZurQ4cOevTRR1WjRg01aNBAo0eP1sKFC7VlyxZJRxdQQJ06dTR9+nR99NFHXEs9ing8Hm3cuFGffPKJJOnDDz/UkCFDJEldunTRgQMHNGbMGElSsWLFJEler1dFihRR8eLF3Rka+e6jjz7S4MGD9eKLL2rr1q3BrwX79+/XgQMHFBcXp7i4OB0+fFilS5fWk08+qR9//FHLli07rrRzzezo4/f7Vbx4cQ0YMEAZGRk67bTTVLx4cfXv31+///67JKl169by+Xzq1q2bzjvvPI0bN049evRQ0aJFXZ4eOHV0KRyLLoXs0KVwMnQpxDq6FI5Fl0J26FI4GbqU+/guHKEOHz6s4sWLy+fzac6cOVm+IJYsWfKEz4KoVauWqlevXkBToiAcPnxYI0aM0KhRo/Tggw+qa9euOvvssyVJZ5xxhrp27aqPP/5Y//nPfyRJBw4c0IwZM1S8eHGVLl3axcmRnzp37qy+ffvqf//7n5577jlt3rxZknTjjTdq4cKFevLJJyWFSvehQ4dUs2ZNJSYm8j9gUSwjI0OSgs+2q127tq6//nrVrVtXd955p/x+v/r166dFixapSpUq+uGHH9SjRw917NhRc+bMUePGjd09ASCM6FIIoEshO3QpZIcuBYTQpRBAl0J26FLIDl0qgrh2kVFkcey1kAPmz59v6tSpY6699lqzatUqs3nzZlOxYkXz0EMPuTAh3LR69WrTqlUr4ziOufvuu7PctmbNGjNgwABTqVIlU6VKFdO0aVNTrlw5M3/+fJemRX666667gtdEN8aYxx57zDRp0sQMGDDAbNy40RhjzLPPPmuKFi1qhgwZYlauXGlWrVplrrzySnPJJZfwButRbMWKFeaGG24w06ZNM/v37w8eHzNmjElOTjb79+8333zzjWnXrp1p27Zt8Nrq2X3/AQojuhROhi6FALoUToQuhVhHl8LJ0KUQQJfCidClIotjzP9fgBeuCVw7fe/evdq0aZP8fr/q1q2rYsWKaeHChbrxxhtVo0YNLV68WB06dND48eOz/D5Er8BjfPjwYV111VVKS0tTiRIldMcdd6hr167B++3bt087d+7UtGnTdPrpp6t58+aqWbOmi5MjP6xcuVKvvfaannjiCRUpUiR4fMSIEZo6daratm2rBx98UOXLl9frr7+uAQMGKCEhQcWLF1e5cuX0ww8/qEiRInztiEJpaWlq37695syZo8TERHXv3l0lSpTQ6NGj5TiOevbsqQYNGmjQoEGaMmWKXn/9de3cuVNvvfWW6tWr5/b4wCmjS+FE6FI4Fl0KJ0KXQqyjS+FE6FI4Fl0KJ0KXijxs/Lks8IVu+fLlwTc0bdWqlQYPHhy8z7x583TTTTfpr7/+0qRJk9SiRQtJR99A+dhrqSM6LViwQA0bNlRcXJyWLl2qJ554Qlu2bFHfvn2zlCzyEFvefvtt+Xw+3XrrrZKOlqxPP/1U7dq1U//+/VWpUiVt3bpVa9asUZEiRXTBBRfI4/EoMzOTN0yOQkeOHNHLL7+sGTNmaOvWrbrtttv06aefavfu3frnP/+pdevWqWjRonrjjTckSR988IEmT56sMWPG8AbrKPToUvg7dClkhy6FY9GlEMvoUvg7dClkhy6FY9GlIpBbLzWECb60ecmSJaZChQpm6NCh5o8//gjePn/+fLN7925jjDELFiwwtWvXNtdff72ZM2eOK/Oi4ARe4rxjxw5zzjnnmIsuusgcOnTIGGPMr7/+am688UZz8cUXm/fee88YY8ywYcPMgw8+aDIyMlybGQVn27Zt5oorrjDNmzc3kyZNCh5/9NFHTePGjc1DDz1kUlJSjvt9mZmZBTkmCkjg68XBgwfNhAkTzEUXXWR69uxpjDHm7bffNvfdd59xHMc4jmMWL14c/H3HXnYBKKzoUjgRuhROhi6FY9GlEMvoUjgRuhROhi6FY9GlIhMbfy7btGmTqVOnjvnXv/6V5fgzzzxjzjjjDDN8+HCza9cuY4wxc+fONfXr1zeXXnqpmTt3rhvjIp8Eyva+ffuyHE9PTzeTJk0yF1xwgWnXrl2WktWzZ09To0YN07JlS1OsWDHz22+/FfjcKBiBb6DHXvN6zpw5pmvXruaiiy4y77//fvD4iBEjTNOmTU2fPn3Mzp07C3xWuOuvv/4yL730kmnYsKHp1atX8H+6ZsyYYb755htjjOF6+og6dCkYQ5fCydGlkFN0KcQiuhSMoUvh5OhSyCm6VOTgYrou+/bbb1WmTBn1798/eGz48OEaOXKkGjVqpOnTp2vChAnavXu3mjZtqtdff11paWmqVKmSi1MjnAKX1Vi4cKFuuOEGbdu2LXhbkSJF1LFjRw0YMEC7d+/WNddco8OHD6tZs2YaNGiQHnnkEbVq1UqLFi3S+eef7+JZIL/4/f7gpTJ27typtLQ0+f1+XXDBBcFLJ4wfP16TJk2SJA0dOlRt2rTRkSNHVLZsWTdHRz4z/3+l7sOHD0uSMjMzVbx4cfXo0UN9+/bVwoUL1bNnT2VkZKhdu3Zq27atJHEdfUQduhToUjgZuhROhC4FHEWXAl0KJ0OXwonQpSIb7/HnsltvvVUrVqzQL7/8Ejw2cOBA/eMf/1CbNm300EMPadasWWrXrp2GDBmiEiVK6MiRIypatKiLUyNcAuVq0aJFatasmYYOHaqrrrpKY8eO1Ztvvhm8X0ZGhj777DM98sgjSk5O1pQpU8hAjBk+fLimTJkij8ejsmXL6vnnn9fZZ5+tRYsWaeTIkdqxY4fuuusudenSRVLo2vqGa+xHpcDj+vXXX+vdd99VSkqK2rZtq2uvvVaNGjXSX3/9pffee0+vvPKK6tatqzfeeCPLG28D0YQuFdvoUsgpuhSORZcCQuhSsY0uhZyiS+FYdKnIx/aqyxITE7V3717t378/eGz06NFq06aNJOmpp55SYmKiVq9erRIlSkiS4uPj3RgVYRYoVytXrlSrVq30+OOPa+jQoZo9e7Zmz56tO+64I3jfIkWKqEOHDrr++uv15Zdf6vLLL9eRI0dcnB75ze/3B//7zTff1Lhx43T//ferT58+iouL00UXXaTPP/9cjRs31oMPPqjKlStrxIgR+uabbySJchXlHMfR1KlTdf3116tKlSpq0aKFfv75Z/Xt21dz585ViRIl1L17d915552aM2eO7rrrLrdHBvINXSp20aVwMnQpnAxdCgihS8UuuhROhi6Fk6FLRb44tweIddWrV9f69ev1zTff6JprrlFcXFzwG6/P55PH41GNGjVUqVIl+Xw+eb1evmBGgcBjvGTJErVp00YHDx7UP/7xD0lSjx495PV6NXHiRPXq1UuvvfaaHMeR4zhq2LChmjdvrlKlSmn79u1KTk52+UyQXwIve//888+1bt06Pffcc7r55pslSXfffbduvvlm3XTTTfr999/VrFkz3Xvvvfrqq6906aWXBv8MvlZEryVLlmjw4MEaN26cbrvtNu3atUv16tVTYmKi+vTpo5dfflnnn3++unTpoiJFiujiiy92e2Qg39ClYhNdCn+HLoWToUsBIXSp2ESXwt+hS+Fk6FKFQIG/qyCOc/7555saNWqYGTNmmL/++it43OfzmUGDBplq1aqZ1atXuzghwinwBqaLFi0yCQkJplevXubmm2825cuXNz/99JMxxpi9e/ea559/3jRp0sTccsstwd87bNgwM2DAALN//35XZkfBmjt3rqlTp45JSEgwb7/9tjHGmCNHjgRvb9y48XFvwG6MMZmZmQU2I9yxcOFC07NnT3PkyBGzYcMGU6tWLXPHHXeYL774wtSoUcM0a9Ys+PXk2DffBqIVXSq20KWQU3QpnAhdCsiKLhVb6FLIKboUToQuFfnY+CtAgZCvXbvWrF+/3mzatMkYY8wff/xhzjnnHFOhQgUzaNAgM2/ePPPWW2+ZO+64wyQlJZkFCxa4OTbywfLly43jOGbYsGHGmKMZuPHGG0358uXNzz//bIw5WrImTJhg6tWrZ8qVK2fatm1rihcvbpYtW+bm6ChA+/btM2PGjDFVq1Y17dq1Cx7PyMgwmZmZ5qqrrjJ33XWXixPCTZs3bzbGGNO9e3fTvXt3k56ebowx5oorrjCnnXaaad26tTl48CAFC1GFLoUAuhRygi6Fk6FLIRbRpRBAl0JO0KVwMnSpyMbGXwH7+OOPTbVq1UylSpVM+/btzfvvv2+MMWbnzp3m2muvNeXKlTOO45iaNWuaa665xvz+++8uT4z8sGfPHjNmzJgsx7IrWQcOHDB//PGHGTRokBk+fLhZvny5C9OiIASecWf/ev/+/Wb8+PHmzDPPNN27d89yn/PPPz/bZ1Yheti5sKWlpZlGjRqZcePGGWOMOXz4sOnVq5d54YUXzPbt2wtiRKDA0aVgDF0Kx6NLITt0KeB4dCkYQ5fC8ehSyA5dqvByjDHG7cuNxoqtW7fq8ssvV//+/VW8eHF9++23+vnnn3XvvfcG3+Byw4YN2rt3r2rUqKEiRYoE3zgZ0SMzM1NxcaG31wxcV12Sli1bphEjRmjmzJmaNm2amjdvnu39EF2OfWwnTpyopUuXateuXbrhhht03XXXKSMjQ6+99poef/xxlStXTnXr1pXX69W8efO0fPnyLHlC9Fi3bp1mzJihjh07qnz58tne5+DBg+rWrZsyMjLUv39/ffPNN/r00081a9YsVa5cuYAnBvIfXQoSXQrHo0shO3Qp4Hh0KUh0KRyPLoXs0KUKObd3HqPdsS9l3bZtm+nRo4c5fPiwMcaYVatWmX/961/mrLPOMs8//3zwfn+3k47oFniGVeXKlc3s2bPdHgcF6MEHHzTly5c3nTt3NldffbXxeDzmvvvuMzt27DCHDh0yL774ojnrrLNMgwYNzDfffBP8fRkZGS5OjfwybNgwU6pUKTN27Fiza9euE95v0qRJpk2bNqZixYrmzDPPNPPnzy/AKYH8R5dCbtGlYhddCseiSwFH0aWQW3Sp2EWXwrHoUoUbr/jLR8YYOY6j6dOn65133lHJkiW1atUqfffdd8H7rF69WuPHj9c333yj3r17q1+/fu4NjIixfPlyPfDAA1qzZo2WLFmiokWLynEct8dCPvr+++/VtWtXTZs2Teeff74k6aOPPtJdd92lPn36aOTIkUpNTdVbb72l9957T+ecc45effVVSZLP55PX63VzfOSTgQMH6sMPP9T999+vm266SWXLls32fmvWrJHjOCpZsqQqVqxYwFMC+YcuhbyiS8UeuhSyQ5dCrKNLIa/oUrGHLoXs0KUKL16fnQ8Ce6mO4+jbb7/V9ddfr0OHDmnp0qX64Ycf9OyzzwbvW7t2bd19991q3ry5Jk2apH379rk0NfJbbvbY69WrpzFjxmjWrFkqVqwY5SoKLV68WNOmTdNPP/0kSTp8+LCKFy+uqlWryufzyRijzp07a8yYMXrmmWe0ePFiJSUl6dZbb1WPHj20aNEi3XjjjZJEuYpCPp9PkjR69Gh16tRJY8eO1dtvv63du3dnud+hQ4c0fPhwff/996pVqxblClGDLoXs0KVwLLoUToYuhVhHl0J26FI4Fl0KJ0OXKvy4AG8+CHwz3LRpk/bu3aunn35a9957rzZu3KjXX39djzzyiLxer+6//35J0plnnql///vfKlGihEqXLu3i5MgPgWfYBT4HnOjZMIH7nXXWWQU5JgrQe++9p2eeeUZnnHGGGjRooJYtW8rr9SolJUW7d+/W6aefriNHjqho0aK65pprVLlyZa1Zs0aNGjVSqVKldOutt+rQoUP6/PPPtW3bNp1++ulunxLC4PDhwypWrJiko6U58DXi6aefliSNHTtWkoLPsDpy5IgGDhyoF154QYsXL3ZtbiA/0KVwLLoUbHQpZIcuBYTQpXAsuhRsdClkhy4VZQrwsqIxJSUlxTiOY8qUKWPGjx8fPL5t2zbz6KOPmlKlSplx48a5OCEKQuBa+v/73/9M//79TZcuXcxTTz1ltm/fnuV2xI633nrLJCQkmEmTJpm9e/cGj2dmZpprr73WNGrUyKxduzZ4fMeOHaZ27drm888/N8aEMrN//36zZ8+eAp0d+Wfz5s2mU6dOZubMmVmOZ2ZmBv/7gQceMMnJyebZZ581W7duNf369TPFixc3CxYsKOhxgQJBl4IxdCkcjy6F7NClgOPRpWAMXQrHo0shO3Sp6MOlPvNJ+fLlNXbsWPl8Pq1cuTJ4vFKlSurTp48GDhyo+++/Xy+99JKLUyK/OY6jTz/9VFdffbWMMSpXrpymTp2qSy65RAcOHOBSCTHmjz/+0FNPPaVx48apS5cuwWdSGmPk9XrVr18/Va5cWW3bttWkSZP0wQcf6NZbb1WpUqXUvn17SQo+S69kyZI67bTTXDwbhNORI0e0efNm/ec//wleZkMKPcNKkp555hl16tRJ48eP19VXX60JEyZo9uzZatKkiVtjA/mKLgWJLoWs6FI4EboUcDy6FCS6FLKiS+FE6FJRyLUtxyiT3TNkDh06ZMaOHWs8Ho956qmnsty2detWM3r0aLNixYqCGhEu2LJli2ncuLF54YUXjDHGbNq0yVSoUMHcfffdWe7HM6xiw9dff21q1KhhVq5cecLHfPHixeauu+4yZcqUMY0bNzZXXnmlSU9PN8ZkfZYNos+qVatM+/btzRVXXGF+/PHH4HG/3298Pl/w19dee61JTk42ixcvdmNMIN/QpZAduhSORZfCydClEOvoUsgOXQrHokvhZOhS0cUxJhfv7Ipsmf+/9vWcOXP0+++/a8uWLerUqZOqVaumUqVKacyYMXrggQc0evRoDRgwIPj7TnQtbRROr776qqpVq6YrrrgimIkVK1boyiuv1MKFC5WamqoWLVron//8p15++WVJ0ldffaWLLrpIJUqUcHl6FIRRo0bp2Wef1c6dOyUpy/X1/X6/PB6Pli9fLr/frxo1aigjI0OJiYlyHEeZmZmKi+NtWaPd6tWrdd9998kYo6FDh6ply5bB2w4ePKjHHntMf/75p4YNG6bk5GQXJwXCiy4FiS6Fv0eXwt+hSyFW0aUg0aXw9+hS+Dt0qSjixm5jNPr4449NyZIlTZs2bczpp59uatasafr16xe8ZvaYMWNM0aJFzSOPPOLypMgPmzdvNv/85z9NvXr1zKxZs4LH169fb9q2bWu+/vprc8YZZ5g77rjDZGRkGGOMWbFihbn99tvNL7/84tLUKGgfffSRSUhIMF9//fUJ7/PQQw+Z22+/PcuzqI59Vg2iX3bPsDpy5Ii55557jOM4ZtGiRS5PCOQPulRso0shJ+hSyAm6FGIVXSq20aWQE3Qp5ARdKjrwHn9hsHLlSvXv319jx47VjBkztHXrVvXu3Vvz58/XM888o4MHD6pv37569NFHNW7cOO3Zs8ftkRFmVapU0eDBg9WoUSPdd999+t///ifp6LXz09LS1L59e1166aV6+eWXg8+OefXVV7Vo0SLVqFHDzdFRgM477zzFx8frlVde0caNG4PHzf+/8DotLU1r167VOeeck+VZlx4PX6pjSe3atTVu3Dg5jqPHHntM3333nYYOHarXXntN8+fPV6NGjdweEQg7uhToUsgJuhRygi6FWESXAl0KOUGXQk7QpaKEyxuPUeHHH380VatWNcuXL89yfMSIEaZmzZomJSXFGGPM4cOHzZ49e9wYEfno2GfAfP3116Znz56mQYMGZvbs2cYYY7Zt22aqV69uLr74YvPRRx+Zzz77zNx7770mMTGRayHHoEmTJpmiRYuabt26mQULFgSPb9myxfzjH/8wLVu2DD77DrFt1apV5qqrrjKnnXaaiY+PN/Pnz3d7JCDf0KViG10KuUGXQk7RpRBL6FKxjS6F3KBLIafoUoUb7/F3Csz/Xwd59uzZ6tatm6ZNm6Zzzz1X6enpio+PlySVKVNGjzzyiO677z6Xp0V+CeTgiy++0FtvvaVt27bpp59+Uv369fXcc8+pbdu2WrNmjXr27Km9e/dKkqpVq6ZnnnmGZ0jEIJ/PpzfeeEN9+/ZVxYoVdfbZZ8vv9ys1NVV+v18//fSTihQpwnstQNLRZ+4+9NBDGjlypBo0aOD2OEDY0aUg0aWQO3Qp5AZdCtGOLgWJLoXcoUshN+hShRcbf7lkjnnT02OPNWjQQJUqVdKMGTOCL5lPS0tT27ZtNXDgQHXs2NGNcVFAfvzxR7Vu3VrPP/+8WrVqpcWLF+udd97R9u3b9dxzz+nSSy9VRkaGduzYoSJFiqhEiRIqWbKk22PDRYsWLdLrr7+ulStXqlq1amrSpInuvPNOeb1e3jAZWWRkZKhIkSJujwGEDV0K2aFLIbfoUsgpuhSiDV0K2aFLIbfoUsgpulThxMZfLgTK1W+//aZffvlF8fHxqlGjhtq3b69ly5apffv2qlmzpkaMGKHixYtr2rRpevnll/Xrr79yvewo98QTT2jmzJnBa6hL0vfff69HH31UO3bs0Msvv6xWrVq5OCEKC55RBSCa0aVwInQphAtdCkA0o0vhROhSCBe6FBAdeGfOXHAcR1OmTFG7du00efJkjR8/XldffbUGDx6s+vXr69tvv1Vqaqp69uypjh076uOPP9ZXX31FuYoBJUqUUEpKinbu3Bk81rp1a3Xr1k3Lly/XjTfeqO+++869ARGRsnveBeUKQDSjS+FE6FLIC7oUgFhDl8KJ0KWQF3QpIHrxmt1cWL16te655x49+eSTuuuuu7Rnzx5Nnz5dt912mzwejx5//HEtXLhQixcvlsfjUcWKFVWhQgW3x0YBqFu3rhzH0ZdffqkbbrhBJUqUkCTVq1dPF198sRo2bKgzzjjD5SkRaezLswBAtKNL4UToUsgLuhSAWEOXwonQpZAXdCkgerHxdwJvvvmmLrnkEiUnJweP7dq1S0lJSbrmmmskHX2D5O7duyszM1N33nmn2rVrp9atW/PGuFEscFmNlStXav/+/dq/f78uueQStW/fXu3atdOgQYPk8/nUrl07ValSRV988YXOOOMMjRgxQqVLl3Z7fAAACgxdCtmhSwEAkDN0KWSHLgUAyAk2/rKxf/9+DRo0SFWqVNG0adNUtWpVSVLRokW1evVqrVmzRlWqVAl+s73ssstUuXJlbdu2zeXJkZ8Cj/fHH3+sfv36KT4+Xjt37lSjRo00duxYjR8/Xh6PR08//bQefvhhVatWTUuXLtWcOXMoVwCAmEKXQnboUgAA5AxdCtmhSwEAcsox2V3MF9q0aZPat2+vkiVL6uOPP1a1atW0f/9+devWTcWKFdPDDz8cfAbVkSNH1KJFC91///266aabXJ4c+WnOnDm64oorNHbsWF1wwQWKi4tTly5d5PP59Pbbb6thw4b6/vvvtW7dOh08eFBXXHGFzjzzTLfHBgCgwNGlkB26FAAAOUOXQnboUgCAnGDj7//5/X55PB75/X4ZY+T1erV582a1bdtWSUlJmjJliqpWraqPPvpIzz//vMqWLas777xTycnJeuutt/TGG2/o119/VfXq1d0+FYRB4FlU9q8nTJigt99+W999952KFCkij8ejI0eO6IILLlCZMmU0c+ZMF6cGAMA9dCkciy4FAEDu0KVwLLoUAOBUcKlPhcrVqlWrNG7cOG3ZskUtWrTQgAED9O2336pt27a6/vrr9dlnn6lz587yeDyaNGmS/vnPf6pu3brKyMjQV199RbmKEoE87Ny5UykpKXIcR+edd54kafv27UpNTVXRokUlSYcOHVJCQoLeeOMNXXLJJZo3b57OO+883hwXABBT6FI4Fl0KAIDcoUvhWHQpAMCp8rg9gNsC30wXL16sVq1aacuWLSpWrJgGDx6sp59+WlWrVtW3336r/fv368orr9S2bdt0ww036IMPPtAff/yhjz/+WD///LOaNGni9qkgDAJ5WLZsma6//noNHTpUI0eOVGZmpiSpQ4cO2rhxo/7zn/9IkhISEiRJGRkZKleunBITEylXAICYQpfCsehSAADkDl0Kx6JLAQDCIaZf8Rf4ZrpkyRI1b95c/fr10xNPPCG/369y5cppy5YtOnjwoKpWrapvvvlG1157ra688kp99tlnqlq1qurVq+f2KSCMjDHyeDz6448/1KpVK/Xt21d9+vRR1apV5fF4ZIxR7dq19dBDD2n8+PEyxujBBx9UWlqavvjiC8XFxfFmyQCAmEKXwrHoUgAA5A5dCseiSwEAwiXm3+Nv06ZNOvfcc3XJJZfoo48+Ch7v0qWLVqxYocOHD6t69eq69dZb1aJFC1177bU6cOCAZs2apSpVqrg4OfLDnj17dO211+rcc8/V2LFjg8cDZVySUlJS9O6772rUqFEqX768kpKStG3bNk2fPl3nnnuuW6MDAOAKuhSORZcCACB36FI4Fl0KABAOMX+pT5/Ppxo1aujIkSP66aefJElPPvmk/vvf/+qGG27QgAEDlJKSoocfflh//fWXPv30U5UuXVoZGRkuT478sH37dm3btk0dO3aU3+8PHg+UK2OMkpOT9dBDD2n+/Pnq16+fBg0apDlz5lCuAAAxiS6FY9GlAADIHboUjkWXAgCEQ8y/4k+SVq9erfvuu0/x8fGqUKGCPvvsM73zzju6/PLLJUkbN25U9erVNX78eN15553KzMxUXFxMXyU1ar3//vu6+eablZ6eLsdxsjyjKuDgwYP6/fff1axZM5emBAAgstClEECXAgAg9+hSCKBLAQDCIeZf8SdJtWvX1tixY3Xo0CG99957euihh3T55ZfLGKOMjAx5vV6dc845KleunCRRrqJY9erVFRcXp08++USSjitXkvT666/r4YcfVnp6ekGPBwBARKJLIYAuBQBA7tGlEECXAgCEAxt//69OnTqaMGGCLrroIv3vf//T7Nmz5TiOihQpopdffln79+/XBRdc4PaYyGfJyclKTEzU22+/rZSUlODxY18Yu2HDBp133nkqUqSIGyMCABCR6FKQ6FIAAOQVXQoSXQoAEB5c6tMSuLyCMUajRo3SN998o+HDh+vnn39WkyZN3B4PBeCTTz5Rt27d1LlzZw0aNEj169eXdPRSCo8//rjef/99zZgxQ3Xq1HF5UgAAIg9dCnQpAADyji4FuhQA4FSx8ZeN1atXq3///vrtt9+0d+9e/fLLLzrvvPPcHgsFxO/3a+LEibrnnnt05plnqnnz5ipWrJi2bNmiOXPm6KuvvqJsAwBwEnSp2EaXAgDg1NClYhtdCgBwqtj4O4GVK1fqoYce0siRI9WgQQO3x4ELfvvtNz399NNas2aNSpUqpRYtWqh3796qXbu226MBABDx6FKgSwEAkHd0KdClAAB5xcbfSWRkZHC97Bjn8/nk9XrdHgMAgEKJLgW6FAAAeUeXAl0KAJAXbPwBJ2GMkeM4x/03AAAA/h5dCgAAIO/oUgCAvGDjDwAAAAAAAAAAAIgCHrcHAAAAAAAAAAAAAHDq2PgDAAAAAAAAAAAAogAbfwAAAAAAAAAAAEAUYOMPAAAAAAAAAAAAiAJs/AEAAAAAAAAAAABRgI0/AAAAAAAAAAAAIAqw8QcAAAAAAAAAAABEATb+AES9Nm3a6F//+pckqXr16nruuedcnQcAAKAwoUsBAADkHV0KQEFj4w8AAAAAAAAAAACIAmz8AQAAAAAAAAAAAFGAjT8AMc1xHL388su66qqrVLx4cdWrV0+//PKL1qxZozZt2qhEiRJq0aKF1q5d6/aoAAAAEYcuBQAAkHd0KQD5gY0/ADHvscce00033aRFixapbt266tatm/r06aN///vfmjdvnowxuueee9weEwAAICLRpQAAAPKOLgUg3Nj4AxDzbr31VnXu3Fl16tTRwIEDtWHDBnXv3l1XXHGF6tWrp/vvv1/fffed22MCAABEJLoUAABA3tGlAIQbG38AYl7Dhg2D/12xYkVJ0jnnnJPl2OHDh5WWllbgswEAAEQ6uhQAAEDe0aUAhBsbfwBiXpEiRYL/7TjOCY/5/f6CHQwAAKAQoEsBAADkHV0KQLix8QcAAAAAAAAAAABEATb+AAAAAAAAAAAAgCjAxh8AAAAAAAAAAAAQBRxjjHF7CAAAAAAAAAAAAACnhlf8AQAAAAAAAAAAAFGAjT8AAAAAAAAAAAAgCrDxBwAAAAAAAAAAAEQBNv4AAAAAAAAAAACAKMDGHwAAAAAAAAAAABAF2PgDAAAAAAAAAAAAogAbfwAAAAAAAAAAAEAUYOMPAAAAAAAAAAAAiAJs/AEAAAAAAAAAAABRgI0/AAAAAAAAAAAAIAqw8QcAAAAAAAAAAABEATb+AAAAAAAAAAAAgCjwfzSh0Z4TRw0JAAAAAElFTkSuQmCC\n" + }, + "metadata": {} + } + ] + }, + { + "cell_type": "code", + "source": [ + "# Define Bloom level labels for clarity\n", + "bloom_level_labels = {\n", + " 1: \"Remember\",\n", + " 2: \"Understand\",\n", + " 3: \"Apply\",\n", + " 4: \"Analyze\",\n", + " 5: \"Evaluate\",\n", + " 6: \"Create\"\n", + "}\n", + "bloom_levels = df['bloom_level'].unique()\n", + "# Create a figure with 6 subplots (2 rows x 3 columns)\n", + "fig, axes = plt.subplots(2, 3, figsize=(18, 12), sharey=True)\n", + "\n", + "# Flatten the axes array for easy iteration\n", + "axes = axes.flatten()\n", + "\n", + "# Set y-axis limits\n", + "y_limits = (-2, 2)\n", + "\n", + "# Iterate over each Bloom level and its corresponding subplot\n", + "for idx, level in enumerate(bloom_levels):\n", + " subset = df[df['bloom_level'] == level]\n", + "\n", + " # Perform ANOVA\n", + " anova = f_oneway(*[subset[subset['llm'] == llm]['bloom_mean'] for llm in subset['llm'].unique()])\n", + " print(f\"ANOVA Results for Bloom Level {level}: \", anova)\n", + "\n", + " # Post-hoc test\n", + " tukey = pairwise_tukeyhsd(endog=subset['bloom_mean'], groups=subset['llm'], alpha=0.05)\n", + " print(f\"\\nPost-hoc test for Bloom Level {level}:\\n\", tukey)\n", + "\n", + " # Plotting\n", + " ax = axes[idx]\n", + " palette = sns.color_palette(\"tab10\")\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n", + "\n", + " ax.set_ylim(y_limits)\n", + " ax.set_yticks([-2, -1, 0, 1, 2])\n", + " ax.set_yticklabels([-2, -1, 0, 1, 2])\n", + " ax.set_xlabel('LLM', fontsize=12)\n", + " ax.set_ylabel('Score', fontsize=12)\n", + "\n", + " # Set x-axis labels\n", + " llm_labels = ['BloomGPT', 'Code Tutor', 'Copilot', 'LLaMa']\n", + " ax.set_xticks(range(len(llm_labels)))\n", + " ax.set_xticklabels(llm_labels, rotation=45, ha='right')\n", + "\n", + " # Set title for each subplot\n", + " # ax.set_title(f'Bloom Alignment by LLM for Level \"{bloom_level_labels[level]}\"', fontsize=14)\n", + "\n", + "# Adjust layout to prevent clipping and show the plot\n", + "plt.tight_layout()\n", + "plt.show()\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 1000 + }, + "id": "qKbdV81JD4SL", + "outputId": "19c05443-0b16-42b0-bfa2-aa2c0f19f2f5" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "ANOVA Results for Bloom Level 1: F_onewayResult(statistic=1.5963337074448187, pvalue=0.21754362517477804)\n", + "\n", + "Post-hoc test for Bloom Level 1:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05\n", + "===================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "---------------------------------------------------\n", + " 1 2 0.05 0.9969 -0.6245 0.7245 False\n", + " 1 3 -0.4 0.5625 -1.2412 0.4412 False\n", + " 1 4 -0.46 0.3799 -1.2388 0.3188 False\n", + " 2 3 -0.45 0.4941 -1.3207 0.4207 False\n", + " 2 4 -0.51 0.3264 -1.3206 0.3006 False\n", + " 3 4 -0.06 0.9981 -1.0138 0.8938 False\n", + "---------------------------------------------------\n", + "ANOVA Results for Bloom Level 2: F_onewayResult(statistic=2.5371540401366386, pvalue=0.08165720744535691)\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":35: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n", + ":35: UserWarning: The palette list has more values (10) than needed (4), which may not be intended.\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "Post-hoc test for Bloom Level 2:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05\n", + "===================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "---------------------------------------------------\n", + " 1 2 0.2861 0.4028 -0.2114 0.7836 False\n", + " 1 3 0.135 0.9179 -0.4486 0.7186 False\n", + " 1 4 -0.265 0.5986 -0.8486 0.3186 False\n", + " 2 3 -0.1511 0.8831 -0.7221 0.4199 False\n", + " 2 4 -0.5511 0.0612 -1.1221 0.0199 False\n", + " 3 4 -0.4 0.3417 -1.0475 0.2475 False\n", + "---------------------------------------------------\n", + "ANOVA Results for Bloom Level 3: F_onewayResult(statistic=2.0195885928003734, pvalue=0.13581965180674502)\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":35: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n", + ":35: UserWarning: The palette list has more values (10) than needed (4), which may not be intended.\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "Post-hoc test for Bloom Level 3:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05\n", + "===================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "---------------------------------------------------\n", + " 1 2 -0.0889 0.9806 -0.7274 0.5496 False\n", + " 1 3 0.0444 0.9975 -0.594 0.6829 False\n", + " 1 4 -0.7333 0.1422 -1.6363 0.1696 False\n", + " 2 3 0.1333 0.9393 -0.5052 0.7718 False\n", + " 2 4 -0.6444 0.2296 -1.5474 0.2585 False\n", + " 3 4 -0.7778 0.1098 -1.6807 0.1252 False\n", + "---------------------------------------------------\n", + "ANOVA Results for Bloom Level 4: F_onewayResult(statistic=6.170470141259958, pvalue=0.0033182252322939594)\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":35: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n", + ":35: UserWarning: The palette list has more values (10) than needed (4), which may not be intended.\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "Post-hoc test for Bloom Level 4:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n", + "====================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "----------------------------------------------------\n", + " 1 2 -0.0944 0.9665 -0.6619 0.473 False\n", + " 1 3 -0.7302 0.0115 -1.3187 -0.1417 True\n", + " 1 4 -0.9444 0.0408 -1.8573 -0.0316 True\n", + " 2 3 -0.6357 0.0369 -1.2401 -0.0313 True\n", + " 2 4 -0.85 0.0784 -1.7732 0.0732 False\n", + " 3 4 -0.2143 0.9194 -1.1506 0.722 False\n", + "----------------------------------------------------\n", + "ANOVA Results for Bloom Level 5: F_onewayResult(statistic=2.398786778521197, pvalue=0.0890654046509717)\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":35: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n", + ":35: UserWarning: The palette list has more values (10) than needed (4), which may not be intended.\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "Post-hoc test for Bloom Level 5:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05\n", + "===================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "---------------------------------------------------\n", + " 1 2 -0.2545 0.7738 -0.9798 0.4707 False\n", + " 1 3 -0.325 0.6763 -1.1124 0.4624 False\n", + " 1 4 -1.0667 0.0576 -2.1593 0.026 False\n", + " 2 3 -0.0705 0.9944 -0.8417 0.7008 False\n", + " 2 4 -0.8121 0.1941 -1.8933 0.269 False\n", + " 3 4 -0.7417 0.2936 -1.8654 0.3821 False\n", + "---------------------------------------------------\n", + "ANOVA Results for Bloom Level 6: F_onewayResult(statistic=2.003861500836659, pvalue=0.14029288297114362)\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":35: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n", + ":35: UserWarning: The palette list has more values (10) than needed (4), which may not be intended.\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "Post-hoc test for Bloom Level 6:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05\n", + "===================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "---------------------------------------------------\n", + " 1 2 0.6167 0.2917 -0.3213 1.5546 False\n", + " 1 3 0.5357 0.3677 -0.3565 1.4279 False\n", + " 1 4 -0.2167 0.9598 -1.4276 0.9942 False\n", + " 2 3 -0.081 0.9964 -1.1246 0.9627 False\n", + " 2 4 -0.8333 0.3293 -2.1598 0.4932 False\n", + " 3 4 -0.7524 0.3957 -2.0469 0.5421 False\n", + "---------------------------------------------------\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":35: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n", + ":35: UserWarning: The palette list has more values (10) than needed (4), which may not be intended.\n", + " sns.boxplot(x='llm', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "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\n" + }, + "metadata": {} + } + ] + }, + { + "cell_type": "code", + "source": [ + "llm_labels = {\n", + " 1: \"BloomGPT\",\n", + " 2: \"Code Tutor\",\n", + " 3: \"Copilot\",\n", + " 4: \"LLaMa\"\n", + "}\n", + "\n", + "\n", + "# Define LLMs and Bloom levels\n", + "llms = df['llm'].unique()\n", + "bloom_levels = sorted(df['bloom_level'].unique()) # Ensure levels are sorted\n", + "\n", + "# Create a figure with 2 rows and 3 columns of subplots\n", + "fig, axes = plt.subplots(2, 2, figsize=(13, 10), sharey=True)\n", + "\n", + "# Flatten the axes array for easy iteration\n", + "axes = axes.flatten()\n", + "\n", + "# Define a color palette\n", + "# palette = sns.color_palette(\"seismic\")\n", + "palette = ['indigo', 'blue', 'green', 'yellow', 'orange', 'red']\n", + "\n", + "\n", + "# Set y-axis limits\n", + "y_limits = (-2, 2)\n", + "\n", + "# Iterate over each LLM and its corresponding subplot\n", + "for idx, llm in enumerate(llms):\n", + " subset = df[df['llm'] == llm]\n", + "\n", + " # Perform ANOVA\n", + " anova = f_oneway(*[subset[subset['bloom_level'] == level]['bloom_mean'] for level in bloom_levels])\n", + " print(f\"ANOVA Results for LLM {llm}: \", anova)\n", + "\n", + " # Post-hoc test\n", + " tukey = pairwise_tukeyhsd(endog=subset['bloom_mean'], groups=subset['bloom_level'], alpha=0.05)\n", + " print(f\"\\nPost-hoc test for LLM {llm}:\\n\", tukey)\n", + "\n", + " # Plotting\n", + " ax = axes[idx]\n", + " sns.boxplot(x='bloom_level', y='bloom_mean', data=subset, palette=palette, ax=ax)\n", + "\n", + " ax.set_ylim(y_limits)\n", + " ax.set_yticks([-2, -1, 0, 1, 2])\n", + " ax.set_yticklabels([-2, -1, 0, 1, 2])\n", + " ax.set_xlabel('Bloom Level', fontsize=10)\n", + " ax.set_ylabel('Score', fontsize=10)\n", + "\n", + " # Set x-axis labels\n", + " bloom_labels = ['Remember', 'Understand', 'Apply', 'Analyze', 'Evaluate', 'Create']\n", + " ax.set_xticks(range(len(bloom_levels)))\n", + " ax.set_xticklabels(bloom_labels, rotation=45, ha='right')\n", + "\n", + " # Set title for each subplot\n", + " # ax.set_title(f'Bloom Alignment for \"{llm_labels[llm]}\"', fontsize=14)\n", + "\n", + "# Adjust layout to prevent clipping and show the plot\n", + "plt.tight_layout()\n", + "plt.show()\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 1000 + }, + "id": "WSFOkZ81FTgh", + "outputId": "21356997-a7a2-4f47-8de1-f442b56a45a4" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "ANOVA Results for LLM 1: F_onewayResult(statistic=5.456765553913385, pvalue=0.000413823816268813)\n", + "\n", + "Post-hoc test for LLM 1:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n", + "====================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "----------------------------------------------------\n", + " 1 2 -0.275 0.8949 -1.0429 0.4929 False\n", + " 1 3 0.2333 0.9374 -0.5104 0.9771 False\n", + " 1 4 0.3444 0.7443 -0.3993 1.0882 False\n", + " 1 5 -0.1 0.9985 -0.8239 0.6239 False\n", + " 1 6 -0.75 0.0268 -1.4431 -0.0569 True\n", + " 2 3 0.5083 0.4066 -0.2783 1.2949 False\n", + " 2 4 0.6194 0.2009 -0.1671 1.406 False\n", + " 2 5 0.175 0.984 -0.5929 0.9429 False\n", + " 2 6 -0.475 0.4126 -1.2139 0.2639 False\n", + " 3 4 0.1111 0.998 -0.652 0.8742 False\n", + " 3 5 -0.3333 0.7694 -1.0771 0.4104 False\n", + " 3 6 -0.9833 0.0021 -1.6972 -0.2695 True\n", + " 4 5 -0.4444 0.4948 -1.1882 0.2993 False\n", + " 4 6 -1.0944 0.0005 -1.8083 -0.3806 True\n", + " 5 6 -0.65 0.0779 -1.3431 0.0431 False\n", + "----------------------------------------------------\n", + "ANOVA Results for LLM 2: F_onewayResult(statistic=1.8901211372007738, pvalue=0.11493763845043799)\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":41: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='bloom_level', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "Post-hoc test for LLM 2:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05\n", + "===================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "---------------------------------------------------\n", + " 1 2 -0.0389 1.0 -0.7338 0.656 False\n", + " 1 3 0.0944 0.9985 -0.6005 0.7893 False\n", + " 1 4 0.2 0.9598 -0.5151 0.9151 False\n", + " 1 5 -0.4045 0.4688 -1.0691 0.26 False\n", + " 1 6 -0.1833 0.9802 -0.9557 0.589 False\n", + " 2 3 0.1333 0.9913 -0.5408 0.8075 False\n", + " 2 4 0.2389 0.9077 -0.456 0.9338 False\n", + " 2 5 -0.3657 0.5435 -1.0084 0.2771 False\n", + " 2 6 -0.1444 0.9925 -0.8982 0.6093 False\n", + " 3 4 0.1056 0.9975 -0.5893 0.8005 False\n", + " 3 5 -0.499 0.2114 -1.1418 0.1438 False\n", + " 3 6 -0.2778 0.8801 -1.0315 0.476 False\n", + " 4 5 -0.6045 0.0936 -1.2691 0.06 False\n", + " 4 6 -0.3833 0.6802 -1.1557 0.389 False\n", + " 5 6 0.2212 0.9427 -0.5046 0.947 False\n", + "---------------------------------------------------\n", + "ANOVA Results for LLM 3: F_onewayResult(statistic=1.8350943829712325, pvalue=0.13224279995948046)\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":41: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='bloom_level', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "Post-hoc test for LLM 3:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05\n", + "===================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "---------------------------------------------------\n", + " 1 2 0.26 0.9815 -0.8726 1.3926 False\n", + " 1 3 0.6778 0.3546 -0.3368 1.6924 False\n", + " 1 4 0.0143 1.0 -1.044 1.0726 False\n", + " 1 5 -0.025 1.0 -1.0589 1.0089 False\n", + " 1 6 0.1857 0.9946 -0.8726 1.244 False\n", + " 2 3 0.4178 0.7616 -0.524 1.3595 False\n", + " 2 4 -0.2457 0.9739 -1.2343 0.7429 False\n", + " 2 5 -0.285 0.9454 -1.2475 0.6775 False\n", + " 2 6 -0.0743 0.9999 -1.0629 0.9143 False\n", + " 3 4 -0.6635 0.2014 -1.5144 0.1874 False\n", + " 3 5 -0.7028 0.1287 -1.5232 0.1176 False\n", + " 3 6 -0.4921 0.5128 -1.3429 0.3588 False\n", + " 4 5 -0.0393 1.0 -0.9131 0.8346 False\n", + " 4 6 0.1714 0.9921 -0.7311 1.0739 False\n", + " 5 6 0.2107 0.9771 -0.6631 1.0846 False\n", + "---------------------------------------------------\n", + "ANOVA Results for LLM 4: F_onewayResult(statistic=0.9812065068037764, pvalue=0.4608482814833007)\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":41: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='bloom_level', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "Post-hoc test for LLM 4:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05\n", + "===================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "---------------------------------------------------\n", + " 1 2 -0.08 0.9999 -1.1798 1.0198 False\n", + " 1 3 -0.04 1.0 -1.31 1.23 False\n", + " 1 4 -0.14 0.9995 -1.595 1.315 False\n", + " 1 5 -0.7067 0.4896 -1.9767 0.5633 False\n", + " 1 6 -0.5067 0.7827 -1.7767 0.7633 False\n", + " 2 3 0.04 1.0 -1.23 1.31 False\n", + " 2 4 -0.06 1.0 -1.515 1.395 False\n", + " 2 5 -0.6267 0.6088 -1.8967 0.6433 False\n", + " 2 6 -0.4267 0.8773 -1.6967 0.8433 False\n", + " 3 4 -0.1 0.9999 -1.6875 1.4875 False\n", + " 3 5 -0.6667 0.6546 -2.0866 0.7532 False\n", + " 3 6 -0.4667 0.8866 -1.8866 0.9532 False\n", + " 4 5 -0.5667 0.8486 -2.1541 1.0208 False\n", + " 4 6 -0.3667 0.9718 -1.9541 1.2208 False\n", + " 5 6 0.2 0.997 -1.2199 1.6199 False\n", + "---------------------------------------------------\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":41: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='bloom_level', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "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\n" + }, + "metadata": {} + } + ] + }, + { + "cell_type": "code", + "source": [ + "llm_labels = {\n", + " 1: \"BloomGPT\",\n", + " 2: \"Code Tutor\",\n", + " 3: \"Copilot\",\n", + " 4: \"LLaMa\"\n", + "}\n", + "\n", + "\n", + "# Define LLMs and Bloom levels\n", + "llms = df['llm'].unique()\n", + "bloom_levels = sorted(df['bloom_level'].unique()) # Ensure levels are sorted\n", + "\n", + "# Create a figure with 2 rows and 3 columns of subplots\n", + "fig, axes = plt.subplots(2, 2, figsize=(13, 10), sharey=True)\n", + "\n", + "# Flatten the axes array for easy iteration\n", + "axes = axes.flatten()\n", + "\n", + "# Define a color palette\n", + "# palette = sns.color_palette(\"seismic\")\n", + "# palette = ['indigo', 'blue', 'green', 'yellow', 'orange', 'red']\n", + "\n", + "palette = sns.cubehelix_palette(rot=-.2)\n", + "\n", + "\n", + "\n", + "# Set y-axis limits\n", + "y_limits = (-2, 2)\n", + "\n", + "# Iterate over each LLM and its corresponding subplot\n", + "for idx, llm in enumerate(llms):\n", + " subset = df[df['llm'] == llm]\n", + "\n", + " # Perform ANOVA\n", + " anova = f_oneway(*[subset[subset['bloom_level'] == level]['bloom_mean'] for level in bloom_levels])\n", + " print(f\"ANOVA Results for LLM {llm}: \", anova)\n", + "\n", + " # Post-hoc test\n", + " tukey = pairwise_tukeyhsd(endog=subset['bloom_mean'], groups=subset['bloom_level'], alpha=0.05)\n", + " print(f\"\\nPost-hoc test for LLM {llm}:\\n\", tukey)\n", + "\n", + " # Plotting\n", + " ax = axes[idx]\n", + " sns.boxplot(x='bloom_level', y='bloom_mean', data=subset, palette=palette, ax=ax)\n", + "\n", + " ax.set_ylim(y_limits)\n", + " ax.set_yticks([-2, -1, 0, 1, 2])\n", + " ax.set_yticklabels([-2, -1, 0, 1, 2])\n", + " ax.set_xlabel('Bloom Level', fontsize=10)\n", + " ax.set_ylabel('Score', fontsize=10)\n", + "\n", + " # Set x-axis labels\n", + " bloom_labels = ['Remember', 'Understand', 'Apply', 'Analyze', 'Evaluate', 'Create']\n", + " ax.set_xticks(range(len(bloom_levels)))\n", + " ax.set_xticklabels(bloom_labels, rotation=45, ha='right')\n", + "\n", + " # Set title for each subplot\n", + " # ax.set_title(f'Bloom Alignment for \"{llm_labels[llm]}\"', fontsize=14)\n", + "\n", + "# Adjust layout to prevent clipping and show the plot\n", + "plt.tight_layout()\n", + "plt.show()\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 1000 + }, + "id": "-BpP1YEDJ8PS", + "outputId": "7f0851ba-8f5c-4486-ac9c-05a95cfa0b51" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "ANOVA Results for LLM 1: F_onewayResult(statistic=5.456765553913385, pvalue=0.000413823816268813)\n", + "\n", + "Post-hoc test for LLM 1:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05 \n", + "====================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "----------------------------------------------------\n", + " 1 2 -0.275 0.8949 -1.0429 0.4929 False\n", + " 1 3 0.2333 0.9374 -0.5104 0.9771 False\n", + " 1 4 0.3444 0.7443 -0.3993 1.0882 False\n", + " 1 5 -0.1 0.9985 -0.8239 0.6239 False\n", + " 1 6 -0.75 0.0268 -1.4431 -0.0569 True\n", + " 2 3 0.5083 0.4066 -0.2783 1.2949 False\n", + " 2 4 0.6194 0.2009 -0.1671 1.406 False\n", + " 2 5 0.175 0.984 -0.5929 0.9429 False\n", + " 2 6 -0.475 0.4126 -1.2139 0.2639 False\n", + " 3 4 0.1111 0.998 -0.652 0.8742 False\n", + " 3 5 -0.3333 0.7694 -1.0771 0.4104 False\n", + " 3 6 -0.9833 0.0021 -1.6972 -0.2695 True\n", + " 4 5 -0.4444 0.4948 -1.1882 0.2993 False\n", + " 4 6 -1.0944 0.0005 -1.8083 -0.3806 True\n", + " 5 6 -0.65 0.0779 -1.3431 0.0431 False\n", + "----------------------------------------------------\n", + "ANOVA Results for LLM 2: F_onewayResult(statistic=1.8901211372007738, pvalue=0.11493763845043799)\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":44: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='bloom_level', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "Post-hoc test for LLM 2:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05\n", + "===================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "---------------------------------------------------\n", + " 1 2 -0.0389 1.0 -0.7338 0.656 False\n", + " 1 3 0.0944 0.9985 -0.6005 0.7893 False\n", + " 1 4 0.2 0.9598 -0.5151 0.9151 False\n", + " 1 5 -0.4045 0.4688 -1.0691 0.26 False\n", + " 1 6 -0.1833 0.9802 -0.9557 0.589 False\n", + " 2 3 0.1333 0.9913 -0.5408 0.8075 False\n", + " 2 4 0.2389 0.9077 -0.456 0.9338 False\n", + " 2 5 -0.3657 0.5435 -1.0084 0.2771 False\n", + " 2 6 -0.1444 0.9925 -0.8982 0.6093 False\n", + " 3 4 0.1056 0.9975 -0.5893 0.8005 False\n", + " 3 5 -0.499 0.2114 -1.1418 0.1438 False\n", + " 3 6 -0.2778 0.8801 -1.0315 0.476 False\n", + " 4 5 -0.6045 0.0936 -1.2691 0.06 False\n", + " 4 6 -0.3833 0.6802 -1.1557 0.389 False\n", + " 5 6 0.2212 0.9427 -0.5046 0.947 False\n", + "---------------------------------------------------\n", + "ANOVA Results for LLM 3: F_onewayResult(statistic=1.8350943829712325, pvalue=0.13224279995948046)\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":44: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='bloom_level', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "Post-hoc test for LLM 3:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05\n", + "===================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "---------------------------------------------------\n", + " 1 2 0.26 0.9815 -0.8726 1.3926 False\n", + " 1 3 0.6778 0.3546 -0.3368 1.6924 False\n", + " 1 4 0.0143 1.0 -1.044 1.0726 False\n", + " 1 5 -0.025 1.0 -1.0589 1.0089 False\n", + " 1 6 0.1857 0.9946 -0.8726 1.244 False\n", + " 2 3 0.4178 0.7616 -0.524 1.3595 False\n", + " 2 4 -0.2457 0.9739 -1.2343 0.7429 False\n", + " 2 5 -0.285 0.9454 -1.2475 0.6775 False\n", + " 2 6 -0.0743 0.9999 -1.0629 0.9143 False\n", + " 3 4 -0.6635 0.2014 -1.5144 0.1874 False\n", + " 3 5 -0.7028 0.1287 -1.5232 0.1176 False\n", + " 3 6 -0.4921 0.5128 -1.3429 0.3588 False\n", + " 4 5 -0.0393 1.0 -0.9131 0.8346 False\n", + " 4 6 0.1714 0.9921 -0.7311 1.0739 False\n", + " 5 6 0.2107 0.9771 -0.6631 1.0846 False\n", + "---------------------------------------------------\n", + "ANOVA Results for LLM 4: F_onewayResult(statistic=0.9812065068037764, pvalue=0.4608482814833007)\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":44: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='bloom_level', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "Post-hoc test for LLM 4:\n", + " Multiple Comparison of Means - Tukey HSD, FWER=0.05\n", + "===================================================\n", + "group1 group2 meandiff p-adj lower upper reject\n", + "---------------------------------------------------\n", + " 1 2 -0.08 0.9999 -1.1798 1.0198 False\n", + " 1 3 -0.04 1.0 -1.31 1.23 False\n", + " 1 4 -0.14 0.9995 -1.595 1.315 False\n", + " 1 5 -0.7067 0.4896 -1.9767 0.5633 False\n", + " 1 6 -0.5067 0.7827 -1.7767 0.7633 False\n", + " 2 3 0.04 1.0 -1.23 1.31 False\n", + " 2 4 -0.06 1.0 -1.515 1.395 False\n", + " 2 5 -0.6267 0.6088 -1.8967 0.6433 False\n", + " 2 6 -0.4267 0.8773 -1.6967 0.8433 False\n", + " 3 4 -0.1 0.9999 -1.6875 1.4875 False\n", + " 3 5 -0.6667 0.6546 -2.0866 0.7532 False\n", + " 3 6 -0.4667 0.8866 -1.8866 0.9532 False\n", + " 4 5 -0.5667 0.8486 -2.1541 1.0208 False\n", + " 4 6 -0.3667 0.9718 -1.9541 1.2208 False\n", + " 5 6 0.2 0.997 -1.2199 1.6199 False\n", + "---------------------------------------------------\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + ":44: FutureWarning: \n", + "\n", + "Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.\n", + "\n", + " sns.boxplot(x='bloom_level', y='bloom_mean', data=subset, palette=palette, ax=ax)\n" + ] + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "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\n" + }, + "metadata": {} + } + ] + }, + { + "cell_type": "code", + "source": [ + "average_per_llm = df.groupby('llm')[['bloom_mean', 'subject_mean', 'persian_mean']].mean().reset_index()\n", + "\n", + "# Display the result\n", + "print(average_per_llm)" + ], + "metadata": { + "id": "4Z5G35DH3P_H" + }, + "execution_count": null, + "outputs": [] + }, + { + "cell_type": "code", + "source": [ + "import pandas as pd\n", + "import pingouin as pg\n", + "\n", + "# Assuming df is your full DataFrame\n", + "aspects = ['bloom', 'subject', 'persian']\n", + "icc_results = {}\n", + "\n", + "for aspect in aspects:\n", + " # Reshape the DataFrame for each aspect\n", + " aspect_data = df[[f'expert{i}_{aspect}' for i in range(1, 6)]]\n", + "\n", + " # Calculate ICC(3) for each aspect\n", + " icc = pg.intraclass_corr(data=pd.melt(aspect_data.reset_index(), id_vars=['index']),\n", + " targets='index',\n", + " raters='variable',\n", + " ratings='value')\n", + " icc_value = icc[icc['Type'] == 'ICC3'].iloc[0]['ICC']\n", + " icc_results[aspect] = icc_value\n", + "\n", + "print(\"ICC(3) for Bloom Aspect: \", icc_results['bloom'])\n", + "print(\"ICC(3) for Subject Aspect: \", icc_results['subject'])\n", + "print(\"ICC(3) for Persian Aspect: \", icc_results['persian'])\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "sQtUUCJgNykE", + "outputId": "96d8d288-e257-4f2f-e693-a507cde9c44e" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "ICC(3) for Bloom Aspect: 0.24766316392528162\n", + "ICC(3) for Subject Aspect: 0.03564926079673736\n", + "ICC(3) for Persian Aspect: 0.28197912500238065\n" + ] + } + ] + }, + { + "cell_type": "code", + "source": [ + "import pandas as pd\n", + "import pingouin as pg\n", + "from itertools import combinations\n", + "\n", + "# Assuming df is your full DataFrame\n", + "aspects = ['bloom', 'subject', 'persian']\n", + "num_experts = 5\n", + "icc_results = {}\n", + "\n", + "for aspect in aspects:\n", + " # Store ICC results for each aspect\n", + " aspect_icc_results = {}\n", + "\n", + " # Generate all combinations of experts with one expert removed\n", + " for i in range(1, num_experts + 1):\n", + " experts_to_keep = [j for j in range(1, num_experts + 1) if j != i]\n", + " aspect_data = df[[f'expert{j}_{aspect}' for j in experts_to_keep]]\n", + "\n", + " # Calculate ICC(3) for each combination of experts\n", + " icc = pg.intraclass_corr(data=pd.melt(aspect_data.reset_index(), id_vars=['index']),\n", + " targets='index',\n", + " raters='variable',\n", + " ratings='value')\n", + " icc_value = icc[icc['Type'] == 'ICC3'].iloc[0]['ICC']\n", + "\n", + " # Store the ICC value with information about which expert was removed\n", + " aspect_icc_results[f'removed_expert_{i}'] = icc_value\n", + "\n", + " # Store the results for the current aspect\n", + " icc_results[aspect] = aspect_icc_results\n", + "\n", + "# Display the results\n", + "for aspect, results in icc_results.items():\n", + " print(f\"\\nICC(3) for {aspect.capitalize()} Aspect:\")\n", + " for removed_expert, icc_value in results.items():\n", + " print(f\" - {removed_expert}: {icc_value}\")\n" + ], + "metadata": { + "id": "74DSs9o-X-Ab", + "colab": { + "base_uri": "https://localhost:8080/" + }, + "outputId": "6c00f8d0-4e21-4a86-b434-b49b6b19bac8" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "ICC(3) for Bloom Aspect:\n", + " - removed_expert_1: 0.20272054511636603\n", + " - removed_expert_2: 0.23314443621217876\n", + " - removed_expert_3: 0.23185539534700264\n", + " - removed_expert_4: 0.30623775570990786\n", + " - removed_expert_5: 0.2577864885356637\n", + "\n", + "ICC(3) for Subject Aspect:\n", + " - removed_expert_1: 0.03162602574573157\n", + " - removed_expert_2: -0.04670226719130603\n", + " - removed_expert_3: 0.042087447108603626\n", + " - removed_expert_4: 0.0329998474247476\n", + " - removed_expert_5: 0.08219289175435389\n", + "\n", + "ICC(3) for Persian Aspect:\n", + " - removed_expert_1: 0.22273981298371529\n", + " - removed_expert_2: 0.25655846231109886\n", + " - removed_expert_3: 0.27915491267030773\n", + " - removed_expert_4: 0.365293867756639\n", + " - removed_expert_5: 0.2702629990765585\n" + ] + } + ] + }, + { + "cell_type": "code", + "source": [ + "import pandas as pd\n", + "import pingouin as pg\n", + "from itertools import combinations\n", + "\n", + "# Assuming df is your full DataFrame\n", + "aspects = ['bloom', 'subject', 'persian']\n", + "num_experts = 5\n", + "icc_results = {}\n", + "\n", + "for aspect in aspects:\n", + " # Store ICC results for each aspect\n", + " aspect_icc_results = {}\n", + "\n", + " # Generate all combinations of experts with two experts removed\n", + " for combo in combinations(range(1, num_experts + 1), num_experts - 2):\n", + " experts_to_keep = list(combo)\n", + " aspect_data = df[[f'expert{j}_{aspect}' for j in experts_to_keep]]\n", + "\n", + " # Calculate ICC(3) for each combination of experts\n", + " icc = pg.intraclass_corr(data=pd.melt(aspect_data.reset_index(), id_vars=['index']),\n", + " targets='index',\n", + " raters='variable',\n", + " ratings='value')\n", + " icc_value = icc[icc['Type'] == 'ICC3'].iloc[0]['ICC']\n", + "\n", + " # Store the ICC value with information about which experts were removed\n", + " removed_experts = [j for j in range(1, num_experts + 1) if j not in experts_to_keep]\n", + " aspect_icc_results[f'removed_experts_{removed_experts}'] = icc_value\n", + "\n", + " # Store the results for the current aspect\n", + " icc_results[aspect] = aspect_icc_results\n", + "\n", + "# Display the results\n", + "for aspect, results in icc_results.items():\n", + " print(f\"\\nICC(3) for {aspect.capitalize()} Aspect:\")\n", + " for removed_experts, icc_value in results.items():\n", + " print(f\" - {removed_experts}: {icc_value}\")\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "xskjM1_bWmiW", + "outputId": "d50aa742-e900-42fa-bb04-b6a57d123df5" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "ICC(3) for Bloom Aspect:\n", + " - removed_experts_[4, 5]: 0.3418077474892396\n", + " - removed_experts_[3, 5]: 0.2354240419449059\n", + " - removed_experts_[3, 4]: 0.3061646888997388\n", + " - removed_experts_[2, 5]: 0.2832413415084961\n", + " - removed_experts_[2, 4]: 0.2583339931902764\n", + " - removed_experts_[2, 3]: 0.2051602796417968\n", + " - removed_experts_[1, 5]: 0.157390767157138\n", + " - removed_experts_[1, 4]: 0.3140818976206799\n", + " - removed_experts_[1, 3]: 0.16720084787966236\n", + " - removed_experts_[1, 2]: 0.1756050306482049\n", + "\n", + "ICC(3) for Subject Aspect:\n", + " - removed_experts_[4, 5]: 0.10181736928347998\n", + " - removed_experts_[3, 5]: 0.1146634716441758\n", + " - removed_experts_[3, 4]: 0.036641740312712515\n", + " - removed_experts_[2, 5]: 0.004855005904736855\n", + " - removed_experts_[2, 4]: -0.0777073999691023\n", + " - removed_experts_[2, 3]: -0.07834549878345497\n", + " - removed_experts_[1, 5]: 0.06697267407526683\n", + " - removed_experts_[1, 4]: 0.009849157054126556\n", + " - removed_experts_[1, 3]: 0.04220795017758979\n", + " - removed_experts_[1, 2]: -0.030314120158585457\n", + "\n", + "ICC(3) for Persian Aspect:\n", + " - removed_experts_[4, 5]: 0.3885221681553873\n", + " - removed_experts_[3, 5]: 0.25724211158045374\n", + " - removed_experts_[3, 4]: 0.3983084928769027\n", + " - removed_experts_[2, 5]: 0.20752219802311947\n", + " - removed_experts_[2, 4]: 0.36369575869766446\n", + " - removed_experts_[2, 3]: 0.23505958511107844\n", + " - removed_experts_[1, 5]: 0.20016380016380045\n", + " - removed_experts_[1, 4]: 0.27921201048534394\n", + " - removed_experts_[1, 3]: 0.2019454329774613\n", + " - removed_experts_[1, 2]: 0.2147839752957072\n" + ] + } + ] + }, + { + "cell_type": "code", + "source": [ + "import pandas as pd\n", + "import pingouin as pg\n", + "from itertools import combinations\n", + "\n", + "# Assuming df is your full DataFrame\n", + "aspects = ['bloom', 'subject', 'persian']\n", + "num_experts = 5\n", + "icc_results = {}\n", + "\n", + "for aspect in aspects:\n", + " # Store ICC results for each aspect\n", + " aspect_icc_results = {}\n", + "\n", + " # Generate all combinations of experts with three experts removed\n", + " for combo in combinations(range(1, num_experts + 1), num_experts - 3):\n", + " experts_to_keep = list(combo)\n", + " aspect_data = df[[f'expert{j}_{aspect}' for j in experts_to_keep]]\n", + "\n", + " # Calculate ICC(3) for each combination of experts\n", + " icc = pg.intraclass_corr(data=pd.melt(aspect_data.reset_index(), id_vars=['index']),\n", + " targets='index',\n", + " raters='variable',\n", + " ratings='value')\n", + " icc_value = icc[icc['Type'] == 'ICC3'].iloc[0]['ICC']\n", + "\n", + " # Store the ICC value with information about which experts were removed\n", + " removed_experts = [j for j in range(1, num_experts + 1) if j not in experts_to_keep]\n", + " aspect_icc_results[f'removed_experts_{removed_experts}'] = icc_value\n", + "\n", + " # Store the results for the current aspect\n", + " icc_results[aspect] = aspect_icc_results\n", + "\n", + "# Display the results\n", + "for aspect, results in icc_results.items():\n", + " print(f\"\\nICC(3) for {aspect.capitalize()} Aspect:\")\n", + " for removed_experts, icc_value in results.items():\n", + " print(f\" - {removed_experts}: {icc_value}\")\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "6LxiZbmIW7MX", + "outputId": "5a39f75f-6136-4c09-8658-0a575cfeb764" + }, + "execution_count": null, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\n", + "ICC(3) for Bloom Aspect:\n", + " - removed_experts_[3, 4, 5]: 0.3605138421378032\n", + " - removed_experts_[2, 4, 5]: 0.3488821418713774\n", + " - removed_experts_[2, 3, 5]: 0.2870111840771187\n", + " - removed_experts_[2, 3, 4]: 0.1867044307712342\n", + " - removed_experts_[1, 4, 5]: 0.3047538200339555\n", + " - removed_experts_[1, 3, 5]: 0.017228043034948216\n", + " - removed_experts_[1, 3, 4]: 0.3685734841078049\n", + " - removed_experts_[1, 2, 5]: 0.18416397877432106\n", + " - removed_experts_[1, 2, 4]: 0.2437865190528111\n", + " - removed_experts_[1, 2, 3]: 0.11481782443502522\n", + "\n", + "ICC(3) for Subject Aspect:\n", + " - removed_experts_[3, 4, 5]: 0.1754522567416755\n", + " - removed_experts_[2, 4, 5]: 0.03777681285280071\n", + " - removed_experts_[2, 3, 5]: -0.005209813393956866\n", + " - removed_experts_[2, 3, 4]: -0.19275608375778155\n", + " - removed_experts_[1, 4, 5]: 0.03261370484597021\n", + " - removed_experts_[1, 3, 5]: 0.11998310097169401\n", + " - removed_experts_[1, 3, 4]: -0.0037428651632829143\n", + " - removed_experts_[1, 2, 5]: -0.017602371687974357\n", + " - removed_experts_[1, 2, 4]: -0.018231292517007204\n", + " - removed_experts_[1, 2, 3]: -0.04574627884746704\n", + "\n", + "ICC(3) for Persian Aspect:\n", + " - removed_experts_[3, 4, 5]: 0.46856860206293544\n", + " - removed_experts_[2, 4, 5]: 0.3503985064981688\n", + " - removed_experts_[2, 3, 5]: 0.10053064958828935\n", + " - removed_experts_[2, 3, 4]: 0.4091008969057752\n", + " - removed_experts_[1, 4, 5]: 0.2829679231018469\n", + " - removed_experts_[1, 3, 5]: 0.1377817853922452\n", + " - removed_experts_[1, 3, 4]: 0.279812869626622\n", + " - removed_experts_[1, 2, 5]: 0.20218745914854697\n", + " - removed_experts_[1, 2, 4]: 0.27015250544662295\n", + " - removed_experts_[1, 2, 3]: 0.19148868039322386\n" + ] + } + ] + } + ] +} \ No newline at end of file